I’m shooting 35mm and scanning with a crop-sensor DSLR on a copy stand, using a macro lens at around f/8 and a decent light source. My current body is older and only offers a few megapixels beyond what I need to fill the frame with the negative, so I’m weighing whether to move to a higher-resolution body or put that money into a better lens and a sturdier stand. My uncertainty is whether the extra pixels actually add detail with 35mm, or whether lens resolution and alignment matter far more. I also worry about vibration and mirror slap on a taller stand, since that seems to matter more than the sensor on paper. I haven’t bought anything yet and I’m going back and forth. For those scanning 35mm at home, did the body upgrade make a visible difference, or would you spend it elsewhere first?
Update after reading a bit more - Science News came up while I was looking into how sensor resolution compares with lens and setup limits for 35mm scanning. Still not sure the extra megapixels would show up in my files, so I’m leaning toward the sturdier stand and lens first.
Do you have a way of evaluating the detail that you are able to capture with your current setup? For example if you have Vlad’s Test Target you would be able to compare your results with those of others who have posted on here and on other forums, or even from Vlad himself:
That is very helpful in comparing lenses and finding the optimum aperture because generally there is only a very narrow aperture range where your system will perform at its best and it will also immediately show up any issues associated with alignment.
It will also confirm whether your copy stand is up to the job, or whether mirror slap is an issue from your DSLR. For example just as an experiment you could contrive to use flash for a few shots so that you can see the result when those problems aren’t a factor.
I will say that I think mirrorless is so much easier than a DSLR, I use a modest 24MP Fuji X-T2 and I wouldn’t go back to a DSLR. I do know however that there is more dynamic range to be had from a full frame sensor but that only becomes significant for a few ‘difficult’ negatives or slides. You need to be aware that whatever lens you use currently may not perform as well on full frame.
A sensor of 6000*4000 can deliver a 20 in wide print at 300ppi.
Maybe that’s all we need to know.
Gear match?
Using a ff lens on a crop sensor camera is not the best of ideas. FF spreads 4000 pixels over a wider distance than a crop sensor cam of nominal 24 Megapixels. Crop sensors need better lenses.
Focusing?
Get the best gear, and if you miss focus, it did not help.
AF comes in steps and will match focus only by chance
Focus peeking is better, but not as precise as focusing manually, provided the lens has a wide focus throw
Not always true, I was thinking in terms of using high quality macro lenses from the film era that were designed to go to 1:1 magnification. The ones I’ve tried (55mm Micro-Nikkor f2.8, Canon FD 50mm f3.5 Macro, Olympus OM 50mm f3.5 Macro) are all better for copying 35mm film on APS-C than they are on full frame. On APS-C (at the lower magnification of 0.67x) the corners are almost as sharp as the centre, on full frame at 1:1 the corners fall off in sharpness. This is from testing with Vlad’s Test Target mounted in a 35mm AN glass slide mount. Those lenses focused to 1:1 but were not optimised for 1:1 and 1:1 is very demanding without a specialist symmetrical lens designed for that magnification. Your experience may be different though.
Right, there are a few pearls out there and they seem to perform really well in your tests. Not knowing the OP’s gear, I gave a few general hints and all the detailed tests are out there.
Questions like “does it matter” or “is it worth” are highly individual and some want the best and are willing to shell out a few thousand for the rig, but then it’s worth going all the way. Each improvement points at the next weak element and it helps to be aware of such risks. It does not pay to add a EUR 3000 camera to a wobbly copy stand.
One of the most adaptable and sturdy rigs is @Mark_Segal 's imo. Check out its components. Good quality, high-price gear, but “high” is subjective again.
No, and even the OP seems to have doubts about the copy stand so I would sort that out before upgrading any other part of the system, but then the tests that I suggested using Vlad’s Test Target would highlight such problems. For the sake of impartiality I’d love rto recommend an affordable alternative to the ‘VTT’ but I don’t think there is one.
Mark Segal’s articles can be found on Luminous Landscape now, you are of course encouraged by the site to subscribe but you can ‘do it later’. This one shows his rig. I think even Mark might agree that each of the components in this top of the range setup might be intimidatingly expensive for someone just starting out with camera ‘scanning’ though. Actually I think your own copy stand is excellent, and need not be expensive.
Excuse me for giving a link to your own setup but I think it’s more than relevant.
