Thanks. Very useful. There's variation in copies of the 200-600. I had two and was underwhelmed by both, forcing a step up to the Sony 400 2.8 and TCs.
Alan, Special thanks for the time & effort you put forth creating this very detailed video. Such a great comparison of the three lens. Given the $$ point of the Nikon Z 180-600mm and it coupled with the 1.4 TC, the images at the end of your video were remarkable in my view. Again, THANK YOU!
Thank you for the time and effort. You convinced me to order one for my Z7 II. I am stepping up from my D7500 + Tanron 18-400mm zoom. I can use the Tamron with the Z adapter, but I'm back to DX rsolution. Unfortunately, I'm going to double the weight and size to haul around.
Very well done, thanks. I really appreciate that you did your tests in well controlled conditions. I'm expecting to pick up my 180-600 this week. Planning to use it in those conditions where the 800mm is just too long.
Super helpful, thanks so much for taking the time to do this excellent comparison. I’m one of the many waiting anxiously for my 180-600 lens to arrive and watching this pacifies me for the time being.
This is the best test video I've seen on the 180-600 and 200-600 so far. When I got my z6 II two years ago I was considering getting the 200-500 and using my FTZ adapter. Then I found out there was a 200-600 (180-600 actually) on Nikon's roadmap. I placed my order the morning it was announced. The first orders that came to my camera store went out to NPS members. I hope mine will be in the next shipment. I can't wait to try it. I know where an Osprey best is that I want to try it at.
Oh my goodness that was an ultimate relaxed review I've saw with my morning coffee. Thank you for that. Also all what you say about the optics. The sony 200-600 mm sleep now in my Dry cabinet, I click him on only on my aps-c bodies. And for me the results are good enough for me. But, maybe there comming a point in time that the upgrade to the topper from Sony under my arm ,sleep out from the shop to my study room. BTW, My wife now noting of this action in the future.So that said,thank you for sharing the birds on the end in your video. Special the Nectar lovers ❤ Greetz from Thailand.
Great review! That doll is perfect - has a test pattern surrounded with feathers. Also I like that the image is shot in still air. Unfortunately the image files are altered in camera even in raw mode so the only way to truly test a lens in isolation is with an scientific camera, and the Nikon is actually sharper all things being equal, however the camera is of course part of the equation. VR actually blurs the image if switched on with both brands. (By the way Nikkor is not pronounced that way, the word is Japanese in origin with the high fast "ee". Neekor)
The ambition behind the video is great and for bird-shooters relevant, but as a trained scientist, uni teacher of AI software engineering and also having gone to a BSc level photography school, I see a couple essential flaws in the video. It's clear that Alan doesn't deeply understand the "Bayer paradigm" kept under a veil of ignorance by marketers and fluencers, thus becoming the Bayer conspiracy (p/fun intended). What Alan pronounced as "moray" and I would call "Moiré" (wmahray) is not caused by either cameras or lenses, but purely a digital artefact of "raw processing" (raw conversion). The Bayer paradigm is based on a colour-blind sensor that sees the entire humanly visible spectrum of light (more) and this is called "panchromatic" (pan=all, chromatic-colours) in old B&W film terms. In order to get colour images, a filter grid is laid over that sensor that precisely aligns with the photocells in the sensor and each filter in the grid reduces the light travelling to a photocell so only the light of one colour band passes. Essentially each individual cell now sees one colour band - like early emulsions of B&W film and papers - and this makes each corresponding data element in the raw file "monochromatic" (mono=single,one). Which means, when we think of colour vision in three spectral bands, that each data element misses exposure values in two spectral bands. While the data raw elements are positive, they are monochrome, and if you could depict the raw file as is, you would see an extremely ugly, noisy image. For the sake of the argument, I call that "Bayer noise" as it comes with the paradigm. But the paradigm doesn't end there, as it prescribes that some software shall make guesses about the missing colours and correct the image for that. As the raw image is a Bayer image or "Bayerised", we call the first (naive) step in mathematically precise and repeatable but wild-assed guessing conceptually "deBayerisation". The, now classical, approach to deBayerisation is to extrapolate the missing colours for each raw monochrome data element [x,y] by a weighted interpolation of the exposure values of direct and indirect neighbours of [x,y]. And this is where a digital artefact is created: "mosaicking". Moiré is a very recognisable artefact generated by such naive deBayerisation that reminds me of the Newton rings we might see in projecting (between glass mounted) slides in the film past. But there are other mosaicking artefacts like crinkly lines (my thoughts go out to Adobe Camera Raw (ACR) ), or cross-bleeding of colour along the edge between two differently coloured, bordering, blobs in your image (my thoughts go out to JPEGs of not too long ago). Every 14-bits (“deep” - gradation resolution) raw data element must become RGB: R must become RGB, G must become RGB, G must become RGB. ACR does that converting to a 16 bits-per-channel RGB pixel (48 per pixel - bpp) in ProPhoto colour space. For Lightroom to display this, the image needs to be compressed again to 8bpc or 24 bpp. We can had over this ProPhoto image from Lightroom to Photoshop and convert it to what it already is under the hood, or even double that to 32 bpc (96bpp). As your monitor/display probably has limited colour space and may fake 8 bpc from using 7 bpc, assessing what, say, a printed image will look like from that monitor/display is challenging. Note that some colour printers can render more than Adobe RGB colour space In the early years of digital photography - the photocells and the circuit with each that makes the cell plus circuit a photosite are all analogue - the low resolution provided camouflage for implied issues. Better raw processing would need more powerful computers that are big, heavy, power hungry, and expensive. A supercomputer of about 1990 like a Cray One - probably with its own building and support organisation - is as powerful as two NVIDIA GTX 1080ti graphics adapters, of now some years ago. Moore's Law.
