Here I'll summarize the comments section for ya: Most comments: "Hey cool thanks man" or "here are some helpful tips to get better results with you DSLR scan" A few comments: "You suck, your camera sucks, you didn't even use a camera with pixel shift, and you're an idiot" I guess we can't all still be friends 😂 On a more positive note: THANK YOU to the many folks who made a contribution! I'm very encouraged and surprised by the number of viewers who were kind enough to pitch in a couple bucks to make more videos like this possible. I look forward to bringing you more entertaining and informative videos this year!
Hi Nick, Sure we can still be friends, I guess I fall into the second group? My point was that your assumption you were "bringing you more entertaining and informative videos" is a little off the mark. Entertaining YES. Informative NOT So much. First a couple of points, yes I watched the whole video, have worked in prepress with drum scanners, own an Epson and do not own an A7RIV, and shoot both film and digital, for over 25 years, yes I am old. I get the pretend vomiting at the beginning was a joke, I watch the whole bit, it was actually kind of funny, and was intended to be self deprecating. I get that. That said, the techniques shown in the video showed a profound bias, perhaps unconscious on your part, but there none the less. Let me elaborate. The statements you made about the quality of drum scans are correct. They are very good, and will give you excellent results. The issues there are cost, if you have to pay someone else, and the ageing nature of the technology, as no one makes new ones, and service is becoming problematic. You also presents a good case for flat bed scans, and show good techniques on how to get the best results out of this gear with fluid mounting etc. And yes I have an EpsonV700 that I use with fluid mount, I agree you can get excellent results with careful technique. When it comes to the camera scanning though, you go massively off the rails, and basically do everything you could possibly think of wrong, and then complain it is frustrating and gives poor results. Where to start. The camera, an ageing mid to entry level DSLR. First mistake. And yes I know it was what you had on hand, but if you are going to test these a technique to assess its potential for quality, at least use a descent model. Even if you have to borrow one. Especially if you are then going to make a RUclips video about it, do some research and use the best gear possible. C'mon dude. If I was doing this test I would have borrowed a Sony A7RIV as others have mentioned, which can do multi shot exposures and produce a 240MP image. This would have eliminated most, if not all of your most vigorous complaints related to stitching. As with a file that size, it becomes unnecessary, especially if you are "not a sharpness whore". Secondly, you go to so much trouble to carefully mount the film and mask it for the scan. Then just ignore all that good technique when you do the camera scan. I would fluid mount exactly as you do for the scan, and mask as you do, for the camera scan, and for gods sake put a lens hood on. Lens flare is not your friend! Thirdly, If you really do feel the need to stitch (remember, you are "not a sharpness whore") then put your camera on a copy stand and MOVE THE FILM, not the camera. I could go on but you get the point. You obviously, although I suspect unconsciously, predetermined the results. If you take the time to develop good technique, as you have done with scanning, then I suspect the results would have been quite different. If you see saying these things as "being a dick" I am sorry, but if you put stuff out there, expect feed back, and sometimes it will not be glowing. If it is however intended to be constructive (my first post could have been kinder I guess) then take a step back and reflect on it. Kind regards, Jason Berge. www.shoot.co.nz
@@nickcarverphoto Hi Nick, Sorry you felt I was being patronising, I guess that comes from being old enough to be your dad. I don't know "everything", not yet. Which is why I was interested in the video. I can always learn something and try to, every day. Cheers, Jason.
I find your act disgusting, no matter what the useful info you were going to present, I could not stand you longer than 59 second. Down-vote well deserved. Better luck the next time.
@@jasmb570 cool comment dude, whens your motion picture documentay about the soja 7 VII scanner and general ass-holyness coming out? :P Kinder regard A dude who uses a dedicated Fuji film scanner :D
Why would you hate on Nick? He's trying to help us out. How about the haters create some useful You Tube content for a change, and do something constructive with their time.
I'm ridiculously selfish in that I wish RUclips was your main gig. In a world of endless repetitive photography gear review videos that all say the same thing, you're actually dolling out some useful information. Your videos inspired me to dive back into film and it 100% saved me from the boredom I developed with digital. Thanks for all that you do. Oh, and cheers!
Thanks Nick for doing this for all of us. I have been waiting for a comprehensive test like this for a long time. Certainly a lot of factors to consider going forward, and things to weigh back and forth. Your hard work is appreciated.
Sir, you've done great work! Good methodology, good presentation, and a very good assessment based on what's important to you. You hit all the right points and you've done it in really clear and concise manner. Kudos! I also appreciate how much work it took to do all of these tests and analyze the results. Thank you.
Strickland's work is amazing. I'm glad he reached out to you. He's a film photographer so he gets what we are up against when it comes to scanning our film. Really nice video, Nick. Thanks....
Now that is what I call a thorough, clear, concise and even scientific comparison of three methods of scanning. Well done and seriously thank you! That must have been a serious amount of time and effort. You’re a good man and I hope you know that we all really appreciate your hard work.
I'm the complete opposite although I respect that opinion. Don't care about color so much as long as it's sharp. I guess that's why he said that part of the video is divisive.
Real Life Ryan same here! So much valuable insight gained from his months of experience in those 36 min! It was like a massage to my mental stress of theoretically debating over this exact same question! 💆♂️
Nick, you are so bloody good at these. No umms and errs, no jump cuts, no unnecessary words just on the money start to finish! Another great one. Cheers
I've tried dslr scanning with single frame 35mm and I pretty much ran into the same issues with colours and workflow too. I'm now shooting mostly 6x6 and starting to go for 6x12 & 4x5 and F stitching 20frames together. I went with the "baby" v550 and I now enjoy photography again
I've done a lot of DSLR scanning, but the real secret to a stress-free process is stick to B&W and don't stitch anything ever. I've tried to stitch-scan 4x5 and even when you think it's right, it's probably still wrong you just haven't noticed yet (images magically growing width somehow). :)
@@nickcarverphoto I've found that it works well with highly textural/detailed images IF your camera and film are perfectly parallel, but fails terribly on things like abstract still life or anything where the stitch software can't find seams. In those cases, I suppose it's possible to manually align the images, but quite painful. I've also found the colours to be quite difficult to get right when scanning colour negative with my DSLR, but I thought I was just bad at colour, so it's kind of reassuring to see that you had a difficult experience as well. Don't know if youtube will let me link, but this is an example of an image I recently scanned with the DSLR stitch method, and a crop of approximately 35% instagram.com/p/B7fX3euAN-M/?igshid=vlpmey2revpf The resolution and sharpness is very nice, but your video has made me consider whether an Epson might be better purely for the ease of getting nice colours...
@@nickcarverphoto I use this free software from Microsoft called Image Composite Editor. It rarely fails to stitch images together. It was written to stitch hundreds of photos. www.microsoft.com/en-us/research/product/computational-photography-applications/image-composite-editor/
What a video, I can't believe I watched 30 minutes of film scanning knowing I never going to do it😄, much less with that attention to detail. Good job man!
What a FANTASTIC video. Amazing information presented in a straight forward manner. I appreciate the many hours it must have taken to produce this video.
Great video and comparison. Two things to digitalize with a digital camera. 1. Masking the light board and only leaving the film surface can immensely increase color accuracy. 2. Every color channel has a dynamic range. Because of the orange mask of negative film the red channel exposes differently. With a blue filter you can even out those differences and have more beef for your inversion. I don't know actually how good the 6D sensor is, but not every Canon is known for having the best dynamic range, probably a Sony made sensor in a Nikon or Pentax or even a Faveon sensor might help also in this regard. So quick improvements, record with a blue filter and mask the stray light from your lightpad
Plus 1 on the Foveon Sensor for colours however this is personal and subjective and also Microsoft Ice for Stitching - great job and best presentation style as usual
I was thinking the mask might help for DSLR scanning, but does it apply both in “one shot” and “multi partial shots for stitching”? The film has a slight frame around it already. (I’m just theorizing here without testing) As for 20MP Canon 6D (which I still shoot), I think that only affects number of stitches (MP) and colors (camera sensor)... other cameras will perform differently.
