I owned a $400,000 Kodak Pro Photo CD film scanning system from 1993 to 2007 which used true RGB linear scanning arrays and Schneider apochromatic lenses . We could scan 35mm frames up to 4k x 6k resolution (24MP) at 16bit Lab colour but we did this rarely as scans of our studio calibration photos (taken on EVERY film stock) showed that only Kodachrome 25 ASA (ISO now) transparencies shot in ideal studio lit conditions with the very best lenses and apertures could provide enough film detail to justify this, 24MP resolution. We did over two million scans for archive producers and movie companies (all the stills for Star Wars etc.) and nobody ever wanted more that the 4k x 6k Pro resolution. It was the 16bit tonal range and the hundreds of custom negative and positive film terms developed by Kodak and us that made the scanning quality the best in the world! David Myers, Digital Masters Australasia.
Very impressive. So I think you are saying that anything over 24MP is unnecessary. I've compared my D600 (24MP) images to my D850 (45MP) images using a Tamron 90mm f2.8 Macro 1:1 lens on each and I really can't see a difference but I don't have a trained eye.
Great comment. Did you work with medium format negatives and scans though? I think that perceived resolution and detail would be a lot higher than 35mm hence the purpose of Darryls video. Or do you think the 24mp limit still holds true for MF 645? Personally, in my experience 24 MP more than handles 35mm negative scans and adequately handles 645. For 6x6 up to 6x9 I think it is worth it to go up to 47 megapixels. For large format I think you should be capturing those negatives with a digital medium format sensor of at least 50 MP as realistically the amount of resolution is absurd and 100 MP images are required to fully capture an 8x10 negative and match it's print dimension capability. But at that point it's not practical to DSLR scan a large format negative because it will be very hard to keep the negative flat and thus you are better off using a lab scanner. That way they can keep it perfectly flat. That or drum scanning.
Also, this shows insane details at 709Mpx. I guess film’s limit is pretty high (especially when you go beyond medium format) th-cam.com/video/sqN7n9bXgtU/w-d-xo.html
Adding the timestamp at the beginning of the video is such a good-natured generous gesture (losing watch time-an important ad revenue metric), that it deserves this user engagement datapoint post and a thumbs up. Wish all content creators were this awesome.
Paused at 2:44. There are conflicting forces. Let's start with the 35mm negatives and their resolution. In the film days, resolution was expressed in a linear unit because that directly relates with human perception of detail resolution: LinePairs per millimeter LP/mm. A pair being a black and a white line. Two times the LP/mm means we see that as twice as sharp (ceteris paribus, assuming parts in both comparison tests can reproduce both). You might argue that a 36mm x 24mm negative needs at least 7,200 LP * 4,800 LP and this gets us to an area size of 34.56 million. Why that number? Because 100 LP/mm was attainable with good film and professional grade lenses. Nikon glass would get better in the center and a bit less towards the edges when stopped down 2 to 3 f-stops from fully opened. With Leica glass you might see some higher values in the center of the frame and a bit more fall-off from that towards edges/corners. So a 36MP digital camera would equate that resolution? Well, if the black and white lines in the subject precisely align with the photosites in the sensor. But there is a problem with a digital camera. You see, film has RGB at all coordinates because the color is in three layers on top of each other with subtractive filtering between them. But a digital camera has not. The sensor is colorblind and in order to get color in the raw file, there is a filter grid over the sensor that has repeating filter patterns of 4 photosites that thus build an R.G.G.B quartet next to each other. The problem then is to convert this into RGB pixels and this is done in raw processing. This happens in camera for JPEG and MPEG shots, and in your raw processing software for your raw shots. Your camera comparisons in Lightroom tell more about Lightroom than about the camera. The 14 bits gradation resolution you had at the R.G.G.B photosite data in your raw file have been degraded to less than 9 in raw processing. That raw processing is a mathematically precise and repeatable guessing of missing color for the R.G.G.B quartets in the raw file so as to change R to RGB, G to RGB and B to RGB. This creates artifacts like Moiré if we recognize it and "noise" if we do not recognize it. Application of basic computer vision type AI can help a bit, but without "deep knowledge" in the raw processing software about what is in the photo, raw processing is not helped a lot. Lower resolution (than 36MP) cameras make raw processing easier with a so-called Anti-Aliasing filter (AA) or low-pass filter. These filters disperse the light for the R cell to its surrounding neighbors. Hence I call it the fuzzy filter. It helps raw processing and the idea already was applied in the 1970s to aid image processing out of data from scanning tunneling electron microscopes. Or, the presence of a fuzzy filter must be part of the reasoning. We see that higher than 35MP resolution cameras generally have done away with the fuzzy filter - making lenses sharper and depth of field shallower. The problem with digital reproduction of film is the grain in the film. With increasing resolution, this generates a raw file that has more detail of the grain and this is not likely understood by your raw processing software. So we have to find an optimum here between these conflicting forces. But my negative is black and white - unless you shoot a Leica Monochrom that is even more expensive for not placing the R.G.G.B filter layer and it still may have the fuzzy filter by the way because Leica's old lens designs are not well adapted to sensors, you always have an R.G.G.B raw file that needs raw processing. A scanner can run a multipass scan and build an RGB.RGB.RGB.RGB file that is better than raw and does not need raw processing. If it can create more than 8 bits per color channel TIFF, you basically outperform the raw file. So 16 bit TIFF is way more better than 14 bits raw when at the digital level there is a factor of only 4 between them. Some people would say that 24MP is enough. Well it might be the optimum, bottom line, after these conflicting forces got sorted out. The problem with digital is not only that we look at the results of raw processing (where we lost a lot of quality) but also look at digitally upscaled or even upsampled representations of our images, when we depict them larger than 100%. This gives an illusion of image quality. Even if we depict our images at smaller scales there is some digital processing going on, to the images. Like anti-aliasing for the display, depending on your software. Some defunct guy would call your digital photo on your computer "fake" and here he would have a point. I have tested Topaz's Gigapixel AI app last year as a way to get control over what is sent to the printer. Its ability to upsample and in that make guesses about what detail to add in blowing up a raw file, was incredible. A year later and Adobe have improved their Enhance Super Resolution features by a bit. The magic of a 12MP Nikon D700 is in these aspects: for human perception, two times as sharp as 12MP is 48MP and all wow conversations about in between values is extremely naive. It explains people's remarks that the gain from 12 to 24 is not as great as they had expected - it's only 40%. The magic is also in its discrete analog to digital circuit (i.e. not bundled with, stacked on, the sensor) and it is in the relation between Nikon F-glass's qualities and the fuzzy filter. To circumvent raw processing, we could mimic the multipass of scanners by shifting the original a tiny bit and shoot again, later match/stack the layers in Photoshop. But our raw images already have had raw processing at that moment and damage was done. What about sensor shift - cameras have IBIS so that should be easy, right? Well, the wild assed guessing of RGB colors in raw processing gets stuck in the edges of the image where there are no neighboring cell values to make guess from. The difficult solution is to have two algorithms in raw processing of which one is specific to the edge problem. The other solution is to have a sensor with more columns and rows in the edges and record the data thereof but never allow these columns and rows in the displayed image - they only serve as aid in RGB guessing. Well, throw in a couple more, and these rows and columns can also be used in IBIS without moving the sensor. If there are enough rows and columns you might only move the sensor, say with 5 photosite units, when you detect 5 units shift at image level. Or, I don't expect the IBIS systems to be able to make single photosite unit steps that we would need for RGB shots of 14.14.14 bits natively. Or, if you make a living in scanning negatives and want to speed up - the promise of the digital camera - then I would look at a Pentax K-something that can do this sensor shift actually. I would not use it anywhere else, by the way. All this still leaves us with the film grain problem. With increasing resolution the raw file has more grain detail and the software may have difficulties abstracting that grain noise away. When we stick with the regular Bayer-filter filtered sensor and its raw files, a comparative test is required indeed. And it will be valid until somebody develops a raw processing program for film scans.
This was one of the most well put comments about film scanning that I've ever witnessed. I've read through it thrice and my man has it been a journey. I'd love to learn more from you on this topic. Thanks!
Two commenters, thank you for the compliments. The summary is that we think we know what we are looking at when we compare shots in Adobe Camera Raw (ACR, the raw processor of Lightroom Classic as well as Photoshop), but we see ACR inn the first place. IN another YT channel I discussed if Topaz Gigapixel AI could deal with film grain and the host of that channel says he tested it and the short answer is, no. That's a pity, but I still can blow up my digital images to very large sizes with it, just not film. Note that my write-up applies to "digital photography" in general and that scanning or reproducing film images into digital merely is a use case in there. Finally, I'm obsessed with the truth and removing veils of ignorance, not "sharpness" as a goal in itself, or "beautiful bokeh" and I'm old-school about the decisive moment. But add that face/eye detect AF help me shoot portraits with a different interaction than I would in the film days. And I can obsess about that. To paraphrase a photographer that is a lot more famous than I am, people obsessing about x, y, z probably have less in the artistic department to talk about or show off. Let's go back to that. Take my write up as a way to put your mind at ease. Shoot, review, shoot, review, consciously, 10,000 times - to train your brain. Worry about all these details here a bit later.
@@jpdj2715 will you suggest a better way to digitize our film is by scanning ( using epson, nikon coolscan, frontier, noritus etc.. ) rather than a dslr (A scanner can run a multipass scan and build an RGB.RGB.RGB.RGB file that is better than raw and does not need raw processing)?
@@joshmcdzz6925 - that's what I wrote. The question to me is unanswered how well the scanner can do the RGB.RGB.RGB.RGB where a Bayer sensor would do R.G.B.G instead, that in raw processing must be converted into the RGB.RGB.RGB.RGB. The scanner depends on software to drive it and pull the scan data out, next convert them into an image file. The R.G.B.G is at best 14 bits each and I would expect the RGB in RGB.RGB.RGB.RGB to be 16/16/16 for each coordinate - 48 native bits per coordinate with no guessing of missing colours. But nobody is transparent in this digital photography market. I have the CoolScan and the Epson flatbed, but bottom line will grab my Nikon Z 7ii with 105/2.8S macro, plus raw process these images. IIRC the Nikon D850 had a profile for reversing shots of negatives. "Bayer" keeps my processing workflow standard, and my files and their filing too. Lazy? Probably. Yes, it's a lot faster.
Back when digital cameras we're barely more than a glimmer in kodak's eye, I read an article in one of the photography magazines that said that 24 megapixels would be enough to capture all the data that a 35 mm slide contains. This was all based on math, of course, because no one had ever seen in a 24 megapixel sensor at the time.
@@johnwinter6061 That is who I have decided to treat my 42 megapixel Sony a7r2 as if it were a medium format camera, and get a Sony a6700 to use as my daily driver.
Really interesting to see these comparisons, thank you. I do disagree on the grain comparison though - I felt the one on the left (50MP) was more pleasing and the grain to me looks smoother and more "shaped" than the garin on the right photo which looked more pixellated to me (although I'm looking at it from a compressed video). Guess it all boils down to testing for yourself and seeing what your own preference is.
I agree with you. Images taken with 50 MP camera are sharper and you can see more detail of the grain of the film. I'm using my 24 MP Nikon D610 for scanning of the film but I would prefer camera with more MP and no anti-aliasing filter like Nikon D810, Nikon D850 or similar.
@@MrCROBosanceros - I'm just getting started scanning with my D850 but would like to buy a D600 or D610 (24MP) as a dedicated copy camera. I have a D5500 (24MP) that I will try but it's a cropped sensor so I don't expect it to be as good as Darryl's 24MP scans.
For 35mm, I’ve been really happy to use 24 megapixels. I have an older Nikon D600 that I use exclusively for my 35mm scans so I don’t waste shutter count on my newer cameras and it works great for that purpose. I’m glad to learn that 24 works well for medium format. I have quite a few medium format negatives to convert. Great video. Looking forward to your video on the Kaiser system
I did the same thing. I own a D850 but bought a D600 just for film negative "scanning". I'm extremely pleased with the results for both my 35mm and 120 (6x7) negatives. BTW, I've just started taking two images of each 6x7 negative to increase the resolution. I haven't compared the single image to double image yet but I'm hoping I see a marked improvement to make it worth the extra step of creating a "pano" image of the 6x7 in Lightroom.