Hardware help Somewhere in the history of implementing the Bayer paradigm, hardware help was thrown in to make raw processing easier: the Optical Low-Pass Filter (OLPF). This OLPF makes the wild-assed colour-guessing easier and the idea of such a thing was not new, as it already, conceptually, was applied by developers of the Scanning Tunnelling Electron Microscope (STEM) in the 1970s. As deBayerisation does not work at the edges of the raw image (i.e. transitively the sensor) here hardware help was added as well to remove the requirement for a separate "edge algorithm" in the deBayerisation: by adding rows and columns of photosites that have their data recorded in the raw file but are not shown to the use after raw processing. The Nikon Z 9 (and Z 8) add a lot more even to that and my assumption/hypothesis is that this is used for IBIS in software that becomes part of the raw processing pipeline and hence does not add latency of its own - but it only works for motion in the sensor plane. So with naive, brute force, simple algorithm, deBayerisation generating digital artefacts, with increasing resolution with the help of Moore's Law's consequences, we got served with "demosaicking" that was added to raw processing after deBayerisation: helix [1] up first and repair later. This again started as old school algorithmic "artificial" intelligence that dominates roughly the 1950s-2000 era. So when we talk about "Moray" in the context of discussing the resolution of a camera or lens, then my teacher pencil starts to draw an F spontaneously. As Alan in the video of him testing, or as illustration, shows a long Nikon lens on a Nikon Z 9, we now also know that this camera does not have an OLPF [2]. Yes, the OLPF made raw processing easier, but it has cons [3] that increasingly get in the way with increasing resolution. As the subconscious understanding of how to do the Bayer paradigm evolved to the assumption that the OLPF would always be there, we consequently have been served with a decade of bad raw processing software that was not adapted to its absence. Nikon first Eliminated the OLPF in the D800E version of the D800 that are otherwise identical ("ceteris paribus" in methodology jargon).
Observations As DxO Mark (a different legal entity than DxO) use a simplified deBayerisation raw processing tool that does not do demosaicking, this poses some validity questions within the de facto assumptions of the Bayer paradigm. However, when sharpness is considered in isolation based on their observations, I am inclined to set my doubt aside. The problem with lens and camera tests is that people frequently violate the "ceteris paribus" without being aware of doing so. Comparing a 20MP Nikon D500 (APS-C format) to a 24MP D or Z and saying that the APS-C is actually very good, violates ceteris paribus because the D500 has no OLPF and the 24MP full frame has one. The latter's sharpness is negatively impacted, and the effect of larger photosites on low light sensitivity is also impacted by the LPF in the full frame.
Bayer noise If we understand all this, then we still have to address "Bayer noise". In images from cameras without OLPF, I would call the presence of Bayer noise plain "failed deBayerisation" or "inadequate deBayerisation". What I call Bayer noise is called by marketers and fluencers "luminance noise" or "colour noise" but it merely follows from the Bayer paradigm and inadequate computer vision [4] AI. When we see noise in the blurry zones, the darker zones, and/or low contrast zones of our images, then this is Bayer noise and inadequate raw processing. You may not detect this in images of A3 size, but larger or at 100% or more on your display would easily see it. If we treat the Bayer paradigm as a social contract between camera manufacturer and software developers, then we have to attribute all this Bayer noise to software not being good enough - if we don;t then it takes 10 years for the Mudbricks to come up with something better and spend our contribution (investment) in developing the same old on new platforms, introducing new code streams that raise the L in P&L. We could blame Mr. Bayer, but attempts to make a different architecture sensor in the colour film paradigm have not been commercially viable, so far. I would shoot large format for retaining colour saturation at larger magnifications - with enlargements you enlarge the grains as well as the empty space between the grains and thus enlargement dilutes colour saturation [5]. But in digital we have no empty space. When we go larger than 100% the software invents new pixels or dots and we can get as much saturation as the inks in our printers can give (most monitor/displays do less than excellent colour printers). Removal of Bayer noise is just playing with different software tools. For years, in answer to the Mudbricks asking me if I would recommend their software to others, my answer is "no, not as long as I need third party software to do what I expect your product to do."
Relevance to non bird shooters and validity As I am reserved about the relevance of the test to others than bird photographers, I must add that the reference test target that Alan shows as his old subject in testing, allows for comparing centre to edge/corner sharpness of a lens. Lenses typically have better centre sharpness and away from the centre, sharpness becomes less. So as the corners are farthest away from the centre that is where sharpness is least. But the old target does not inform on actual resolution in a number. Photography relies on LinePairs per millimetre (LP/mm) for measuring sharpness - a linear unit (1 dimension) and for numbers to be meaningful they need to relate LP to mm sensor. As we have seen above, this is not the end, as raw processing is applied. What that does we don't know exactly and if it violates ceteris paribus: today's ACR has 1,195 "camera profiles" that process your raw image towards an image of RGB pixels in the "Adobe Standard style [6]. If you think that digital images "all look alike" then here's why. While the Mudbricks chose a flattish, neutral, rendition to process and open your raw files, Capture One (C1) chooses a vivid approach, but in all cases, we as users of the raw processing tools are at the helm of the destiny of our images. Not the camera. The people blabbering about not enough photons or colour science of a camera - sic - only help to camouflage the real issue in the Bayer paradigm. I'm satisfied with my camera and lenses, it does not have an OLPF, and I occasionally - when there is an occasion to do so - use 3rd party tools. In the Nikon domain, I cannot say that Nikon has done a much faster and better job in raw processing than the Mudbricks, though. They worked with the Mudbricks, explaining the responsibility, but that was it. For SOOC raw converted images, this may all be different. Different software, different powers at processor level, different developers.