Are you suggesting an actual “blue filter” in front of the lens, or a layer in Photoshop? Have you done this to demo? How do you choose the proper color of “blue” for your filter? 1) shooting through a physical filter may affect sharpness 2) I presume PS can do the same color inversion with or without an additional blue filter
To even out the red channel I just adjust the camera's white balance to where each channel is somewhat the same. That way I don't have to worry about the red channel clipping when "scanning" with my DSLR. Also, to get the camera and the film to perfectly level I use a mirror that I place on the surface and once I see the center of my macro lens I know the camera is set up right.
I've been a photographer for many years and am always learning - thank you for your detailed video. In my experience (I've printed around 1000 various prints ranging larger ones about 3 feet by 5 feet) and never once did I require anything close to the resolution you are talking about - even for exhibition quality prints. I've even done tests to where I have a 16 megapixel image (These were 4 feet by 3 feet) and not one person even noticed a difference between my 40 megapixel shots and my 16's. When people view prints, they don't view them from 6 inches away - it's the sharpness and quality of glass and camera sensor that makes the difference. I've done tests where I have literally large prints at 20 megapixels sitting on a table and have had dozens of people look at them and they can't tell the difference between that and my high resolution cameras. In fact, many billboards are actually only 3 megapixels. Also, people scanning film at these very high resolutions - when the film itself is no more than 8-12 megapixels for the negative itself doesn't make any sense. This hole resolution thing is completely overblown - I've never had a client come back to me saying "boy, that wasn't 200 megapixels" - I was looking at your art print with my magnifying glass for 3 hours and noticed that... Anyhow, just my 2 cents.
I couldn’t agree more. I’ve told my students for years that they don’t need nearly as many megapixels as they think they do. Real world applications with real-life viewing distances are rarely factored into pixel-peeping file comparisons like this. That being said, a 24x72 at 300ppi does require 155.5 megapixels. I’m not saying you couldn’t up-rez 30 megapixels to 24x72 at 300ppi, but it would still need to be a 155.5 megapixel file before the printer spits it out. Thanks for viewing and for providing your experiences here. The experiment pitting the 16 megapixel prints against the 40 megapixel prints is especially interesting.
Great video and love the attention to detail. For inverting the DSLR scan, look up “Negative Lab Pro” - I’m a bit biased (since I’m the creator) but with it you should be able to get much better colors out of your DSLR scan. For starters, it has raw camera profiles made specifically for scanning negatives with your digital camera (probably some of the color issues you are seeing is a result of using a standard raw profile made for regular digital images, which adds tone curves, hue twists, and other adjustments that corrupt your negative before you even begin to invert it). It’s also a one-click solution that integrates directly into Lightroom, so it is all raw, non-destructive and easy to do things like batch processing. I hope you’ll check it out, and happy to help in any way I can. Cheers!
Hi Nate, Nick gives Negative Lab Pro a 'recommend' at 25:10 and states that he used the plugin however couldn't get the colour rendition to his liking.
Oof! Right you are, Paul. Not sure how I missed that (and he even kindly included a link to Negative Lab Pro in the description). I feel like an idiot, lol.
Thanks for the knowledge sandwich Nick. Will def be contributing. Question: did you have a a chance to compare between silver fast and negative lab pro for the negative film inversion?
Another insightful, easy to understand and well presented video. Good job, Nick, and thanks for the opportunity to help defray the costs... my contribution is on it's way!
Nick, thank you ever so much for an outstanding video and your detailed explanations. In respect of color inversion from the drum scan--I am not aware of SilverFast offering their software for any such device, but you may want to try their HDR Studio software (no, this is NOT about HDR images; they named the software long before the existence of this digital imaging phenomenon), as it offers the same controls as their scanning software, including the NegaFix dialog for negatives. Having proper positive (= inverted) drum scans from Michael Strickland and loading them into SilverFast HDR Studio may offer you the desired control over the color inversion. Again, many thanks for a superb video and for sharing your insights, Michael
Exactly my point as well. If you love NegaFix there is Studio AI version of the software that will load TIFF or DNG files (DNG preferred) and do its work in the same way as NegaFix in the scanner version. I found Negative Lab Pro giving medicore colors too often in comparison.
Great video - I tried out negative Lab pro with scans I had from my epson V700 and was very pleased with the results and just how easy it was. Would love to know if you have given that a go and how it compares to using negafix.
Nick! Man! This! Was! An! Unreal/ most badass! video I've seen about scanning! By far the best quality content you'd find anywhere online! Thank you!!!!!
1:23 flatbed scanner's linear CCD sensor has three separate arrays for red, green and blue, digital camera has bayer CFA which requires interpolation, so there is a difference
"Scanning" with a DSLR takes a long time to set up but i prefer it compared to my flatbed scanner when scanning an entire roll of 35mm film. I use a old Minolta macro bellows/slide copier from the late 60s or early 70s in combination with a Nikon D7100 and a external flash. It takes me 10-20 minutes to set it up and to dial in the exposure and after that i can "scan" a roll in a minute if the roll is not cut in pieces. I adjust the exposure by changing the flash output and fine tweaking with ISO and aperture. If the next roll is the same film stock and age i can continue with that without any adjustments. I check the histogram and if good i scan the next film. I reverse the images with negative lab pro in lightroom and have a good starting point for adjustments.
Thomas Svedin thats what he did, but he said he still wasn’t happy with the colour. I guess you are happy with the colour you are getting from this set-up.
Ah ha... I got a similar set up in waiting from Pentax... haven't used it yet. Thanks for the tip! I suspect using a DSLR with 35mm will be easier than medium/large format.
@@slimnics Do you import them into Lightroom? If so, there's a plugin called Negative Lab Pro that makes converting a 2 click breeze. I honestly used to dread scanning my colour negatives because messing around with the tone curves to get colours I was happy with was a huge pain, but the previously mentioned plugin has proven to be worth the price.
Haha I'm 28 seconds in and already gave it a thumbs up for the DSLR vomit. Haha nice work Nick. I love your videos, content and humor. Thanks for all you do for film!
@@nickcarverphoto Just finished the video and brilliant and thorough comparison, thank you! It has me thinking I might try to scan some of my negatives in Silverfast again with Negafix. I started moving over to Silverfast RAW + Negative Lab Pro for that with generally decent results. It seems like Ektar in sizes larger than 35mm is a hot mess to convert on anything for me. Gorgeous results once I'm done but a huge pain. I end up with like 3 completely different images from the same negative... Portra has been easy-peasy with NLP for me. 400H ok-ish but man Ektar...oof. One of these days I'd like to see about printing some of my larger C-41 negatives in a darkroom to see how that compares - particularly Ektar for the reasons above. I don't have access to that. I know there's places to have that done though I don't know their cost (guessing it will approach or exceed a drum scan). Finally +1 for Michael Strickland. He scanned a 4x5 Provia 100 shot on a trip I took last year and it was staggeringly gorgeous.
I've had pretty decent success scanning with a camera. I'm starting with 35mm film, so most digital cameras will be sufficient to overkill for a single 35mm frame. I'm using an Olympus OM-D E-M1 Mark II -- yeah, Micro Four Thirds -- with a Nikon ES-2 film/slide holder attached to a macro lens. This completely solves the alignment problems. And I'm scanning in 8-shot sensor shift mode, so it's actually sampling twice as much light as a 6D would (my old camera), but also sampling RGGB per pixel, so there's no Bayer interpolation. Definitely outclasses my Epson V500, and at a tiny fraction of the time. That's the other thing... I've been archiving decades of film, not simply trying to make that one great print. Negative Lab is pretty good, but no, not as automatic as SilverFast. It's a shame they don't make a traditional plug-in out of that, rather than just a scanner front-end. Though if Lightroom itself understood inversions first hand, things would be dramatically easier to adjust after the Negative Lab inversion.
