Quick tips/info: 10:28 when importing negative scans into camera raw (or any other software), setting the color profile to Adobe Neutral instead of Adobe Color is recommended. This will give you more accurate colors, especially greens as Adobe Neutral produces a flatter color profile while Adobe Color will exaggerate colors which are then further exaggerated when inverted. Obviously Adobe Color can be a deliberate choice as well, but if you want more accurate colors Adobe Neutral does not oversaturate them. 15:39 This will also fix the color consistency issues Will mentioned having. It's not the scanning setup it's the Default Color Profile. I've tried the subtractive process for the color mask in the past and it works generally on most images, however if there is a genuine color within the negative that matches the subtracted color, it will create really weird banding or something that looks like Solarization (Man Ray) on those parts of the image. You can kind of see this happening at 15:07 with the rim of the bottom blue (sea anemone?) and around the opening of the blue one above the lowest one to the left. Imacon scan has more blue in this area while the GFX scan has far less blue due to the subtraction. You can also see this happening at 16:20 which I believe is also related to the Adobe Color profile clipping/oversaturating the green channel. I just generally click the button to recalculate the histogram and manually adjust each color layer R, G, B while holding alt to just before the layer starts to clip in the actual image (Clipping the frame is fine). Also default Camera Raw noise reduction and sharpening are extremely aggressive. Personally I set NR to 0, detail 100, color NR to between 50-75 and sharpness to between 10-20 with all other sharpness settings, such as radius, to 0 or as low as possible. Grain looks far nicer this way while retaining more detail as NR is basically off besides color noise, so the comparison would be skewed even more to the GFX if Adobe's default Raw settings weren't so aggressive. For flatness of the negative, I've been experimenting with just placing it between 2 sheets of glass from Builders Warehouse. This has produced the flattest most consistent scans for me. Newton rings are a thing, however, they only seem to be appearing when there is a large amount of pressure placed on the glass/negative, i.e from a clamp to keep the sheets together. Otherwise, I've had no issues with Newton rings, but I will probably end up creating a better holder using a sheet of glass and a diffusion material such as a flat polycarbonate sheet mounted above tungsten lightbulbs.
8 месяцев назад+9
A even better tip is to create a true linear profile by using Adobes free DNG Profile Editor. Fuji files are using a pretty close to linear tone curve. For other brands it's a way bigger deal.
lol thank you. so much wisdom here i'm taking notes. could you share some tips on how to properly meter the negative? thanks a bunch mate!
8 месяцев назад+2
@@chris_jorge make sure that there is no clipping in the red channel. You don't have to overexpose. Exposing just right is ok and no clipping in the red channel and everything works find. I don't think that you need a large amount of dynamic range because the tonal range of the negative is pretty narrow. Between the brightest part and the darks much difference. Form a technical stand point I do not understand why William stacks the exposure.
Brilliant comment! I’m also heading towarda sandwitching my film between to pieces of glass. To avoid newton rings i think I’ll end up getting some scaner fluid.
That's very interesting comparison to see! As someone who's been dlsr scanning for years, I have a few suggestions that may or may not help you or anyone else: Negative Lab Pro is amazing and worth every penny ~ manually converting negatives is not fun Making sure your sensor is parallel to the film is very important. I saw you using a bubble at some point, so maybe you already know this, but the weight of that camera is definitely making the copy stand lean a noticeable bit. Changing to an adjustable mounting system or getting a sturdier copy stand would help. It's important to scan in total darkness (again, you might already know this, but your video shows you scanning next to a window so I just wanted to be sure) DSLR scanning has saved me SO much time, so I can't see myself every going back to using a dedicated scanner. Glad to see you sharing videos again! Love the new content
Thank you so much Noyel! Some excellent advice here, I tried NLP with imacon scans a number of years ago and didn't find the results to my liking, but I imagine things have changed and been refined a decent amount by now so will definitely revisit it if the occasion arises to try out inverting things again! The valoi kit actually came with a custom little mirror that I used in conjunction with the spirit level, the base for the setup could definitely have been more steady though, my main issue was that the valoi clamps were too small to fit on any of my sturdier desks so I ended up using a little bookshelf to mount things to haha! The scans were done in total darkness thankfully, I just made use of the window for the sake of having light for the video portion! Thanks so much for sharing your insights, so useful to be able to put things out and get feedback so that I know what to watch out for, the appeal of this DSLR business is indeed immense haha, if my imacon starts giving up I reckon I'll swiftly transition into the camera side of things! Also very glad to hear you're enjoying the new things!
Thanks a ton Marc! Still feel like I get lost in it all sometimes and have to stop and revisit things at a later date, but I enjoy the challenge of making things sit nicely with each other and bringing out what the negative has in it haha!
Been scanning with the gfx100s for a good couple of years now; with the gf120 you need fuji's macro adapter (the bigger one, the 45) to get it to 1:1 and then you can fill the frame entirely; I even manage to almost fill it up with a 35mm frame+sprockets. On the holder side - the Valoi is the best amongst the few I tried; they also have a plastic-y fresnel-y thingy to put on the light source and it gives you an extra stop of light. In my somewhat limited experience (including a Hassy X5 though) the gfx is currently the best scanning setup.
Amazing! So glad to hear from someone with experience with the setup! I had no idea I needed the macro adapter ahah, although that does make a lot of sense! I was flabbergasted when it came to trying to scan a 35mm frame and seeing how small the frame was with the focus from this lens alone haha. Thanks so much again for your insights!
@@willysheepskin same, when I got the 120 I had no idea it was 1:2 and it was a nasty surprise, haha. But the 45 macro adapter fixed that, now it lives glued to the lens.
Great video; I saw you were having problems with gradients at 16:00. It could be a banding issue; you can try switching the bit rate of your gfx file by going to image > mode > and change it from 8 to 16 bits/channel to see if anything changes.
Amazing comparison, big thanks. I'm using a Sony A7R3 and it usually smokes Epson 850V. Both machines are a level down to your setup, but still... It is amazing to see these cameras outperforming dedicated scanners. You earned a new subscriber. Keep up.
Thanks Marjan! I love the Sony A7R system, I can definitely see it being excellent for a camera scanning setup! Might have to make it the one I use for the inevitable comparison between the GFX and a different camera haha. Thanks for the sub!
Thanks for this. I have also tested both methods, but with my camera (Panasonic S5) I still sometimes have problems with color, especially of course with negatives (strange colors or color casts that are hard to correct). I have only rented a Flextight so far, but I feel like I got pleasing colors way quicker (just with Flexcolor) than with DSLR scans where I had to tinker more with the image. I also have the impression from your examples that there is simply more color information in the Flextight scans, which makes subtle color differences visible. I don't know what it is, maybe the CCD sensor, the Bayer filter or the internal post-processing, but I have the feeling that some information is not captured with the DSLR which is crucial for a good conversion. I guess it kind of makes sense that the Flextight performs so well in terms of color if it was designed for that exact purpose. But since the Flextight scanners are still so expensive, repair is difficult, and camera technology is getting better and cheaper, the DSLR route is probably the more sensible option for the future. Besides, the DSLR can even take photos and videos :D
Thank you for your insights! I think we’re on the same page regarding our findings haha, the flextight is definitely a more refined system but there’s so much going for a dslr setup that I’m very keen on trying to figure out a workflow that’d result in comparable color rendering! Some really good insights in the comments here with tips for getting more out of files
Those weird color issues may come from two (or more) things: 1. CameraRaw doing some internal weird stuff. Adobe makes input camera profiles themselves (by means of spectral sensitivity measurement of the sensors iirc), but what ends up in the tiff has a adobe specific output rendering applied. The latter part might do things which look good on normal images but not good on inverted and contrast enhanced negative scans. 2. The manual inversion in PS is prone to what I would call "wrong math". There are long threads in open source development forums for programs like RawTherapee and Darktable about the math that should be applied here (ideally in camera colorspace not what ACR spits out to PS). 3. color smoothing on the RAF to TIFF conversion inside ACR. At least to my eyes it looks like there is some heavy smoothing applied there. easy to check with an open source raw-developer where color-smoothing can be turned off in the debayering step. Maybe the glorious colored grain can be recovered. So IMHO there is a very good chance that the raw files from the GFX are actually better and contain more info without the wierd color issues you presented here. Thanks for the awesome comparison to the flextight! Very nerdy tech stuff, but if it enables you to keep shooting analog then I am not going to complain! Cheers from Germany!
Thanks sooo much for this man! Want to respond in advance to your footnote and say that it's awesome that you have this depth of knowledge but are still more invested in just shooting analog than nerding out on tech stuff haha! I myself am also more of a fan of making things than getting too techy around here, but have had many requests from folks to talk more about flextight things since the resources on the platform are quite limited when it comes to Imacons, that and the fact that I find the comment sections on videos like these tend to end up being immensely helpful resources (for the most part, some folks just want to be mean haha) So thank you very very much for contributing such marvellous insights to the "youtube comment encyclopedia" here haha. Also I should mention that the odds of Camera Raw being the culprit here are very high as I completely overlooked the importance of making sure it wasn't doing anything funky since I basically never use it. REALLY fascinating to know that PS itself puts out wrong math potentially with manual inversion, that's also something I never really considered, I like to imagine such expensive software as being infallible but clearly not. Your third observation also seems to have a high potential of being correct haha, I couldn't offer an insight since it's beyond my understanding but all of what you've said makes sense to me and thus the fact that all three elements could potentially be culprits makes it seem very likely that the fault is not in fact with the camera. Thanks so much again for your insights, so great to be able to learn new things from folks with knowledge like yours! Cheers to Germany from windy Cape Town haha!
@@willysheepskin You're most welcome! I was a bit surprised for the techy-nerdy video from you and thought I could finally share/contribute something. I think not being able to service the Imacon is a huge single point of failure for you as an analog photographer. So when I want to see more work from you in the future I gotta make sure you still can be happy even if the Imacon stops working (knock on wood that it doesn't happen too soon!).😁 PS assumes a color and workingspace by default. The "wrong math" I alluded to arises when this is not taken into account. PS doesn't know that you're trying to invert a negative, so it assumes whatever ACR assigned as output colorspace and "gamma". So the good news is, and it needs to be reiterated: scanning the negatives with *a* camera, not necessarily the GFX100, can most probably be done to a quality level that is almost ballpark imacon level. Needs locking down a certain workflow which might be tedious, but it's doable. (Also: I watch your channel because it's NOT focussing (pun intended) too much on the techy stuff. Me dipping my toes back into analog changes my photography process and the results that is very very hard to replicate with digital. Digital is convenient, Analog isn't and because of that the approach of taking and making a picture is dramatically different. I *think* that is what I see in your process too. And your results are just SO so good!) Much love.
Interesting video. I have to say, out of so many film photographers, I always liked your colors the most. That was certainly due to how the Imacon interpreted colors and what you developed from it. The GFX has a pixel shift mode so theoretically, you can use 400MP for scanning. I believe using exposure bracketing isn't really necessary with today's sensors. Usually, +1 overexposure is enough to scan everything perfectly; at least it has been for me for years with Fuji X cameras. The strange color bending transitions might even be caused by bracketing? That would be my first guess.
Thanks so much man! My initial tests with the GFX were single exposures but I noticed things clipping in the highlights since I was aiming for a bit over (out of luck, not experience ahah). I'd definitely easily believe I'd gone astray somewhere with that though since my experience with camera scanning is so limited. With regards to the bracketing causing the blending issues, the raw example I showed was a single exposure, without blending, and still had the problem so it's something outside of that I'm afraid. Could very well still be something caused by me though haha!
The head to head comparison I never knew I needed. Thanks Will! Surprising results, and comprehensive breakdown. I can’t afford to part from a flatbed just yet, but dslr scanning is very enticing. Cheers 🥂
Thanks for checking it out Justin! Nice new profile pic haha! I'm quite keen on a flatbed at the moment tbh, really want to try sort out my at home workflow and having quick contact sheets at home as opposed to using someone elses scanner would be subliiiiime! Cheers!!
Hello from colorado, usa! I love this comparison, i didnt realize how tough it would be to do the camera setup till i looked it up the price tag on the fuji 😢. I would be interested in a more low buck (sub 1000$) camera to see how much better the medium format is for doing print work. Been missing these videos, glad to see them again 🎉
Hello from Cape Town, South Africa! Yeah the camera is the catch, sigh, but you can do similar things with the more economical cameras from what I've heard in the comments here, so will definitely revisit this with a more budget friendly setup in the future! Thanks for checking this one out! Glad to be putting them out again haha, cheers man!
MY GUY! Thanks so much for making this video. I've been flip flopping between Imacon or GFX setup and I really wanted your take as a long-term Imacon user. Another pro to using a GFX setup is you also get a really damn good stills camera out of it.
Thanks so much for watching it Evan! Having the camera is indeed a pretty huge benefit haha, honestly who knows maybe you end up getting addicted to the speed of it and bypassing the analog bit altogether lolol (Can you tell I'm bored of dusting scans at the moment haha). Seriously though, the GFX is insane and I reckon if I bought one and went through the comments here to make sure I wasn't falling into some of the errors I likely made that resulted in the sketchy occurrences I shared here then it would bring nothing but joy across the board. Very glad to have put out the right thing at the right time for you haha, let me know how you go!