A super detailed comparison video. Yes of course there is a point that each lens has to be processed through their respective camera sensors. So perhaps an issue with making total side by side comparisons. But you make a very clear analysis of how (optically) these lenses generally perform. I had never really taken on board the atmospheric factors that you so very clearly highlighted. Very well done. More videos of this type please!
Well done!!!! Thanks for the thorough review. In the TC comparison, it looked like the 800 was sharpest by a bit if you look at the circular wood piece. The 180-600 a bit front focused because it had the sharpest hands on the figure. Sony looked great. Three awesome options for photographers.
Dear Sir, words are falling short to appreciate this fantastic video. I gained a lot of knowledge about bird photography from this extremely informative video. Actually, nobody would have ever considered the factors that you mentioned in the introduction of this video but I realized how much these factors contribute while comparing resolution of lenses. Many many thanks for such a nice video.❤❤❤❤
Thank you for the video, it was an interesting and helpful to watch! Really liked that you took the time to do the 1.4TC comparison as well. Two items where improvements could be made: 1) It would have been nice to see the camera settings under each comparison image. 2) Take into consideration that many lenses(especially zooms) do no perform at their best when close to minimum focus distance. This can impact your results as much if not more than atmospheric distortions.
Great review thank you ! I'm on the verge of acquiring one, but still debating about the 400mm f/4.5, therefore waiting a bit longer for more field results and sharpness comparisons to make a choice. Very timely ! Loved your bird pics and videos at the end. Beautiful results and tack sharp
great detailed work, Sounds like a lab test, do you not plan to take the lenses out in the wild to shoot ? only indoors/backyard ? new perspective to look at lenses i didn't' realize...
Thanks for doing this comparison and posting...your results appear to be in alignment with some early reports I've seen. As an owner of the Z9 and Sony 200-600 plus both TC's and Megadap adapter, my problem is whether it's worth buying the 180-600 at all -- the Sony is really hard to beat in terms of sharpness and contrast. VR on the Nikon would be better of course, but I only use the zoom on a tripod so there's no advantage there.
Great test, thank you for all the effort you put into it! I also have the Z9+ 800mm pf & the A1+200-600mm AND using the Megadap etz21 to adapt the 200-600 to the z9 sometimes , and wondering if going to the 180-600 Nikkor is worth it. ?? Your test is great, maybe the best so far, but I think a test with the Sony lens adapted to the Z9 and at f6.3 (and f9 with the tc's) will be better to show any differences between the lenses. How do you find the AF speed of the Nikon 180-600 compare to the Sony? From my own usage I found that the Sony 200-600 is a bit faster than the Nikon 800mm (without tc's). again thank you for the test
Test #3 >>>> Look at the fine hairs off the ends of the feathers. On the lower center of the three frame on all three. On my 30" monitor. Center First Place Nikon 600, - Right side Second Place Nikon Z 180-600, - Left side Third Place Sony 200-600. Now that's real pixel peeping. LOL
I also own all three lenses as well as the Z9 and A1. I agree with your conclusions regardless of your flawed methodology. Perceived sharpness of images from a lens can only be accurately compared when rendered by the same sensor. Here you are comparing the Sony A1/200-600 against the Nikon Z9/180-600. This is why one finds drastically different DXO perceived sharpness scores when the same lens is tested on different sensors. Higher resolution sensors will produce higher perceived sharpness on the same lens.
As a Sony user, it is interesting the resolution of the 200-600, although concerning about the breathing at 600mm which looks like it is more like 550mm
about time I found some one who gets it regarding atmosphere and lens testing while i still think test charts is the way to go. but keep up the good work. my tests are same as you indoors consider using your phone conected by wifi asa phone shutter release to elimate camera shake.
It would be good to test sharpness at around 30-50 metres in a sports hall or large indoor space because this is more of a semi-macro test. The Nikon performs quite differently at infinity to close and focus breathing isn't a factor. Unfortunately the pixel size might be different causing resolution to appear different.
20:18, Awsome shot. Thank you for the review. I get confused with the term 100%. What exactly does it mean. Double the image size, match image pixels to screen pixels or something different? In this case, it may not make a lot of difference, but I would want to see a comparison with each image cropped to the same size which would require maybe an additional 10% crop on the Sony. The A1 has 10% more pixels, so it likely won't make a difference. A test that I think has value is the bare lense cropped to match the 1.4TC. Does the TC provide enough quality improvement to give up the additional light and focus performance. When I did that test on my 500PF, it did not.
Alan, thank you for this superbly crafted review and comparison of the new NIkkor Z180-600! As a sharpness "fanatic" myself, I very much appreciate the painstaking process that you went through to ensure a fair and consistent result! I use a Z8 for landscape, product, wildlife and macro photography, and indeed have the 180-600 on preorder, which I hope to get in the very near future. You have reassured me as to the excellent sharpness of this lens. One question if I may...I bought the Wimberly P-30 Plate for the foot of the 180-600, since all my tripods, monopods, etc., are Arca-Swiss style. Is this a good option for this lens? Thank you sir. And BTW, your images are stunning!