@@camilomarulanda6915 Nikon ES-2 (www.nikonusa.com/en/nikon-products/product/miscellaneous/es-2-film-digitizing-adapter-set.html ). It's designed to screw into the filter adapter on a Nikon Macro lens in the 60-100mm range. There's some room for adjustment. I put it on an Olympus 30mm macro lens, using a reducing adapter to fit it. There's a bit of control over the film to lens distance, so you can adjust the crop without too much trouble.
Good tips! (Though without needing to stitch the photos together) The multi shot stitching / pano seems like the challenge software still has issues with. And for some reason the (4hr old) link you posted doesn’t work for me - website says “It doesn’t look like that page is here.” - but I know the device.
@@bthemedia If you look closely the ')' at the end of the link, is blue, which means the YT comment software/algorithm(?) believes it is part of the link. Here is the proper link: www.nikonusa.com/en/nikon-products/product/miscellaneous/es-2-film-digitizing-adapter-set.html ;) Much love, Gjermund
@@DaveHaynie Thanks for sharing your idea and work flow, much appreciated. Now I am more positive into looking at an Olympus for my next camera purchase. Could I also be so bold to ask if you could edit your last comment, to put space between the link for the ES-2 film adapter set and the ')'? Like this: www.nikonusa.com/en/nikon-products/product/miscellaneous/es-2-film-digitizing-adapter-set.html ). Much love, Gjermund
26:13 Silverfast profiled every kind of combination of film types and scanners. When you reproduce with the dslr, you have no profiles for the camera/film combination, so you try to color correct with your eye which will not bring good and reproducable results. This is a common problem. Getting a colorchecker SG and profile the combination on your own would bring you better results. But u need an xrite i1pro2 publish for measuring the colorchecker sg and for making the icc profile. Or basiccolor input. So i don't think that solution is practicable for you. btw. Excellent review...
Interesting... especially considering the variety of “film scanners / sensors” is much less diverse than the number of “digital cameras for scanning” - having a profile for each scanner seems possible (especially if many scanners use the same sensor technology).
A few notes on this... I used to own a drum scanner, I've owned a few Epson scanners and a (better than the Epson) Microtek flatbed scanner, a Nikon VS9000 film scanner, and of course, I've done the digital camera approach. I also used to run scanning workshops for LF shooters who were transitioning to the digital age. The Epson scanners top out at about 2200 lppm for delivered resolution if I recall correctly. I don't think I tested the V750, but that is where they all essentially ended up. Beyond that, you aren't really getting more information when you scan above the 2400 SPI setting, but it's probably no worse than up-sampling in PS, so choose your poison for that. The drum scanner I owned tested out to about 3500 lppm, so it was clearly superior to the Epson (and the Microtek as well). The Nikon also came out over 3000lppm, but it was limited to 6x9 film if I recall correctly. This becomes much more important with smaller film sizes (35mm, for example), because they magnification ratio typically becomes larger. For example, a 12x magnification from 35mm film yields approximately a 12"x18" image, which isn't very large by many people's expectations. It's certainly not a wall-filling image at that size. 20x magnifications are pretty common at those film sizes. There, the sharpness differences become much more apparent and important for the final results. The drum scanners (all that I'm aware of, anyway) have an aperture to them. This aperture interacts with the image in positive or sometimes negative ways, so you really need someone who knows scanning well (and cares to do some prescan testing) to ensure that you will get the 'best' result. Some of this is subjective as well, so there really isn't a 'right' answer; just what you may like the best or that meets your aesthetic preferences. This is important for my next comment... It looks to me like there was a good bit of grain aliasing in the drum scan from the negative. The grain pops out more than it should (at least to my preferences, and from the sound of it, to your preferences as well). This may have been reduced somewhat by changing the aperture of the scanner from a smaller one to a larger one since I presume they didn't run a full resolution scan of the film for you. Again, this is aesthetic-driven; some people will like the crunchier look, others will like a smoother, less film-grain look. This may have been exacerbated by the sharpening you did to the scan because once you have the grain aliasing in the scan, it can be hard to not accentuate it without turning the image to mush first. The grain aliasing is dependent on the type of film so it becomes a bit of a witch's brew when comparing different film types across different scanning options (especially when some of them don't have any aperture options at the sensor, as is the case with the flatbed and the film approaches). It also comes down to the operator. In the slide scan, the Aztek scan could have been made more gritty-looking by changing the aperture, but most scanner operators try to balance the aesthetic, and I suspect that they did that a little more than was done in the negative scan. I suggest that if you are going to shoot with a digital camera for this, you probably should use one that does not have an AA filter built into the sensor to achieve a bit better results. If that's the only camera you have, then you use what you have, but it won't really be a fair comparison to the other options. Second, if you didn't do it when setting up your shooting arrangement for the camera and the light table, you should make a mask to ensure that the lens is only seeing the portion of the area that is recorded on the sensor and no more to reduce internal lens flare and reduced contrast in the shadow zones that can cause difficulty making an image that has proper shadow contrast. You may have done that, but didn't show it in your setup. It will have an impact in how the system renders the image. There's lots more, but overall, it appears you did a good, thorough job with this comparison, but I think that people need to understand that some of the decisions you made are very dependent on the source material and conditions as it is on the scanning process. Your answers may be different given different film sizes and other variables that will impact the results.
Great post. I was considering using my sony A7iii or A6300 but not so sure now. Definitely the workflow issues mentioned could be ironed out but colour and ending up with something that looks digital is a real worry.
@@SnapographyAP For what it's worth, I have been using a Sony a7r3 for 35mm slide scanning for a few years. The quality easily exceeds what I have seen from flatbed scans, and in most cases, appears to be recording film grain reasonably well (such that it is unlikely to produce a sharper result using a drum scanner, for example). It's even possible to make multiple exposures with Pixel Shift with this approach and a bit better color fidelity is achieved, but frankly it isn't worth the extra work based on what I saw. Scanning odd film sizes like the 6x17 he does is a rel problem if you need a very high resolution final file for very large reproductions. At that point, I would pull out the flatbed (Microtek) and use that, but honestly, that is so rare for most people that it might just be better to find a good (competent) scanning shop to do the work rather than deal with it internally (but I completely understand the desire to control all the variables). For things like 4x5 film, I'd stick to the digital camera approach unless a huge file is needed. At that point, I'd use the flatbed and wet-mount or go out to a shop.
@@mjmdiver1137 cheers that's a big help. I think the video was very useful because it highlights issues but trying to draw conclusions when there are so many variables, including your own experience is tricky. What do you use to hold your negatives?
@@nicolacammisa5205 Yes, I would add that the mask should be for the area of coverage of the shot, not the area of the image/frame of film (unless it is a non-stitch full-frame capture). This masking is something that LF photographers would do a lot because having image circle light into the lens that isn't part of forming the actual frame cuts contrast and can cause flare and other odd behavior. It's actually pretty easy to see at times, especially with digital where you can apply some curves to the file.
4 года назад+6
Hej Nick! Did you tried to stitch the panorama with PTgui? I have stitched around 1000 full equirectangular panoramas. My worst mistake was always the camera position on the panoramic head. If it's shifted 1mm or more, my panorama wouldn't stitch perfectly(I know this is not your case). Also Photoshop's auto align panorama tool is a baby tool compared to the much more detailed softs out there. The reason why photoshop could not do the job is because the algorithm is for the 'sensor in one place' .. and you did moved the sensor, so basically it is not a panorama stitch what you were trying to do there. It is some kind of mapping thing which would not work with photoshop's basic image alignment. If you would bend the film somehow and photograph from the same sensor starting point then PS would be able to do the job pretty good. The other thing is PTgui. That software is absolutely amazing doing stitching individual frames. You can control the control points (you can match the power lines in your case, etc.). One more tip. The sweet spot of your macro lens is f8 not f/11. F/10 is probably the least usable smallest aperture that you can use on the USM version on your 6D. Your 6D has a pixel size of 5.96 μm and after f/10 the diameter of your diffraction airy disk becomes larger than you pixel size = img quality loss. One more tip. Hide the light table's not needed light sources. Darken but the negative. Try to forget the light table, use flash instead - it absolutely avoids camera shake. Also I believe that the power of the 14 bit raw is underestimated. One more tip... wait, that's enough.