Very nice video, and beautiful photos! Great comparison, I just got the negative supply 120 kit and the vairo easy 135 kit. Going to pair it with the gfx 50sii. So this video makes me feel good about my choice lol. Didn't think about stacking exposures to get more dynamic range. Very useful to know, to get the most out of the negative. Guess I will take two photos and merge the same way you did in PS. You were using a stop or half a stop of difference between your two photos in the gfx?
Thank you kindly! I'd feel very good about my choice in your shoes too haha! I used 3 images here with a full stop between them, actually noticed while putting this together that the shadows and highlights still sometimes were crushed/clipped so if I were you I'd tinker with even further ranges potentially! This GFX had a setting where you could set it up to bracket however you like, really is a marvellous bit of kit haha!
@@willysheepskin Thanks alot for the reply with confirming the 3 exposures/stops and extra info on the clipping. Very useful to know, hopefully I'm able to pull out imacon level images like yourself. Keep up the awesome videos and photos. Always excited and happy to see a new video of yours pop up in my feed.
Very cool comparison and great explanation as to your hypothesis regarding specifically green color rendering between the Imacon and Fuji GFX 100S color science.
good comparison. thanks for the infos. especially liked the "color editing" technique! have you tried the Digitaliza (Lomography) medium format holder?
Wassup Willy! What I've done for a while with camera scans (after being schooled by Mr. Simon Maddock) is to insure that when importing a raw that any and all noise reduction (color and luminance) is turned off completely. I'm not sure what Photoshop / Camera Raw does by default, but both Lightroom and Capture 1 have defaults that aren't great if you want to reproduce the film's grain. After NR is turned off some light sharpening with a small radius brings back the grain really nicely. I'm also curious if the weirds greens you got were a result of combining exposures in the way you did. Were the colors still sometimes weird when working with just one shot?
Ahoy Karl! Yung Simon actually messaged me just as this video to similarly school me, I think you're both absolutely correct in that Camera Raw did some funky things here, I completely overlooked that being a potential thing since I never use Camera Raw so didn't realise it had the capacity to make things funky. Those colors though were indeed weird when working with just one shot, the raw scan example I used in the viddy was just one of the 3 bracketed exposures and it still had the funk going on so it's unlikely to have been caused by the stacking business, although it could've been camera raw up to something for sure potentially since I definitely didn't keep any sort of eye on what it was up to, Thanks for the insights though man, glad to be learning as I go here haha!
First of all, you look great man. Glad to see. Second of all, awesome video and very cool to hear your insights on this. And finally, thanks for touching on your color process, I’ve been dying to hear about it 🤍
Thanks so much Sander! Have been making a very concerted effort to build up my strength and such as much as possible and am feeling better and better for it. So glad you enjoyed the video too!
If you would be interested in hires images form gfx, you can try pixel shift (if you have a gfx that supports it). I would be very much interested in such a comparison. As for the weird green, it is weird. Fuji had their woes with xtrans sensors that have been know to have mushi greens because of how the green subpixels are arrayed, but afain gfx should be using a bayer sensor which has no reason to respond in such a way to green color.
Cheers Michal! If I get the chance to I'll definitely try some pixel shift things next time, would be very interested to see about it also! Interesting to hear that there is precedent for the Fuji struggling with greens, so many potential causes and solutions have been identified in the comments here haha, could make a whole video just trying them out!
We have made a home-brew (aka cheap) dslr scanning setup in our community dark room. Old enlarger stand to hold the camera, 3D printed film holders, Kaiser light source and two planes (enlarger base and a light source base) with threaded feet in the corners to make everything somewhat level. Works well enough with Sony A7II and Sigma 70mm f2.8 ART Macro lens. Tho the post processing had always been a massive drag for me... B&W is fine, but colour broke me. I seldom got decent results and I blame it for my kind of a photography burn-out (cause I still like taking pictures, they just end up developped and accupying space on our drying racks or as RAW files in my PC...) The way you do it seems much less complicated then what I used (even tried the neg. lab pro thing and didn't end up liking it), so I look forward to give it a try. So this is in a way a thank you. There might be hope for me still :D (or I just return to our ol'Epson 500)
Ah fret not Tomáš! I myself am a large fan of making things B&W when the color isn't working out haha, there are some images that just don't want to render properly on Portra but look great in B&W ahah! I try to zen out when doing post processing things, but will admit that I do sometimes rage quit fi things aren't going my way, in fact the latest bunch of negs I've got aren't working out well so I'm taking a break from em haha. I believe in you though, sometimes things make more sense once you come back to them after a while, the key is to just keep going and eventually things make more sense lol!
awesome video, thanks for the detailed comparison! at 21:07, i wonder if that is lens aberration on the GFX? i.e., if you were to use an APO lens, would that go away?
The imacon isn't really something one would want to move around haha! The DSLR scanning setup also would be best suited to just staying put. I think the going rate for an Imacon 343 is around $4000 these days, and a GFX 100S body will set you back about the same. If you got a fuji lens for it - which I used here but isn't actually the best lens for the job - it'd be around $7000 total I think. So more expensive but the equipment is about 30 years newer and has a warranty haha! If you just got GFX body and adapted a lens to it to work - which would likely mean you get it to focus closer too, it'd cost about the same as the 343, and substantially less than the highest end flextight X5 scanner!
That probably would've been a good thing to try hey, although I'm not sure if it would achieve the same level of flatness if it weren't being held under tension, could've been very interesting though! Perhaps next time haha!
II have had good experiences with compensating the orange mask of the negatives using an 80b blue filter at camera "scanning". As a result, the RAW files have to be less tortured when compensating. In my opinion, bracketing is not necessary with the wide dynamic range of your camera. Colour negatives are so flat in terms of contrast that one exposure is sufficient. I convert with the Negmaster plugin for Photoshop. I find the results much more balanced in terms of colour, modulation and the highlights are also much better defined. Did you try the highres mode of the GFX to get higher resolution?
I reckon the end result was similar, but that HDR is more likely to make things get funky since as a single process I'd imagine it being more likely to blend things in strange ways. The exposure stacking just basically takes the lightest bits and the darkest bits from each frame and mixes them together in a clean manner to make one image with the best of both whilst retaining detail.
You say the GFX smoked the imacon in color rendering on that anemone photo, but the GFX rendered the whole bottom section of anemones as pink whereas the imacon showed that they're actually different colors. There was a blue one up top that the GFX really desaturated too. Were you talking specifically about color noise?
Hey Will! I currently scan on a Fuji GFX 50R mounted to a Pentax 645 120mm F/4 Macro, with the exact Valoi setup you have, and do the conversion in Lightroom with Negative Lab Pro. My biggest struggle with this setup has always been that the focus depth is incredibly shallow, and often the negative can be out of focus, somewhere on the frame. What I do to ensure my negative is as parallel as possible to the front of the lens is with a level app on my phone that shows degrees of angle to the nearest hundredth decimal point, and line up the holder with the negative and the camera that way. It's possible that going to a smaller sensor camera could solve that problem, as APS-C and Full Frame will both have more focus depth. I suppose the question is, is the potential downgrade in sensor quality worth it for the increase in focus depth?
I definitely agree that sometimes it needs a lot of contrast, but that's a good thing because it means it's a flat image and there's a lot of detail.@@Xinbaby999
I'm very jealous of your setup Eric haha! Very interesting to hear that your struggle is mainly with depth of field, I feel like if the negative was perfectly flat the depth of field wouldn't be as much of an issue though, I alluded to there being means to keep things flatter in the video, my main guess would be mounting with some kind of anti newton glass or even fluid mounting but that adds a bunch of hassle to what is really a lovely process with the camera scanning setup. Have you given anything outside of the valoi holder a go yet?
It's true that the Valoi holder doesn't hold the negative perfectly flat, but *usually* it's enough for me, plus the speed of being able to scan a negative, and then push the roll through to the next negative, to scan again, is very nice. With that said, while I haven't tried a different mounting setup yet, it's been something I've been thinking about for a while now. It's also worth noting that with 35mm, this isn't really a problem as the negative sits a lot flatter than 120, and with that Pentax lens plus an extension tube, it's possible to get beyond 1:1 magnification. @@willysheepskin
Moooofy! That'd be interesting to look into for sure! And noo I mean RA-4 since I imagine they'd use darkroom colors for reference in terms of how the negative is expected to render once inverted! All conjecture though haha, although I'd looooove to meet the folks behind the Imacon color science to give them a hug and pick their brains!
What an upgrade you lucky soul! What lens have you paired with it? I was using the fuji macro here but it maxed out at like 9000 pixels of actual negative, I've seen folks use adapted setups that allow for more of the frame filled which would be quite useful lol!
@@willysheepskin I sold my X-T4 and XF lenses and took the money from another gig to switch to GFX. Wasn't able to afford anything but the 35-70 "kit" lens, but I have a ton of M645 and Nikon glass that I adapt for different subjects. For scanning, I'm using the Mamiya 120mm F4, no focal reducer, just a dummy adapter. With that you can go full edge to edge on anything 645 and up as it's a 1:1 macro (and all those formats are larger than 44x33mm). If you're willing, I'd love to try inverting that bee photo with Negative Lab Pro and see if I get the same posterization in the greens. I'd want to try with both a RAW and the TIFF merged file.
@@nathanjohnson5304 Great kit! Very interested in seeing how NLP handles that posterization. In my experience the color noise you get from Imacon .tiff is reduced when using .fff files straight from Imacon then inverted in NLP.
Most interesting. I think the big point is the one you made. Good scanners are becoming unobtainium. We need to make DSLR scanning work. It looks like there are things to explore -negative flatness, color depth, color space, methodology (exposure stacking vs not, inversion methods) and more. One commenter mentioned stitching areas of the negative together for more resolution. That really seems like it shouldn't work. But I've done it a fair number of times now and damned if it doesn't. At least I haven't caught it out yet. Or maybe I've just been lucky. Might be worth trying if you repeat the experiment with a smaller camera. It's also worth noting that you can actually use the digital camera to take pictures :-). So it can maybe amortize it's cost that way.
Ahahah it is most definitely good to keep in mind that the Camera can take pictures alongside scanning lol! Thanks for taking the time to watch and offer some thoughts on this Carl! Really appreciate it!
Hi Will, that's was very interesting to see comparison with high end scanner. I could tell from affordable consumer level perspective: last week I've sold my flatbed Epson V600 Photo because even my aps-c Fujifilm XT3 makes sharper and I feel overall better results. The only downside - resolution is not enough to scan medium format film, so going to upgrade to some high res FF camera.
Thanks Artur! Great to know that things are better than an epson with an XT3! Maybe look into pixel shift manoeuvres, you might get the resolution you're looking for out of that process!
Might try that Alex Burke inversion method, mines a bit convoluted. Are you just treating photoshop colour grading like you would an RA4 darkroom print. I always enjoy a scanning video, I use the x5 at my local facility and pray it doesn’t die.
I do indeed try to treat it just that way! Feels like it'll maintain more of the true rendering of the film that way than trying to get too funky with it haha. I'd hope that the X5 has many more years in it than my poor ole 343, they only stopped selling em in like 2019 after all!
I ended up outsourcing a bunch of STL files to be 3d printed which werw designed by Stuart Jenkins for a film holder. I found his design much more consisted in getting flat negs scanned. I'm only using a 5DS though, but I could still tell they were sharper than previous holders. Also manual inversion is OK, but try out Film Lab beta 3. Because it renders negs linear, I see grain come through a different way than manual inversion in PS. Its the closest I've gotten to a noritsu too with colors
Great video! Thank you, wanted to see a comparison of these two for a moment. I wonder if sony or Panasonic cameras will render greens better than Fuji.
I really enjoy your scanning videos. I, too, am an Imacon scanner. However, my digital media lab , at the Center for Photography Woodstock just installed a Fuji GFX scanning station. I will have to give it to try now.
Hey, Will, just a follow-up. I have not used CPW Kingstons Fuji rig, but I did set up my own Canon 5D3 copy rig. I was able to stitch 4 frames together and copy a 645 negative. Sharpness was only discernable at 100% magnification. Color was pretty close, but I gave the win to the Imacon. Finally, the Imacon is also way easier to use.
Thank you for this vid! Great comparison. I am surprised by this but to me the imacon scans were leagues better. Of course this may be down to your conversion process and corrections but the comparison at 21:32 is a good example of this. The Imacon is wayyyy more life like and "3 dimensional" if that even makes sense. The whole image feels softer and warmer on the Imacon, whereas the Fuji scan seems "crushed" and "brittle".
Thanks for checking it out! I definitely still prefer the imacon in terms of it's rendering too, I think the image you're reffering too might be a bit more crushed because I also couldn't get the shadows to match in detail so just bumped the contrast so the scan wouldn't look too wonky. There are a number of images I scanned where the shadows definitely lose a lot though, there's a signboard image where the camera scan crushes a lot of what the imacon can pick up in the darker bits. I'd be curious to see if I could get more out of it though with more intense bracketing, if I ever get the chance again I'll have to see haha!