Thanks for the comments. I really can't answer your question however I think that plate will work as long as you can add a second screw to the plate. For a lens such as this, you want to be sure it's secure and won't twist. Two screws will make sure it stays solid w/o twisting on you.
@@daycreek Thanks for the reply, Alan! FYI, the Wimberley plate has a raised lip on one edge, so I think it would prevent turning/torquing of the lens.
You got to name that Kachina Doll .. 🙂 !! BTW any chance you can test it will a TC2.0 .. I know i know its not recommended by anyone.. but I found on a good deal. and i am curious..
Thanks for this. First observation is that on my screen (27" Photographers Ben-Q) at about 9:50 into your video, the Sony seems marginally sharper that the Nikon 180-600. At 11:20 I can certainly see the difference in sharpness on my screen. Having said that, All three, according to your example images, look very acceptable. I have the Z800 PF and the Z 100-400 + 1.4xTC and, as yet, am undecided on whether it is worth getting the 180-600 - I certainly can't carry all three around!!!
Great review and photographs.. I noticed some of your shots were at iso 25000! I have never gone that high before, how do you deal with all the noise created without losing image detail? Thank you..
Did you compensate for the difference in sensor resolution in any way? (Sony A1 = 50MP vs Nikon Z9 = 45MP) If not, then the real max focal length of the Sony is even shorter than it looks when comparing the size of the subject at the same magnification.
@@daycreek Thanks for replying. You are saying that both raw images aren't at 100%, even though the image viewer (in the top of the image) says so? If so, this means the Nikon image was upscaled (more) than the Sony, right?
I kind of expect a lens from any of the big manufacturers to be sharp at f/8, the only thing that really mattered to me was the sharpness at f/6.3... Other tests have shown, that it falls behind the Sony quite a bit, wide open. This is a very good test for what it is, it's just not particularly useful, I'd say. I'm not taking anyone's word for anything, it's not impossible to showcase sharpness at f/6.3. The comparison with the 800 f/6.3 was very interesting though, I've been thinking about getting that one.
@@livejames9374 Well, I switched from fuji to Nikon a couple of months ago, if Sony had a better wildlife camera, the sharpness of the 200-600 would've been the deciding factor. Anyway, I do believe it's important, and I don't care if you disagree. There's going to be situations, in which you'll have to shoot wide open, especially if you have to crop in afterwards, sharpness is essential.
@@cy9nvs The 180-600 has a real 600 focal length, the 200-600 has a shorter effective focal length. So, even if the Sony lense ist faster, you have to crop more and therefore you lose this sharpness advantage.
Zoomed in the sony the was clearly sharper, not sure how you couldn't see it. More contrast and resolution. Granted if you zoom out the differences are very small
hi alan great vid would be great if you could send a link to the raw files ..ive been delving into sharpness from lenses for years and the nikon z7 ii and z9 have no aa filter so that means the Nyquist frequency is high if you have dxo photolab try prime dx safe the file as a tiff and you will notice detail that it creates based on the Nyquist frequency i gurantee you you will see a big difference between the noram raw and the dxo deep prime dx
I just dont get the classical obsession with sharpness. at some point, all these are within a margin of error and once you export to JPEG and compress for online consumption, there is very little left. I much rather have low ISO for cleanear images. so I am not replacing my old 600mm f4 for this (as long as I can carry it :))
Hi. I was of that same opinion, until I saw a Photography Life test review of the NIkkor Z 70-200 f/2.8 S...They said that the MTF numbers on that lens (at around f/4-5.6) approach those of some primes in the respective focal lengths. (I don't have that lens, but thought that the test was very interesting in showing just how far zoom lens technology has come in recent years.)
I think it's more about the lower contrast produced by 180-600, which may appear as softness and probably also a bit of back focus which would have to be fine tuned. Notice that the feathers are more or less equally sharp. What's a bit too soft was the face of the figure which is closer to the lens.
![Nikon 180-600 Z vs Sigma 150-600 C [And Sharpness Testing]](http://i.ytimg.com/vi/Q7AUrTrtQMg/mqdefault.jpg)
One of the best comparisons I've seen!
Thanks. Very useful.
There's variation in copies of the 200-600. I had two and was underwhelmed by both, forcing a step up to the Sony 400 2.8 and TCs.
Thank you Alan, you took the time to make the test I was looking for, especially with the TC 1.4!
Now this is what i was looking for Great video 👏👏👏
Alan, Special thanks for the time & effort you put forth creating this very detailed video. Such a great comparison of the three lens. Given the $$ point of the Nikon Z 180-600mm and it coupled with the 1.4 TC, the images at the end of your video were remarkable in my view. Again, THANK YOU!
Thank you for the time and effort. You convinced me to order one for my Z7 II. I am stepping up from my D7500 + Tanron 18-400mm zoom. I can use the Tamron with the Z adapter, but I'm back to DX rsolution. Unfortunately, I'm going to double the weight and size to haul around.
Very well done, thanks. I really appreciate that you did your tests in well controlled conditions. I'm expecting to pick up my 180-600 this week. Planning to use it in those conditions where the 800mm is just too long.
First time watcher. Really excellent video. Thanks for sharing..
Very nice. Can't wait to test the 180-600 against my 400 f4.5
Super helpful, thanks so much for taking the time to do this excellent comparison. I’m one of the many waiting anxiously for my 180-600 lens to arrive and watching this pacifies me for the time being.