4 года назад
Oh yes! Thank you for your amazing videos! Really really helpful and enjoyable. Greetings colleague!
4 года назад
Here is another tip Nick! I know you are not that 'sharp kinda guy' BUT if you really want to see the recorded grain itself on the prints or on the files, then DO NOT RESIZE your image. Make a print as same as the actual file resolution. Do not fool around with pixels. Imagine if you have a prime number of matrixes you need to divide it somehow, so no matter what type of image resizing matrix algorithm you choose (you can find really interesting stuff how image resize works on forums) on your soft, it will be always cheating and adding some crazy shit to it or removing important pixels. I use Epson scanners also, but If you doing it RIGHT I mean RIGHT, I think the DSLR scan is unbeatable, both in color and resolution even if we take the drum scan method as well. Yeah... it's a great shit load of hustle and struggle but if you are making 1 image in a week, I think it is worth it. Cheers! - Franziskaner wheat beer.
Color rendition is exactly why I gave up scanning on my own : when I saw the result my lab was able to get with a Fuji SP3000 (waaaaaaay better than the Noritsu), I couldn't come back.
*YOU'RE TOTALLY WRONG, YOU SCREWED THIS UP SO MUCH!!* There. With that out of the way, I can now proceed to stop the video after not hearing you say exactly what I expected about the method I like best and go back to eating glue while I stare at the sun
You actually could do it in a „single“ exposure with the dslr, so avoid stitching the images. Just zoom in so you have the whole negative in frame and take multiple exposures (the more images, the better, but use the same settings on each of em). Load all your image in a stack into photoshop, convert them to one smart object and select stack mode „average“ or „median“. Once this is done you can just pump up your image resolution without losing much quality.
I think you are referring to the “high resolution” mode of cameras that do “pixel level shifting” of the sensor, so that the pixel resolution (eg 2x2 RGGB Bayer sensor) overlaps and is averaged. I believe Tony Northrop made a video comparing this against “hand held” equivalent (i.e. subtle overlap / shake from hand held) and let PS do a statistical average overlay of the image to remove random sensor noise. That technique may work for landscape photography, but I do not think it applies here to macro photography where the lens is focused on the grain of the film to be “scanned”.
bwvids it would work perfectly. A *tiny* nudge to the holder is all that is needed. Exposing at different stops, essentially bracketing would also really help with the range he’s so desperately needs to avoid “weird banding” artifacts.
Thank you my guy. I've been looking into getting a drum camera as my v700 scans are meh. And lastly I'll never use a digital camera to scan my negs because the cost and I'll probably never use it after
@@EdwinNew If you have a scanning transmission electron microscope you can have an electron energy loss spectrometer (EELS) as a detector to see "colors"
Oh my god man. The effort you go to with this channel is flabbergasting ! It must have taken you hours upon hours....the tech spec slides, the discussions with the labs, setting up the various methods of scanning. Not to mention how much it must of cost you to make it. It's made me tired just watching it! I look forward to your videos man. They're so entertaining but equally so informative. It's obvious you really know your onions; the comical delivery just makes it all the more engaging (though your sick story when travelling back from Europe is still the funniest RUclips I've ever seen!). Don't know if you know but you've been mentioned a few times on podcasts I listen to. It would be great for you to be a guest on one of them, one day (I seem to think I listened to one when you were a guest in fact but I forget which it was - I recall you telling the story of the "Glass behind a glass" title?). Your efforts also remind me again of why I send all my important stuff to a lab to deal with. All that fluid mounting and setting up SLRs on tripods is just too much fuss for my sanity. Re Silverfast and the "NegaFix"...do you know what version that feature comes with? Is it only with the more expensive line or does it come with the standard line (I bought my Epson V550 second hand so it not included). Keep up the great work mate, for the sake of all of us without the commitment to make such great videos. I'll try to contribute when I can, but sadly I am rather hard up. If/when I had more surplus dough, you'd be the first to get some. As for the haters and moaners...dont let them drive you down. RUclips is full of bell ends these days with too much time on their hands and not enough experience to have an opinion on anything, let alone the fine art process of analogue photography.
Here I'll summarize the comments section for ya:
Most comments: "Hey cool thanks man" or "here are some helpful tips to get better results with you DSLR scan"
A few comments: "You suck, your camera sucks, you didn't even use a camera with pixel shift, and you're an idiot"
I guess we can't all still be friends 😂
On a more positive note: THANK YOU to the many folks who made a contribution! I'm very encouraged and surprised by the number of viewers who were kind enough to pitch in a couple bucks to make more videos like this possible. I look forward to bringing you more entertaining and informative videos this year!
Hi Nick,
Sure we can still be friends, I guess I fall into the second group?
My point was that your assumption you were "bringing you more entertaining and informative videos" is a little off the mark. Entertaining YES. Informative NOT So much.
First a couple of points, yes I watched the whole video, have worked in prepress with drum scanners, own an Epson and do not own an A7RIV, and shoot both film and digital, for over 25 years, yes I am old.
I get the pretend vomiting at the beginning was a joke, I watch the whole bit, it was actually kind of funny, and was intended to be self deprecating. I get that.
That said, the techniques shown in the video showed a profound bias, perhaps unconscious on your part, but there none the less. Let me elaborate.
The statements you made about the quality of drum scans are correct. They are very good, and will give you excellent results. The issues there are cost, if you have to pay someone else, and the ageing nature of the technology, as no one makes new ones, and service is becoming problematic.
You also presents a good case for flat bed scans, and show good techniques on how to get the best results out of this gear with fluid mounting etc. And yes I have an EpsonV700 that I use with fluid mount, I agree you can get excellent results with careful technique.
When it comes to the camera scanning though, you go massively off the rails, and basically do everything you could possibly think of wrong, and then complain it is frustrating and gives poor results.
Where to start.
The camera, an ageing mid to entry level DSLR. First mistake. And yes I know it was what you had on hand, but if you are going to test these a technique to assess its potential for quality, at least use a descent model. Even if you have to borrow one. Especially if you are then going to make a RUclips video about it, do some research and use the best gear possible. C'mon dude.
If I was doing this test I would have borrowed a Sony A7RIV as others have mentioned, which can do multi shot exposures and produce a 240MP image.
This would have eliminated most, if not all of your most vigorous complaints related to stitching. As with a file that size, it becomes unnecessary, especially if you are "not a sharpness whore".
Secondly, you go to so much trouble to carefully mount the film and mask it for the scan. Then just ignore all that good technique when you do the camera scan.
I would fluid mount exactly as you do for the scan, and mask as you do, for the camera scan, and for gods sake put a lens hood on. Lens flare is not your friend!
Thirdly, If you really do feel the need to stitch (remember, you are "not a sharpness whore") then put your camera on a copy stand and MOVE THE FILM, not the camera.
I could go on but you get the point.
You obviously, although I suspect unconsciously, predetermined the results.
If you take the time to develop good technique, as you have done with scanning, then I suspect the results would have been quite different.
If you see saying these things as "being a dick" I am sorry, but if you put stuff out there, expect feed back, and sometimes it will not be glowing. If it is however intended to be constructive (my first post could have been kinder I guess) then take a step back and reflect on it.
Kind regards,
Jason Berge.
www.shoot.co.nz
@@nickcarverphoto Hi Nick,
Sorry you felt I was being patronising, I guess that comes from being old enough to be your dad.
I don't know "everything", not yet. Which is why I was interested in the video. I can always learn something and try to, every day.
Cheers,
Jason.
I find your act disgusting, no matter what the useful info you were going to present, I could not stand you longer than 59 second. Down-vote well deserved. Better luck the next time.