Interesting video, Will. Really good point about the Imacon maintaining tension on the negative by curving the holder! Contrary to others, I would not recommend Negative Lab Pro unless you have tried their process and like their results ahead of everything else. Manual inversion is better, it's just tedious. I thought the Imacon was exhibiting some chroma noise in the shadows, typical of CCD sensors. Unfortunate because exposure really shouldn't be an issue for scanning negatives. Regarding the colour differences: Did you encounter any artefacts on shots without exposure bracketing? Simple 1D operations really shouldn't be breaking the colours like that. 'Colour science' is a little bit of a gimmicky marketing term because the colour of a digital image is dictated by the dyes from the CFA and the 3x3 colourimetric matrix. This can be undone, giving you more control but the process is a little tedious. Once tonemapped, most digital cameras look very similar. I also prefer the colours on the Imacon. If you like the starting point of the Imacon ahead of the GFX, you can tonemap them, and if you're willing to shoot a few charts, you can match the colours almost perfectly using data interpolation.
Thanks so much for the insights Amaraldo! That's actually fascinating to know, I really appreciate you sharing your knowledge here! It would be worth the effort if the result was color matched gfx and imacon scans for sure! The banding in the Bee image was still there in the raw scan unfortunately, the example I showed here was one of the three bracketed exposures and it still had the issue, but others have suggested that perhaps it had something to do with the light source setup since it could've impacted the detail with which certain colors were projected. I love that there's so much advice regarding the mystery in this comment section though haha! Cheers!
I'd really love to see some evaluations of how a Sony Mirrorless camera would do. Personally, I scan with an A7RII, but I've struggled immensely getting colors I like.
I’ve been scanning my archives, transparency, and black-and-white negatives for the past three years with the GFX 100s, using a bellow extension with a Schneider 90mm enlarger lens. utilize the pixel shift feature if you want a massive file that competes with any scanner on the market today. With the Fuji not only are you able to do amazing scans, if your properly equipped with a magnetized film holder system, made of heavy gauge metal (not plastic), then you can get through thousands of negatives in a single day. Something no scanner is going to give you at this quality and price range. You also have the added benefit of recording the negative in 4k as you feed the negative through the holder. In actuality any digital camera with the proper professional copy stand and negative holder system, with either a bellows extension or macro lens would suffice in archiving all your film work.
Yeah, I ended up rescanning my entire print archive in like 30 minutes which would've taken me at least a week solid on my imacon, kind of blew my mind!
Really interesting Video ! I use a Sony A7R2 for scanning with 42mp. It's quite sharp, but I have to say even with the lowest ISO setting it introduces a lot of color noise to the image. I tested it against a Panasonic GH5 with 12mp, a Sony A7 4 with I think 32mp and even a Canon R6 (which I think gave me the best results color wise) and all three were significantly cleaner. You can get a A7R2 used for quite cheap, but I would recommend spending just a bit more to get an A7R 3.
Thanks Daniel! Interesting to hear that the GH5 was sharper than the A7r2!My video cam is a GH5 so maybe I could use it for comparison to the GFX when the time comes to see about less expensive cameras haha! I've used th A7Riv a bunch for bts work on commercials, I reaaally like that camera haha, would be high on the list if I ever looked into getting a Digi stills cam!
I use a very simple DSLR scanning setup with my old Canon EOS 7D from 2009. It introduces A LOT of colour noise, but works very well for me as I only shoot black and white. It's not one for the most high res scans with an old 18mp sensor, but it gives me quite good results with b/w.
The camera have a multi-shot mode, where it shifts the sensor and captures 16 shots. By utilizing the Fujifilm Pixel Shift Combiner software in accurate color mode, you can seamlessly combine these images to create a final image without the need for the demosaicing process.
How about creating a LUT to translate the Imacon color space to the GFX? I did it for Frontier colors. I am pretty sure that the color space of the Imacon is smaller for some colors like the greens for example.
Hmmm that'd be fascinating to try! From my experiments though it's always seemed like LUTs rely on consistency in the source frame and having the colors presented match the colors that the LUT is based on, which gets tricky with varying lighting situations. But I also am not particularly experienced with them, so perhaps there are methods to get around that? Keen to give it a go though, if I manage something exciting I'll share it in a video haha!
8 месяцев назад
@@willysheepskin LUTs are like dictionaries. You would only need the original color space of the Flextight and could then translate it to ProPhotoRGB. Fluctuations in the material would not be relevant because the LUT is not created based on a negative, but rather only determines what the Flextight color space should look like translated into ProPhotoRGB. However, it may be that the differences in the color representation of certain colors can also be attributed to the light source. You would have to measure the Flextight and compare it with the options that we can use to take photos. If you take it very seriously, the way the sensor works also plays a role, but that's only the last 0.5%, which can be ignored.
Cheers Rhett! I'd be keen to do a video going in depth on the imacon side of things, but the amount of negative feedback I get every time I do a video like this puts me off it for a whiiiiile ahah, we'll see though maybe I brave it sometime, it is a bit of a safer bet since it is my main workflow method whereas the GFX was a brand new thing
@@willysheepskin again I'd love to see it. Your color balancing is so nice and the flexcolor software isn't that intuitive, so I'd love to hear from a pro :) also I promise to be mean to everyone who's mean to you in the comments 🫡 maybe that helps the decision ahaha
I tested with GFX and saw the same issue. I tested with Canon EOS 5DS and red gradients were *also* borked. Another takeaway I had from experimenting is that found that the grain looks screwy on high ISO film, very bad on black and white especially. It's something to do with the Bayer color filter array. For example, Delta 3200 looks TERRIBLE, *unless* you use a B&W converted camera to convert it -or- you use a Pentax with pixel shift. Pentax does some crazy black magic with their composite mode, where they move the sensor to cancel the Bayer color filter array. I tested Fuji, Sony, Panasonic, and Olympus pixel shift. All of them had issues, whether it was color, or banding, or grain. Pentax? Amazing conversions. What I recommend is an el-cheapo K70 using pixel shift for 35mm or smaller formats. For medium format, if I'm in a hurry, I still use the K70. If not, I use the V850, where I still get 55mp of real optical resolution out of 6x8 or 31mp out of 645. If I need more than that, I send it out. I've tested most of the commercial film holder options - the Valoi Easy35 hands-down wins for easy of setup and perfect alignment. For 120/220, I just use a cheap and effective EFH v3. With either option, I use a Kinetronics Staticvac right before the negative slides through the holder to avoid almost all dust. Awesome video - that Imacon is just Amazing. Super interesting to see the difference between that and the Fuji. EDIT: Forgot, I plant to try one of these for 120 - look on Etsy for "BlackscaleLab" - they make a quick setup film holder like the Easy35, which is probably worth it because it makes the setup time super fast:)
Thanks for this Jeremy! Quite glad to hear that I'm not nuts haha, although a decent amount of potential solutions have been offered here, the fact that you've tested such a variety of setups and have found them all to be lacking is really useful! Also now I have a reason to sing the praises of a modern Pentax in conversation which is a treat haha! Thanks also for the advice regarding the holder, it sounds very useful, and its so great to have someone with your experience offering insights so that we can all learn!
Lovely video and super interesting. Ignore the haters, they mostly have no idea what they're talking about. I'd be bummed if you didn't post unique stuff like this because some basement dwellers were nit picking.
I’m having the same dilemma about how to scan my 120 film except I have X2D (and 500C/M as my film camera). What is the max image resolution you can get from Imacon? My lab offers about 4000x4000 scans with Imacon 🤔
@@willysheepskin that’s insane! I wonder what the grain limit for 120 film is. I’ve seen 4x5 was scanned to the size surpassing 700Mpx (that’s like 30K on the longer side) in a drum scanner and they still had details there. I’d expect 120 film might be a quarter of that 🤔
No one would suggest you’re not able to express your views. However what is remarkable about this piece is all that you leave out - and you do not spend any time on the incredible piece of tech Flexcolor .fff files are (Lab color space, raw Scans), and the incredible sharpening range, noise management at scan (it’s not fixed) the dMax capacity of the scanner etc. etc And etc. And the Fuji comparison - colour space, lens choice, camera settings colour space file type …. And the usefulness of the Fuji in high speed processing. The use of the Imacon can be sorted in about 20 minutes and RUclips references abound. Quite possibly a useful comparison is wet mounting the v850 with your little 343. You need to take 3 exposures of a colour neg - 12-14 stops digital capture (in a single exposure) of a 6 stop piece of film - that doesn’t make sense but again your choice. So - I just don’t get what you’re trying to prove or show here - try scanning and shooting an IT-8 target. Your excercise is simply a reasonable record of not knowing much about either - or - choosing not to explain it. It’s weird. Imacon is not an archaic dinosaur - but both are in the hands of those who don’t know how to use them. Again - not saying that’s you - but you leave so much out how it could be useful for anyone is beyond me.
I've read that some people will sandwich the negative between two pieces of anti-newton ring glass to get ultimate flatness. The only real drawback is your setup must be scrupulously clean. No smudges, and no dust. The GFX 50SII isn't that expensive and some of the older GFX 50R and the 50S are coming down in price on the used market. For some reason though the older 50R demands a higher price than the 50S. I've seen the 50S on ebay for between $1200 and $2000 USD for a body only. The Nikon D850 or the Nikon Z7 are at 45Megapixels and The same goes for the Canon EOS R and EOS R5. So there's that for an alternative to the FujiFilm. Did you get to take and landscape shots with the GFX 50SII? Even the RAW images without processing are impressive. To my mind they are as good as the 6x6 images from my Hasselblad 503CX on E100 film.
I wiiiish I'd had the chance to take the camera outside! It was a 100s so the files would've been sublime I'm sure. Sandwiching the neg as you describe would definitely sort out the flatness issue, but as you say smudges and dust would become a bit of a nightmare most likely haha! I've also somehow frequently managed to get newton rings even from anti newton ring glass somehow, so that'd probably be infuriating too. Thanks so much for your insights though Bob, super useful to have a bit of perspective on what some of the older cameras go for now, those numbers are a bit closer to the realm of reason than this 100S haha! All the best to you and yours as always!
You explain how the Imacon setup guarantees the best possible film flatness, but my question is....how much does this actually matter? For example, if you'd be using the 120/4 macro lens on the GFX, even wide open at F/4, at the minimum focus distance of 45cm, depth of field would be 3mm. I realize the "acceptable sharpness" that depth of field calculators use is not always best, but also consider that you probably won't use the lens wide open, but rather stopped down to get the best balance between lens aberrations and diffraction limiting your resolution. At F/8 depth of field would be 6mm, at F/11 8.5mm. Even if we use a more rigorous definition for "acceptable sharpness", that should still leave us with 1-2mm of depth of field which seems reasonable for a simple border-only film holder to me. Personally I've been using my Pentax K-3 ii 24MP APS-C camera with the 35/2.8 Limited Macro lens set to F/5.6 and the very simple Pixlr-Latr film holder for anything from 16mm to 13x18cm negatives and always been happy with the results - I already had the camera and lens, so getting a scanner with comparable image quality probably would've been at least two times as expensive as getting a copy stand, light source and film holder. I only shoot B/W and also have been perfectly happy with the free and open source darktable software for converting them. To be fair I don't make prints from files, I only view the results on digital screens or print straight from the negative.
Ahoy Leon! Very interesting question actually. I've never thought how much it would matter with the GFX kit, I just know on the imacon even like 1/4 mm discrepancy in flatness throws the grain completely out of focus. Someone commented here about how their main issue is with the depth of field of their GFX resulting in variable flatness irrespective of aperture, so I'd imagine there is the potential for that to be a thing too. They also said they were using the same holder as me so the slight variations in flatness I picked up could be what was causing issues for them. You make a good point though, perhaps there's some kind of math that comes into play when the area that technically counts as being in focus (the film plane) is so thin, but again your points regarding theoretical DOF in relation to apertures you've used make sense to me. That alongside you saying your results are great has me a bit stumped haha, maybe your holder is just super rad lol!
Great video as usual! The bayer filter on the gfx sensor has 2 green pixels, 1 red pixel and 1 blue pixel in each 4x4 pixel window (see wikipedia) and the missing values (each pixel needs all 3 channels) are interpolated by the raw processor (in this case PS), maybe the strange halo you are seeing is an artifact of this interpolation in that particular photo? That would cause artifacts in other photos as well when viewed at 100% though. Linear scanners do not use a bayer filter and does not require interpolation as their linear array is 3 pixels wide (one pixel for each channel) and it moves across the film. Another theory is that this part of the image was overexposed (or close to) and PS treated those values badly. Or maybe your curve changes caused it? Looks like a hue shift or some kind of clamping. I don't quite understand the filmic and AgX tone mappers but they should correct for this and maybe PS does not by default. Will link a video on hue shift here:
@@Alex-c3o1oX-trans and Bayer filters function the same in that both require interpolation for missing color channels, the only difference between the two is the pattern of the filter.
Wow Gustav thanks so much for this insight, exactly the sort of information I was keen to hear about from someone that clearly knows a bunch more about the GFX than I do! Will be checking out this video shortly, really appreciate your insights and help!