Hi Hashem, hope you and the family are well over in OZ. Starting to get back in to my film photography now, been a busy few months shooting videos for clients, will be nice to slow down and shoot film again. How is the Leica M-A going? Not sure I could use a Leica with no light meter 😜😜
@@DarrylCarey hey thanks man, same to you. That's good to hear! I love it. Had gotten used to no meter from years using the M4 :) I'll DM you over on instagram soon
Thanks a lot! Very helpful, as I am looking for a digital camera as only an addition to my analogue ones. I understood that 20 to 24 will be more than enough for my 35 to 6x6 negatives.
I’ve always had this question, ever since the lab I use rejected 2 photos I wanted to blow up to 36” x 36” and another at 60” x 60”, respectively. So I’ve been testing a small inexpensive digital camera with a resolution of 48mp. When the high mpx sensor is used, you can blow up the picture to larger sizes with even greater sharpness and clarity. I think 24mp and even 12mp for up to an 8” x 10” print is fine. If you want larger prints like poster size and so forth, the increased resolution of 48mp, 50mp and so forth really pays off. It’s really all about what size of print you want to end up with.
Exactly this is my observation as the manager of a pro lab. I noticed that this pixel (with a digital appearance) works very well when the file is enlarged and then printed in large format. It's not marginal to scan at 50 MP for large prints, it creates the ideal photo grain blur (after pixel extrapolation) for 44 or 60 inch prints.
I personally saw an image from a 6mp Nikon D70 blown up to 30x40” by a lab in Louisiana. It’s was flawless. Sometimes it’s also about the skill of who is doing the enlarging. I’m sure they weren’t using the same software everyone else was. But damn it was amazing. And 6mp!!!!
This is an amazing video! Would be interesting to test how big of a difference there’s between a (24Mp in my case) full frame and an (24Mp) apsc cameras. Great work nonetheless, as always!
I found it to be better with an apsc sensor and a full frame lens, only because you can minimize lens flaws. Also handy because you can manually use lightroom's lens correction that has been used in the film camera because there is no need for DSLR's lens correction. Image quality clearly depends on the camera, for example d750 was slightly better than d5200 but d7500 was about the same as d750 all tested with Nikon's 60mm, 85mm PC, 105mm, 70-180mm and 200mm macro lenses. Lens makes a very minimal difference, only 200mm masterpiece made a little more visible difference in 6x6, but focusing and aligning with film surface was so hard and time consuming that it wasn't worth the effort.
@@karajohn1234 - This sounds encouraging. I'm finalizing my setup today and would like to dedicate my old D5500 as my copy camera with a Tamron 90mm Ti (full frame) Macro lens. I'll be comparing the scans to my D850 scans and hope I can't tell a difference because I'd love to scan with just the D5500.
It will make no difference, apsc will extract all the same level of detail and since scans are at base ISO and shot raw the files have the same of flexibility in post for color adjustment. What actually matters is the macro lens you use for the scans. Some macro lenses are much better and sharper/resolve more detail than others.
Hmm, nice experiment - I have the same setup and have run many of my own. I think you will find that higher megapixels will excel once you print (if that is your target output). A 6x7 negative, for example, can achieve 1:1 optical resolution at 300dpi all the way up to 30”x45” which is equivalent to the 1:1 output of a 63 megapixel camera sensor. A 24 megapixel sensor is only capable of rendering a 16x24 print at the same (optimal) resolution, so capturing a 6x7 negative with a 24 megapixel sensor will limit your output options. There are some great resources for this online. I keep a chart at my desk. Enjoyed the video! You have a new subscriber .
I agree with groovejunky2549. Targeting the printing output dimensions is the key here. Higher resolution / megapixel = larger printing size possibilities.
Dynamic range is a big factor, too. Digital cameras are getting better, but I'd be curious to see a dynamic range comparison with top end flatbed and drum scanners specifically for dynamic range
@@DarrylCarey Doing good thanks! I do both! For 120 6x6 I take two pictures and stitch them together using lightroom classic in order to get more pixles our of a frame. 6x4.5 is just about 3x2 aspect ratio so I get a nice 18 megapixle file!
@@SinaFarhat Yea that's such a great point. Even people who want 40MP or 50MP scans for 6x6 or 6x9 scans can still use lower resolution sensors and stitch the image. 18 or 24MP is more than enough for home scanning.
More is less! This video confirmed my thoughts about high resolution cameras and scanners used for film scanning. Years ago I began scanning my 120 black and white negs with a Nikon 8000 dedicated film scanner - and I hated the results !!! The grain structure was overly sharp and drew way too much attention. Then during the pandemic I rescanned them all again with a Canon 5d Mark 4 (30 megapixals) and loved the results - slightly softer and way more natural, and they print beautifully :)
The Nikon Coolscan scanners are amazing scanners but starting to their age now …. Still going for crazy prices on EBay. I could wrong but I think the Nikon Coolscan scanners used a CCD sensor, which does not have the dynamic of a CMOS
Love your videos Darryl! I remember asking Ming Thein a few years ago the resolution equivalent to film, and he answered that he estimated it between 10 and 12 megapixels. That is consistent with your findings that 12Mp give very nice results for 35mm negatives.
A high resolution 35mm negative, properly exposed and developed, contains around 12~16 MP of data. A 24 MP digital camera is more than adequate to extract all the information recorded on the negative. The situation is different if one is digitizing MF negs. Here a different technique is required.
This is what I thought before but it is simply not true. 24MP resolves much more detail than a 12 and 16 MP sensors. I also tested 36 MP and it resolved even more detail than the 24MP sensor.
@@ShutterKnack I have been experimenting with digital camera ' scanning ' for a few years. In most cases, I find that even a 20mp MFT sensor produces optimum results. The only time I one needs higher pixel count is when digitising MF negatives or high resolution 35mm copy film, like ADOX CMSII. Here only a Flextight or high res drum scanner can resolve the details this film is capable of recording. More mega pixel makes for bigger files, but if the original neg does not carry enough information, inflated files are the result. It is good to experiment with though.
Very useful info Darryl, thanks! I would add that Nikon shooters can purchase the Nikon ES-1 or ES-2 to copy 35mm slides or film strips. If they already own a 60mm macro, it is a simple, straightforward path to digitize film. D850 owners really have it made as the camera includes a special mode to invert color or monochrome films.
To me a far more interesting comparison is: How does a wet darkroom PRINT of say, 16x20, from a medium format negative compare with a 16x20 PRINT produced from a scanned medium format negative?
Man I’m sooo glad I just found this video, I shoot with an X-Pro3 (26MP) as well as an M6 and a Hassy 501cm and just ordered the 80mm fuji macro to “scan” my film negatives, thank you for this Awesome experiment.
I just started scanning with DSLR and tried this experiment scanning 120 negatives with my Nikon D300s vs my Nikon D810 😮 blown away by the level of detail i get with the D810! such a difference and I never would have thought!! thank you! the D300s is super soft less detail and contrast. it matters
So all of this has been scientifically dealt with. It’s all depending on the film speed and grain size but generally speaking: You need about 8 to 12 Megapixels to resolve the detail on a frame of 35mm film. It takes up to 30 Megapixels to resolve the finest grain color negative and black and white film on the grain level. Velvia 50 in 35mm is somewhere in the high 50s to 60s. You can figure how it’d work for medium format. The general thing is that you don’t really need more than 8-20ish megapixels for most 35mm work. The only thing a higher resolution will afford you, is the ability to print larger or to zoom in further at which point you pictures will show grain pretty quickly. Now Darryl. You will not get any more appearance of grain on a 35mm frame of TMax p3200 when you scan it higher than 24 megapixel and only zoom in so little or even at all. It’s true than grain can be more defined when scanning at a higher resolution. But we’re talking finer grain film like Portra and a resolution jump from like 6 to 13 megapixel. With a grainy film like p3200 you’ll have crossed that threshold in the low teens. The reason the 24 megapixel scan appears to be less grainy is because it has some slight motion blur from the scanning! You can clearly see how it’s smeared a little horizontally! You can scan p3200 at 100 megapixel and it will not magically appear to be grainier just because you’ve zoomed in a little. That’s not how resolution works. Also it’s completely fine to scan 120 with a 12 megapixel camera, you’ll still get more detailed results and fine tonality than with a 35mm frame. Sure you’re not utilizing all the resolution the film has to offer. But if you shoot 120 for the resolution you’d probably also have a scanning method to benefit from that. And don’t forget CMOS color sensors (with a Bayer filter) only actually deliver around 70% of the resolution they say they have. I want to encourage anyone nerdy enough and interested in how resolution actually works to watch Steve Yedlin’s Resolution Demo on his website. It’s for cinema but the same principles mostly apply to photography and print. Especially part two of the demo goes deep into fine detail and how much resolution you’d actually need to crop into an image. For all of you who just want to scan their film at home: Use what ever you have, most digital cameras with a halfway decent macro lens will give you more than enough resolution and finer looking scans than all the entire level home scanners. If you don’t have a camera with a macro lens you might want to consider one of the more higher end home scanners.
Thank you for this video! The thing I love about my Nikon D850 is the option to choose between 47, 24, and 11mp, but the downside is 24 and 11mp are limited to 12-bit colour, while the full 47mp has 14-bit.
The D850 is still Nikons best camera. I shot with one for about a year with a D5, the D50 was always the camera I would grab first even for sports. The different raw settings is a great feature with the D850, lets hope Nikon can make a Mirrorless version of the D850 😉
Very helpful. I was going to stitch two D810 images of a 645 negative together. This tells me that I’m better off just taking one image of the negative and crop as much as I need.
Thank you. I am shooting my film archive, 50 years of my film photography. A lot to get thru. I am using an Olympus om1 III with the Olympus 60mm macro lens. Not ideal because of the mismatch of aspect ratios, which reduces the effective megapixels. After watching this I will stick to standard raw files for my 35mm, but use the handheld hi-res mode (not actually handheld, obviously) for my medium format negatives. One thing I also wonder about is the increase in dynamic range when you go to the hi-res modes on the camera. I have also noticed the issue of obnoxious film grain, dust and scratches particularly when comparing digital camera scans to actual scanner scans (epson v850). The epson scans often appear sharper, but the dust and scratches stand out like neon signs. You have no choice but to retouch them if you want to enjoy/display the image. With the film image, you can see them, but they are more muted and minor negative damage can just be ignored. Saves a lot of time.
I gave up using my Epson scanner mostly because of the dust issue and I found the scanning software was adding sharpness to my negatives… Which was driving me crazy
Okay, a bit (2 years) late, but here's my guess (from my own experience and some math): 24 MP is fine for 135 film, >40 MP for medium format, >60 MP for large format (4x5). Technically, 24 MP is enough for nearly every use of the scanned image, even with 4x5, but sometimes one want's to retain all the information in the negative (including grain structure), and then MP count matters ;-). Now I continue to watch … Edit: I disagree about the grain becoming harsh, but this might be my camera - an Olympus MFT with 80 MP pixel-shift hi-res mode, and it gives less "crisp" resolution on pixel-level, so that the grain actually looks quite natural. So, going significantly higher than 50 MP might "give back" the smoothness of the grain and more data to work with. As you're using Sony cameras, I made the experience that the images it gives (normal photography use) look very detailed and extremely sharp on the pixel level, and this might be the problem here (interference between the grain structure and the pixels on the sensor, and the Bayer filter/demosaicing process).
The grains on 50MP are much more detailed and defined but may appear "noisy" but that's what it really looks like when zoomed in. The 24MP just cannot capture enough details of grains so they come out as "smooth."
In 135, and with negative color film, 50MP is about the highest resolution you can achieve from a 35mm negative. On the other hand, slide film can go all the way up to 80MP, which no Digital Full Frame camera is currently able to achieve. I would have loved to see some slide film included in this comparison. Also, keep in mind that while the 24MP vs. 50MP in 135 did not reveal more detail from the negatives, it does not mean that it’s not worth scanning at 50MP. For Digital publishing, 12MP is more than enough, but for printing, you will still be able to print a 50MP file to 20x30 at 280 dpi which not quite 300 dpi, but very close. In comparison the 24MP file printing it at 20x30 would need to be at 190dpi, and will look considerably less detailed when examining them side by side. Maybe printing them and show the side by side comparison, and adding slides would be a good follow up video to this one.
don’t confuse grain for visual data. Scanning a slide above 12MP will resolve the grain better but it won’t extract any more subject data. Edges won’t be any more sharp.