This is the best test video I've seen on the 180-600 and 200-600 so far. When I got my z6 II two years ago I was considering getting the 200-500 and using my FTZ adapter. Then I found out there was a 200-600 (180-600 actually) on Nikon's roadmap. I placed my order the morning it was announced. The first orders that came to my camera store went out to NPS members. I hope mine will be in the next shipment. I can't wait to try it. I know where an Osprey best is that I want to try it at.
Very nice review! All the 3 lenses seems to perform very well, no meaningful differences in terms of sharpness.
Oh my goodness that was an ultimate relaxed review I've saw with my morning coffee. Thank you for that.
Also all what you say about the optics. The sony 200-600 mm sleep now in my Dry cabinet, I click him on only on my aps-c bodies.
And for me the results are good enough for me. But, maybe there comming a point in time that the upgrade to the topper from Sony under my arm ,sleep out from the shop to my study room. BTW, My wife now noting of this action in the future.So that said,thank you for sharing the birds on the end in your video.
Special the Nectar lovers ❤
Greetz from Thailand.
Great Video Alan, thanks for sharing your results.
Great review! That doll is perfect - has a test pattern surrounded with feathers. Also I like that the image is shot in still air. Unfortunately the image files are altered in camera even in raw mode so the only way to truly test a lens in isolation is with an scientific camera, and the Nikon is actually sharper all things being equal, however the camera is of course part of the equation. VR actually blurs the image if switched on with both brands. (By the way Nikkor is not pronounced that way, the word is Japanese in origin with the high fast "ee". Neekor)
You are singing my story bro. Indeed outdoors is so damn challenging with the weather issues
This was an exceptional review! I, as a trained scientist, give you a lot of appreciation for the manner in which you did this comparison. Thank you!
The ambition behind the video is great and for bird-shooters relevant, but as a trained scientist, uni teacher of AI software engineering and also having gone to a BSc level photography school, I see a couple essential flaws in the video.
It's clear that Alan doesn't deeply understand the "Bayer paradigm" kept under a veil of ignorance by marketers and fluencers, thus becoming the Bayer conspiracy (p/fun intended).
What Alan pronounced as "moray" and I would call "Moiré" (wmahray) is not caused by either cameras or lenses, but purely a digital artefact of "raw processing" (raw conversion).
The Bayer paradigm is based on a colour-blind sensor that sees the entire humanly visible spectrum of light (more) and this is called "panchromatic" (pan=all, chromatic-colours) in old B&W film terms.
In order to get colour images, a filter grid is laid over that sensor that precisely aligns with the photocells in the sensor and each filter in the grid reduces the light travelling to a photocell so only the light of one colour band passes. Essentially each individual cell now sees one colour band - like early emulsions of B&W film and papers - and this makes each corresponding data element in the raw file "monochromatic" (mono=single,one). Which means, when we think of colour vision in three spectral bands, that each data element misses exposure values in two spectral bands.
While the data raw elements are positive, they are monochrome, and if you could depict the raw file as is, you would see an extremely ugly, noisy image. For the sake of the argument, I call that "Bayer noise" as it comes with the paradigm. But the paradigm doesn't end there, as it prescribes that some software shall make guesses about the missing colours and correct the image for that. As the raw image is a Bayer image or "Bayerised", we call the first (naive) step in mathematically precise and repeatable but wild-assed guessing conceptually "deBayerisation". The, now classical, approach to deBayerisation is to extrapolate the missing colours for each raw monochrome data element [x,y] by a weighted interpolation of the exposure values of direct and indirect neighbours of [x,y]. And this is where a digital artefact is created: "mosaicking". Moiré is a very recognisable artefact generated by such naive deBayerisation that reminds me of the Newton rings we might see in projecting (between glass mounted) slides in the film past. But there are other mosaicking artefacts like crinkly lines (my thoughts go out to Adobe Camera Raw (ACR) ), or cross-bleeding of colour along the edge between two differently coloured, bordering, blobs in your image (my thoughts go out to JPEGs of not too long ago).
Every 14-bits (“deep” - gradation resolution) raw data element must become RGB: R must become RGB, G must become RGB, G must become RGB. ACR does that converting to a 16 bits-per-channel RGB pixel (48 per pixel - bpp) in ProPhoto colour space. For Lightroom to display this, the image needs to be compressed again to 8bpc or 24 bpp. We can had over this ProPhoto image from Lightroom to Photoshop and convert it to what it already is under the hood, or even double that to 32 bpc (96bpp). As your monitor/display probably has limited colour space and may fake 8 bpc from using 7 bpc, assessing what, say, a printed image will look like from that monitor/display is challenging. Note that some colour printers can render more than Adobe RGB colour space
In the early years of digital photography - the photocells and the circuit with each that makes the cell plus circuit a photosite are all analogue - the low resolution provided camouflage for implied issues.
Better raw processing would need more powerful computers that are big, heavy, power hungry, and expensive. A supercomputer of about 1990 like a Cray One - probably with its own building and support organisation - is as powerful as two NVIDIA GTX 1080ti graphics adapters, of now some years ago. Moore's Law.
Hardware help
Somewhere in the history of implementing the Bayer paradigm, hardware help was thrown in to make raw processing easier: the Optical Low-Pass Filter (OLPF). This OLPF makes the wild-assed colour-guessing easier and the idea of such a thing was not new, as it already, conceptually, was applied by developers of the Scanning Tunnelling Electron Microscope (STEM) in the 1970s.
As deBayerisation does not work at the edges of the raw image (i.e. transitively the sensor) here hardware help was added as well to remove the requirement for a separate "edge algorithm" in the deBayerisation: by adding rows and columns of photosites that have their data recorded in the raw file but are not shown to the use after raw processing. The Nikon Z 9 (and Z 8) add a lot more even to that and my assumption/hypothesis is that this is used for IBIS in software that becomes part of the raw processing pipeline and hence does not add latency of its own - but it only works for motion in the sensor plane.