@@willysheepskinThanks, but I'm by no means an expert (especially on the GFX)! I'm just a curious hobbyist interested in how digital RAW processors work under-the-hood as they decide so much of the look of digital pictures nowadays. I've thought about it some more and I don't think it's the Bayer/X-trans filter that cause it (my first hunch) as the effect of interpolation should only really be visible at very small scales (4-6 pixels wide) and the artifact in your picture is much larger. It could still be the second theory, that the raw editor does not handle overexposure well and introduces a hue shift. Could you try looking at the raw image in Darktable with the Filmic modifier and see if you can still spot it? What program are you using to turn the raw image into a tiff, by the way? Noticed now that the GFX image was a tiff.
How could 16:40 be an issue with with GFX rendering greens when it's recording the negative? My guess would be that it's more-so the merging method of the bracketed images
yo. few things about 17:05.there are multiple reasons for that. a) it looks more like over-exposure problem, i have that sometimes with my D850 when I scan films. b) it might happen because you merge layers using "multiply". not the best way to do it. to understand whats happening - double click on layer, see the blend if beneath. hold alt and grad the pointing arrows on the right side (it should split). one for shadows, one for highlights. try to do the same without holding alt - thats what multiply does. c). you`re working in 8bit. using gfx and keeping at 8bit - is a waste of gfx DR, sometimes this falling off may be fixed just by changing bits :). sometimes you can just manually change to 16 bit and that gradation will be smooth again, if you imported the file in 16bit in first place c)as someone mentioned in comments, Adobe Color profiles are not the best. Try Capture One. waaay better for start (especially if you work just with curves, you need a good starting base), and super easy to change raw profiles. Than export it in 16bit to PS/
Thanks so much for the advice Tim! Great to learn more about merging things, haven't ever really used it outside of trying to get more out of these scans hah! Really appreciate you sharing your knowledge here, so great to have resources like this to refer back to if I get stuck with the camera scanning business again!
@@willysheepskin no problem. its a technique for high end retouchers. but so that you know - curves perform different in 8bit and 16bit. if you have color fallout - just change to 16 bit, it will make gradation smooth "magically" dissapear :) you can see it on seemless backgrounds, sky and greys.
No need to apologise! You do indeed unfortunately, with Windows the OS thats worked the best for me, flawlessly for years now, has been windows XP. It depends on the scanner but in general a computer that matches the age of the scanner is a good bet, but older doesn't hurt since they'll all run on XP. With regards to Apple machines, G4 and G5 computers are good for the older ones, but modern machines running up to OS Mojave will still work, any OS after Mojave (Think Mojave might not work either?) is too new for the software the machines run on (Flexcolor) to work. In general the best bet for an older one is a Win XP machine, and newer ones run nicely on Macs so long as the OS isn't too new. Thanks for checking this out!
One way you may be able to get better results from the GFX is to try different RAW processors. If you going to pixel peep Adobe is not the best processor.
If you're being snarky then it probably stands for listening RRReally closely to what I say at 9:45 ? If you're honestly curious then it stands for Reflex, and indicates that the camera operates with a prism/ mirror (Unless its a TLR) and that you're actually looking at the image that will be exposed as opposed to an approximation through a viewfinder. Which doesn't actually apply to the camera I'm using in this video, but someone who felt like being snarky for no reason might comment on it, ignoring the fact that DSLR scanning is a descriptor of the process and not the specific equipment and then get told to listen to what I said at 9:45
Ahaha you and me both my friend, even if I sold all my cameras I'd still only be able to afford a body for it loool, but I'm happy with what I've got going, just good to know there's an option in case of the worst happening lol!
DSLR Scanning allow for a lot of flexibility over most types of scanning. You can do multishot noise reduction, or create ever larger images with higher magnification lenses. 1:1 on a GFX might be 50/100MP but if you used a higher magnification lens, 2:1, 5:1 etc you can easily stitch 300MP to over 800MP. Scanners can't do that. You are limited in terms of your maximum resolution. Do you need a 25,000x25,000 image? No. But if you did, you could create that natively with a DSLR scan and print an image at 300 DPI that's the size of a soccer field. Or you could go even crazier and use filters for per channel images that you can combine together later. Of course you are limited by the size of the grain and the information it retains which varies by format. But 100MP vs a 10MP image will always produce a sharper image regardless. Using a Canon 800D and 60mm 2.8 USM, I've stitched together, 2, 4, 8, 12, and 16 images of a single 6x6 negative and I was still able to physically see more detail in the negative the closer I got to 1:1 and the more images I used. It produced 6GB images in Photoshop but I am able to zoom into an image of a house I was standing over 20-30 meters away from and see tiny cracks in the sandstone wall. And sure color bit depth might be less than the Imacon, but 14 bit per pixel images are the standard and current technology is barely able to support 10 bit color anyway, so that more to do with long term archival or specialized printing processes that don't have a finite color range.
It's so exciting to hear about the marvels this scanning medium has to offer if one gets invested in it haha! Natively printing something that size is insane, honestly have always been mega stoked with A0 prints natively but the potential for sizes like you mention is awesome haha! Glad to hear from someone with your experiences that some of my observations make sense, and thanks in advance for the supremely informative comment you left a bit after this one!
Biggest issue I see with your comparison is the aperture. F16 is way too much, you should have used f8 max. on the GFX. Also using ISO400 is not a good idea as that's the 'worst' dynamic range you can get from the GFX100 sensor before the second native ISO kicks in. You basically lose 2 stops of photographic dynamic range. Even so it would have been enough for the negatives but they were still underexposed by at least 1.5 stops. So you've made it way harder for yourself by stopping down to f16 because that lowered the resolution, reduced the available light, making it necessary to increase the ISO, reducing the dynamic range, making it necessary to make brackets.
3:00 If you have a proper macro lens (1:1 ratio) than you can achieve similar functionality to your Imacon scanner. When I scan 120 film using my mirrorless camera, it's usually right after scanning 35mm frames. I will leave everything as I had it for 35mm and take 6 shots of the 120 film by sliding the light source (which the holder is on) around on my table. I will then convert everything in lightroom using negative lab pro and stich the 6 files (3 at a time) together as a panorama. A simple square crop (I shoot 6x6) and I have a 10k image file (my camera is APS-c ~6k horizontal) with amazing sharpness. I have run into problems with this technique when scanning images that have large brightness differences across the 6 frames. For example, I shot a hydro dam which was getting lit by the sunset (~4pm in winter) and the foreground was dark and lower in the frame. When I did the conversion, there was severe color differences between the frames, even the parts that overlapped so I couldn't stitch them together. For images with more consistent light throughout, this technique has worked fantastic and doesn't give me problems in B&W either. Similar to the Imacon, it takes more time lol. With your inversion process, you might not run into the same color/brightness issues that I run into with negative lab pro. NLP is also worth every penny, I highly, highly recommend you use it for digital scanning. They have a free trial so you can try before you buy. I think it would fix the color issues you are having with the gfx currently. 7:50 You don't label your archive sheets.... you psycho😆 21:51 Look into the Essential Film holder. It will be a better option than the Valorie holder but it does take quite a awhile to ship from the seller. If you got money to spend, Negative Supply supposedly make the best holders on the market. They are made of steel and you can move film through with a winding knob. You might need to sell a kidney for one of them though, they are quite expensive and probably over engineered.
The EFH isn't a better option, both of them aren't great for 120 because of film flatness, also your moving around stitching shows that flatness of field is not something your scan is concerned with because those micro movements would be enough to mess with sharpness.
@@Alex-c3o1o I mean I know from doing it you are wrong and I know from being a macro photographer on digital that my DoF is enough to cover very small changes. As for the EFH, I haven't found a single review anywhere that states it's not better than a Valorie so....
@@ZeLoShady the depth of field isn't enough if you are using a macro lens that is optimised at around f4, which you would need because at 1 to 1 f4 is f8. And with any decently high pixel sensor you would otherwise lose sharpness due to the airy disk being bigger than the pixels, which you should know as "macro photographer". And just because you can't do proper research about the EFH does not mean that a film holder where you can literally slide the negative around, meaning there is no tension, can hold the 120 negative flat enough. Also when you aren't using pixel shift you will have the debayering errors anyway, so even there your setup is suboptimal. Also also even the negative supply film holder has issues with tension if you use cut negatives like will here. Just because something is good enough for you does not mean it is optimal or can even come remotely close to a flextight.
@@Alex-c3o1o @user-vp5yo5ys5r LOL all I did was suggest a better holder than a Valorie, not a flextight. Obviously something that clamps the film and then bends it to apply tention is better but I am not saying my setup is superior, not sure why you think I'm making that claim.. Someone is a bit of a snob eh. Unfortunately for you, my setup is sharpest at F11 and because I use APS-c and not medium format, I get a lot more DoF in macro. As far as airy disc goes, who gives a fuck, that's a problem I'm not looking to solve. If I make a print of a film image scanned with my setup vs scanned with an Imacon, no one will be able see any signs of Airy Disc anyways! Maybe if the print was 10 feet long it could matter but there won't be enough print resolution on an 8x10 to see it which makes that a non-issue that only snobs would care about. There's the theoretical world and then there's the real world bud.
Thanks for the advice Alex! Yeah if I were to use this setup again I'd be sure to source a macro collar for the lens since I didn't actually know it was a thing ahah, I kind of just figured the macro lens would just macro looool! Will also revisit NLP if I come back to this topic sometime as a lot of people have said good things so I'm figuring its come a long way since I last tried it. I doooo label my negatives, that folder just isn't labelled since I got behind at some point and haven't bothered catching up yet ahah, although it's been multiple years now so maybe I have in fact simply transitioned from being a labler to not being a labler, and am in fact a bit nuts ahah! Will be looking into that holder for sure for future reference! Thanks again for the advice, appreciate you taking the time!
Fuji GFX is a Mirrorles medium format camera NOT a DSLR. The video seem to be a review of the imacon scanner mashed together with the comparison intended rather than focusing on the comparison intended. The results from a mirrorless/dslr scan have the respective camera colour science interpretation. A purpose scanner with a dedicated software which has the correct film colour profile will yield the more true to the negative colour interpretation. That's why all scanners use CCD sensor instead of CMOS sensor. At the end it's all boiling down to using the right negative for the right job. If you use Kodak portra for landscape then yes you need to manipulate the colours and going thru the hassle of a ccd scan is pointless as the colours will not be great.
Quick tips/info: 10:28 when importing negative scans into camera raw (or any other software), setting the color profile to Adobe Neutral instead of Adobe Color is recommended. This will give you more accurate colors, especially greens as Adobe Neutral produces a flatter color profile while Adobe Color will exaggerate colors which are then further exaggerated when inverted. Obviously Adobe Color can be a deliberate choice as well, but if you want more accurate colors Adobe Neutral does not oversaturate them. 15:39 This will also fix the color consistency issues Will mentioned having. It's not the scanning setup it's the Default Color Profile.
I've tried the subtractive process for the color mask in the past and it works generally on most images, however if there is a genuine color within the negative that matches the subtracted color, it will create really weird banding or something that looks like Solarization (Man Ray) on those parts of the image. You can kind of see this happening at 15:07 with the rim of the bottom blue (sea anemone?) and around the opening of the blue one above the lowest one to the left. Imacon scan has more blue in this area while the GFX scan has far less blue due to the subtraction. You can also see this happening at 16:20 which I believe is also related to the Adobe Color profile clipping/oversaturating the green channel. I just generally click the button to recalculate the histogram and manually adjust each color layer R, G, B while holding alt to just before the layer starts to clip in the actual image (Clipping the frame is fine).
Also default Camera Raw noise reduction and sharpening are extremely aggressive. Personally I set NR to 0, detail 100, color NR to between 50-75 and sharpness to between 10-20 with all other sharpness settings, such as radius, to 0 or as low as possible. Grain looks far nicer this way while retaining more detail as NR is basically off besides color noise, so the comparison would be skewed even more to the GFX if Adobe's default Raw settings weren't so aggressive.
For flatness of the negative, I've been experimenting with just placing it between 2 sheets of glass from Builders Warehouse. This has produced the flattest most consistent scans for me. Newton rings are a thing, however, they only seem to be appearing when there is a large amount of pressure placed on the glass/negative, i.e from a clamp to keep the sheets together. Otherwise, I've had no issues with Newton rings, but I will probably end up creating a better holder using a sheet of glass and a diffusion material such as a flat polycarbonate sheet mounted above tungsten lightbulbs.
A even better tip is to create a true linear profile by using Adobes free DNG Profile Editor. Fuji files are using a pretty close to linear tone curve. For other brands it's a way bigger deal.
Thank you for taking the time to write this post - very helpful
lol thank you. so much wisdom here i'm taking notes. could you share some tips on how to properly meter the negative? thanks a bunch mate!
@@chris_jorge make sure that there is no clipping in the red channel. You don't have to overexpose. Exposing just right is ok and no clipping in the red channel and everything works find. I don't think that you need a large amount of dynamic range because the tonal range of the negative is pretty narrow. Between the brightest part and the darks much difference. Form a technical stand point I do not understand why William stacks the exposure.
Brilliant comment!
I’m also heading towarda sandwitching my film between to pieces of glass.
To avoid newton rings i think I’ll end up getting some scaner fluid.
Great to see you uploading regularly again bro
Thanks G! Going to try keep it rolling for a while, feels very good haha!
That's very interesting comparison to see!