This is very helpful. I am struggling to buy a scanner (Epson v850 pro) or to use my camera Sony A7iii to do it. Since my Sony A7iii has 24.2 MP, I think it is more than sufficient to do the job!
I reckon anything more then 16 megapixels will only bring up more film grain and noise (35mm Film) this also depends on the film obviously the quality and speed of film will give different results, I found Fuji Reala really good with low grain. I do a lot of scanning and 12-16megapixels is heaps. also when using a digital camera to scan negs, u have to factor in the lenses centre vs edge sharpens, cause unlike a CCD Scanner which scans and moves across the film, a digital camera relies on the lenses resolution to capture in one take, & as we know lenses vary in sharpens from centre to the edges, so What I do is use a canon 30 megapixel full frame sensor and macro lens on approx f11 and i crop into the centre of the image so I'm not using the outer edges of the lens which are not going to be as sharp, generally I crop in till I get about a 15 megapixel image, which is perfect for 35mm, but importantly my final image will be sharp from edge to edge as I am only using the centre of the lens (the best part of the lens)
Yes, Darryl I hate you for it but I think I probably have some good shots of you from Snake Alley and I might incorporate them in a future video. 😜 It's still weird to see you using Sony cameras. I bet you already have the new 14/1.8 mm on order. BTW... The new studio set up looks great!
CRAP ! I forgot you had some photos of me 😂😱😵 Switch to Sony about 6 weeks ago, can not believe how good the AF is on the cameras. Will only get the 14mm when we catch up in Taiwan next, so we can do YT videos together 🤣🤣🤣 You still in Mex? Hope your keeping safe
@@DarrylCarey Hell yeah, I hope we'll be able to do that asap! I'll be back home tomorrow. México was amazing and I'll probably be back at some point in the future. 🙏
Keep in mind he is using full frame cameras. With a 24MP APSC camera I get a lot more detail on 6x7 by stitching 4 images together. 24MP apsc is just not enough for one shot on negs that big.
This is actually a very good and informative video. I appreciate your efforts here. Though, I am a bit surprised the 645 MF image didn't show better results past 24MP. This would mean that 645 doesn't perform any better than the 35mm if "camera scanning" is a part of the workflow. My mind goes immediately to the LP/mm resolving capacity of the Sony lens (as well as differences in lens resolving capacity of the original film cameras) being a limiting factor. I have always assumed that 35mm film would hit the wall around 20MP, 645 around 50MP, and 6x7 around 80MP, when scanned with a dedicated film scanner like a Nikon 9000ED--even higher with drum scans. But we all know what happens when making assumptions! 😂 As an aside on lens resolving capacity, I know that Mamiya/Phase One had "excellent lenses" back in the day when a Leaf 22MP back was all the rage. But though the years, as the MP capacity of their backs went past 40, then to 80, then to 100, and now 150MP, the glass up front has simply struggled to keep up. This has forced them to constantly upgrade their lenses to ever more expensive versions. I think they have just released their 4th generation 80mm prime Schneider at this point--$6,000 USD for anyone interested. The same thing has to be happening to Sony/Nikon/Canon as they are increasing the megapixels on their models. Lenses with higher resolution are usually larger, heavier (more metal less plastic), and tighter per copy QC at the factory. All this makes them more and more expensive. Back to the subject at hand. It may be that a 90mm lens that was good for 12 & 24MP cameras in 2017 has hit its own wall somewhere along the way to 50MP in 2021. Just a thought.
I enjoyed it. Thank you. Did almost the same experiment before seeing this video. Compared my Canon 6D vs 5DSr cameras for copying 35mm negatives while using the same 100mm/2.8 macro. Other than double the hard drive space I didn't see discernable differences.
The short answer is, it depends on the grain size of the film and whether or not your lens is actually capable of the resolution necessary to make a difference. Additionally, it depends on the inidividual sensors; some lower-resolution sensors are better than some high-resolution sensors.
You can extract more detail from medium format if it is there :) shoot something with closed aperture and fine details. And it depends what film is used and how it is developed. Fine grain modern negatives can give 50mp detail even on 645 easily. And about the grain - it depends how you sharpen the image. Sharpening is different for 12, 24 and 50mp sensors because the size of the detail is different. That is all from my experience. I was even able to get 24 usable megapixels of detail from 35mm Ektar 100 shots. BTW what is the film scanning device you were using, can we get a link where it is selling?
In the mid to late 90's I would have my negatives scanned to the Kodak Photo CD's (Not the consumer grade Picture CD). The standard photo CD images scanned at around 6 Mega Pixels per image and the Master Photo CD's were around 25 Megapixels per image. I just started experimenting with my Canon R7 taking a few photos of negatives on my portable light table I bought way back in the 90s at Ritz Camera. I don't have a fancy set up yet. Just hand holding my camera and not quite filling the whole frame with the negative and my early results were good but not great. I hope to get a better set up in the future. But I was wondering about the mefapixels and how it affects the grain of 35mm film and 120 medium format. So I'm excited to see your results in this video
Hmmmm... I wonder how the gfx would do. The Sony 50MB sensor used much smaller pixel than its 24MB sibling. I also would wonder how the results would be with 67 and 69 chromes, color negs and bw negs. I recently shot some 1960 Color realist chromes. These images were side side by side 1 inch square. (Amazing to see the 3D impact in the viewer). I shot this with a 120mm GF on the gfx 50S. What was really amazing was the ability to recover shadows in completely underexposed areas. Things that appeared to be just black turned out to have lots of detail!! The latitude of chromes has always been about 5 stops but digital sensors may be able to rescue bad exposures from long ago! Thanks for doing this. Very interesting !
ı have a gfx50 r and recently ı tried to scan my 35mm negatives. now ı think no more scan epson v700 :D ı use analog m42 macro lens 1to1 ıts minimum focus distance 5 cm. ı dıdnt try my 120 mm macro gfx lens because minimum fıus distance 45 cm . gfx50 r working realy good !
Great to see this practical experiment. I expect as one moves up in film size (6x7 or even 4x5 film) the high megapixel camera would become more useful.
The top of the line Minolta scanner Dimage 5400 from the early 2000s outputs 40MP files (or life-size 24x36mm frames at effective 5400dpi). This was considered overkill by some and this kind of resolution wasn't exactly worth it unless you were shooting film with high-resolving power with a very sharp lens on a stable platform. From your test, I definitely prefer the 50MP files showing the grain structure - those are actual silver halide crystals you are seeing. Although I agree that 24MP seems more than enough for most use cases. But if you are shooting extremely precise work on, say, slide film and with a razor-sharp lens, that's where the benefit of the high-res sensor will show through. I remember a few years ago, people used to say digital doesn't quite have the resolving power of film. Funny how things turned out.
I think your conclusion is pretty spot on. 35mm film has a nominal resolution of around 18MP given its size and typical grain density. That number varies MASSIVELY depending on the film stock, but it's roughly a figure people seem to agree on these days. So any camera with around that many MP will preserve the maximum amount of detail, and even then 12MP is clearly enough unless you REALLY want to also preserve the grain pattern. Very interesting results and now that companies are making better camera scanning solutions it's great to see research like this helping inform people.
@@DarrylCarey TH-cam home page recommended it. Been watching a lot of film related photography content lately. Former photographer turned video editor because of COVID
Very cool video. thanks for taking the time to look into these things. I've recently moved into film (away from digital) - but I'm curious to know how folks catalog, organise, folder keyword etc their scanned negatives into LR or Cap1. I wonder if you've made a video about that yet? I'm going to check out your playlists etc. Thanks. :)
Depends on the film stock/size, I did LOTS of tests on a high end drum scanner many years back and a LOT of 35mm film often wasn't worth scanning above 5-6k (and/or past the visible grain).....motion film stock was even lower being grainier (and running at 24 fps, giving that organic 'film' feeling) Bit depth makes a difference too, not all resolution is down to pixel count, which is why properly downsampled pixels result in better pixels (as the art world and artists drawing bigger than the end printed result has shown us for a long time and more than a while back).
Flaw in experiment. 3200 Kodak shows nice sharp grain at higher ISO.This is what you want as it proves it is resolving more detail. When enlarging in darkroom, a grain focuser is used because when the grain is sharp, the image is sharp.. If the neg was made with best lens like Otus or APO Summicron , is higher MP copy camera required? I can visually see with plain eye on the neg APO lenses sharper. I submit the best copy lens is required and high MP.. Also when looking at one shot at 200% is by nature going to be less sharp than one at 100%.
Thanks for the feedback and yes an Otus lens would give you a sharper image than a Sony or Canon macro lenses, but sadly most people do not have the budget for an Zeiss Otus lens. Been looking at getting a set of Otus lenses for my RED cameras, but they are still selling for crazy prices.
Most flatbed scanning software talks in DPI so at 2400 DPI a 35mm scan would be about 3600 X 2400 = about 8.6 MP which is pretty adequate for most family photo type applications. The file size will depend on the bit depth, of course. Occasionally I’ll go to 3200 DPI.
Follow up.. put the baseboard, backstand, Kaiser light, lomo holders with my SL2 and a Sigma 100 2.8 macro. delightful... Tethers right into the software.. super easy
Interesting comparison but I think you miss a point worth mentioning. Even though apparently the 50MP is no different than 24MP in terms of capturing fine detail, it does give you the opportunity to print double the size. So for someone who wants to print as big as possible from a 645 negative without post-interpolation, the 50MP scan does make a difference.
Hi Zisis, yes you are right that you can print larger from the 50mp. I have printed A2 prints from the 24mp files and the detail was very good. So I'm guessing you can print A1 or even A0 from the 50mp. The problem with printing larger prints, is you need to stand further away from the print to appreciate it so the extra detail if any will be lost.
The resolution can be calculated using Fourier optics. The result for an ideal circular aperture, and green light with a wavelength of about 0.55 µm can be expressed conveniently in terms of the f-ratio of the lens. Thus, it turns out that at the focal plane of the camera, the resolution is very approximately 4f/3 µm, or half this length for the Nyquist-Shannon sampling limit. Solving this for f/5.6 and a 36 x 24 mm frame, we find the N-S sampling limit is 9643 x 6429 pixels, or about 62 megapixels, assuming the lens is optically perfect. It is physically impossible to do better than this.
This might be correct for digital sensor, but not for film. As what is captured on the frame is totally irrelevant in scanning. Even if you shoot with a Holga, you are still going to want to scan the image at the end use size. You don't adjust your scanning resolution based on the resolution of the lens used. In scanning, you are not scanning some abstract information on the frame. But replicating the structure of the film, that you then print out using inkjet. And for non print purposes, you do not benefit from extra information as none of its retained after down scaling. Its simply decreases the image quality, because now it has been scaled through an algorytm. In every case, you want to scan an image that is the end use size. So this approach is wrong, most people do not understand what scanning is as they constantly approach it like it was a digital format.
@@Nobody-Nowhere I think you may have misunderstood: Fourier optics gives the optical resolution, regardless of how the image is recorded. This is the theoretical limit determined purely by the finite wavelength of light and physical dimensions of the optical system. You can find a detailed explanation of the theory in any standard advanced textbook on optics. There is an added complication with film versus digital. Digital sensors usually employ a Bayer filter (or similar) plus a so-called demosaicing algorithm to interpolate between pixels and reconstruct the image, while scanned film measures the film density in each recorded channel equally at every pixel, without demosaicing or interpolation, which has further implications for resolution. Nevertheless, both methods ultimately yield a digital representation of the image, where the Nyquist-Shannon sampling theorem applies.
@@cdl0 Yes, and its totally irrelevant for scanning. As when you scan a film, you are not scanning an image on the film. Its totally irrelevant what is recorder on the film.