So with naive, brute force, simple algorithm, deBayerisation generating digital artefacts, with increasing resolution with the help of Moore's Law's consequences, we got served with "demosaicking" that was added to raw processing after deBayerisation: helix [1] up first and repair later. This again started as old school algorithmic "artificial" intelligence that dominates roughly the 1950s-2000 era.
So when we talk about "Moray" in the context of discussing the resolution of a camera or lens, then my teacher pencil starts to draw an F spontaneously.
As Alan in the video of him testing, or as illustration, shows a long Nikon lens on a Nikon Z 9, we now also know that this camera does not have an OLPF [2]. Yes, the OLPF made raw processing easier, but it has cons [3] that increasingly get in the way with increasing resolution.
As the subconscious understanding of how to do the Bayer paradigm evolved to the assumption that the OLPF would always be there, we consequently have been served with a decade of bad raw processing software that was not adapted to its absence.
Nikon first Eliminated the OLPF in the D800E version of the D800 that are otherwise identical ("ceteris paribus" in methodology jargon).
Observations
As DxO Mark (a different legal entity than DxO) use a simplified deBayerisation raw processing tool that does not do demosaicking, this poses some validity questions within the de facto assumptions of the Bayer paradigm. However, when sharpness is considered in isolation based on their observations, I am inclined to set my doubt aside.
The problem with lens and camera tests is that people frequently violate the "ceteris paribus" without being aware of doing so. Comparing a 20MP Nikon D500 (APS-C format) to a 24MP D or Z and saying that the APS-C is actually very good, violates ceteris paribus because the D500 has no OLPF and the 24MP full frame has one. The latter's sharpness is negatively impacted, and the effect of larger photosites on low light sensitivity is also impacted by the LPF in the full frame.
Bayer noise
If we understand all this, then we still have to address "Bayer noise". In images from cameras without OLPF, I would call the presence of Bayer noise plain "failed deBayerisation" or "inadequate deBayerisation".
What I call Bayer noise is called by marketers and fluencers "luminance noise" or "colour noise" but it merely follows from the Bayer paradigm and inadequate computer vision [4] AI. When we see noise in the blurry zones, the darker zones, and/or low contrast zones of our images, then this is Bayer noise and inadequate raw processing. You may not detect this in images of A3 size, but larger or at 100% or more on your display would easily see it.
If we treat the Bayer paradigm as a social contract between camera manufacturer and software developers, then we have to attribute all this Bayer noise to software not being good enough - if we don;t then it takes 10 years for the Mudbricks to come up with something better and spend our contribution (investment) in developing the same old on new platforms, introducing new code streams that raise the L in P&L.
We could blame Mr. Bayer, but attempts to make a different architecture sensor in the colour film paradigm have not been commercially viable, so far.
I would shoot large format for retaining colour saturation at larger magnifications - with enlargements you enlarge the grains as well as the empty space between the grains and thus enlargement dilutes colour saturation [5]. But in digital we have no empty space. When we go larger than 100% the software invents new pixels or dots and we can get as much saturation as the inks in our printers can give (most monitor/displays do less than excellent colour printers).
Removal of Bayer noise is just playing with different software tools. For years, in answer to the Mudbricks asking me if I would recommend their software to others, my answer is "no, not as long as I need third party software to do what I expect your product to do."
Relevance to non bird shooters and validity
As I am reserved about the relevance of the test to others than bird photographers, I must add that the reference test target that Alan shows as his old subject in testing, allows for comparing centre to edge/corner sharpness of a lens. Lenses typically have better centre sharpness and away from the centre, sharpness becomes less. So as the corners are farthest away from the centre that is where sharpness is least. But the old target does not inform on actual resolution in a number. Photography relies on LinePairs per millimetre (LP/mm) for measuring sharpness - a linear unit (1 dimension) and for numbers to be meaningful they need to relate LP to mm sensor. As we have seen above, this is not the end, as raw processing is applied. What that does we don't know exactly and if it violates ceteris paribus: today's ACR has 1,195 "camera profiles" that process your raw image towards an image of RGB pixels in the "Adobe Standard style [6].
If you think that digital images "all look alike" then here's why. While the Mudbricks chose a flattish, neutral, rendition to process and open your raw files, Capture One (C1) chooses a vivid approach, but in all cases, we as users of the raw processing tools are at the helm of the destiny of our images. Not the camera. The people blabbering about not enough photons or colour science of a camera - sic - only help to camouflage the real issue in the Bayer paradigm. I'm satisfied with my camera and lenses, it does not have an OLPF, and I occasionally - when there is an occasion to do so - use 3rd party tools. In the Nikon domain, I cannot say that Nikon has done a much faster and better job in raw processing than the Mudbricks, though. They worked with the Mudbricks, explaining the responsibility, but that was it. For SOOC raw converted images, this may all be different. Different software, different powers at processor level, different developers.
A super detailed comparison video. Yes of course there is a point that each lens has to be processed through their respective camera sensors. So perhaps an issue with making total side by side comparisons. But you make a very clear analysis of how (optically) these lenses generally perform. I had never really taken on board the atmospheric factors that you so very clearly highlighted. Very well done. More videos of this type please!