As someone who's been dlsr scanning for years, I have a few suggestions that may or may not help you or anyone else:
Negative Lab Pro is amazing and worth every penny ~ manually converting negatives is not fun
Making sure your sensor is parallel to the film is very important. I saw you using a bubble at some point, so maybe you already know this, but the weight of that camera is definitely making the copy stand lean a noticeable bit. Changing to an adjustable mounting system or getting a sturdier copy stand would help.
It's important to scan in total darkness (again, you might already know this, but your video shows you scanning next to a window so I just wanted to be sure)
DSLR scanning has saved me SO much time, so I can't see myself every going back to using a dedicated scanner. Glad to see you sharing videos again! Love the new content
A cheap flat mirror is great for getting things parallel
Thank you so much Noyel! Some excellent advice here, I tried NLP with imacon scans a number of years ago and didn't find the results to my liking, but I imagine things have changed and been refined a decent amount by now so will definitely revisit it if the occasion arises to try out inverting things again!
The valoi kit actually came with a custom little mirror that I used in conjunction with the spirit level, the base for the setup could definitely have been more steady though, my main issue was that the valoi clamps were too small to fit on any of my sturdier desks so I ended up using a little bookshelf to mount things to haha!
The scans were done in total darkness thankfully, I just made use of the window for the sake of having light for the video portion!
Thanks so much for sharing your insights, so useful to be able to put things out and get feedback so that I know what to watch out for, the appeal of this DSLR business is indeed immense haha, if my imacon starts giving up I reckon I'll swiftly transition into the camera side of things! Also very glad to hear you're enjoying the new things!
The happiest cat ❤
The noisiest cat I reckon
Your colour balancing is super impressive.
Thanks a ton Marc! Still feel like I get lost in it all sometimes and have to stop and revisit things at a later date, but I enjoy the challenge of making things sit nicely with each other and bringing out what the negative has in it haha!
Been scanning with the gfx100s for a good couple of years now; with the gf120 you need fuji's macro adapter (the bigger one, the 45) to get it to 1:1 and then you can fill the frame entirely; I even manage to almost fill it up with a 35mm frame+sprockets. On the holder side - the Valoi is the best amongst the few I tried; they also have a plastic-y fresnel-y thingy to put on the light source and it gives you an extra stop of light. In my somewhat limited experience (including a Hassy X5 though) the gfx is currently the best scanning setup.
Amazing! So glad to hear from someone with experience with the setup! I had no idea I needed the macro adapter ahah, although that does make a lot of sense! I was flabbergasted when it came to trying to scan a 35mm frame and seeing how small the frame was with the focus from this lens alone haha. Thanks so much again for your insights!
@@willysheepskin same, when I got the 120 I had no idea it was 1:2 and it was a nasty surprise, haha. But the 45 macro adapter fixed that, now it lives glued to the lens.
Have you tried Negative Supply?
Great video; I saw you were having problems with gradients at 16:00. It could be a banding issue; you can try switching the bit rate of your gfx file by going to image > mode > and change it from 8 to 16 bits/channel to see if anything changes.
Amazing comparison, big thanks. I'm using a Sony A7R3 and it usually smokes Epson 850V. Both machines are a level down to your setup, but still... It is amazing to see these cameras outperforming dedicated scanners. You earned a new subscriber. Keep up.
Thanks Marjan! I love the Sony A7R system, I can definitely see it being excellent for a camera scanning setup! Might have to make it the one I use for the inevitable comparison between the GFX and a different camera haha. Thanks for the sub!
Doing the lord's work man - these tests are invaluable for the community! Thank you
Thanks Kevin! Glad to be able to contribute in some small way haha!
Thanks for this. I have also tested both methods, but with my camera (Panasonic S5) I still sometimes have problems with color, especially of course with negatives (strange colors or color casts that are hard to correct). I have only rented a Flextight so far, but I feel like I got pleasing colors way quicker (just with Flexcolor) than with DSLR scans where I had to tinker more with the image. I also have the impression from your examples that there is simply more color information in the Flextight scans, which makes subtle color differences visible. I don't know what it is, maybe the CCD sensor, the Bayer filter or the internal post-processing, but I have the feeling that some information is not captured with the DSLR which is crucial for a good conversion. I guess it kind of makes sense that the Flextight performs so well in terms of color if it was designed for that exact purpose. But since the Flextight scanners are still so expensive, repair is difficult, and camera technology is getting better and cheaper, the DSLR route is probably the more sensible option for the future. Besides, the DSLR can even take photos and videos :D
Thank you for your insights! I think we’re on the same page regarding our findings haha, the flextight is definitely a more refined system but there’s so much going for a dslr setup that I’m very keen on trying to figure out a workflow that’d result in comparable color rendering! Some really good insights in the comments here with tips for getting more out of files
Those weird color issues may come from two (or more) things:
1. CameraRaw doing some internal weird stuff. Adobe makes input camera profiles themselves (by means of spectral sensitivity measurement of the sensors iirc), but what ends up in the tiff has a adobe specific output rendering applied. The latter part might do things which look good on normal images but not good on inverted and contrast enhanced negative scans.
2. The manual inversion in PS is prone to what I would call "wrong math". There are long threads in open source development forums for programs like RawTherapee and Darktable about the math that should be applied here (ideally in camera colorspace not what ACR spits out to PS).
3. color smoothing on the RAF to TIFF conversion inside ACR. At least to my eyes it looks like there is some heavy smoothing applied there. easy to check with an open source raw-developer where color-smoothing can be turned off in the debayering step. Maybe the glorious colored grain can be recovered.
So IMHO there is a very good chance that the raw files from the GFX are actually better and contain more info without the wierd color issues you presented here.
Thanks for the awesome comparison to the flextight! Very nerdy tech stuff, but if it enables you to keep shooting analog then I am not going to complain!
Cheers from Germany!
Thanks sooo much for this man! Want to respond in advance to your footnote and say that it's awesome that you have this depth of knowledge but are still more invested in just shooting analog than nerding out on tech stuff haha!
I myself am also more of a fan of making things than getting too techy around here, but have had many requests from folks to talk more about flextight things since the resources on the platform are quite limited when it comes to Imacons, that and the fact that I find the comment sections on videos like these tend to end up being immensely helpful resources (for the most part, some folks just want to be mean haha)
So thank you very very much for contributing such marvellous insights to the "youtube comment encyclopedia" here haha.
Also I should mention that the odds of Camera Raw being the culprit here are very high as I completely overlooked the importance of making sure it wasn't doing anything funky since I basically never use it.
REALLY fascinating to know that PS itself puts out wrong math potentially with manual inversion, that's also something I never really considered, I like to imagine such expensive software as being infallible but clearly not.
Your third observation also seems to have a high potential of being correct haha, I couldn't offer an insight since it's beyond my understanding but all of what you've said makes sense to me and thus the fact that all three elements could potentially be culprits makes it seem very likely that the fault is not in fact with the camera.
Thanks so much again for your insights, so great to be able to learn new things from folks with knowledge like yours!
Cheers to Germany from windy Cape Town haha!
@@willysheepskin You're most welcome!
I was a bit surprised for the techy-nerdy video from you and thought I could finally share/contribute something. I think not being able to service the Imacon is a huge single point of failure for you as an analog photographer.
So when I want to see more work from you in the future I gotta make sure you still can be happy even if the Imacon stops working (knock on wood that it doesn't happen too soon!).😁
PS assumes a color and workingspace by default. The "wrong math" I alluded to arises when this is not taken into account. PS doesn't know that you're trying to invert a negative, so it assumes whatever ACR assigned as output colorspace and "gamma".
So the good news is, and it needs to be reiterated: scanning the negatives with *a* camera, not necessarily the GFX100, can most probably be done to a quality level that is almost ballpark imacon level.
Needs locking down a certain workflow which might be tedious, but it's doable.
(Also: I watch your channel because it's NOT focussing (pun intended) too much on the techy stuff. Me dipping my toes back into analog changes my photography process and the results that is very very hard to replicate with digital. Digital is convenient, Analog isn't and because of that the approach of taking and making a picture is dramatically different. I *think* that is what I see in your process too. And your results are just SO so good!)
Much love.
Interesting video. I have to say, out of so many film photographers, I always liked your colors the most. That was certainly due to how the Imacon interpreted colors and what you developed from it. The GFX has a pixel shift mode so theoretically, you can use 400MP for scanning. I believe using exposure bracketing isn't really necessary with today's sensors. Usually, +1 overexposure is enough to scan everything perfectly; at least it has been for me for years with Fuji X cameras. The strange color bending transitions might even be caused by bracketing? That would be my first guess.
Thanks so much man! My initial tests with the GFX were single exposures but I noticed things clipping in the highlights since I was aiming for a bit over (out of luck, not experience ahah). I'd definitely easily believe I'd gone astray somewhere with that though since my experience with camera scanning is so limited.
With regards to the bracketing causing the blending issues, the raw example I showed was a single exposure, without blending, and still had the problem so it's something outside of that I'm afraid. Could very well still be something caused by me though haha!
Babe come quick new Will Sheepskin vid just dropped
Legeeeeend!
I always wondered how the GFX would perform, thank you for this
Thank you kindly for checking it out!
@@willysheepskin appreciate you man, glad you're doing well
The head to head comparison I never knew I needed. Thanks Will! Surprising results, and comprehensive breakdown. I can’t afford to part from a flatbed just yet, but dslr scanning is very enticing. Cheers 🥂
Thanks for checking it out Justin! Nice new profile pic haha! I'm quite keen on a flatbed at the moment tbh, really want to try sort out my at home workflow and having quick contact sheets at home as opposed to using someone elses scanner would be subliiiiime! Cheers!!
Hello from colorado, usa!
I love this comparison, i didnt realize how tough it would be to do the camera setup till i looked it up the price tag on the fuji 😢.
I would be interested in a more low buck (sub 1000$) camera to see how much better the medium format is for doing print work.
Been missing these videos, glad to see them again 🎉
Hello from Cape Town, South Africa! Yeah the camera is the catch, sigh, but you can do similar things with the more economical cameras from what I've heard in the comments here, so will definitely revisit this with a more budget friendly setup in the future!
Thanks for checking this one out! Glad to be putting them out again haha, cheers man!
MY GUY! Thanks so much for making this video. I've been flip flopping between Imacon or GFX setup and I really wanted your take as a long-term Imacon user. Another pro to using a GFX setup is you also get a really damn good stills camera out of it.
Thanks so much for watching it Evan! Having the camera is indeed a pretty huge benefit haha, honestly who knows maybe you end up getting addicted to the speed of it and bypassing the analog bit altogether lolol (Can you tell I'm bored of dusting scans at the moment haha).
Seriously though, the GFX is insane and I reckon if I bought one and went through the comments here to make sure I wasn't falling into some of the errors I likely made that resulted in the sketchy occurrences I shared here then it would bring nothing but joy across the board.
Very glad to have put out the right thing at the right time for you haha, let me know how you go!
Fantastic video - i never really considered bracketing film, will definitely do it now!
Thanks Sindri, glad to hear that you enjoyed it and got something out of it!
A good follow up to this would be to compare slide film scans for colour accuracy because you can compare the digitised images to the actual positive.
Very nice video, and beautiful photos! Great comparison, I just got the negative supply 120 kit and the vairo easy 135 kit. Going to pair it with the gfx 50sii. So this video makes me feel good about my choice lol.
Didn't think about stacking exposures to get more dynamic range. Very useful to know, to get the most out of the negative. Guess I will take two photos and merge the same way you did in PS. You were using a stop or half a stop of difference between your two photos in the gfx?
Thank you kindly! I'd feel very good about my choice in your shoes too haha! I used 3 images here with a full stop between them, actually noticed while putting this together that the shadows and highlights still sometimes were crushed/clipped so if I were you I'd tinker with even further ranges potentially! This GFX had a setting where you could set it up to bracket however you like, really is a marvellous bit of kit haha!
@@willysheepskin Thanks alot for the reply with confirming the 3 exposures/stops and extra info on the clipping. Very useful to know, hopefully I'm able to pull out imacon level images like yourself.
Keep up the awesome videos and photos. Always excited and happy to see a new video of yours pop up in my feed.
Loved this comparison. Thanks!
So glad you enjoyed it Parker! Cheers!
Very cool comparison and great explanation as to your hypothesis regarding specifically green color rendering between the Imacon and Fuji GFX 100S color science.
good comparison. thanks for the infos. especially liked the "color editing" technique! have you tried the Digitaliza (Lomography) medium format holder?
Love this video. Best photography channel by far.
Wassup Willy!
What I've done for a while with camera scans (after being schooled by Mr. Simon Maddock) is to insure that when importing a raw that any and all noise reduction (color and luminance) is turned off completely. I'm not sure what Photoshop / Camera Raw does by default, but both Lightroom and Capture 1 have defaults that aren't great if you want to reproduce the film's grain. After NR is turned off some light sharpening with a small radius brings back the grain really nicely.
I'm also curious if the weirds greens you got were a result of combining exposures in the way you did. Were the colors still sometimes weird when working with just one shot?
Ahoy Karl!
Yung Simon actually messaged me just as this video to similarly school me, I think you're both absolutely correct in that Camera Raw did some funky things here, I completely overlooked that being a potential thing since I never use Camera Raw so didn't realise it had the capacity to make things funky.