@@cdl0 Lets say you would go by these numbers, and scan that 62megapixels. That gives you a print size of 42x27cm @ 600ppi. But what if you want to print larger? Will you simply upscale that image? Or will you scan more information from the film, to accomodate for a larger print size? Guess which one produces better results? Its not the image you are scanning, you are replicating the structure of the film on to the paper. This is exactly why most people misunderstand scanning, as they constantly approach it from this digital mindset where there is some set amount of "data" on the film that can be digitized until there is nothing left to extract. You are scanning film, literally. You want to resolve as much information from the film, as its needed for given print size.
@@Nobody-Nowhere Hmmmm, I can see you don't understand the theory. There _is_ a maximum amount of data that can be extracted from an image that made by a lens of a given size, regardless of how that image is recorded. Using a higher resolution sensor, or scanning a film at higher resolution cannot retrieve more information than this limit. This is true for any optical system owing to the finite wavelength of light, which you can think of as the brush used to paint the picture. Once you stand close enough to see the artist's brush strokes, standing closer will not reveal more detail in the painting. The film or sensor is only the canvas, not the brush. Thus, in your example, enlarging an image by mathematical interpolation is equivalent to scanning it at a higher resolution.
IMHO, the larger the negative the faster you reach the limit of any 'scanning' camera.. especially if you are 6X9 or large format .. as its always the same resolution irrespective of the negative size .. however I can agree that there is a theoretical max dpi for each film
It depends upon the film format. Of course you want all the megapixels of your camera, if you're "scanning" 6x9 or even 4x5 negatives because these huge negatives can have more than 100mp resolution. For 35mm I think 24mp is plenty. However while "scanning" the negatives with your camera you often have to crop the edges. That means a 40mp sensor is perfect to get 24mp "scans".
Great video! I am curious is you've tried taking multiple images of 120 film and then stitch them together for more resolution? I'm theorizing this would help the 12 mp camera do better on medium format film.
Nice video! Regarding the 12 vs 24mp comparison: seems obvious that the diiference is less with a 35mm negative (the photo with the helmet). This is logical, too: a 6x6 negative holds more information, which is why you used the format in the first place. I did my 35mm color slide conversion with a Nikon D90 and Sigma 70mm/2.8, which is 13mp. My guess is: if you distinguish the grains of the original, you are probably maxing out the camera sensor! because all of my color and BW negative material is 35mm, I probably stick with the D90. A higher resolution will only result in larger files, not better ones. One thing I would like some opinion on: the light source. I have used a setup with a professional copy stand (heavy, stable) and I use a flash unit (a Nikon SB400 connected to the hot shoe by a SC17 cable). All very old stuff, but it doesn’t fail. The flash output is soft (bounced, ofcourse) of constant light temperature and there is no vibration due to the very short duration. The camera is triggered using an IR remote, no me touching the camera! Do you think this is good setup?
Excellent demonstration. Any wonder why the D6 is only 20.5 megapixels. I bought a Z6 with 24.5 megapixels because the added cost to buying a more costly higher megapixel camera just isn’t worth the investment.
The Nikon D6, was my last pro Nikon camera I owned. I shot IRONMAN races around the world and nobody ever asked for larger images, 20 - 24mp is the sweet spot for me …
Interesting exercise, however I am not totally convinced that is anywhere extensive enough to draw conclusions; what I mean is that there are too many variables that aren't accounted for. First of all different negatives gives different resolutions, all the negatives used are 100 ISO and above, what about 64,50,25 ISO, the lenses used have an influence, different optical linear resolution, and the aperture value used when shooting also have an impact; there might be little to no difference between 24 and 50 MP on a 160 ISO shoot at f4 on a cheapish optic whilst there might be quite a bit on a 25 ISO shot at f8 on a Zeiss, and this differences will undoubtedly be more noticeable once you go to medium and large format negatives. If we look at the face value and just consider the linear resolution of a negative a Kodachrome 25 ISO 35mm is around 8MP, however when scanning you need at least 4 times that, this is to have enough square pixel for each roundish dot on the negative and factor in the Bayer configuration. A traditional Drum scan at 5-6000 DPI for a 35mm is around 50MP. I haven;t used negatives/transparencies since Kodak stopped Kodachrome, but if I well remember in the old days the best way to scan for a pleasing grain rendition was to use the highest resolution possible.
Darryl, thanks for the video. My scanning workflow is like yours and saves me a lot of time while producing what I think are great results; I'm using the a7R3 through the 2.8/90mm Macro G and processing with NLP for both 135 and 120 film. I haven't found confirmation, but presume the A1 does not have an anti aliasing filter like the other high MP Sony cameras; do you think this is a factor in your results? Also, did you try the medium 21MP and small 12MP file size options on the A1? Thanks
Not sure about the different options on the Sony A1, need to look into that. I thought it was a good idea to test 3 different cameras, so people with 12mp or 24mp will have a good idea of the results they will get from their camera. Was looking at a A7R3 but went with the A7c because of the weight and size.
Very interesting comparison and a nicely done video, I didn't expect the limit to be 24mp for the medium format. I was wondering what will happen with sheet films using the same cameras. If you shoot any please make a video.
Please, kindly compare scanned and photographed film images side-by-side! Are there diffrences in resolution, color depth, visual appearence, noise etc.?
This pixel peeping is fun but, the average photo enthusiast scanning his silver negatives is going to print them on paper. Beyond roughly a 12 mpx scan you are whistling in the wind because you are getting 300 dpi on the printed picture no matter how many zillion pixels of information there is/was on the silver negative, 300 ppi (pixels per inch) is all you get.. If you are going to the trouble and expense of making an analogue silver gelatin negative you have to print it as an analogue silver gelatin print to appreciate the image. And if you want Smooth, sharp, detailed images, why are you shooting a coarse (high speed) film in the first place? The final point I want to make is why are you digitizing your film? AI will make a fantastic image that will be more striking than anything you or I can create.
Might want to consider the resolving power of the lens used on the 50mp since glass also has limitations with regards to how much detail is transmitted if the glass was designed in a time period where the peak mp of cameras was around 24 you will likely start seeing glass imperfections the higher you go on resolving power with the sensor.
Depends what you want to achieve: 11x14" or large format 36x48" prints ? Defined grain is fine if you want sharp chemical/analogue grain. Also dynamic range is important - newer sensors have higher DR - helpful esp with compensation developers eg divided D-23, some pyro etc - revealing all the shadow and highlight detail. Each photographer has different requirements - do your own tests with your personality developed negs and several hired cameras eg GFX100 etc. Also, you can save up your negs, hire a GFX100 or Phase One and capture negs for a day! :)
I've done 24mp but used pixel shift resolution. So it's basically free of noise etc. artifacts. I should probably compare it to standard shooting without pixel shift since processing those is a real pain.
@@DarrylCarey Word of warning if you want to test the pixel shift. Afaik, there is only 2 softwares that process the files properly. One is pentax DCU5 and the other is rawtherapee. DCU5 is almost impossible to even install, and if you get it to work you are greeted with the worst image editing program I know. It is slow, uses proparly just single core and no gpu acceleration. It's UI is like it was made in the 90's. Rawtherapee on the other hand is confusing as hell and I gave up on that. There is a free software the converts the pixel shift pictures to dng's but it does only automatic processing and doesn't let you choose any sharpening / noise reduction etc.
What happens if you use 24 mp sensor but use it with sensor shift high res to make it 96 mp? I think this test i am gonna have to do myself, because i have a camera that can do this and i recently started to mess around with film. My work table will of course be much more simple and cheaper and for the lens i will find some vintage adapted macro capable.
It depends on the film grain, which is related to ISO. For ISO 100 I'd definitely use the 50 Mpixel. Did once a scanner test on medium format, not a best film, not the best objective and a "semi-pro" scanner. Started at 125 Mpixel, at 75 Mpixel many (very) visible details, like small flowers in the far field vanished in a continuum. So I can tell that medium format requires something like 100 Mpixels average, maybe even more for high definition shots.
I think the grain in the 3200 speed film wasn't that bad on the 50mp. It was just a lot sharper than the 24. The man's eyes were soft on the 24 while quite sharp in the 50mp image
I did a side by side grain-peeping comparison of the files you uploaded and found that the "A7c 6.jpg" 24MP scan seems to have had some sort of camera shake/focus issue? (The area to the top left of the man's face) It's even less sharp than the 12MP A7S II. Do you think this might've coincidentally contributed to the aesthetically pleasing/soft quality of the grains as you mentioned at 10:30?
I think that a lot depends on the capability of specific lens to support high megapixel camera. For example, if this macro lens on 50mpix camera have max resolution about 32mpix, than of cause will be hard to see the difference between images made at 24mpix resolution. Second, it is important to have high resolution lens on film camera. Old lenses frequently not delivering more than 24mpix on 35mm camera. Third argument - for medium format film, if old lens has good resolution and we compare image scanned by 50 mpix Sony camera with an image scanned by 50 mpix Fujifilm GFX camers. The same image from GFX camera will looks better because optically it is easier to receive high resolution image on bigger sensor.
![ILLSLICK - KILLSHOT REMIX [FULL VERSION]](http://i.ytimg.com/vi/RIK8RrqIAzE/mqdefault.jpg)
I owned a $400,000 Kodak Pro Photo CD film scanning system from 1993 to 2007 which used true RGB linear scanning arrays and Schneider apochromatic lenses . We could scan 35mm frames up to 4k x 6k resolution (24MP) at 16bit Lab colour but we did this rarely as scans of our studio calibration photos (taken on EVERY film stock) showed that only Kodachrome 25 ASA (ISO now) transparencies shot in ideal studio lit conditions with the very best lenses and apertures could provide enough film detail to justify this, 24MP resolution. We did over two million scans for archive producers and movie companies (all the stills for Star Wars etc.) and nobody ever wanted more that the 4k x 6k Pro resolution. It was the 16bit tonal range and the hundreds of custom negative and positive film terms developed by Kodak and us that made the scanning quality the best in the world! David Myers, Digital Masters Australasia.
Very impressive. So I think you are saying that anything over 24MP is unnecessary. I've compared my D600 (24MP) images to my D850 (45MP) images using a Tamron 90mm f2.8 Macro 1:1 lens on each and I really can't see a difference but I don't have a trained eye.
Great comment. Did you work with medium format negatives and scans though? I think that perceived resolution and detail would be a lot higher than 35mm hence the purpose of Darryls video. Or do you think the 24mp limit still holds true for MF 645? Personally, in my experience 24 MP more than handles 35mm negative scans and adequately handles 645.
For 6x6 up to 6x9 I think it is worth it to go up to 47 megapixels. For large format I think you should be capturing those negatives with a digital medium format sensor of at least 50 MP as realistically the amount of resolution is absurd and 100 MP images are required to fully capture an 8x10 negative and match it's print dimension capability. But at that point it's not practical to DSLR scan a large format negative because it will be very hard to keep the negative flat and thus you are better off using a lab scanner. That way they can keep it perfectly flat. That or drum scanning.
Looks like this guy can see the difference th-cam.com/video/1thY_Az14bA/w-d-xo.html
Also, this shows insane details at 709Mpx. I guess film’s limit is pretty high (especially when you go beyond medium format)
th-cam.com/video/sqN7n9bXgtU/w-d-xo.html
This is it. This is the final say in the 35mm film resolution debate
Adding the timestamp at the beginning of the video is such a good-natured generous gesture (losing watch time-an important ad revenue metric), that it deserves this user engagement datapoint post and a thumbs up. Wish all content creators were this awesome.
Paused at 2:44. There are conflicting forces. Let's start with the 35mm negatives and their resolution. In the film days, resolution was expressed in a linear unit because that directly relates with human perception of detail resolution: LinePairs per millimeter LP/mm. A pair being a black and a white line. Two times the LP/mm means we see that as twice as sharp (ceteris paribus, assuming parts in both comparison tests can reproduce both). You might argue that a 36mm x 24mm negative needs at least 7,200 LP * 4,800 LP and this gets us to an area size of 34.56 million. Why that number? Because 100 LP/mm was attainable with good film and professional grade lenses. Nikon glass would get better in the center and a bit less towards the edges when stopped down 2 to 3 f-stops from fully opened. With Leica glass you might see some higher values in the center of the frame and a bit more fall-off from that towards edges/corners. So a 36MP digital camera would equate that resolution? Well, if the black and white lines in the subject precisely align with the photosites in the sensor.