Well done!!!! Thanks for the thorough review. In the TC comparison, it looked like the 800 was sharpest by a bit if you look at the circular wood piece. The 180-600 a bit front focused because it had the sharpest hands on the figure. Sony looked great. Three awesome options for photographers.
Dear Sir, words are falling short to appreciate this fantastic video. I gained a lot of knowledge about bird photography from this extremely informative video. Actually, nobody would have ever considered the factors that you mentioned in the introduction of this video but I realized how much these factors contribute while comparing resolution of lenses. Many many thanks for such a nice video.❤❤❤❤
Thanks for this useful comparison of birding lenses. Lovely photos and footage at the end.
Thank you for the video, it was an interesting and helpful to watch! Really liked that you took the time to do the 1.4TC comparison as well.
Two items where improvements could be made:
1) It would have been nice to see the camera settings under each comparison image.
2) Take into consideration that many lenses(especially zooms) do no perform at their best when close to minimum focus distance. This can impact your results as much if not more than atmospheric distortions.
Great review thank you ! I'm on the verge of acquiring one, but still debating about the 400mm f/4.5, therefore waiting a bit longer for more field results and sharpness comparisons to make a choice. Very timely ! Loved your bird pics and videos at the end. Beautiful results and tack sharp
I'm in the same debate, possibly magical image rendition around 400 with the 4.5, then tc's to make up the rest to equal the 180-600?
Thanks for sharing Alan. Still patiently waiting for my copy to show up.
great detailed work, Sounds like a lab test, do you not plan to take the lenses out in the wild to shoot ? only indoors/backyard ? new perspective to look at lenses i didn't' realize...
Thanks for doing this comparison and posting...your results appear to be in alignment with some early reports I've seen. As an owner of the Z9 and Sony 200-600 plus both TC's and Megadap adapter, my problem is whether it's worth buying the 180-600 at all -- the Sony is really hard to beat in terms of sharpness and contrast. VR on the Nikon would be better of course, but I only use the zoom on a tripod so there's no advantage there.
Great test, thank you for all the effort you put into it! I also have the Z9+ 800mm pf & the A1+200-600mm AND using the Megadap etz21 to adapt the 200-600 to the z9 sometimes , and wondering if going to the 180-600 Nikkor is worth it. ?? Your test is great, maybe the best so far, but I think a test with the Sony lens adapted to the Z9 and at f6.3 (and f9 with the tc's) will be better to show any differences between the lenses. How do you find the AF speed of the Nikon 180-600 compare to the Sony? From my own usage I found that the Sony 200-600 is a bit faster than the Nikon 800mm (without tc's). again thank you for the test
Test #3 >>>> Look at the fine hairs off the ends of the feathers. On the lower center of the three frame on all three. On my 30" monitor. Center First Place Nikon 600, - Right side Second Place Nikon Z 180-600, - Left side Third Place Sony 200-600. Now that's real pixel peeping. LOL
I also own all three lenses as well as the Z9 and A1. I agree with your conclusions regardless of your flawed methodology. Perceived sharpness of images from a lens can only be accurately compared when rendered by the same sensor. Here you are comparing the Sony A1/200-600 against the Nikon Z9/180-600. This is why one finds drastically different DXO perceived sharpness scores when the same lens is tested on different sensors. Higher resolution sensors will produce higher perceived sharpness on the same lens.
Thank goodness the youtube expert is here
Thanks for sharing your results
As a Sony user, it is interesting the resolution of the 200-600, although concerning about the breathing at 600mm which looks like it is more like 550mm
At less than 30 feet it is more like 435 feet
about time I found some one who gets it regarding atmosphere and lens testing while i still think test charts is the way to go. but keep up the good work. my tests are same as you indoors consider using your phone conected by wifi asa phone shutter release to elimate camera shake.
Wow, really well done. Congrats
Thanks for the vid. Now I need to see some autofocus speed examples for racing!
Very informative video! Thanks for posting. Looks like those high iso images clean up pretty nicely?
It would be good to test sharpness at around 30-50 metres in a sports hall or large indoor space because this is more of a semi-macro test. The Nikon performs quite differently at infinity to close and focus breathing isn't a factor. Unfortunately the pixel size might be different causing resolution to appear different.
Thank you for sharing this, I had to subscribe! 👏👍😀
Excellent video ! Thank you !
20:18, Awsome shot. Thank you for the review. I get confused with the term 100%. What exactly does it mean. Double the image size, match image pixels to screen pixels or something different? In this case, it may not make a lot of difference, but I would want to see a comparison with each image cropped to the same size which would require maybe an additional 10% crop on the Sony. The A1 has 10% more pixels, so it likely won't make a difference. A test that I think has value is the bare lense cropped to match the 1.4TC. Does the TC provide enough quality improvement to give up the additional light and focus performance. When I did that test on my 500PF, it did not.
Alan, thank you for this superbly crafted review and comparison of the new NIkkor Z180-600! As a sharpness "fanatic" myself, I very much appreciate the painstaking process that you went through to ensure a fair and consistent result! I use a Z8 for landscape, product, wildlife and macro photography, and indeed have the 180-600 on preorder, which I hope to get in the very near future. You have reassured me as to the excellent sharpness of this lens. One question if I may...I bought the Wimberly P-30 Plate for the foot of the 180-600, since all my tripods, monopods, etc., are Arca-Swiss style. Is this a good option for this lens? Thank you sir. And BTW, your images are stunning!
Thanks for the comments. I really can't answer your question however I think that plate will work as long as you can add a second screw to the plate. For a lens such as this, you want to be sure it's secure and won't twist. Two screws will make sure it stays solid w/o twisting on you.