Those colors though were indeed weird when working with just one shot, the raw scan example I used in the viddy was just one of the 3 bracketed exposures and it still had the funk going on so it's unlikely to have been caused by the stacking business, although it could've been camera raw up to something for sure potentially since I definitely didn't keep any sort of eye on what it was up to,
Thanks for the insights though man, glad to be learning as I go here haha!
First of all, you look great man. Glad to see. Second of all, awesome video and very cool to hear your insights on this. And finally, thanks for touching on your color process, I’ve been dying to hear about it 🤍
Thanks so much Sander! Have been making a very concerted effort to build up my strength and such as much as possible and am feeling better and better for it. So glad you enjoyed the video too!
If you would be interested in hires images form gfx, you can try pixel shift (if you have a gfx that supports it). I would be very much interested in such a comparison.
As for the weird green, it is weird. Fuji had their woes with xtrans sensors that have been know to have mushi greens because of how the green subpixels are arrayed, but afain gfx should be using a bayer sensor which has no reason to respond in such a way to green color.
Cheers Michal! If I get the chance to I'll definitely try some pixel shift things next time, would be very interested to see about it also! Interesting to hear that there is precedent for the Fuji struggling with greens, so many potential causes and solutions have been identified in the comments here haha, could make a whole video just trying them out!
We have made a home-brew (aka cheap) dslr scanning setup in our community dark room. Old enlarger stand to hold the camera, 3D printed film holders, Kaiser light source and two planes (enlarger base and a light source base) with threaded feet in the corners to make everything somewhat level.
Works well enough with Sony A7II and Sigma 70mm f2.8 ART Macro lens.
Tho the post processing had always been a massive drag for me... B&W is fine, but colour broke me. I seldom got decent results and I blame it for my kind of a photography burn-out (cause I still like taking pictures, they just end up developped and accupying space on our drying racks or as RAW files in my PC...)
The way you do it seems much less complicated then what I used (even tried the neg. lab pro thing and didn't end up liking it), so I look forward to give it a try.
So this is in a way a thank you. There might be hope for me still :D (or I just return to our ol'Epson 500)
Ah fret not Tomáš! I myself am a large fan of making things B&W when the color isn't working out haha, there are some images that just don't want to render properly on Portra but look great in B&W ahah!
I try to zen out when doing post processing things, but will admit that I do sometimes rage quit fi things aren't going my way, in fact the latest bunch of negs I've got aren't working out well so I'm taking a break from em haha.
I believe in you though, sometimes things make more sense once you come back to them after a while, the key is to just keep going and eventually things make more sense lol!
awesome video, thanks for the detailed comparison! at 21:07, i wonder if that is lens aberration on the GFX? i.e., if you were to use an APO lens, would that go away?
Another very interesting presentation. Wonderful to have your regular videos again. ❤🔥🔥
Thank you so much! Very glad to be putting them out regularly again!
Thank you for this informations!
Are there differences between a IMACON 343 and the IMACON Photo?
So , so interesting. How does price compare between the two options? And portability?
The imacon isn't really something one would want to move around haha! The DSLR scanning setup also would be best suited to just staying put. I think the going rate for an Imacon 343 is around $4000 these days, and a GFX 100S body will set you back about the same. If you got a fuji lens for it - which I used here but isn't actually the best lens for the job - it'd be around $7000 total I think. So more expensive but the equipment is about 30 years newer and has a warranty haha!
If you just got GFX body and adapted a lens to it to work - which would likely mean you get it to focus closer too, it'd cost about the same as the 343, and substantially less than the highest end flextight X5 scanner!
@@willysheepskin yowzers.. that's not cheap...but then you do get a pretty presentable digital camera in the deal as well....
Can you use the Imacon's holder for your dslr scanning?
That probably would've been a good thing to try hey, although I'm not sure if it would achieve the same level of flatness if it weren't being held under tension, could've been very interesting though! Perhaps next time haha!
16:05 Try using 16bit in adobe camera raw instead of 8bit, it should fix those wired bokeh color
Will have a go at that for sure! Thanks!
II have had good experiences with compensating the orange mask of the negatives using an 80b blue filter at camera "scanning". As a result, the RAW files have to be less tortured when compensating. In my opinion, bracketing is not necessary with the wide dynamic range of your camera. Colour negatives are so flat in terms of contrast that one exposure is sufficient.
I convert with the Negmaster plugin for Photoshop. I find the results much more balanced in terms of colour, modulation and the highlights are also much better defined.
Did you try the highres mode of the GFX to get higher resolution?
Good to see you back posting again!
Also just getting in to camera scanning so perfect timing! Was your focus stacking akin to HDR?
I reckon the end result was similar, but that HDR is more likely to make things get funky since as a single process I'd imagine it being more likely to blend things in strange ways. The exposure stacking just basically takes the lightest bits and the darkest bits from each frame and mixes them together in a clean manner to make one image with the best of both whilst retaining detail.
@@willysheepskin thanks for the explanation! I've spent hours looking for someone who could explain that!
You say the GFX smoked the imacon in color rendering on that anemone photo, but the GFX rendered the whole bottom section of anemones as pink whereas the imacon showed that they're actually different colors. There was a blue one up top that the GFX really desaturated too. Were you talking specifically about color noise?
Lovely video brother, so nice to be getting regular uploads now. Looking forward to more! Print shop update soon? #sheepsquad
Thanks so much ya legend, new one is up now haha! Print shop update veeeery soon just need to make some more bangers lolol!
The never ending hunt for bangers hahahaha@@willysheepskin
Thank you for a great comparison video! It is exactly at the time when I'm loking for a setup for 135/120 film scanning.
Glad to hear that it was helpful! Cheers!
Hey Will! I currently scan on a Fuji GFX 50R mounted to a Pentax 645 120mm F/4 Macro, with the exact Valoi setup you have, and do the conversion in Lightroom with Negative Lab Pro. My biggest struggle with this setup has always been that the focus depth is incredibly shallow, and often the negative can be out of focus, somewhere on the frame. What I do to ensure my negative is as parallel as possible to the front of the lens is with a level app on my phone that shows degrees of angle to the nearest hundredth decimal point, and line up the holder with the negative and the camera that way.
It's possible that going to a smaller sensor camera could solve that problem, as APS-C and Full Frame will both have more focus depth. I suppose the question is, is the potential downgrade in sensor quality worth it for the increase in focus depth?
hi my friend! Im using the same lens and same camera, I always add a lot contrast on lightroom...
I definitely agree that sometimes it needs a lot of contrast, but that's a good thing because it means it's a flat image and there's a lot of detail.@@Xinbaby999
I'm very jealous of your setup Eric haha! Very interesting to hear that your struggle is mainly with depth of field, I feel like if the negative was perfectly flat the depth of field wouldn't be as much of an issue though, I alluded to there being means to keep things flatter in the video, my main guess would be mounting with some kind of anti newton glass or even fluid mounting but that adds a bunch of hassle to what is really a lovely process with the camera scanning setup. Have you given anything outside of the valoi holder a go yet?
It's true that the Valoi holder doesn't hold the negative perfectly flat, but *usually* it's enough for me, plus the speed of being able to scan a negative, and then push the roll through to the next negative, to scan again, is very nice. With that said, while I haven't tried a different mounting setup yet, it's been something I've been thinking about for a while now. It's also worth noting that with 35mm, this isn't really a problem as the negative sits a lot flatter than 120, and with that Pentax lens plus an extension tube, it's possible to get beyond 1:1 magnification. @@willysheepskin
Willy when you say RA-4 don't you mean C-41?
Also, would be curious as to how scanning with a digital camera using the imacon holders would turn out.
Moooofy! That'd be interesting to look into for sure! And noo I mean RA-4 since I imagine they'd use darkroom colors for reference in terms of how the negative is expected to render once inverted! All conjecture though haha, although I'd looooove to meet the folks behind the Imacon color science to give them a hug and pick their brains!
Ooooo! I moved from a Nikon LS-8000 to GFX100S this month. Awesome to see the difference between it and the Flextight!
What an upgrade you lucky soul! What lens have you paired with it? I was using the fuji macro here but it maxed out at like 9000 pixels of actual negative, I've seen folks use adapted setups that allow for more of the frame filled which would be quite useful lol!
@@willysheepskin I sold my X-T4 and XF lenses and took the money from another gig to switch to GFX. Wasn't able to afford anything but the 35-70 "kit" lens, but I have a ton of M645 and Nikon glass that I adapt for different subjects.
For scanning, I'm using the Mamiya 120mm F4, no focal reducer, just a dummy adapter. With that you can go full edge to edge on anything 645 and up as it's a 1:1 macro (and all those formats are larger than 44x33mm).
If you're willing, I'd love to try inverting that bee photo with Negative Lab Pro and see if I get the same posterization in the greens. I'd want to try with both a RAW and the TIFF merged file.
@@nathanjohnson5304 Great kit! Very interested in seeing how NLP handles that posterization. In my experience the color noise you get from Imacon .tiff is reduced when using .fff files straight from Imacon then inverted in NLP.
Most interesting. I think the big point is the one you made. Good scanners are becoming unobtainium. We need to make DSLR scanning work. It looks like there are things to explore -negative flatness, color depth, color space, methodology (exposure stacking vs not, inversion methods) and more. One commenter mentioned stitching areas of the negative together for more resolution. That really seems like it shouldn't work. But I've done it a fair number of times now and damned if it doesn't. At least I haven't caught it out yet. Or maybe I've just been lucky. Might be worth trying if you repeat the experiment with a smaller camera. It's also worth noting that you can actually use the digital camera to take pictures :-). So it can maybe amortize it's cost that way.
Ahahah it is most definitely good to keep in mind that the Camera can take pictures alongside scanning lol! Thanks for taking the time to watch and offer some thoughts on this Carl! Really appreciate it!
Hi Will, that's was very interesting to see comparison with high end scanner. I could tell from affordable consumer level perspective: last week I've sold my flatbed Epson V600 Photo because even my aps-c Fujifilm XT3 makes sharper and I feel overall better results. The only downside - resolution is not enough to scan medium format film, so going to upgrade to some high res FF camera.
Thanks Artur! Great to know that things are better than an epson with an XT3! Maybe look into pixel shift manoeuvres, you might get the resolution you're looking for out of that process!
Might try that Alex Burke inversion method, mines a bit convoluted. Are you just treating photoshop colour grading like you would an RA4 darkroom print. I always enjoy a scanning video, I use the x5 at my local facility and pray it doesn’t die.
I do indeed try to treat it just that way! Feels like it'll maintain more of the true rendering of the film that way than trying to get too funky with it haha. I'd hope that the X5 has many more years in it than my poor ole 343, they only stopped selling em in like 2019 after all!
I ended up outsourcing a bunch of STL files to be 3d printed which werw designed by Stuart Jenkins for a film holder. I found his design much more consisted in getting flat negs scanned. I'm only using a 5DS though, but I could still tell they were sharper than previous holders. Also manual inversion is OK, but try out Film Lab beta 3. Because it renders negs linear, I see grain come through a different way than manual inversion in PS. Its the closest I've gotten to a noritsu too with colors
Great video! Thank you, wanted to see a comparison of these two for a moment. I wonder if sony or Panasonic cameras will render greens better than Fuji.
Thank you Andrey! Sounds like something worth investigating I reckon haha!
Oh, I've noticed that you work within a 8 bit file. GFX RAWs give you 16 bit data. Use it, if it's there.
I really enjoy your scanning videos. I, too, am an Imacon scanner. However, my digital media lab , at the Center for Photography Woodstock just installed a Fuji GFX scanning station. I will have to give it to try now.
Thank you Ralph! I'm keen to hear your thoughts as a fellow Imacon aficionado once you've had a go with it!
Hey, Will, just a follow-up. I have not used CPW Kingstons Fuji rig, but I did set up my own Canon 5D3 copy rig. I was able to stitch 4 frames together and copy a 645 negative. Sharpness was only discernable at 100% magnification. Color was pretty close, but I gave the win to the Imacon. Finally, the Imacon is also way easier to use.
Thank you for this vid! Great comparison.
I am surprised by this but to me the imacon scans were leagues better. Of course this may be down to your conversion process and corrections but the comparison at 21:32 is a good example of this. The Imacon is wayyyy more life like and "3 dimensional" if that even makes sense. The whole image feels softer and warmer on the Imacon, whereas the Fuji scan seems "crushed" and "brittle".
Thanks for checking it out! I definitely still prefer the imacon in terms of it's rendering too, I think the image you're reffering too might be a bit more crushed because I also couldn't get the shadows to match in detail so just bumped the contrast so the scan wouldn't look too wonky.
There are a number of images I scanned where the shadows definitely lose a lot though, there's a signboard image where the camera scan crushes a lot of what the imacon can pick up in the darker bits. I'd be curious to see if I could get more out of it though with more intense bracketing, if I ever get the chance again I'll have to see haha!
Interesting video, Will. Really good point about the Imacon maintaining tension on the negative by curving the holder! Contrary to others, I would not recommend Negative Lab Pro unless you have tried their process and like their results ahead of everything else. Manual inversion is better, it's just tedious.
I thought the Imacon was exhibiting some chroma noise in the shadows, typical of CCD sensors. Unfortunate because exposure really shouldn't be an issue for scanning negatives.