But there is a problem with a digital camera. You see, film has RGB at all coordinates because the color is in three layers on top of each other with subtractive filtering between them. But a digital camera has not. The sensor is colorblind and in order to get color in the raw file, there is a filter grid over the sensor that has repeating filter patterns of 4 photosites that thus build an R.G.G.B quartet next to each other. The problem then is to convert this into RGB pixels and this is done in raw processing. This happens in camera for JPEG and MPEG shots, and in your raw processing software for your raw shots. Your camera comparisons in Lightroom tell more about Lightroom than about the camera. The 14 bits gradation resolution you had at the R.G.G.B photosite data in your raw file have been degraded to less than 9 in raw processing.
That raw processing is a mathematically precise and repeatable guessing of missing color for the R.G.G.B quartets in the raw file so as to change R to RGB, G to RGB and B to RGB. This creates artifacts like Moiré if we recognize it and "noise" if we do not recognize it. Application of basic computer vision type AI can help a bit, but without "deep knowledge" in the raw processing software about what is in the photo, raw processing is not helped a lot.
Lower resolution (than 36MP) cameras make raw processing easier with a so-called Anti-Aliasing filter (AA) or low-pass filter. These filters disperse the light for the R cell to its surrounding neighbors. Hence I call it the fuzzy filter. It helps raw processing and the idea already was applied in the 1970s to aid image processing out of data from scanning tunneling electron microscopes.
Or, the presence of a fuzzy filter must be part of the reasoning. We see that higher than 35MP resolution cameras generally have done away with the fuzzy filter - making lenses sharper and depth of field shallower.
The problem with digital reproduction of film is the grain in the film. With increasing resolution, this generates a raw file that has more detail of the grain and this is not likely understood by your raw processing software. So we have to find an optimum here between these conflicting forces.
But my negative is black and white - unless you shoot a Leica Monochrom that is even more expensive for not placing the R.G.G.B filter layer and it still may have the fuzzy filter by the way because Leica's old lens designs are not well adapted to sensors, you always have an R.G.G.B raw file that needs raw processing.
A scanner can run a multipass scan and build an RGB.RGB.RGB.RGB file that is better than raw and does not need raw processing. If it can create more than 8 bits per color channel TIFF, you basically outperform the raw file. So 16 bit TIFF is way more better than 14 bits raw when at the digital level there is a factor of only 4 between them.
Some people would say that 24MP is enough. Well it might be the optimum, bottom line, after these conflicting forces got sorted out.
The problem with digital is not only that we look at the results of raw processing (where we lost a lot of quality) but also look at digitally upscaled or even upsampled representations of our images, when we depict them larger than 100%. This gives an illusion of image quality. Even if we depict our images at smaller scales there is some digital processing going on, to the images. Like anti-aliasing for the display, depending on your software. Some defunct guy would call your digital photo on your computer "fake" and here he would have a point.
I have tested Topaz's Gigapixel AI app last year as a way to get control over what is sent to the printer. Its ability to upsample and in that make guesses about what detail to add in blowing up a raw file, was incredible. A year later and Adobe have improved their Enhance Super Resolution features by a bit.
The magic of a 12MP Nikon D700 is in these aspects: for human perception, two times as sharp as 12MP is 48MP and all wow conversations about in between values is extremely naive. It explains people's remarks that the gain from 12 to 24 is not as great as they had expected - it's only 40%.
The magic is also in its discrete analog to digital circuit (i.e. not bundled with, stacked on, the sensor) and it is in the relation between Nikon F-glass's qualities and the fuzzy filter.
To circumvent raw processing, we could mimic the multipass of scanners by shifting the original a tiny bit and shoot again, later match/stack the layers in Photoshop. But our raw images already have had raw processing at that moment and damage was done.
What about sensor shift - cameras have IBIS so that should be easy, right? Well, the wild assed guessing of RGB colors in raw processing gets stuck in the edges of the image where there are no neighboring cell values to make guess from. The difficult solution is to have two algorithms in raw processing of which one is specific to the edge problem. The other solution is to have a sensor with more columns and rows in the edges and record the data thereof but never allow these columns and rows in the displayed image - they only serve as aid in RGB guessing. Well, throw in a couple more, and these rows and columns can also be used in IBIS without moving the sensor. If there are enough rows and columns you might only move the sensor, say with 5 photosite units, when you detect 5 units shift at image level. Or, I don't expect the IBIS systems to be able to make single photosite unit steps that we would need for RGB shots of 14.14.14 bits natively.
Or, if you make a living in scanning negatives and want to speed up - the promise of the digital camera - then I would look at a Pentax K-something that can do this sensor shift actually. I would not use it anywhere else, by the way.
All this still leaves us with the film grain problem. With increasing resolution the raw file has more grain detail and the software may have difficulties abstracting that grain noise away.
When we stick with the regular Bayer-filter filtered sensor and its raw files, a comparative test is required indeed. And it will be valid until somebody develops a raw processing program for film scans.
That was an incredibly insightful comment, thanks!
This was one of the most well put comments about film scanning that I've ever witnessed. I've read through it thrice and my man has it been a journey. I'd love to learn more from you on this topic.
Thanks!
Two commenters, thank you for the compliments. The summary is that we think we know what we are looking at when we compare shots in Adobe Camera Raw (ACR, the raw processor of Lightroom Classic as well as Photoshop), but we see ACR inn the first place. IN another YT channel I discussed if Topaz Gigapixel AI could deal with film grain and the host of that channel says he tested it and the short answer is, no. That's a pity, but I still can blow up my digital images to very large sizes with it, just not film.
Note that my write-up applies to "digital photography" in general and that scanning or reproducing film images into digital merely is a use case in there.
Finally, I'm obsessed with the truth and removing veils of ignorance, not "sharpness" as a goal in itself, or "beautiful bokeh" and I'm old-school about the decisive moment. But add that face/eye detect AF help me shoot portraits with a different interaction than I would in the film days. And I can obsess about that. To paraphrase a photographer that is a lot more famous than I am, people obsessing about x, y, z probably have less in the artistic department to talk about or show off.
Let's go back to that. Take my write up as a way to put your mind at ease. Shoot, review, shoot, review, consciously, 10,000 times - to train your brain. Worry about all these details here a bit later.
@@jpdj2715 will you suggest a better way to digitize our film is by scanning ( using epson, nikon coolscan, frontier, noritus etc.. ) rather than a dslr (A scanner can run a multipass scan and build an RGB.RGB.RGB.RGB file that is better than raw and does not need raw processing)?
@@joshmcdzz6925 - that's what I wrote. The question to me is unanswered how well the scanner can do the RGB.RGB.RGB.RGB where a Bayer sensor would do R.G.B.G instead, that in raw processing must be converted into the RGB.RGB.RGB.RGB.
The scanner depends on software to drive it and pull the scan data out, next convert them into an image file. The R.G.B.G is at best 14 bits each and I would expect the RGB in RGB.RGB.RGB.RGB to be 16/16/16 for each coordinate - 48 native bits per coordinate with no guessing of missing colours.
But nobody is transparent in this digital photography market.
I have the CoolScan and the Epson flatbed, but bottom line will grab my Nikon Z 7ii with 105/2.8S macro, plus raw process these images. IIRC the Nikon D850 had a profile for reversing shots of negatives.
"Bayer" keeps my processing workflow standard, and my files and their filing too.
Lazy? Probably. Yes, it's a lot faster.
prefer the sharper grain on the higher res
Back when digital cameras we're barely more than a glimmer in kodak's eye, I read an article in one of the photography magazines that said that 24 megapixels would be enough to capture all the data that a 35 mm slide contains. This was all based on math, of course, because no one had ever seen in a 24 megapixel sensor at the time.
Other videos suggest this ratio.
@@johnwinter6061 That is who I have decided to treat my 42 megapixel Sony a7r2 as if it were a medium format camera, and get a Sony a6700 to use as my daily driver.
Really interesting to see these comparisons, thank you. I do disagree on the grain comparison though - I felt the one on the left (50MP) was more pleasing and the grain to me looks smoother and more "shaped" than the garin on the right photo which looked more pixellated to me (although I'm looking at it from a compressed video). Guess it all boils down to testing for yourself and seeing what your own preference is.
I agree with you. Images taken with 50 MP camera are sharper and you can see more detail of the grain of the film.
I'm using my 24 MP Nikon D610 for scanning of the film but I would prefer camera with more MP and no anti-aliasing filter like Nikon D810, Nikon D850 or similar.
@@MrCROBosanceros - I'm just getting started scanning with my D850 but would like to buy a D600 or D610 (24MP) as a dedicated copy camera. I have a D5500 (24MP) that I will try but it's a cropped sensor so I don't expect it to be as good as Darryl's 24MP scans.
For 35mm, I’ve been really happy to use 24 megapixels. I have an older Nikon D600 that I use exclusively for my 35mm scans so I don’t waste shutter count on my newer cameras and it works great for that purpose. I’m glad to learn that 24 works well for medium format. I have quite a few medium format negatives to convert. Great video. Looking forward to your video on the Kaiser system
The Nikon D600 was an awesome camera, not the best AF but still a great camera. I must get around to reviewing the Kaiser system
I did the same thing. I own a D850 but bought a D600 just for film negative "scanning". I'm extremely pleased with the results for both my 35mm and 120 (6x7) negatives. BTW, I've just started taking two images of each 6x7 negative to increase the resolution. I haven't compared the single image to double image yet but I'm hoping I see a marked improvement to make it worth the extra step of creating a "pano" image of the 6x7 in Lightroom.
Great video Darryl! Really did us all a service with the comparison, and nice production value 👍
Hi Hashem, hope you and the family are well over in OZ. Starting to get back in to my film photography now, been a busy few months shooting videos for clients, will be nice to slow down and shoot film again. How is the Leica M-A going? Not sure I could use a Leica with no light meter 😜😜
@@DarrylCarey hey thanks man, same to you. That's good to hear! I love it. Had gotten used to no meter from years using the M4 :) I'll DM you over on instagram soon
Thanks a lot! Very helpful, as I am looking for a digital camera as only an addition to my analogue ones. I understood that 20 to 24 will be more than enough for my 35 to 6x6 negatives.
Hi Robert, the 24mp will be fine for 6x6. I test it with some of my Hasselblad negatives and the results are the same.
Such a good comparison! Was so looking forward to this.
Thanks Anupam
I’ve always had this question, ever since the lab I use rejected 2 photos I wanted to blow up to 36” x 36” and another at 60” x 60”, respectively. So I’ve been testing a small inexpensive digital camera with a resolution of 48mp. When the high mpx sensor is used, you can blow up the picture to larger sizes with even greater sharpness and clarity. I think 24mp and even 12mp for up to an 8” x 10” print is fine. If you want larger prints like poster size and so forth, the increased resolution of 48mp, 50mp and so forth really pays off. It’s really all about what size of print you want to end up with.
Exactly this is my observation as the manager of a pro lab. I noticed that this pixel (with a digital appearance) works very well when the file is enlarged and then printed in large format. It's not marginal to scan at 50 MP for large prints, it creates the ideal photo grain blur (after pixel extrapolation) for 44 or 60 inch prints.
I personally saw an image from a 6mp Nikon D70 blown up to 30x40” by a lab in Louisiana. It’s was flawless. Sometimes it’s also about the skill of who is doing the enlarging. I’m sure they weren’t using the same software everyone else was. But damn it was amazing. And 6mp!!!!
This is an amazing video! Would be interesting to test how big of a difference there’s between a (24Mp in my case) full frame and an (24Mp) apsc cameras. Great work nonetheless, as always!
I used a G9 in my video on converting negatives and that did a really good job with detail, so an APSC sensor would be fine
Really happy that someone finally addressed this and actually did a test-great vid I agree!