@@jasmb570 cool comment dude, whens your motion picture documentay about the soja 7 VII scanner and general ass-holyness coming out? :P
Kinder regard
A dude who uses a dedicated Fuji film scanner :D
🤣 Can please everyone.
My favorite thing about Nick is that he hates on his haters before they can even hate
I also try to hate on myself before they can hate on me. Haha
@@nickcarverphoto That's actually the reason why I couldn't watch the video to th end.
Why would you hate on Nick? He's trying to help us out. How about the haters create some useful You Tube content for a change, and do something constructive with their time.
Right. Because when people don't like something it means that they never do anything useful.
Thats not going to get him a good group of people. Grow up, focus on the content and the shared love of photography.
I followed your instructions and scanned my film with a Etch-a-Sketch. I got nominated for a pulitzer prize, thanks!
I'm ridiculously selfish in that I wish RUclips was your main gig. In a world of endless repetitive photography gear review videos that all say the same thing, you're actually dolling out some useful information.
Your videos inspired me to dive back into film and it 100% saved me from the boredom I developed with digital.
Thanks for all that you do.
Oh, and cheers!
Damn son you clearly put a ton of effort into testing all these methods, appreciate ya 👍
Wow, it's you!
Thank you all the way from Tel Aviv for making such a brilliant video on this fascinating subject matter!
All of the questions I had in my head while watching Nick's video - were all answered after the video was completed. Thanks Nick!
Thanks Nick for doing this for all of us. I have been waiting for a comprehensive test like this for a long time. Certainly a lot of factors to consider going forward, and things to weigh back and forth. Your hard work is appreciated.
This is exactly what I wanted to see, thank you Nick! Love your channel and your work!
Dude, thats great... So much work in just one Video... That alone deserves a thumbs up :)
Sir, you've done great work! Good methodology, good presentation, and a very good assessment based on what's important to you. You hit all the right points and you've done it in really clear and concise manner. Kudos!
I also appreciate how much work it took to do all of these tests and analyze the results. Thank you.
Such a great video! Really appreciate the time and effort that went into this. Great humour too. Two thumbs up!
My god, the difference between colors with Aztek and Strickland scans 😍
Fantastic video as always! I come for the knowledge and stay for the charisma
Yeah, and Michael Strickland was more than happy to look at my Epson scan to try and match it. Standup guy.
Had to watch the video until this made sense. Was about to start looking for a "used Strickland scanner".
Outstanding analysis! Thank you for sharing your data and thoughts.
Strickland's work is amazing. I'm glad he reached out to you. He's a film photographer so he gets what we are up against when it comes to scanning our film. Really nice video, Nick. Thanks....
Wow! Amazing video! Thanks for posting it.
Not only is this wonderfully organized and presented, it’s entertaining af. Kudos to you Nick 👏👏👏
exactly the video i was looking for these last couple days
Boom!
ahh finally, man of quality is back
Thank you, sir!
Holy crap this was a good video! So much useful information and so well presented. Thank you!
Now that is what I call a thorough, clear, concise and even scientific comparison of three methods of scanning. Well done and seriously thank you! That must have been a serious amount of time and effort. You’re a good man and I hope you know that we all really appreciate your hard work.
Colour is more important than sharpness. Amen to that.
I'm the complete opposite although I respect that opinion. Don't care about color so much as long as it's sharp. I guess that's why he said that part of the video is divisive.
Well done Nick! Didn’t even feel like a 36 minute video. It was interesting and kept my attention.
Real Life Ryan same here! So much valuable insight gained from his months of experience in those 36 min! It was like a massage to my mental stress of theoretically debating over this exact same question! 💆♂️
Nick, you are so bloody good at these. No umms and errs, no jump cuts, no unnecessary words just on the money start to finish! Another great one. Cheers
Good film helps make a good drum scan. Great shot, and happy to help out! Looking forward to seeing more of your work on my drum.
I've been experimenting with stitch-scanning 5x4 film with my DSLR, but haven't had any point of comparison so this vid is just what I needed
I hope your stitching goes more smoothly than mine.
I've tried dslr scanning with single frame 35mm and I pretty much ran into the same issues with colours and workflow too. I'm now shooting mostly 6x6 and starting to go for 6x12 & 4x5 and F stitching 20frames together. I went with the "baby" v550 and I now enjoy photography again
I've done a lot of DSLR scanning, but the real secret to a stress-free process is stick to B&W and don't stitch anything ever. I've tried to stitch-scan 4x5 and even when you think it's right, it's probably still wrong you just haven't noticed yet (images magically growing width somehow). :)
@@nickcarverphoto I've found that it works well with highly textural/detailed images IF your camera and film are perfectly parallel, but fails terribly on things like abstract still life or anything where the stitch software can't find seams. In those cases, I suppose it's possible to manually align the images, but quite painful.
I've also found the colours to be quite difficult to get right when scanning colour negative with my DSLR, but I thought I was just bad at colour, so it's kind of reassuring to see that you had a difficult experience as well.
Don't know if youtube will let me link, but this is an example of an image I recently scanned with the DSLR stitch method, and a crop of approximately 35% instagram.com/p/B7fX3euAN-M/?igshid=vlpmey2revpf
The resolution and sharpness is very nice, but your video has made me consider whether an Epson might be better purely for the ease of getting nice colours...
@@nickcarverphoto I use this free software from Microsoft called Image Composite Editor. It rarely fails to stitch images together. It was written to stitch hundreds of photos. www.microsoft.com/en-us/research/product/computational-photography-applications/image-composite-editor/
What a video, I can't believe I watched 30 minutes of film scanning knowing I never going to do it😄, much less with that attention to detail.
Good job man!
I watched the last 3 minutes and that was enough.
Nice to see Nick back. Great comparison and very informative.
What a FANTASTIC video. Amazing information presented in a straight forward manner. I appreciate the many hours it must have taken to produce this video.
Great video and comparison. Two things to digitalize with a digital camera.
1. Masking the light board and only leaving the film surface can immensely increase color accuracy.
2. Every color channel has a dynamic range. Because of the orange mask of negative film the red channel exposes differently. With a blue filter you can even out those differences and have more beef for your inversion. I don't know actually how good the 6D sensor is, but not every Canon is known for having the best dynamic range, probably a Sony made sensor in a Nikon or Pentax or even a Faveon sensor might help also in this regard.
So quick improvements, record with a blue filter and mask the stray light from your lightpad
Plus 1 on the Foveon Sensor for colours however this is personal and subjective and also Microsoft Ice for Stitching - great job and best presentation style as usual
I was thinking the mask might help for DSLR scanning, but does it apply both in “one shot” and “multi partial shots for stitching”? The film has a slight frame around it already. (I’m just theorizing here without testing)
As for 20MP Canon 6D (which I still shoot), I think that only affects number of stitches (MP) and colors (camera sensor)... other cameras will perform differently.
Are you suggesting an actual “blue filter” in front of the lens, or a layer in Photoshop? Have you done this to demo? How do you choose the proper color of “blue” for your filter?
1) shooting through a physical filter may affect sharpness
2) I presume PS can do the same color inversion with or without an additional blue filter
To even out the red channel I just adjust the camera's white balance to where each channel is somewhat the same. That way I don't have to worry about the red channel clipping when "scanning" with my DSLR.
Also, to get the camera and the film to perfectly level I use a mirror that I place on the surface and once I see the center of my macro lens I know the camera is set up right.