Regarding the colour differences:
Did you encounter any artefacts on shots without exposure bracketing? Simple 1D operations really shouldn't be breaking the colours like that. 'Colour science' is a little bit of a gimmicky marketing term because the colour of a digital image is dictated by the dyes from the CFA and the 3x3 colourimetric matrix. This can be undone, giving you more control but the process is a little tedious. Once tonemapped, most digital cameras look very similar.
I also prefer the colours on the Imacon. If you like the starting point of the Imacon ahead of the GFX, you can tonemap them, and if you're willing to shoot a few charts, you can match the colours almost perfectly using data interpolation.
Thanks so much for the insights Amaraldo! That's actually fascinating to know, I really appreciate you sharing your knowledge here! It would be worth the effort if the result was color matched gfx and imacon scans for sure!
The banding in the Bee image was still there in the raw scan unfortunately, the example I showed here was one of the three bracketed exposures and it still had the issue, but others have suggested that perhaps it had something to do with the light source setup since it could've impacted the detail with which certain colors were projected. I love that there's so much advice regarding the mystery in this comment section though haha! Cheers!
I'd really love to see some evaluations of how a Sony Mirrorless camera would do. Personally, I scan with an A7RII, but I've struggled immensely getting colors I like.
Sony is top of the list for when the time comes to revisit this with a more reasonable camera!
interesting comparison pity you didn't try the 400mb multi shot of the GFX 100S too
At least I know now where to start if I get the chance to revisit this setup haha!
"long in the tooth" is a very old school phrase for a kid such as yourself...
I’ve been scanning my archives, transparency, and black-and-white negatives for the past three years with the GFX 100s, using a bellow extension with a Schneider 90mm enlarger lens. utilize the pixel shift feature if you want a massive file that competes with any scanner on the market today. With the Fuji not only are you able to do amazing scans, if your properly equipped with a magnetized film holder system, made of heavy gauge metal (not plastic), then you can get through thousands of negatives in a single day. Something no scanner is going to give you at this quality and price range. You also have the added benefit of recording the negative in 4k as you feed the negative through the holder. In actuality any digital camera with the proper professional copy stand and negative holder system, with either a bellows extension or macro lens would suffice in archiving all your film work.
Yeah, I ended up rescanning my entire print archive in like 30 minutes which would've taken me at least a week solid on my imacon, kind of blew my mind!
What lens would you recommend for negative scanning for the fuji?
Really interesting Video ! I use a Sony A7R2 for scanning with 42mp. It's quite sharp, but I have to say even with the lowest ISO setting it introduces a lot of color noise to the image. I tested it against a Panasonic GH5 with 12mp, a Sony A7 4 with I think 32mp and even a Canon R6 (which I think gave me the best results color wise) and all three were significantly cleaner. You can get a A7R2 used for quite cheap, but I would recommend spending just a bit more to get an A7R 3.
Thanks Daniel! Interesting to hear that the GH5 was sharper than the A7r2!My video cam is a GH5 so maybe I could use it for comparison to the GFX when the time comes to see about less expensive cameras haha! I've used th A7Riv a bunch for bts work on commercials, I reaaally like that camera haha, would be high on the list if I ever looked into getting a Digi stills cam!
I use a very simple DSLR scanning setup with my old Canon EOS 7D from 2009. It introduces A LOT of colour noise, but works very well for me as I only shoot black and white. It's not one for the most high res scans with an old 18mp sensor, but it gives me quite good results with b/w.
The camera have a multi-shot mode, where it shifts the sensor and captures 16 shots. By utilizing the Fujifilm Pixel Shift Combiner software in accurate color mode, you can seamlessly combine these images to create a final image without the need for the demosaicing process.
Thanks for the advice! Will definitely keep it in mind for future endeavours!
A very interesting video! Thanks for sharing!
Thanks for checking it out Chris!
How about creating a LUT to translate the Imacon color space to the GFX? I did it for Frontier colors. I am pretty sure that the color space of the Imacon is smaller for some colors like the greens for example.
Please! I'd pay you to create a LUT for my Sony A7RII to get Imacon-esque colors.
Hmmm that'd be fascinating to try! From my experiments though it's always seemed like LUTs rely on consistency in the source frame and having the colors presented match the colors that the LUT is based on, which gets tricky with varying lighting situations. But I also am not particularly experienced with them, so perhaps there are methods to get around that? Keen to give it a go though, if I manage something exciting I'll share it in a video haha!
@@willysheepskin LUTs are like dictionaries. You would only need the original color space of the Flextight and could then translate it to ProPhotoRGB. Fluctuations in the material would not be relevant because the LUT is not created based on a negative, but rather only determines what the Flextight color space should look like translated into ProPhotoRGB. However, it may be that the differences in the color representation of certain colors can also be attributed to the light source. You would have to measure the Flextight and compare it with the options that we can use to take photos. If you take it very seriously, the way the sensor works also plays a role, but that's only the last 0.5%, which can be ignored.
Lovely video, thanks william
I've used pixel shift to get better color accuracy with the gfx... so far only drum scanners got better results in my case.
I'd love to see an imacon editing video. I find the flexcolor software kinda hard to work and there aren't really any good tutorials out there
Cheers Rhett! I'd be keen to do a video going in depth on the imacon side of things, but the amount of negative feedback I get every time I do a video like this puts me off it for a whiiiiile ahah, we'll see though maybe I brave it sometime, it is a bit of a safer bet since it is my main workflow method whereas the GFX was a brand new thing
@@willysheepskin again I'd love to see it. Your color balancing is so nice and the flexcolor software isn't that intuitive, so I'd love to hear from a pro :) also I promise to be mean to everyone who's mean to you in the comments 🫡 maybe that helps the decision ahaha
I tested with GFX and saw the same issue. I tested with Canon EOS 5DS and red gradients were *also* borked.
Another takeaway I had from experimenting is that found that the grain looks screwy on high ISO film, very bad on black and white especially. It's something to do with the Bayer color filter array. For example, Delta 3200 looks TERRIBLE, *unless* you use a B&W converted camera to convert it -or- you use a Pentax with pixel shift.
Pentax does some crazy black magic with their composite mode, where they move the sensor to cancel the Bayer color filter array. I tested Fuji, Sony, Panasonic, and Olympus pixel shift. All of them had issues, whether it was color, or banding, or grain. Pentax? Amazing conversions. What I recommend is an el-cheapo K70 using pixel shift for 35mm or smaller formats. For medium format, if I'm in a hurry, I still use the K70. If not, I use the V850, where I still get 55mp of real optical resolution out of 6x8 or 31mp out of 645. If I need more than that, I send it out.
I've tested most of the commercial film holder options - the Valoi Easy35 hands-down wins for easy of setup and perfect alignment. For 120/220, I just use a cheap and effective EFH v3. With either option, I use a Kinetronics Staticvac right before the negative slides through the holder to avoid almost all dust.
Awesome video - that Imacon is just Amazing. Super interesting to see the difference between that and the Fuji.
EDIT: Forgot, I plant to try one of these for 120 - look on Etsy for "BlackscaleLab" - they make a quick setup film holder like the Easy35, which is probably worth it because it makes the setup time super fast:)
Thanks for this Jeremy! Quite glad to hear that I'm not nuts haha, although a decent amount of potential solutions have been offered here, the fact that you've tested such a variety of setups and have found them all to be lacking is really useful! Also now I have a reason to sing the praises of a modern Pentax in conversation which is a treat haha!
Thanks also for the advice regarding the holder, it sounds very useful, and its so great to have someone with your experience offering insights so that we can all learn!
I wonder if pixel shift or exposure bracketing could also help reduce the problems that happened with the gfx scans
A number of people have suggested that it just might've! If I ever have the chance to revisit this I'll have to give it a go!
Lovely video and super interesting. Ignore the haters, they mostly have no idea what they're talking about. I'd be bummed if you didn't post unique stuff like this because some basement dwellers were nit picking.
I’m having the same dilemma about how to scan my 120 film except I have X2D (and 500C/M as my film camera).
What is the max image resolution you can get from Imacon? My lab offers about 4000x4000 scans with Imacon 🤔
17k pixels on the long end is the biggest mine can go!
@@willysheepskin that’s insane! I wonder what the grain limit for 120 film is. I’ve seen 4x5 was scanned to the size surpassing 700Mpx (that’s like 30K on the longer side) in a drum scanner and they still had details there. I’d expect 120 film might be a quarter of that 🤔
How about using the imacon film holder for the fuji setup?
I'll most certainly do so if the opportunity to revisit this arises!
No one would suggest you’re not able to express your views. However what is remarkable about this piece is all that you leave out - and you do not spend any time on the incredible piece of tech Flexcolor .fff files are (Lab color space, raw
Scans), and the incredible sharpening range, noise management at scan (it’s not fixed) the dMax capacity of the scanner etc. etc
And etc. And the Fuji comparison - colour space, lens choice, camera settings colour space file type ….
And the usefulness of the Fuji in high speed processing. The use of the Imacon can be sorted in about 20 minutes and RUclips references abound. Quite possibly a useful comparison is wet mounting the v850 with your little 343. You need to take 3 exposures of a colour neg - 12-14 stops digital capture (in a single exposure) of a 6 stop piece of film - that doesn’t make sense but again your choice. So - I just don’t get what you’re trying to prove or show here - try scanning and shooting an IT-8 target. Your excercise is simply a reasonable record of not knowing much about either - or - choosing not to explain it. It’s weird. Imacon is not an archaic dinosaur - but both are in the hands of those who don’t know how to use them. Again - not saying that’s you - but you leave so much out how it could be useful for anyone is beyond me.
Its interesting how at 17:54 you can clearly see the red berries or leaves on the Imacon scan but this is absent in the Fujifilm.
Indeed! It's ability to discern fine color is for sure still unparalleled!
i would skip led panel and use strobe... better light source improves colors !
I've read that some people will sandwich the negative between two pieces of anti-newton ring glass to get ultimate flatness. The only real drawback is your setup must be scrupulously clean. No smudges, and no dust. The GFX 50SII isn't that expensive and some of the older GFX 50R and the 50S are coming down in price on the used market. For some reason though the older 50R demands a higher price than the 50S. I've seen the 50S on ebay for between $1200 and $2000 USD for a body only. The Nikon D850 or the Nikon Z7 are at 45Megapixels and The same goes for the Canon EOS R and EOS R5. So there's that for an alternative to the FujiFilm. Did you get to take and landscape shots with the GFX 50SII? Even the RAW images without processing are impressive. To my mind they are as good as the 6x6 images from my Hasselblad 503CX on E100 film.
I wiiiish I'd had the chance to take the camera outside! It was a 100s so the files would've been sublime I'm sure. Sandwiching the neg as you describe would definitely sort out the flatness issue, but as you say smudges and dust would become a bit of a nightmare most likely haha! I've also somehow frequently managed to get newton rings even from anti newton ring glass somehow, so that'd probably be infuriating too.
Thanks so much for your insights though Bob, super useful to have a bit of perspective on what some of the older cameras go for now, those numbers are a bit closer to the realm of reason than this 100S haha!
All the best to you and yours as always!
You can do pixel shift 400 megapixel on GFX 100 series
Will have to try that next time!
what size is the negative here? 6x7 or 4x5?
You explain how the Imacon setup guarantees the best possible film flatness, but my question is....how much does this actually matter?
For example, if you'd be using the 120/4 macro lens on the GFX, even wide open at F/4, at the minimum focus distance of 45cm, depth of field would be 3mm.
I realize the "acceptable sharpness" that depth of field calculators use is not always best, but also consider that you probably won't use the lens wide open, but rather stopped down to get the best balance between lens aberrations and diffraction limiting your resolution.
At F/8 depth of field would be 6mm, at F/11 8.5mm. Even if we use a more rigorous definition for "acceptable sharpness", that should still leave us with 1-2mm of depth of field which seems reasonable for a simple border-only film holder to me.
Personally I've been using my Pentax K-3 ii 24MP APS-C camera with the 35/2.8 Limited Macro lens set to F/5.6 and the very simple Pixlr-Latr film holder for anything from 16mm to 13x18cm negatives and always been happy with the results - I already had the camera and lens, so getting a scanner with comparable image quality probably would've been at least two times as expensive as getting a copy stand, light source and film holder. I only shoot B/W and also have been perfectly happy with the free and open source darktable software for converting them. To be fair I don't make prints from files, I only view the results on digital screens or print straight from the negative.
Ahoy Leon! Very interesting question actually. I've never thought how much it would matter with the GFX kit, I just know on the imacon even like 1/4 mm discrepancy in flatness throws the grain completely out of focus.
Someone commented here about how their main issue is with the depth of field of their GFX resulting in variable flatness irrespective of aperture, so I'd imagine there is the potential for that to be a thing too. They also said they were using the same holder as me so the slight variations in flatness I picked up could be what was causing issues for them.
You make a good point though, perhaps there's some kind of math that comes into play when the area that technically counts as being in focus (the film plane) is so thin, but again your points regarding theoretical DOF in relation to apertures you've used make sense to me. That alongside you saying your results are great has me a bit stumped haha, maybe your holder is just super rad lol!
Great video as usual!