I found it to be better with an apsc sensor and a full frame lens, only because you can minimize lens flaws. Also handy because you can manually use lightroom's lens correction that has been used in the film camera because there is no need for DSLR's lens correction. Image quality clearly depends on the camera, for example d750 was slightly better than d5200 but d7500 was about the same as d750 all tested with Nikon's 60mm, 85mm PC, 105mm, 70-180mm and 200mm macro lenses. Lens makes a very minimal difference, only 200mm masterpiece made a little more visible difference in 6x6, but focusing and aligning with film surface was so hard and time consuming that it wasn't worth the effort.
@@karajohn1234 - This sounds encouraging. I'm finalizing my setup today and would like to dedicate my old D5500 as my copy camera with a Tamron 90mm Ti (full frame) Macro lens. I'll be comparing the scans to my D850 scans and hope I can't tell a difference because I'd love to scan with just the D5500.
It will make no difference, apsc will extract all the same level of detail and since scans are at base ISO and shot raw the files have the same of flexibility in post for color adjustment. What actually matters is the macro lens you use for the scans. Some macro lenses are much better and sharper/resolve more detail than others.
Hmm, nice experiment - I have the same setup and have run many of my own.
I think you will find that higher megapixels will excel once you print (if that is your target output). A 6x7 negative, for example, can achieve 1:1 optical resolution at 300dpi all the way up to 30”x45” which is equivalent to the 1:1 output of a 63 megapixel camera sensor. A 24 megapixel sensor is only capable of rendering a 16x24 print at the same (optimal) resolution, so capturing a 6x7 negative with a 24 megapixel sensor will limit your output options. There are some great resources for this online. I keep a chart at my desk.
Enjoyed the video! You have a new subscriber .
I agree with groovejunky2549. Targeting the printing output dimensions is the key here. Higher resolution / megapixel = larger printing size possibilities.
Good video, decent length, info packed & well fleshed out thanks!
Thanks Izzet 😀
Dynamic range is a big factor, too. Digital cameras are getting better, but I'd be curious to see a dynamic range comparison with top end flatbed and drum scanners specifically for dynamic range
Thanks for a great comparison!
My 18 megapixel aps-c sensor does a great job for me, lots of resolution for my photos!
Have a good week!
Hi Sina, hope you are well. You converting 35mm or medium format?
@@DarrylCarey Doing good thanks!
I do both!
For 120 6x6 I take two pictures and stitch them together using lightroom classic in order to get more pixles our of a frame.
6x4.5 is just about 3x2 aspect ratio so I get a nice 18 megapixle file!
@@SinaFarhat Yea that's such a great point. Even people who want 40MP or 50MP scans for 6x6 or 6x9 scans can still use lower resolution sensors and stitch the image. 18 or 24MP is more than enough for home scanning.
More is less! This video confirmed my thoughts about high resolution cameras and scanners used for film scanning. Years ago I began scanning my 120 black and white negs with a Nikon 8000 dedicated film scanner - and I hated the results !!! The grain structure was overly sharp and drew way too much attention. Then during the pandemic I rescanned them all again with a Canon 5d Mark 4 (30 megapixals) and loved the results - slightly softer and way more natural, and they print beautifully :)
The Nikon Coolscan scanners are amazing scanners but starting to their age now …. Still going for crazy prices on EBay. I could wrong but I think the Nikon Coolscan scanners used a CCD sensor, which does not have the dynamic of a CMOS
Love your videos Darryl! I remember asking Ming Thein a few years ago the resolution equivalent to film, and he answered that he estimated it between 10 and 12 megapixels. That is consistent with your findings that 12Mp give very nice results for 35mm negatives.
I was a little surprised by how well 12mp did with 35mm negatives.
A high resolution 35mm negative, properly exposed and developed, contains around 12~16 MP of data. A 24 MP digital camera is more than adequate to extract all the information recorded on the negative. The situation is different if one is digitizing MF negs. Here a different technique is required.
This is what I thought before but it is simply not true. 24MP resolves much more detail than a 12 and 16 MP sensors. I also tested 36 MP and it resolved even more detail than the 24MP sensor.
@@ShutterKnack I have been experimenting with digital camera ' scanning ' for a few years. In most cases, I find that even a 20mp MFT sensor produces optimum results. The only time I one needs higher pixel count is when digitising MF negatives or high resolution 35mm copy film, like ADOX CMSII. Here only a Flextight or high res drum scanner can resolve the details this film is capable of recording. More mega pixel makes for bigger files, but if the original neg does not carry enough information, inflated files are the result. It is good to experiment with though.
Everything matters bud
@@crystalous I beg to disagree. Not megapixels for sure.
Really enjoyed this. I was looking at getting a high resolution camera for scanning negatives recently. You’ve just saved me a lot of money!
Glad the video helped, check out my other video on using a Mirrorless camera to convert your negatives.
Very useful info Darryl, thanks! I would add that Nikon shooters can purchase the Nikon ES-1 or ES-2 to copy 35mm slides or film strips. If they already own a 60mm macro, it is a simple, straightforward path to digitize film. D850 owners really have it made as the camera includes a special mode to invert color or monochrome films.
Hi JK, I remember when the ES-2 camera out, never got around to using one. Will see if I can get my hands on one for a review.
To me a far more interesting comparison is: How does a wet darkroom PRINT of say, 16x20, from a medium format negative compare with a 16x20 PRINT produced from a scanned medium format negative?
The wet darkroom print wins hands down as long as the enlarger optics and operator skills are up to par.
Thanks Darryl. I am about to start converting, but have to wade through the choices of converters.
Let me know how you go 😁
Man I’m sooo glad I just found this video, I shoot with an X-Pro3 (26MP) as well as an M6 and a Hassy 501cm and just ordered the 80mm fuji macro to “scan” my film negatives, thank you for this Awesome experiment.
I just started scanning with DSLR and tried this experiment scanning 120 negatives with my Nikon D300s vs my Nikon D810 😮 blown away by the level of detail i get with the D810! such a difference and I never would have thought!! thank you! the D300s is super soft less detail and contrast. it matters
So all of this has been scientifically dealt with. It’s all depending on the film speed and grain size but generally speaking:
You need about 8 to 12 Megapixels to resolve the detail on a frame of 35mm film. It takes up to 30 Megapixels to resolve the finest grain color negative and black and white film on the grain level. Velvia 50 in 35mm is somewhere in the high 50s to 60s. You can figure how it’d work for medium format. The general thing is that you don’t really need more than 8-20ish megapixels for most 35mm work. The only thing a higher resolution will afford you, is the ability to print larger or to zoom in further at which point you pictures will show grain pretty quickly.
Now Darryl. You will not get any more
appearance of grain on a 35mm frame of TMax p3200 when you scan it higher than 24 megapixel and only zoom in so little or even at all. It’s true than grain can be more defined when scanning at a higher resolution. But we’re talking finer grain film like Portra and a resolution jump from like 6 to 13 megapixel. With a grainy film like p3200 you’ll have crossed that threshold in the low teens. The reason the 24 megapixel scan appears to be less grainy is because it has some slight motion blur from the scanning! You can clearly see how it’s smeared a little horizontally! You can scan p3200 at 100 megapixel and it will not magically appear to be grainier just because you’ve zoomed in a little. That’s not how resolution works.
Also it’s completely fine to scan 120 with a 12 megapixel camera, you’ll still get more detailed results and fine tonality than with a 35mm frame. Sure you’re not utilizing all the resolution the film has to offer. But if you shoot 120 for the resolution you’d probably also have a scanning method to benefit from that.
And don’t forget CMOS color sensors (with a Bayer filter) only actually deliver around 70% of the resolution they say they have.
I want to encourage anyone nerdy enough and interested in how resolution actually works to watch Steve Yedlin’s Resolution Demo on his website. It’s for cinema but the same principles mostly apply to photography and print. Especially part two of the demo goes deep into fine detail and how much resolution you’d actually need to crop into an image.
For all of you who just want to scan their film at home:
Use what ever you have, most digital cameras with a halfway decent macro lens will give you more than enough resolution and finer looking scans than all the entire level home scanners. If you don’t have a camera with a macro lens you might want to consider one of the more higher end home scanners.
Thanks sharing and some really good info, yes you are right about using whatever camera you have at home, a 12mp sensor will give you good results.
Thank you for this video! The thing I love about my Nikon D850 is the option to choose between 47, 24, and 11mp, but the downside is 24 and 11mp are limited to 12-bit colour, while the full 47mp has 14-bit.
The D850 is still Nikons best camera. I shot with one for about a year with a D5, the D50 was always the camera I would grab first even for sports. The different raw settings is a great feature with the D850, lets hope Nikon can make a Mirrorless version of the D850 😉
12-bit is fine if you don’t do really intensive creative color grading. You shouldn’t run into any problems when converting negs.
(Edit: spelling)
Very helpful. I was going to stitch two D810 images of a 645 negative together. This tells me that I’m better off just taking one image of the negative and crop as much as I need.
I have boxes of my grandfather's negatives to scan and I've been wondering if my 26MP camera is enough, now I know! Thank you!
Hi, please let me know how you get on with converting the negatives 😁
Thank you. I am shooting my film archive, 50 years of my film photography. A lot to get thru. I am using an Olympus om1 III with the Olympus 60mm macro lens. Not ideal because of the mismatch of aspect ratios, which reduces the effective megapixels. After watching this I will stick to standard raw files for my 35mm, but use the handheld hi-res mode (not actually handheld, obviously) for my medium format negatives.
One thing I also wonder about is the increase in dynamic range when you go to the hi-res modes on the camera.
I have also noticed the issue of obnoxious film grain, dust and scratches particularly when comparing digital camera scans to actual scanner scans (epson v850). The epson scans often appear sharper, but the dust and scratches stand out like neon signs. You have no choice but to retouch them if you want to enjoy/display the image. With the film image, you can see them, but they are more muted and minor negative damage can just be ignored. Saves a lot of time.
I gave up using my Epson scanner mostly because of the dust issue and I found the scanning software was adding sharpness to my negatives… Which was driving me crazy
Okay, a bit (2 years) late, but here's my guess (from my own experience and some math): 24 MP is fine for 135 film, >40 MP for medium format, >60 MP for large format (4x5). Technically, 24 MP is enough for nearly every use of the scanned image, even with 4x5, but sometimes one want's to retain all the information in the negative (including grain structure), and then MP count matters ;-). Now I continue to watch … Edit: I disagree about the grain becoming harsh, but this might be my camera - an Olympus MFT with 80 MP pixel-shift hi-res mode, and it gives less "crisp" resolution on pixel-level, so that the grain actually looks quite natural. So, going significantly higher than 50 MP might "give back" the smoothness of the grain and more data to work with. As you're using Sony cameras, I made the experience that the images it gives (normal photography use) look very detailed and extremely sharp on the pixel level, and this might be the problem here (interference between the grain structure and the pixels on the sensor, and the Bayer filter/demosaicing process).
The grains on 50MP are much more detailed and defined but may appear "noisy" but that's what it really looks like when zoomed in. The 24MP just cannot capture enough details of grains so they come out as "smooth."
In 135, and with negative color film, 50MP is about the highest resolution you can achieve from a 35mm negative. On the other hand, slide film can go all the way up to 80MP, which no Digital Full Frame camera is currently able to achieve. I would have loved to see some slide film included in this comparison.
Also, keep in mind that while the 24MP vs. 50MP in 135 did not reveal more detail from the negatives, it does not mean that it’s not worth scanning at 50MP. For Digital publishing, 12MP is more than enough, but for printing, you will still be able to print a 50MP file to 20x30 at 280 dpi which not quite 300 dpi, but very close. In comparison the 24MP file printing it at 20x30 would need to be at 190dpi, and will look considerably less detailed when examining them side by side.
Maybe printing them and show the side by side comparison, and adding slides would be a good follow up video to this one.
Hi Abdon, thanks for the amazing info. Planning in the near future to do a video with printing from the digital files from all 3 cameras.
don’t confuse grain for visual data. Scanning a slide above 12MP will resolve the grain better but it won’t extract any more subject data. Edges won’t be any more sharp.
Many modern digital cameras offer high resolution mode. Film scanning is actually a perfect application for this mode.
This is very helpful. I am struggling to buy a scanner (Epson v850 pro) or to use my camera Sony A7iii to do it. Since my Sony A7iii has 24.2 MP, I think it is more than sufficient to do the job!