@@Verdoux007 that with the mirror is clever, the wide balance though will not better handle physically the orange mask, but the filter will
I've been a photographer for many years and am always learning - thank you for your detailed video. In my experience (I've printed around 1000 various prints ranging larger ones about 3 feet by 5 feet) and never once did I require anything close to the resolution you are talking about - even for exhibition quality prints. I've even done tests to where I have a 16 megapixel image (These were 4 feet by 3 feet) and not one person even noticed a difference between my 40 megapixel shots and my 16's. When people view prints, they don't view them from 6 inches away - it's the sharpness and quality of glass and camera sensor that makes the difference. I've done tests where I have literally large prints at 20 megapixels sitting on a table and have had dozens of people look at them and they can't tell the difference between that and my high resolution cameras. In fact, many billboards are actually only 3 megapixels. Also, people scanning film at these very high resolutions - when the film itself is no more than 8-12 megapixels for the negative itself doesn't make any sense. This hole resolution thing is completely overblown - I've never had a client come back to me saying "boy, that wasn't 200 megapixels" - I was looking at your art print with my magnifying glass for 3 hours and noticed that... Anyhow, just my 2 cents.
I couldn’t agree more. I’ve told my students for years that they don’t need nearly as many megapixels as they think they do. Real world applications with real-life viewing distances are rarely factored into pixel-peeping file comparisons like this. That being said, a 24x72 at 300ppi does require 155.5 megapixels. I’m not saying you couldn’t up-rez 30 megapixels to 24x72 at 300ppi, but it would still need to be a 155.5 megapixel file before the printer spits it out. Thanks for viewing and for providing your experiences here. The experiment pitting the 16 megapixel prints against the 40 megapixel prints is especially interesting.
You are awesome! A test that we all wanted to see. As thorough as I could have wished for. Thank you, sir!
Top content. Thank you so much Nick. Well done, again. It´s so rare in these days to find a photographer sharing his knowledge the way you do.
Great video and love the attention to detail. For inverting the DSLR scan, look up “Negative Lab Pro” - I’m a bit biased (since I’m the creator) but with it you should be able to get much better colors out of your DSLR scan. For starters, it has raw camera profiles made specifically for scanning negatives with your digital camera (probably some of the color issues you are seeing is a result of using a standard raw profile made for regular digital images, which adds tone curves, hue twists, and other adjustments that corrupt your negative before you even begin to invert it). It’s also a one-click solution that integrates directly into Lightroom, so it is all raw, non-destructive and easy to do things like batch processing. I hope you’ll check it out, and happy to help in any way I can. Cheers!
Hi Nate, Nick gives Negative Lab Pro a 'recommend' at 25:10 and states that he used the plugin however couldn't get the colour rendition to his liking.
Oof! Right you are, Paul. Not sure how I missed that (and he even kindly included a link to Negative Lab Pro in the description). I feel like an idiot, lol.
Thanks for the knowledge sandwich Nick. Will def be contributing.
Question: did you have a a chance to compare between silver fast and negative lab pro for the negative film inversion?
Another insightful, easy to understand and well presented video. Good job, Nick, and thanks for the opportunity to help defray the costs... my contribution is on it's way!
Great video Nick, thanks so much for putting it together.
this is the best channel on youtube if we're being honest
Close!
I agree
Nick, thank you ever so much for an outstanding video and your detailed explanations.
In respect of color inversion from the drum scan--I am not aware of SilverFast offering their software for any such device, but you may want to try their HDR Studio software (no, this is NOT about HDR images; they named the software long before the existence of this digital imaging phenomenon), as it offers the same controls as their scanning software, including the NegaFix dialog for negatives. Having proper positive (= inverted) drum scans from Michael Strickland and loading them into SilverFast HDR Studio may offer you the desired control over the color inversion.
Again, many thanks for a superb video and for sharing your insights, Michael
Exactly my point as well. If you love NegaFix there is Studio AI version of the software that will load TIFF or DNG files (DNG preferred) and do its work in the same way as NegaFix in the scanner version. I found Negative Lab Pro giving medicore colors too often in comparison.
I really appreciate your level of detail and intentionality in your methods. Thank you!
After just the first 6minutes I needed to drop a thumbs up just for the amount of work.
Congratulations and thank you very much, Nick!
Hey, my best prints come from etch-a-sketch scans. Much underrated.
Great video - I tried out negative Lab pro with scans I had from my epson V700 and was very pleased with the results and just how easy it was. Would love to know if you have given that a go and how it compares to using negafix.
Superbly informative and helpful, as always. Many thanks, Nick.
Thanks for all the effort Nick, really appreciated it!
Do you ever have your images printed without converting to digital?
"If it's for the web, you can scan it with an etchasketch for all I care." Brutal, but it's true.
Holy Crap. Now that’s comprehensive. Loved it. Thanks for the amazing video.
You are a legend!
Thank you for all your hard work
You're the best, dude! 😂
Love your videos, keep them coming!
Nick! Man! This! Was! An! Unreal/ most badass! video I've seen about scanning! By far the best quality content you'd find anywhere online! Thank you!!!!!
1:23 flatbed scanner's linear CCD sensor has three separate arrays for red, green and blue, digital camera has bayer CFA which requires interpolation, so there is a difference
"Scanning" with a DSLR takes a long time to set up but i prefer it compared to my flatbed scanner when scanning an entire roll of 35mm film.
I use a old Minolta macro bellows/slide copier from the late 60s or early 70s in combination with a Nikon D7100 and a external flash.
It takes me 10-20 minutes to set it up and to dial in the exposure and after that i can "scan" a roll in a minute if the roll is not cut in pieces. I adjust the exposure by changing the flash output and fine tweaking with ISO and aperture.
If the next roll is the same film stock and age i can continue with that without any adjustments. I check the histogram and if good i scan the next film.
I reverse the images with negative lab pro in lightroom and have a good starting point for adjustments.
Thomas Svedin thats what he did, but he said he still wasn’t happy with the colour. I guess you are happy with the colour you are getting from this set-up.
Ah ha... I got a similar set up in waiting from Pentax... haven't used it yet. Thanks for the tip! I suspect using a DSLR with 35mm will be easier than medium/large format.
@@slimnics Do you import them into Lightroom? If so, there's a plugin called Negative Lab Pro that makes converting a 2 click breeze. I honestly used to dread scanning my colour negatives because messing around with the tone curves to get colours I was happy with was a huge pain, but the previously mentioned plugin has proven to be worth the price.
This video (and a lot of your videos) are presented like a well taught class. I really appreciate it, and enjoy watching and learning!
Great video! Thanks for taking the time go test them all and then share your thoughts!
Haha I'm 28 seconds in and already gave it a thumbs up for the DSLR vomit. Haha nice work Nick. I love your videos, content and humor. Thanks for all you do for film!
Thank you, sir!
@@nickcarverphoto Just finished the video and brilliant and thorough comparison, thank you! It has me thinking I might try to scan some of my negatives in Silverfast again with Negafix. I started moving over to Silverfast RAW + Negative Lab Pro for that with generally decent results. It seems like Ektar in sizes larger than 35mm is a hot mess to convert on anything for me. Gorgeous results once I'm done but a huge pain. I end up with like 3 completely different images from the same negative... Portra has been easy-peasy with NLP for me. 400H ok-ish but man Ektar...oof.
One of these days I'd like to see about printing some of my larger C-41 negatives in a darkroom to see how that compares - particularly Ektar for the reasons above. I don't have access to that. I know there's places to have that done though I don't know their cost (guessing it will approach or exceed a drum scan).
Finally +1 for Michael Strickland. He scanned a 4x5 Provia 100 shot on a trip I took last year and it was staggeringly gorgeous.
Got me laughing out the gate too! 🤢=😆
Yes me too, hilarious. And towards the end, the real question: why shoot film, if it looks like Canon ditital?
I've had pretty decent success scanning with a camera. I'm starting with 35mm film, so most digital cameras will be sufficient to overkill for a single 35mm frame. I'm using an Olympus OM-D E-M1 Mark II -- yeah, Micro Four Thirds -- with a Nikon ES-2 film/slide holder attached to a macro lens. This completely solves the alignment problems. And I'm scanning in 8-shot sensor shift mode, so it's actually sampling twice as much light as a 6D would (my old camera), but also sampling RGGB per pixel, so there's no Bayer interpolation. Definitely outclasses my Epson V500, and at a tiny fraction of the time. That's the other thing... I've been archiving decades of film, not simply trying to make that one great print. Negative Lab is pretty good, but no, not as automatic as SilverFast. It's a shame they don't make a traditional plug-in out of that, rather than just a scanner front-end. Though if Lightroom itself understood inversions first hand, things would be dramatically easier to adjust after the Negative Lab inversion.