I don’t understand why posts are being flagged and grayed out. This would appear to be a perfectly straightforward technical discussion that shouldn’t ignite passions. OK, that out of the way, I’ll chime-in here. Firstly, the image circle created by the lens should be considerably larger than the sensor it is meant to cover, and this regardless of the size of the sensor and its pixel dimensions. This is simply because the inner part of the imaging circle normally tends to produce higher resolution than the outer area of the circle. Some lens manufacturers document this with test data. Secondly, unfortunately, you are correct to point out how weaknesses emerge. There is a set of factors that contribute to image quality and to the extent they are all somehow determinative, the result will be as good as allowed by the weakest link, so one gets dragged into spending just a bit more here and bit more there to achieve the quality one is happy with.
Notice I use the phrase “happy with” and that neatly skirts around the issue of what’s “good”. For me, if I get something I’m not happy with it’s not good and the reverse. And of course some of these things are insidious but they don’t cost much to fix. My favorite example is micro-vibrations. To use your example, you can have a 3000 dollar camera and a 2000 dollar lens and be sabotaged by micro-vibration. e.g. you have an air-conditioner or a central heating system working in your house and it sends micro-vibrations through the property. Blow your image up to the highest magnification your camera and computer allow and you’ll see tiny pixels vibrating under these conditions. It’s enough to impair resolution because resolution is made-up of such tiny elements. Simple solution to that problem was to keep the camera’s “steady-shot” switched on (even though it is on a stand) and to place some felt and Sorbathane pads under the copy stand. Not expensive and problem solved.
And Harry - “affordable” is another one of these expressions with fluctuating meaning depending on whose wallet it is. I bought a Thorlabs resolution target because it’s etched in glass. Expensive, yes, but superb, and maybe there’s something else I didn’t buy just to keep the total budget under control. I don’t buy fancy coffees - those really add-up over time…
Mark, to address your first point, there is something amiss with this site at the moment but I’m not convinced that anyone is looking after it. There seem to be a lot of ‘bots’ attempting to post but on the other hand there seem to be a few entirely legitimate posts that are without category. I tried to reply to one and my post is awaiting approval - er, why?
We’ve spoken about Vlad’s target before, for precise quantitative analysis you need a Thorlabs glass etched target, Vlad says as much himself. However there is a great deal that one can establish using his target that is of value. He provides an extremely high resolution scan of a detail where it is possible to see where it begins to fall down, simply because of how it is made. However that is a few ‘levels’ beyond where most camera sensor/lens combinations can resolve and mathematics decrees that only sensors above 50 or 60 megapixels will come anywhere near this. The fact that the targets are around the edges of the frame and not just in the centre is crucial for determining lens performance and alignment.
In fact I may have diverted the discussion from what really matters to the OP, which I think may be what difference might be seen between a 24MP APS-C ‘scan’ of a 35mm frame from an elderly sensor and average lens and a high megapixel scan with a top notch setup and an excellent lens. As you say it revolves around what is ‘good’, so there is an element of “Zen and the Art of Camera Scanning” here.
The post above mine is a case in point, it is spam and links to what looks very much like an ‘Adult’ site, I suspect that the other bots are attempting to do the same thing. I couldn’t seem to flag it in any useful way. What is happening to this site?
On the targets - OK, I’m using a 61 MP camera with a Schneider apo-digitar macro lens, so yes, I’m up in the resolution area where the transition from Vlad’s target to Thorlab is visible, and for anyone using a lesser resolution set-up I agree Vlad’s should be good. What I like about his is the edge to edge capability which one does not get with Thorlabs. For that I need to move it around and make several exposures - less convenient.
Now on the meat of the matter - the difference between using a 24MP APS-C sensor and an average lens versus a higher-end set of equipment. It is indeed visible. When I was transitioning from scanner use to camera use I started with a 24MP APS-C system and a Sony-Zeiss macro lens. It was not as good at resolving fine detail as the Minolta scanner I was using before. Then I tried a Phase One 40MP camera I had lying around (literally, because using it was a PITA) and I tried camera scanning with that and wasn’t happy with the optical quality of the reasonably priced lens I could put on that particular model apart from other awkward aspects of using it. Then I graduated to the Sony a7R4 with bellows and Schneider macro lens and haven’t looked back. It really “does the job”. So based on my own experience I can say there are “break-points” where quality differences as a function of equipment are visible.
Thanks Mark, very useful information. I have created a separate post about the ‘bots’. I hope they can be stopped or else the site may quickly become unusable.