@@daycreek Thanks for the reply, Alan! FYI, the Wimberley plate has a raised lip on one edge, so I think it would prevent turning/torquing of the lens.
You got to name that Kachina Doll .. 🙂 !! BTW any chance you can test it will a TC2.0 .. I know i know its not recommended by anyone.. but I found on a good deal. and i am curious..
Thanks for this. First observation is that on my screen (27" Photographers Ben-Q) at about 9:50 into your video, the Sony seems marginally sharper that the Nikon 180-600. At 11:20 I can certainly see the difference in sharpness on my screen. Having said that, All three, according to your example images, look very acceptable. I have the Z800 PF and the Z 100-400 + 1.4xTC and, as yet, am undecided on whether it is worth getting the 180-600 - I certainly can't carry all three around!!!
Great video
Great review and photographs.. I noticed some of your shots were at iso 25000! I have never gone that high before, how do you deal with all the noise created without losing image detail? Thank you..
Thanks for a very interesting test. Did you do any bokeh tests ?
I am actually 'gob smacked' how good these images look with 1.4TC at what are 'high' IOS values. Many thanks
Thank you for this comparison. Done Well!
Very nice review. Thank you!
Excellent video. I noticed the Sony seemed slightly darker on the TC comparison...might be less light transmission the the Nikons ?
The Sony appears to have better contrast -- that's a weak spot for the Nikon.
Did you compensate for the difference in sensor resolution in any way? (Sony A1 = 50MP vs Nikon Z9 = 45MP)
If not, then the real max focal length of the Sony is even shorter than it looks when comparing the size of the subject at the same magnification.
bingo! :)
Yes, I did. Both were cropped to 8 x 12 format at 800 px per inch. This upscaled the images (slightly) and created two images of equal size.
@@daycreek Thanks for replying. You are saying that both raw images aren't at 100%, even though the image viewer (in the top of the image) says so? If so, this means the Nikon image was upscaled (more) than the Sony, right?
Slightly. Yes.
Still would have loved to see it at 600@6.3 for the Nikon,
I kind of expect a lens from any of the big manufacturers to be sharp at f/8, the only thing that really mattered to me was the sharpness at f/6.3... Other tests have shown, that it falls behind the Sony quite a bit, wide open. This is a very good test for what it is, it's just not particularly useful, I'd say.
I'm not taking anyone's word for anything, it's not impossible to showcase sharpness at f/6.3. The comparison with the 800 f/6.3 was very interesting though, I've been thinking about getting that one.
They are both plenty sharp enough. Who cares. No Nikon shooter will choose to adapt the 200-600 over the 180-600
@@livejames9374 Well, I switched from fuji to Nikon a couple of months ago, if Sony had a better wildlife camera, the sharpness of the 200-600 would've been the deciding factor.
Anyway, I do believe it's important, and I don't care if you disagree. There's going to be situations, in which you'll have to shoot wide open, especially if you have to crop in afterwards, sharpness is essential.
@@cy9nvsyou didn’t want to go for the A1? It’s as good as it gets for wildlife
@@livejames9374 costs like 3000€ more than the z8 where I live
@@cy9nvs
The 180-600 has a real 600 focal length, the 200-600 has a shorter effective focal length. So, even if the Sony lense ist faster, you have to crop more and therefore you lose this sharpness advantage.
Thank you!
Could you please advise me on your Nikon lens foot mount. I’m having a problem with the lens wanting to spin while on a ARCA Swiss plate.
I'm not quite following your question. Can't you tighten the ring to keep it from turning? I use a video ball head for both video and stills.
Zoomed in the sony the was clearly sharper, not sure how you couldn't see it. More contrast and resolution. Granted if you zoom out the differences are very small
Great comparison! Could you do another one comparing the bokeh quality?
hi alan great vid would be great if you could send a link to the raw files ..ive been delving into sharpness from lenses for years and the nikon z7 ii and z9 have no aa filter so that means the Nyquist frequency is high if you have dxo photolab try prime dx safe the file as a tiff and you will notice detail that it creates based on the Nyquist frequency i gurantee you you will see a big difference between the noram raw and the dxo deep prime dx
hey, just wondering. how tall are you? the camera looks tiny in your hands lol
I just dont get the classical obsession with sharpness. at some point, all these are within a margin of error and once you export to JPEG and compress for online consumption, there is very little left. I much rather have low ISO for cleanear images. so I am not replacing my old 600mm f4 for this (as long as I can carry it :))
No zoom lens is ever going to be as good as a good prime lens.
Hi. I was of that same opinion, until I saw a Photography Life test review of the NIkkor Z 70-200 f/2.8 S...They said that the MTF numbers on that lens (at around f/4-5.6) approach those of some primes in the respective focal lengths. (I don't have that lens, but thought that the test was very interesting in showing just how far zoom lens technology has come in recent years.)
I have the 70-200 f2.8, and it's very good.@@stevetqp9152
I think you should have magnified the images a few times more. Even with 4k it is hard to understand what you are talking about.
Very nice work and great review. Thanks for posting.
All 3 are great lenses - The 180-600 is a little softer than the other 2 but has great VR and fast AF ... 🦘
The 160-800 has no focus breathing, unlike the Sony.
Not on my 2-30" monitors. All 3 are great lenses. N,N,S
I think it's more about the lower contrast produced by 180-600, which may appear as softness and probably also a bit of back focus which would have to be fine tuned. Notice that the feathers are more or less equally sharp. What's a bit too soft was the face of the figure which is closer to the lens.