@@camilomarulanda6915 Nikon ES-2 (www.nikonusa.com/en/nikon-products/product/miscellaneous/es-2-film-digitizing-adapter-set.html ). It's designed to screw into the filter adapter on a Nikon Macro lens in the 60-100mm range. There's some room for adjustment. I put it on an Olympus 30mm macro lens, using a reducing adapter to fit it. There's a bit of control over the film to lens distance, so you can adjust the crop without too much trouble.
Good tips! (Though without needing to stitch the photos together) The multi shot stitching / pano seems like the challenge software still has issues with.
And for some reason the (4hr old) link you posted doesn’t work for me - website says “It doesn’t look like that page is here.” - but I know the device.
@@bthemedia If you look closely the ')' at the end of the link, is blue, which means the YT comment software/algorithm(?) believes it is part of the link.
Here is the proper link:
www.nikonusa.com/en/nikon-products/product/miscellaneous/es-2-film-digitizing-adapter-set.html
;)
Much love,
Gjermund
@@DaveHaynie Thanks for sharing your idea and work flow, much appreciated. Now I am more positive into looking at an Olympus for my next camera purchase.
Could I also be so bold to ask if you could edit your last comment, to put space between the link for the ES-2 film adapter set and the ')'? Like this:
www.nikonusa.com/en/nikon-products/product/miscellaneous/es-2-film-digitizing-adapter-set.html ).
Much love,
Gjermund
@@akirafan28 Done.. works now!
new vid!!! omgso excited! love nick!
So much respect. You’re a total role model to look up to.
26:13 Silverfast profiled every kind of combination of film types and scanners. When you reproduce with the dslr, you have no profiles for the camera/film combination, so you try to color correct with your eye which will not bring good and reproducable results. This is a common problem. Getting a colorchecker SG and profile the combination on your own would bring you better results. But u need an xrite i1pro2 publish for measuring the colorchecker sg and for making the icc profile. Or basiccolor input. So i don't think that solution is practicable for you.
btw. Excellent review...
Interesting... especially considering the variety of “film scanners / sensors” is much less diverse than the number of “digital cameras for scanning” - having a profile for each scanner seems possible (especially if many scanners use the same sensor technology).
That's cool. I can no longer focus at 10" without my glasses and I can never find them anyhow :).
You're gnarly. Thanks for putting all the work into this.
What a video, lot of work behind it, really appreciate man!
A few notes on this... I used to own a drum scanner, I've owned a few Epson scanners and a (better than the Epson) Microtek flatbed scanner, a Nikon VS9000 film scanner, and of course, I've done the digital camera approach. I also used to run scanning workshops for LF shooters who were transitioning to the digital age.
The Epson scanners top out at about 2200 lppm for delivered resolution if I recall correctly. I don't think I tested the V750, but that is where they all essentially ended up. Beyond that, you aren't really getting more information when you scan above the 2400 SPI setting, but it's probably no worse than up-sampling in PS, so choose your poison for that. The drum scanner I owned tested out to about 3500 lppm, so it was clearly superior to the Epson (and the Microtek as well). The Nikon also came out over 3000lppm, but it was limited to 6x9 film if I recall correctly.
This becomes much more important with smaller film sizes (35mm, for example), because they magnification ratio typically becomes larger. For example, a 12x magnification from 35mm film yields approximately a 12"x18" image, which isn't very large by many people's expectations. It's certainly not a wall-filling image at that size. 20x magnifications are pretty common at those film sizes. There, the sharpness differences become much more apparent and important for the final results.
The drum scanners (all that I'm aware of, anyway) have an aperture to them. This aperture interacts with the image in positive or sometimes negative ways, so you really need someone who knows scanning well (and cares to do some prescan testing) to ensure that you will get the 'best' result. Some of this is subjective as well, so there really isn't a 'right' answer; just what you may like the best or that meets your aesthetic preferences.
This is important for my next comment... It looks to me like there was a good bit of grain aliasing in the drum scan from the negative. The grain pops out more than it should (at least to my preferences, and from the sound of it, to your preferences as well). This may have been reduced somewhat by changing the aperture of the scanner from a smaller one to a larger one since I presume they didn't run a full resolution scan of the film for you. Again, this is aesthetic-driven; some people will like the crunchier look, others will like a smoother, less film-grain look. This may have been exacerbated by the sharpening you did to the scan because once you have the grain aliasing in the scan, it can be hard to not accentuate it without turning the image to mush first.
The grain aliasing is dependent on the type of film so it becomes a bit of a witch's brew when comparing different film types across different scanning options (especially when some of them don't have any aperture options at the sensor, as is the case with the flatbed and the film approaches). It also comes down to the operator. In the slide scan, the Aztek scan could have been made more gritty-looking by changing the aperture, but most scanner operators try to balance the aesthetic, and I suspect that they did that a little more than was done in the negative scan.
I suggest that if you are going to shoot with a digital camera for this, you probably should use one that does not have an AA filter built into the sensor to achieve a bit better results. If that's the only camera you have, then you use what you have, but it won't really be a fair comparison to the other options. Second, if you didn't do it when setting up your shooting arrangement for the camera and the light table, you should make a mask to ensure that the lens is only seeing the portion of the area that is recorded on the sensor and no more to reduce internal lens flare and reduced contrast in the shadow zones that can cause difficulty making an image that has proper shadow contrast. You may have done that, but didn't show it in your setup. It will have an impact in how the system renders the image.
There's lots more, but overall, it appears you did a good, thorough job with this comparison, but I think that people need to understand that some of the decisions you made are very dependent on the source material and conditions as it is on the scanning process. Your answers may be different given different film sizes and other variables that will impact the results.
Great post. I was considering using my sony A7iii or A6300 but not so sure now. Definitely the workflow issues mentioned could be ironed out but colour and ending up with something that looks digital is a real worry.
@@SnapographyAP For what it's worth, I have been using a Sony a7r3 for 35mm slide scanning for a few years. The quality easily exceeds what I have seen from flatbed scans, and in most cases, appears to be recording film grain reasonably well (such that it is unlikely to produce a sharper result using a drum scanner, for example). It's even possible to make multiple exposures with Pixel Shift with this approach and a bit better color fidelity is achieved, but frankly it isn't worth the extra work based on what I saw.
Scanning odd film sizes like the 6x17 he does is a rel problem if you need a very high resolution final file for very large reproductions. At that point, I would pull out the flatbed (Microtek) and use that, but honestly, that is so rare for most people that it might just be better to find a good (competent) scanning shop to do the work rather than deal with it internally (but I completely understand the desire to control all the variables).
For things like 4x5 film, I'd stick to the digital camera approach unless a huge file is needed. At that point, I'd use the flatbed and wet-mount or go out to a shop.
@@mjmdiver1137 cheers that's a big help. I think the video was very useful because it highlights issues but trying to draw conclusions when there are so many variables, including your own experience is tricky. What do you use to hold your negatives?
Great tip on masking out the area surrounding the film portion, if using a light box and DSLR. Thank you!
@@nicolacammisa5205 Yes, I would add that the mask should be for the area of coverage of the shot, not the area of the image/frame of film (unless it is a non-stitch full-frame capture).
This masking is something that LF photographers would do a lot because having image circle light into the lens that isn't part of forming the actual frame cuts contrast and can cause flare and other odd behavior. It's actually pretty easy to see at times, especially with digital where you can apply some curves to the file.