The bayer filter on the gfx sensor has 2 green pixels, 1 red pixel and 1 blue pixel in each 4x4 pixel window (see wikipedia) and the missing values (each pixel needs all 3 channels) are interpolated by the raw processor (in this case PS), maybe the strange halo you are seeing is an artifact of this interpolation in that particular photo? That would cause artifacts in other photos as well when viewed at 100% though. Linear scanners do not use a bayer filter and does not require interpolation as their linear array is 3 pixels wide (one pixel for each channel) and it moves across the film.
Another theory is that this part of the image was overexposed (or close to) and PS treated those values badly. Or maybe your curve changes caused it? Looks like a hue shift or some kind of clamping. I don't quite understand the filmic and AgX tone mappers but they should correct for this and maybe PS does not by default. Will link a video on hue shift here:
On tone mapping ruclips.net/video/bTQMKvpDTYU/видео.html
About your GFX sensor claim, it doesn't look at the X Trans sensor.
@@Alex-c3o1oX-trans and Bayer filters function the same in that both require interpolation for missing color channels, the only difference between the two is the pattern of the filter.
Wow Gustav thanks so much for this insight, exactly the sort of information I was keen to hear about from someone that clearly knows a bunch more about the GFX than I do!
Will be checking out this video shortly, really appreciate your insights and help!
@@willysheepskinThanks, but I'm by no means an expert (especially on the GFX)! I'm just a curious hobbyist interested in how digital RAW processors work under-the-hood as they decide so much of the look of digital pictures nowadays.
I've thought about it some more and I don't think it's the Bayer/X-trans filter that cause it (my first hunch) as the effect of interpolation should only really be visible at very small scales (4-6 pixels wide) and the artifact in your picture is much larger.
It could still be the second theory, that the raw editor does not handle overexposure well and introduces a hue shift. Could you try looking at the raw image in Darktable with the Filmic modifier and see if you can still spot it? What program are you using to turn the raw image into a tiff, by the way? Noticed now that the GFX image was a tiff.
How could 16:40 be an issue with with GFX rendering greens when it's recording the negative? My guess would be that it's more-so the merging method of the bracketed images
The raw file I checked still had it though and it wasn’t merged which is what’s confusing!
ah. Fle*x*light. whew. I thought this was a very different kind of video at first.
Ahahahah I can honestly say that similarity has never popped into my head, but now I'll probably think of it every time, thaaanks lol!
@@willysheepskin sorry I cursed you with this forbidden knowledge 🤣
yo. few things about 17:05.there are multiple reasons for that.
a) it looks more like over-exposure problem, i have that sometimes with my D850 when I scan films. b) it might happen because you merge layers using "multiply". not the best way to do it. to understand whats happening - double click on layer, see the blend if beneath. hold alt and grad the pointing arrows on the right side (it should split). one for shadows, one for highlights. try to do the same without holding alt - thats what multiply does. c). you`re working in 8bit. using gfx and keeping at 8bit - is a waste of gfx DR, sometimes this falling off may be fixed just by changing bits :). sometimes you can just manually change to 16 bit and that gradation will be smooth again, if you imported the file in 16bit in first place c)as someone mentioned in comments, Adobe Color profiles are not the best. Try Capture One. waaay better for start (especially if you work just with curves, you need a good starting base), and super easy to change raw profiles. Than export it in 16bit to PS/
Thanks so much for the advice Tim! Great to learn more about merging things, haven't ever really used it outside of trying to get more out of these scans hah!
Really appreciate you sharing your knowledge here, so great to have resources like this to refer back to if I get stuck with the camera scanning business again!
@@willysheepskin no problem. its a technique for high end retouchers. but so that you know - curves perform different in 8bit and 16bit. if you have color fallout - just change to 16 bit, it will make gradation smooth "magically" dissapear :) you can see it on seemless backgrounds, sky and greys.
Do you need an older computer for the flex tight? Doesn’t it have an older OS requirement now? Thanks.
Sorry just heard you later in video.
No need to apologise! You do indeed unfortunately, with Windows the OS thats worked the best for me, flawlessly for years now, has been windows XP. It depends on the scanner but in general a computer that matches the age of the scanner is a good bet, but older doesn't hurt since they'll all run on XP.
With regards to Apple machines, G4 and G5 computers are good for the older ones, but modern machines running up to OS Mojave will still work, any OS after Mojave (Think Mojave might not work either?) is too new for the software the machines run on (Flexcolor) to work.
In general the best bet for an older one is a Win XP machine, and newer ones run nicely on Macs so long as the OS isn't too new.
Thanks for checking this out!
@@willysheepskin thanks for the help! I appreciate your time.
Lucky for me i can get my imacon serviced in the city i live (Copenhagen). But likely it could also die at some point. Will be a sad day for sure…
You live in the ideal place to be an Imacon owner haha!
I would like to know if this is achievable with full frame camera? something around 40MPx
thank you
I reckon it would be! Will be trying it out sometime not too long from now, just as soon as I've recovered from this one lolol!
WILLYYYY
PAIIIIN!
One way you may be able to get better results from the GFX is to try different RAW processors. If you going to pixel peep Adobe is not the best processor.
do you know what the R in dslr stands for?
If you're being snarky then it probably stands for listening RRReally closely to what I say at 9:45 ?
If you're honestly curious then it stands for Reflex, and indicates that the camera operates with a prism/ mirror (Unless its a TLR) and that you're actually looking at the image that will be exposed as opposed to an approximation through a viewfinder.
Which doesn't actually apply to the camera I'm using in this video, but someone who felt like being snarky for no reason might comment on it, ignoring the fact that DSLR scanning is a descriptor of the process and not the specific equipment and then get told to listen to what I said at 9:45
The gfx is my dream camera unfortunately they wouldn't talke my kidney as a payment method 😔
Ahaha you and me both my friend, even if I sold all my cameras I'd still only be able to afford a body for it loool, but I'm happy with what I've got going, just good to know there's an option in case of the worst happening lol!
honestly isn't the kit plus a macro lens just as expensive as an imacon?
DSLR Scanning allow for a lot of flexibility over most types of scanning. You can do multishot noise reduction, or create ever larger images with higher magnification lenses. 1:1 on a GFX might be 50/100MP but if you used a higher magnification lens, 2:1, 5:1 etc you can easily stitch 300MP to over 800MP. Scanners can't do that. You are limited in terms of your maximum resolution. Do you need a 25,000x25,000 image? No. But if you did, you could create that natively with a DSLR scan and print an image at 300 DPI that's the size of a soccer field. Or you could go even crazier and use filters for per channel images that you can combine together later. Of course you are limited by the size of the grain and the information it retains which varies by format. But 100MP vs a 10MP image will always produce a sharper image regardless.
Using a Canon 800D and 60mm 2.8 USM, I've stitched together, 2, 4, 8, 12, and 16 images of a single 6x6 negative and I was still able to physically see more detail in the negative the closer I got to 1:1 and the more images I used. It produced 6GB images in Photoshop but I am able to zoom into an image of a house I was standing over 20-30 meters away from and see tiny cracks in the sandstone wall. And sure color bit depth might be less than the Imacon, but 14 bit per pixel images are the standard and current technology is barely able to support 10 bit color anyway, so that more to do with long term archival or specialized printing processes that don't have a finite color range.
It's so exciting to hear about the marvels this scanning medium has to offer if one gets invested in it haha! Natively printing something that size is insane, honestly have always been mega stoked with A0 prints natively but the potential for sizes like you mention is awesome haha!
Glad to hear from someone with your experiences that some of my observations make sense, and thanks in advance for the supremely informative comment you left a bit after this one!
Biggest issue I see with your comparison is the aperture. F16 is way too much, you should have used f8 max. on the GFX. Also using ISO400 is not a good idea as that's the 'worst' dynamic range you can get from the GFX100 sensor before the second native ISO kicks in. You basically lose 2 stops of photographic dynamic range. Even so it would have been enough for the negatives but they were still underexposed by at least 1.5 stops. So you've made it way harder for yourself by stopping down to f16 because that lowered the resolution, reduced the available light, making it necessary to increase the ISO, reducing the dynamic range, making it necessary to make brackets.
3:00 If you have a proper macro lens (1:1 ratio) than you can achieve similar functionality to your Imacon scanner. When I scan 120 film using my mirrorless camera, it's usually right after scanning 35mm frames. I will leave everything as I had it for 35mm and take 6 shots of the 120 film by sliding the light source (which the holder is on) around on my table. I will then convert everything in lightroom using negative lab pro and stich the 6 files (3 at a time) together as a panorama. A simple square crop (I shoot 6x6) and I have a 10k image file (my camera is APS-c ~6k horizontal) with amazing sharpness. I have run into problems with this technique when scanning images that have large brightness differences across the 6 frames. For example, I shot a hydro dam which was getting lit by the sunset (~4pm in winter) and the foreground was dark and lower in the frame. When I did the conversion, there was severe color differences between the frames, even the parts that overlapped so I couldn't stitch them together. For images with more consistent light throughout, this technique has worked fantastic and doesn't give me problems in B&W either. Similar to the Imacon, it takes more time lol.
With your inversion process, you might not run into the same color/brightness issues that I run into with negative lab pro. NLP is also worth every penny, I highly, highly recommend you use it for digital scanning. They have a free trial so you can try before you buy. I think it would fix the color issues you are having with the gfx currently.
7:50 You don't label your archive sheets.... you psycho😆
21:51 Look into the Essential Film holder. It will be a better option than the Valorie holder but it does take quite a awhile to ship from the seller. If you got money to spend, Negative Supply supposedly make the best holders on the market. They are made of steel and you can move film through with a winding knob. You might need to sell a kidney for one of them though, they are quite expensive and probably over engineered.
The EFH isn't a better option, both of them aren't great for 120 because of film flatness, also your moving around stitching shows that flatness of field is not something your scan is concerned with because those micro movements would be enough to mess with sharpness.
@@Alex-c3o1o I mean I know from doing it you are wrong and I know from being a macro photographer on digital that my DoF is enough to cover very small changes. As for the EFH, I haven't found a single review anywhere that states it's not better than a Valorie so....
@@ZeLoShady the depth of field isn't enough if you are using a macro lens that is optimised at around f4, which you would need because at 1 to 1 f4 is f8. And with any decently high pixel sensor you would otherwise lose sharpness due to the airy disk being bigger than the pixels, which you should know as "macro photographer".
And just because you can't do proper research about the EFH does not mean that a film holder where you can literally slide the negative around, meaning there is no tension, can hold the 120 negative flat enough.
Also when you aren't using pixel shift you will have the debayering errors anyway, so even there your setup is suboptimal.
Also also even the negative supply film holder has issues with tension if you use cut negatives like will here.
Just because something is good enough for you does not mean it is optimal or can even come remotely close to a flextight.
@@Alex-c3o1o @user-vp5yo5ys5r LOL all I did was suggest a better holder than a Valorie, not a flextight. Obviously something that clamps the film and then bends it to apply tention is better but I am not saying my setup is superior, not sure why you think I'm making that claim.. Someone is a bit of a snob eh. Unfortunately for you, my setup is sharpest at F11 and because I use APS-c and not medium format, I get a lot more DoF in macro. As far as airy disc goes, who gives a fuck, that's a problem I'm not looking to solve. If I make a print of a film image scanned with my setup vs scanned with an Imacon, no one will be able see any signs of Airy Disc anyways! Maybe if the print was 10 feet long it could matter but there won't be enough print resolution on an 8x10 to see it which makes that a non-issue that only snobs would care about. There's the theoretical world and then there's the real world bud.
Thanks for the advice Alex! Yeah if I were to use this setup again I'd be sure to source a macro collar for the lens since I didn't actually know it was a thing ahah, I kind of just figured the macro lens would just macro looool!
Will also revisit NLP if I come back to this topic sometime as a lot of people have said good things so I'm figuring its come a long way since I last tried it.
I doooo label my negatives, that folder just isn't labelled since I got behind at some point and haven't bothered catching up yet ahah, although it's been multiple years now so maybe I have in fact simply transitioned from being a labler to not being a labler, and am in fact a bit nuts ahah!
Will be looking into that holder for sure for future reference! Thanks again for the advice, appreciate you taking the time!
Flextight?
Fuji GFX is a Mirrorles medium format camera NOT a DSLR. The video seem to be a review of the imacon scanner mashed together with the comparison intended rather than focusing on the comparison intended. The results from a mirrorless/dslr scan have the respective camera colour science interpretation. A purpose scanner with a dedicated software which has the correct film colour profile will yield the more true to the negative colour interpretation. That's why all scanners use CCD sensor instead of CMOS sensor. At the end it's all boiling down to using the right negative for the right job. If you use Kodak portra for landscape then yes you need to manipulate the colours and going thru the hassle of a ccd scan is pointless as the colours will not be great.
9:45 , Check out my friend Kent Andreasen's work to see what landscapes/everything can look like on Portra 400 scanned on an Imacon.
As soon as you said the Imacon takes 20 minutes to scan a single frame, the comparison was over sorry
A fair conclusion haha!
you doing it so wrong and you have such big lack of knowledge. if you need answers you can ask me
Do a vid on scanning pls!!
9:45
I've got the 343. For me the experience is nicer. Rodenstock is top of the line too.