Extremely helpful info for my project - I have piles of old 35MM B&W negatives that might be decent digital images someday. Thanks!
I reckon anything more then 16 megapixels will only bring up more film grain and noise (35mm Film) this also depends on the film obviously the quality and speed of film will give different results, I found Fuji Reala really good with low grain. I do a lot of scanning and 12-16megapixels is heaps. also when using a digital camera to scan negs, u have to factor in the lenses centre vs edge sharpens, cause unlike a CCD Scanner which scans and moves across the film, a digital camera relies on the lenses resolution to capture in one take, & as we know lenses vary in sharpens from centre to the edges, so What I do is use a canon 30 megapixel full frame sensor and macro lens on approx f11 and i crop into the centre of the image so I'm not using the outer edges of the lens which are not going to be as sharp, generally I crop in till I get about a 15 megapixel image, which is perfect for 35mm, but importantly my final image will be sharp from edge to edge as I am only using the centre of the lens (the best part of the lens)
Really useful information, thank your for the experiment you conducted. I will tune in to your future videos!
Thanks Alan and welcome to the channel.
wow great video. I just bought a Canon 5D mkii for scanning and was worried that it only has 21MP. Now i feel alot better, thank you!
Yes, Darryl I hate you for it but I think I probably have some good shots of you from Snake Alley and I might incorporate them in a future video. 😜
It's still weird to see you using Sony cameras. I bet you already have the new 14/1.8 mm on order.
BTW... The new studio set up looks great!
CRAP ! I forgot you had some photos of me 😂😱😵 Switch to Sony about 6 weeks ago, can not believe how good the AF is on the cameras. Will only get the 14mm when we catch up in Taiwan next, so we can do YT videos together 🤣🤣🤣 You still in Mex? Hope your keeping safe
@@DarrylCarey Hell yeah, I hope we'll be able to do that asap!
I'll be back home tomorrow. México was amazing and I'll probably be back at some point in the future. 🙏
Thanks much, I was worried my older 24MP wouldn't hold up to archiving decades of slides and film.
Thanks for the comparison!! Really good info to know.
Thanks 😀
Keep in mind he is using full frame cameras. With a 24MP APSC camera I get a lot more detail on 6x7 by stitching 4 images together. 24MP apsc is just not enough for one shot on negs that big.
This is actually a very good and informative video. I appreciate your efforts here. Though, I am a bit surprised the 645 MF image didn't show better results past 24MP. This would mean that 645 doesn't perform any better than the 35mm if "camera scanning" is a part of the workflow. My mind goes immediately to the LP/mm resolving capacity of the Sony lens (as well as differences in lens resolving capacity of the original film cameras) being a limiting factor. I have always assumed that 35mm film would hit the wall around 20MP, 645 around 50MP, and 6x7 around 80MP, when scanned with a dedicated film scanner like a Nikon 9000ED--even higher with drum scans. But we all know what happens when making assumptions! 😂
As an aside on lens resolving capacity, I know that Mamiya/Phase One had "excellent lenses" back in the day when a Leaf 22MP back was all the rage. But though the years, as the MP capacity of their backs went past 40, then to 80, then to 100, and now 150MP, the glass up front has simply struggled to keep up. This has forced them to constantly upgrade their lenses to ever more expensive versions. I think they have just released their 4th generation 80mm prime Schneider at this point--$6,000 USD for anyone interested. The same thing has to be happening to Sony/Nikon/Canon as they are increasing the megapixels on their models. Lenses with higher resolution are usually larger, heavier (more metal less plastic), and tighter per copy QC at the factory. All this makes them more and more expensive.
Back to the subject at hand. It may be that a 90mm lens that was good for 12 & 24MP cameras in 2017 has hit its own wall somewhere along the way to 50MP in 2021. Just a thought.
Thank you Darryl. Very informative video indeed. Love your work!
Thanks Paul
I enjoyed it. Thank you. Did almost the same experiment before seeing this video. Compared my Canon 6D vs 5DSr cameras for copying 35mm negatives while using the same 100mm/2.8 macro. Other than double the hard drive space I didn't see discernable differences.
It’s. Nice to see this video is still helping people after all this time 😁
The short answer is, it depends on the grain size of the film and whether or not your lens is actually capable of the resolution necessary to make a difference. Additionally, it depends on the inidividual sensors; some lower-resolution sensors are better than some high-resolution sensors.
You can extract more detail from medium format if it is there :) shoot something with closed aperture and fine details. And it depends what film is used and how it is developed. Fine grain modern negatives can give 50mp detail even on 645 easily. And about the grain - it depends how you sharpen the image. Sharpening is different for 12, 24 and 50mp sensors because the size of the detail is different. That is all from my experience.
I was even able to get 24 usable megapixels of detail from 35mm Ektar 100 shots.
BTW what is the film scanning device you were using, can we get a link where it is selling?
Hi Simeon, the system I'm using is the Kasier FilmCopy Vario
Great video, nice to see the different megapixels!!!
Thanks for watching Brian
In the mid to late 90's I would have my negatives scanned to the Kodak Photo CD's (Not the consumer grade Picture CD). The standard photo CD images scanned at around 6 Mega Pixels per image and the Master Photo CD's were around 25 Megapixels per image. I just started experimenting with my Canon R7 taking a few photos of negatives on my portable light table I bought way back in the 90s at Ritz Camera. I don't have a fancy set up yet. Just hand holding my camera and not quite filling the whole frame with the negative and my early results were good but not great. I hope to get a better set up in the future. But I was wondering about the mefapixels and how it affects the grain of 35mm film and 120 medium format. So I'm excited to see your results in this video
I have a couple of film scanners which have just released, should have a review on my channel soon
Hmmmm... I wonder how the gfx would do. The Sony 50MB sensor used much smaller pixel than its 24MB sibling. I also would wonder how the results would be with 67 and 69 chromes, color negs and bw negs.
I recently shot some 1960 Color realist chromes. These images were side side by side 1 inch square. (Amazing to see the 3D impact in the viewer). I shot this with a 120mm GF on the gfx 50S. What was really amazing was the ability to recover shadows in completely underexposed areas. Things that appeared to be just black turned out to have lots of detail!!
The latitude of chromes has always been about 5 stops but digital sensors may be able to rescue bad exposures from long ago!
Thanks for doing this. Very interesting !
ı have a gfx50 r and recently ı tried to scan my 35mm negatives. now ı think no more scan epson v700 :D
ı use analog m42 macro lens 1to1 ıts minimum focus distance 5 cm. ı dıdnt try my 120 mm macro gfx lens because minimum fıus distance 45 cm . gfx50 r working realy good !
Great to see this practical experiment. I expect as one moves up in film size (6x7 or even 4x5 film) the high megapixel camera would become more useful.
Not quite 'future generation' just yet after only 2 years, but this is already a really helpful archive video. :)
Thanks Darryl. 👍
The top of the line Minolta scanner Dimage 5400 from the early 2000s outputs 40MP files (or life-size 24x36mm frames at effective 5400dpi). This was considered overkill by some and this kind of resolution wasn't exactly worth it unless you were shooting film with high-resolving power with a very sharp lens on a stable platform. From your test, I definitely prefer the 50MP files showing the grain structure - those are actual silver halide crystals you are seeing. Although I agree that 24MP seems more than enough for most use cases. But if you are shooting extremely precise work on, say, slide film and with a razor-sharp lens, that's where the benefit of the high-res sensor will show through. I remember a few years ago, people used to say digital doesn't quite have the resolving power of film. Funny how things turned out.
Absolutely fantastic video!! Thanks so much for sharing your findings!
I think your conclusion is pretty spot on. 35mm film has a nominal resolution of around 18MP given its size and typical grain density. That number varies MASSIVELY depending on the film stock, but it's roughly a figure people seem to agree on these days. So any camera with around that many MP will preserve the maximum amount of detail, and even then 12MP is clearly enough unless you REALLY want to also preserve the grain pattern. Very interesting results and now that companies are making better camera scanning solutions it's great to see research like this helping inform people.
Thanks for watching, may I ask where you found my video?
@@DarrylCarey TH-cam home page recommended it. Been watching a lot of film related photography content lately. Former photographer turned video editor because of COVID
Very cool video. thanks for taking the time to look into these things. I've recently moved into film (away from digital) - but I'm curious to know how folks catalog, organise, folder keyword etc their scanned negatives into LR or Cap1. I wonder if you've made a video about that yet? I'm going to check out your playlists etc. Thanks. :)
You only scan the negatives you use, you keep your negatives in folders on the shelf.
Well done. Thanks!
What it really goes to show you is how much more information medium format film has than any of the other formats haha
Thanks to the YT algorithm I found the video I have been looking for a long time, thanks a lot!
Depends on the film stock/size, I did LOTS of tests on a high end drum scanner many years back and a LOT of 35mm film often wasn't worth scanning above 5-6k (and/or past the visible grain).....motion film stock was even lower being grainier (and running at 24 fps, giving that organic 'film' feeling) Bit depth makes a difference too, not all resolution is down to pixel count, which is why properly downsampled pixels result in better pixels (as the art world and artists drawing bigger than the end printed result has shown us for a long time and more than a while back).
Sadly nobody in NZ is doing drum scans anymore, would be nice to do a side by side of a drum scan and 24mp image
I like sharp grain! But thanks for the comparison, very helpful.
Flaw in experiment. 3200 Kodak shows nice sharp grain at higher ISO.This is what you want as it proves it is resolving more detail. When enlarging in darkroom, a grain focuser is used because when the grain is sharp, the image is sharp.. If the neg was made with best lens like Otus or APO Summicron , is higher MP copy camera required? I can visually see with plain eye on the neg APO lenses sharper. I submit the best copy lens is required and high MP.. Also when looking at one shot at 200% is by nature going to be less sharp than one at 100%.
Thanks for the feedback and yes an Otus lens would give you a sharper image than a Sony or Canon macro lenses, but sadly most people do not have the budget for an Zeiss Otus lens. Been looking at getting a set of Otus lenses for my RED cameras, but they are still selling for crazy prices.
Most flatbed scanning software talks in DPI so at 2400 DPI a 35mm scan would be about 3600 X 2400 = about 8.6 MP which is pretty adequate for most family photo type applications. The file size will depend on the bit depth, of course. Occasionally I’ll go to 3200 DPI.
Follow up.. put the baseboard, backstand, Kaiser light, lomo holders with my SL2 and a Sigma 100 2.8 macro. delightful... Tethers right into the software.. super easy
Thats a nice setup
Interesting comparison but I think you miss a point worth mentioning. Even though apparently the 50MP is no different than 24MP in terms of capturing fine detail, it does give you the opportunity to print double the size. So for someone who wants to print as big as possible from a 645 negative without post-interpolation, the 50MP scan does make a difference.
Hi Zisis, yes you are right that you can print larger from the 50mp. I have printed A2 prints from the 24mp files and the detail was very good. So I'm guessing you can print A1 or even A0 from the 50mp. The problem with printing larger prints, is you need to stand further away from the print to appreciate it so the extra detail if any will be lost.
This is just the video I needed to see ❤ thank you very much
Glad I could help!
Thanks Darryl, I sold my 50mp 5ds and now only have a 20mp R6. Looks like I’ll be just fine to scan the negatives with the 20mp R6
The resolution can be calculated using Fourier optics. The result for an ideal circular aperture, and green light with a wavelength of about 0.55 µm can be expressed conveniently in terms of the f-ratio of the lens. Thus, it turns out that at the focal plane of the camera, the resolution is very approximately 4f/3 µm, or half this length for the Nyquist-Shannon sampling limit. Solving this for f/5.6 and a 36 x 24 mm frame, we find the N-S sampling limit is 9643 x 6429 pixels, or about 62 megapixels, assuming the lens is optically perfect. It is physically impossible to do better than this.
This might be correct for digital sensor, but not for film. As what is captured on the frame is totally irrelevant in scanning. Even if you shoot with a Holga, you are still going to want to scan the image at the end use size. You don't adjust your scanning resolution based on the resolution of the lens used.
In scanning, you are not scanning some abstract information on the frame. But replicating the structure of the film, that you then print out using inkjet.