Hej Nick! Did you tried to stitch the panorama with PTgui? I have stitched around 1000 full equirectangular panoramas. My worst mistake was always the camera position on the panoramic head. If it's shifted 1mm or more, my panorama wouldn't stitch perfectly(I know this is not your case). Also Photoshop's auto align panorama tool is a baby tool compared to the much more detailed softs out there. The reason why photoshop could not do the job is because the algorithm is for the 'sensor in one place' .. and you did moved the sensor, so basically it is not a panorama stitch what you were trying to do there. It is some kind of mapping thing which would not work with photoshop's basic image alignment. If you would bend the film somehow and photograph from the same sensor starting point then PS would be able to do the job pretty good.
The other thing is PTgui. That software is absolutely amazing doing stitching individual frames. You can control the control points (you can match the power lines in your case, etc.).
One more tip. The sweet spot of your macro lens is f8 not f/11. F/10 is probably the least usable smallest aperture that you can use on the USM version on your 6D. Your 6D has a pixel size of 5.96 μm and after f/10 the diameter of your diffraction airy disk becomes larger than you pixel size = img quality loss.
One more tip. Hide the light table's not needed light sources. Darken but the negative. Try to forget the light table, use flash instead - it absolutely avoids camera shake. Also I believe that the power of the 14 bit raw is underestimated.
One more tip... wait, that's enough.
Oh yes! Thank you for your amazing videos! Really really helpful and enjoyable. Greetings colleague!
Here is another tip Nick! I know you are not that 'sharp kinda guy' BUT if you really want to see the recorded grain itself on the prints or on the files, then DO NOT RESIZE your image. Make a print as same as the actual file resolution. Do not fool around with pixels. Imagine if you have a prime number of matrixes you need to divide it somehow, so no matter what type of image resizing matrix algorithm you choose (you can find really interesting stuff how image resize works on forums) on your soft, it will be always cheating and adding some crazy shit to it or removing important pixels.
I use Epson scanners also, but If you doing it RIGHT I mean RIGHT, I think the DSLR scan is unbeatable, both in color and resolution even if we take the drum scan method as well.
Yeah... it's a great shit load of hustle and struggle but if you are making 1 image in a week, I think it is worth it.
Cheers! - Franziskaner wheat beer.
Man what I can say…amazing video with all the infos I needed. Thanks a lot!
Thanks for the great info. Always entertaining and informative.
Color rendition is exactly why I gave up scanning on my own : when I saw the result my lab was able to get with a Fuji SP3000 (waaaaaaay better than the Noritsu), I couldn't come back.
Are you sure the Technician is not a major factor?
*YOU'RE TOTALLY WRONG, YOU SCREWED THIS UP SO MUCH!!*
There. With that out of the way, I can now proceed to stop the video after not hearing you say exactly what I expected about the method I like best and go back to eating glue while I stare at the sun
Muy interesante, gracias por compartir Nick !
You're the best, Nick. Love your videos.
theres a program called "Hugin" for panorama stitching. it rarely misses and even so, it has a manual control provision.
You actually could do it in a „single“ exposure with the dslr, so avoid stitching the images. Just zoom in so you have the whole negative in frame and take multiple exposures (the more images, the better, but use the same settings on each of em). Load all your image in a stack into photoshop, convert them to one smart object and select stack mode „average“ or „median“. Once this is done you can just pump up your image resolution without losing much quality.
So you're doing a HDR-type exposure with the DSLR? (multiple exposures, combining etc.)
I think you are referring to the “high resolution” mode of cameras that do “pixel level shifting” of the sensor, so that the pixel resolution (eg 2x2 RGGB Bayer sensor) overlaps and is averaged. I believe Tony Northrop made a video comparing this against “hand held” equivalent (i.e. subtle overlap / shake from hand held) and let PS do a statistical average overlay of the image to remove random sensor noise. That technique may work for landscape photography, but I do not think it applies here to macro photography where the lens is focused on the grain of the film to be “scanned”.
bwvids it would work perfectly. A *tiny* nudge to the holder is all that is needed.
Exposing at different stops, essentially bracketing would also really help with the range he’s so desperately needs to avoid “weird banding” artifacts.
Johnny McMillan multiple exposures with exactly the same settings. So just multi-image, not hdr.
Thank you for the excellent film printing review. Good to see!
Great video, as always!
Thank you my guy. I've been looking into getting a drum camera as my v700 scans are meh. And lastly I'll never use a digital camera to scan my negs because the cost and I'll probably never use it after
Guy: "This video is long"
Me: "Hold my x2 playback speed"
IDK if you'll see this, but I really appreciate people who are hypercompetant and ultra detail oriented. This video is an absolute masterclass.
WOW. This video is amazing Nick.
Cheers from LA.
This video is 36;02 long. It took me one hour to watch it, due to being interrupted every five minutes by my wife.
Send her to hell and get yourself a nice husband
so your average interruption was 3 minutes 20 seconds long. interesting.
Jeff Dahdah 😂😂😂😂😂😂😂
Women do the same thing as soon as your live game of choice begins on TV
I liked right after you almost threw up. Funny I also gagged at the same time... don't know why though :-/ ....... :)
I hope you weren't eating breakfast. Haha
Made a mess of my Bishop’s robes !
This video deserves more views! good info BTW I loved it.
Very nice presentation and discussion. Great food for thought.
Just the intro was worth the like, hahaha!
I would use Microsoft ICE to stitch the photos together. It works perfect nearly every time and much better than Photoshop and much faster + its free
Makes panoramas a breeze
awesome video; absolute beast in analyzing to get to the best result! thanks
Thank you, great information. After all not a huge surprise but good to have the facts covered.
Use a better light source other than a light table. And your dslr scans would be as fast and as simple as negafix.
you never talk about imacon scanners, the best after drum, maybe just as good, but cost???? epson scanning is a joke against them
Thanks for this! I am new to the scanning world. This was an incredibly helpful video to watch
You put an amazing amount of work into this project
Try scanning film with a microscope
I cannot remember using microscope with good color rendition ever.
@@terriplays1726 You need an electron scanning microscope with bayer color filter
@@EdwinNew If you have a scanning transmission electron microscope you can have an electron energy loss spectrometer (EELS) as a detector to see "colors"
Don't knock scanning with a DSLR! They do an incredible job.. at motivating you into buying an Epson v850.
Haha! Yeah that seems to be the case, for me anyways.
Oh my god man. The effort you go to with this channel is flabbergasting ! It must have taken you hours upon hours....the tech spec slides, the discussions with the labs, setting up the various methods of scanning. Not to mention how much it must of cost you to make it. It's made me tired just watching it! I look forward to your videos man. They're so entertaining but equally so informative. It's obvious you really know your onions; the comical delivery just makes it all the more engaging (though your sick story when travelling back from Europe is still the funniest RUclips I've ever seen!). Don't know if you know but you've been mentioned a few times on podcasts I listen to. It would be great for you to be a guest on one of them, one day (I seem to think I listened to one when you were a guest in fact but I forget which it was - I recall you telling the story of the "Glass behind a glass" title?). Your efforts also remind me again of why I send all my important stuff to a lab to deal with. All that fluid mounting and setting up SLRs on tripods is just too much fuss for my sanity. Re Silverfast and the "NegaFix"...do you know what version that feature comes with? Is it only with the more expensive line or does it come with the standard line (I bought my Epson V550 second hand so it not included). Keep up the great work mate, for the sake of all of us without the commitment to make such great videos. I'll try to contribute when I can, but sadly I am rather hard up. If/when I had more surplus dough, you'd be the first to get some. As for the haters and moaners...dont let them drive you down. RUclips is full of bell ends these days with too much time on their hands and not enough experience to have an opinion on anything, let alone the fine art process of analogue photography.
I love your videos! Thanks for diggin into all that options. And thank you for being not calm if something annoys you!!! :)
Good stuff and very much appreciated. Dropped a contribution as would love to see this content keep coming. Cheers fella from Bristol UK.
I just found you and i am in love. I am cracking up over here you are amazing!
Why thank you!
Wow, thank you for this amazing analysis!
Most of this was pretty much over my head but I was still entertained by it and I stuck with it all the way to the end.