And for non print purposes, you do not benefit from extra information as none of its retained after down scaling. Its simply decreases the image quality, because now it has been scaled through an algorytm. In every case, you want to scan an image that is the end use size.
So this approach is wrong, most people do not understand what scanning is as they constantly approach it like it was a digital format.
@@Nobody-Nowhere I think you may have misunderstood: Fourier optics gives the optical resolution, regardless of how the image is recorded. This is the theoretical limit determined purely by the finite wavelength of light and physical dimensions of the optical system. You can find a detailed explanation of the theory in any standard advanced textbook on optics. There is an added complication with film versus digital. Digital sensors usually employ a Bayer filter (or similar) plus a so-called demosaicing algorithm to interpolate between pixels and reconstruct the image, while scanned film measures the film density in each recorded channel equally at every pixel, without demosaicing or interpolation, which has further implications for resolution. Nevertheless, both methods ultimately yield a digital representation of the image, where the Nyquist-Shannon sampling theorem applies.
@@cdl0 Yes, and its totally irrelevant for scanning. As when you scan a film, you are not scanning an image on the film. Its totally irrelevant what is recorder on the film.
@@cdl0 Lets say you would go by these numbers, and scan that 62megapixels. That gives you a print size of 42x27cm @ 600ppi. But what if you want to print larger? Will you simply upscale that image? Or will you scan more information from the film, to accomodate for a larger print size? Guess which one produces better results?
Its not the image you are scanning, you are replicating the structure of the film on to the paper. This is exactly why most people misunderstand scanning, as they constantly approach it from this digital mindset where there is some set amount of "data" on the film that can be digitized until there is nothing left to extract.
You are scanning film, literally. You want to resolve as much information from the film, as its needed for given print size.
@@Nobody-Nowhere Hmmmm, I can see you don't understand the theory. There _is_ a maximum amount of data that can be extracted from an image that made by a lens of a given size, regardless of how that image is recorded. Using a higher resolution sensor, or scanning a film at higher resolution cannot retrieve more information than this limit. This is true for any optical system owing to the finite wavelength of light, which you can think of as the brush used to paint the picture. Once you stand close enough to see the artist's brush strokes, standing closer will not reveal more detail in the painting. The film or sensor is only the canvas, not the brush. Thus, in your example, enlarging an image by mathematical interpolation is equivalent to scanning it at a higher resolution.
Is there a video coming out of the Negative dslr scan set-up?
If you check my channel you will see a vide I did a while back on the setup for converting negatives to digital using Negative Lab Pro
IMHO, the larger the negative the faster you reach the limit of any 'scanning' camera.. especially if you are 6X9 or large format .. as its always the same resolution irrespective of the negative size .. however I can agree that there is a theoretical max dpi for each film
It depends upon the film format. Of course you want all the megapixels of your camera, if you're "scanning" 6x9 or even 4x5 negatives because these huge negatives can have more than 100mp resolution. For 35mm I think 24mp is plenty. However while "scanning" the negatives with your camera you often have to crop the edges. That means a 40mp sensor is perfect to get 24mp "scans".
The more data you can get with your camera
when digitizing film, the better - of course!
Great video! I am curious is you've tried taking multiple images of 120 film and then stitch them together for more resolution? I'm theorizing this would help the 12 mp camera do better on medium format film.
Hey Minsan, thats a great idea. Will do some testing and if it works will do a video on it
@@DarrylCarey cool! Looking forward to it :)
How did you find the results? I am turning between 12 or 24 mox. ..Thank you!
Thanks for that experiment. Very interesting for me.
Thanks for watching Andreas, more film photography videos coming soon
What kit are you using to hold the camera and neg to do these Photo-scans? Thx
Nice video! Regarding the 12 vs 24mp comparison: seems obvious that the diiference is less with a 35mm negative (the photo with the helmet). This is logical, too: a 6x6 negative holds more information, which is why you used the format in the first place.
I did my 35mm color slide conversion with a Nikon D90 and Sigma 70mm/2.8, which is 13mp. My guess is: if you distinguish the grains of the original, you are probably maxing out the camera sensor! because all of my color and BW negative material is 35mm, I probably stick with the D90. A higher resolution will only result in larger files, not better ones.
One thing I would like some opinion on: the light source. I have used a setup with a professional copy stand (heavy, stable) and I use a flash unit (a Nikon SB400 connected to the hot shoe by a SC17 cable). All very old stuff, but it doesn’t fail. The flash output is soft (bounced, ofcourse) of constant light temperature and there is no vibration due to the very short duration. The camera is triggered using an IR remote, no me touching the camera!
Do you think this is good setup?
Very interesting !!! Thanks for the test.
I tested 24MP vs 36MP and the latter resolved more detail than the former. It seems like 36MP is the new upper limit.
Nicely done, thank you. 👍
That's a comparison on point.
Thank you 😁
Excellent demonstration. Any wonder why the D6 is only 20.5 megapixels. I bought a Z6 with 24.5 megapixels because the added cost to buying a more costly higher megapixel camera just isn’t worth the investment.
The Nikon D6, was my last pro Nikon camera I owned. I shot IRONMAN races around the world and nobody ever asked for larger images, 20 - 24mp is the sweet spot for me …
Great stuff, very informative.
Thanks Scott, hope things are going well for you and the family.
Interesting exercise, however I am not totally convinced that is anywhere extensive enough to draw conclusions; what I mean is that there are too many variables that aren't accounted for. First of all different negatives gives different resolutions, all the negatives used are 100 ISO and above, what about 64,50,25 ISO, the lenses used have an influence, different optical linear resolution, and the aperture value used when shooting also have an impact; there might be little to no difference between 24 and 50 MP on a 160 ISO shoot at f4 on a cheapish optic whilst there might be quite a bit on a 25 ISO shot at f8 on a Zeiss, and this differences will undoubtedly be more noticeable once you go to medium and large format negatives.
If we look at the face value and just consider the linear resolution of a negative a Kodachrome 25 ISO 35mm is around 8MP, however when scanning you need at least 4 times that, this is to have enough square pixel for each roundish dot on the negative and factor in the Bayer configuration. A traditional Drum scan at 5-6000 DPI for a 35mm is around 50MP.
I haven;t used negatives/transparencies since Kodak stopped Kodachrome, but if I well remember in the old days the best way to scan for a pleasing grain rendition was to use the highest resolution possible.
Darryl, thanks for the video. My scanning workflow is like yours and saves me a lot of time while producing what I think are great results; I'm using the a7R3 through the 2.8/90mm Macro G and processing with NLP for both 135 and 120 film. I haven't found confirmation, but presume the A1 does not have an anti aliasing filter like the other high MP Sony cameras; do you think this is a factor in your results? Also, did you try the medium 21MP and small 12MP file size options on the A1? Thanks
Not sure about the different options on the Sony A1, need to look into that. I thought it was a good idea to test 3 different cameras, so people with 12mp or 24mp will have a good idea of the results they will get from their camera. Was looking at a A7R3 but went with the A7c because of the weight and size.
Very interesting comparison and a nicely done video, I didn't expect the limit to be 24mp for the medium format. I was wondering what will happen with sheet films using the same cameras. If you shoot any please make a video.
I’m looking at doing a followup video with a 24mp sensor and some different lenses … to see if this makes a difference
Please, kindly compare scanned and photographed film images side-by-side! Are there diffrences in resolution, color depth, visual appearence, noise etc.?
Fantastic video - I had wondered about this for a while before finding your channel - have subscribed
This pixel peeping is fun but, the average photo enthusiast scanning his silver negatives is going to print them on paper. Beyond roughly a 12 mpx scan you are whistling in the wind because you are getting 300 dpi on the printed picture no matter how many zillion pixels of information there is/was on the silver negative, 300 ppi (pixels per inch) is all you get.. If you are going to the trouble and expense of making an analogue silver gelatin negative you have to print it as an analogue silver gelatin print to appreciate the image.
And if you want Smooth, sharp, detailed images, why are you shooting a coarse (high speed) film in the first place?
The final point I want to make is why are you digitizing your film? AI will make a fantastic image that will be more striking than anything you or I can create.
Might want to consider the resolving power of the lens used on the 50mp since glass also has limitations with regards to how much detail is transmitted if the glass was designed in a time period where the peak mp of cameras was around 24 you will likely start seeing glass imperfections the higher you go on resolving power with the sensor.
Depends what you want to achieve: 11x14" or large format 36x48" prints ? Defined grain is fine if you want sharp chemical/analogue grain. Also dynamic range is important - newer sensors have higher DR - helpful esp with compensation developers eg divided D-23, some pyro etc - revealing all the shadow and highlight detail. Each photographer has different requirements - do your own tests with your personality developed negs and several hired cameras eg GFX100 etc. Also, you can save up your negs, hire a GFX100 or Phase One and capture negs for a day! :)
I've done 24mp but used pixel shift resolution. So it's basically free of noise etc. artifacts. I should probably compare it to standard shooting without pixel shift since processing those is a real pain.
Hi Janne, I forgot about pixel shift. I have used that on Panasonic cameras, works really for shooting negatives. What camera are you using?
@@DarrylCarey Pentax k-70. It has bit different type of pixel shift. The image dimensions stay at 4000x6000 but the detail level is enhanced.
@@JanneRanta Never tried a Pentax digital camera, will have to see if I can get my hands on one to test.
@@DarrylCarey Word of warning if you want to test the pixel shift. Afaik, there is only 2 softwares that process the files properly. One is pentax DCU5 and the other is rawtherapee. DCU5 is almost impossible to even install, and if you get it to work you are greeted with the worst image editing program I know. It is slow, uses proparly just single core and no gpu acceleration. It's UI is like it was made in the 90's. Rawtherapee on the other hand is confusing as hell and I gave up on that. There is a free software the converts the pixel shift pictures to dng's but it does only automatic processing and doesn't let you choose any sharpening / noise reduction etc.
What happens if you use 24 mp sensor but use it with sensor shift high res to make it 96 mp?
I think this test i am gonna have to do myself, because i have a camera that can do this and i recently started to mess around with film. My work table will of course be much more simple and cheaper and for the lens i will find some vintage adapted macro capable.
It depends on the film grain, which is related to ISO. For ISO 100 I'd definitely use the 50 Mpixel. Did once a scanner test on medium format, not a best film, not the best objective and a "semi-pro" scanner. Started at 125 Mpixel, at 75 Mpixel many (very) visible details, like small flowers in the far field vanished in a continuum. So I can tell that medium format requires something like 100 Mpixels average, maybe even more for high definition shots.
I think the grain in the 3200 speed film wasn't that bad on the 50mp. It was just a lot sharper than the 24. The man's eyes were soft on the 24 while quite sharp in the 50mp image
Yes you are right, for me the 24mp looks better.
I did a side by side grain-peeping comparison of the files you uploaded and found that the "A7c 6.jpg" 24MP scan seems to have had some sort of camera shake/focus issue? (The area to the top left of the man's face) It's even less sharp than the 12MP A7S II. Do you think this might've coincidentally contributed to the aesthetically pleasing/soft quality of the grains as you mentioned at 10:30?
Cool video, looking forward to see how my 16MP Oly will fare with 120, 110, 35mm and old diapositives when my Kaiser parts arrive.
Great video, this supports my own findings - thanks for that 😊. I also found there is almost no difference between 24mpx APS-C and 24mpx Full Frame.
At lower ISO the sensors would be the same, only when you go above 1600 iso the APSC sensor will show more grain.
i like the way you say twelve
I think that a lot depends on the capability of specific lens to support high megapixel camera. For example, if this macro lens on 50mpix camera have max resolution about 32mpix, than of cause will be hard to see the difference between images made at 24mpix resolution. Second, it is important to have high resolution lens on film camera. Old lenses frequently not delivering more than 24mpix on 35mm camera. Third argument - for medium format film, if old lens has good resolution and we compare image scanned by 50 mpix Sony camera with an image scanned by 50 mpix Fujifilm GFX camers. The same image from GFX camera will looks better because optically it is easier to receive high resolution image on bigger sensor.
Thank you. I was very looking for this video that how much detail can get from a 35mm film using higherpixel cameras. . Appreciate.
Great comparison video. Less is more! 😄
Thanks John, yes less is always more 😉