Scanner Comparison 2019 [Epson Flatbed | Eversmart Flatbed | Drum Scanners]

[Pere Casals]

One thing not mentioned here, is a wet scan on the v700. It does make a huge difference...enough that I will use my v700/wet over my dedicated 35mm scanner with 35mm film.

Peter, it depends on what dedicated roll film scanner you have, but any cheap Plustek 8200 (or 7600) outresolves the V700 by some margin, here it says it resolves group 5.5 : https://www.filmscanner.info/en/Epso...onV750Pro.html, the test made by Pali shows 5.6 and 5.4, verticals bars and Horizontal bars, while a 35mm Plustek makes 6.0 or 6.1.



Pali also has in his web site a side by side Dry vs Wet with a V800: http://www.analogfilm.camera/2017/11...-or-dry-mount/ but is you apply a bit of sharpening to the best point for each samples you will find equal results.

Wet scanning has an remarkable effect with the V700 when film is curled, the V700 has no autofocus so if film is not in the right height then it has a performance loss, this has been measured by several people with similar results, this graph shows what happens is film is not in place:


The highest curve corresponds to Hor axis, it's more sensitive to height, while in the Vert direction we have a flat in the top (possibly) provocated by vibrations in the carriage displacement or similar.

There are other easy ways to place the film flat in the right position without wet mounting... or making the film to be flat.

Well, yes it would be interesting to also scan the samples wet for the V700, but what's for LF my guess is that it will make little difference, because LF lenses/negatives do extinguish contrast totally by some 2500dpi or 50lp/mm in practical shooting conditions, some LF lenses may deliver 80lp/mm lab, but real photography is not a lab with a flat chart.

Let's see how this test goes, IMHO Pali is to make a very good job by taking the samples. Let's see if once in a lifetime we'll be able to have a serene and fruitful debate about dpi , we can get useful information about when drum scanning makes a difference, and at what print size.

[Pali K]

I think that Supreme might need a bit of cleaning up of dust. Vertical lines in USAF 1951 target get messed up well before horizontal lines. This is how it should look like.

Attachment 186459

You're right. I think I need to clean the optics since it is drastically softer than my initial tests when I first got the scanner. This test was done when I first got the supreme. This just highlights one another aspect of owning these higher-end flatbeds that you need to calibrated them once in a while and keep things clean. This may sound like extra work but to me, it's a benefit that you have the ability to clean and calibrate things to get the best performance.

[Pali K]

...

Let's see how this test goes, IMHO Pali is to make a very good job by taking the samples. Let's see if once in a lifetime we'll be able to have a serene and fruitful debate about dpi , we can get useful information about when drum scanning makes a difference, and at what print size.

Thanks Pere! I think I can definitely come back and do some of these tests with wet mounted film on the Flatbeds once I get through the initial tests. It would be helpful to see the performance improvement achievable with common solutions such as wet mounting and using adjustable holders. I am currently only using an adjustable holder for Epson and everything on the flatbeds will be dry mount scans for the 1st round of tests.

[Pere Casals]

hmm, the supreme is close to what a Hassie does in 35mm, but this one can stitch all the bed surface with this quality, while the hassie has an increasing loss with the format size increase...

[Pere Casals]

Thanks Pere! I think I can definitely come back and do some of these tests with wet mounted film on the Flatbeds once I get through the initial tests. It would be helpful to see the performance improvement achievable with common solutions such as wet mounting and using adjustable holders. I am currently only using an adjustable holder and everything on the flatbeds will be dry mount scans.

I agree, if not painful to do, the wet mount would clear the potential effect of any film curling in the focus, so results would be more consistent, and at the same time we find what the wet does...

Suggestion from Peter Carter is not bad... nothing wrong to see what's the best a machine can do.

[ropel]

This is excellent work. Following along with great interest.

Maybe a suggestion: how about including comparison to dslr 'scans'? Clearly this initiative is primarily a scanner comparison, yet I've seen proper results from camera diplication on this forum. Just curious. :-)

Thanks for sharing all your results Pali!

Sent from my SM-T590 using Tapatalk

[Juergen Cullmann]

Pali, thank you very much for your resolving tests with your Eversmart Supreme. I got the same result with my Supreme than you with your actual test ( about 5100 dpi). I am surprised what resolution will be possible when the scanner got cleaned. Have you done this before? I know, there is a cleaning manual in the file section but I am afraid I could damage the scanner.

Especially the ccd-row can easy be damaged during cleaning with the wrong wipes or wrong detergent/alcohol.

Could you give me an advice how you would
do this procedure? Thanks, Jürgen

[Peter De Smidt]

Maybe a suggestion: how about including comparison to dslr 'scans'? Clearly this initiative is primarily a scanner comparison, yet I've seen proper results from camera diplication on this forum. Just curious. :-)

I spent a bit of time on dslr scanner development. In the threads on this in the DIY section, there are scans of USAF1951 targets and Stouffer step wedges. These were done many years ago now, and I don't remember all of the particulars. For instance, some sharpening might've been used. Anyway, with a D600 with a Rodagon D 75mm @f/4 and 1x magnification about 3000 dpi were achieved. With a D810 and a Magnagon, about 4000 dpi was achieved. Density tests showed about a one-stop advantage over a then current Epson flatbed. Caveats. At 1x, scanning 35mm film is easy, with no combining of exposures involved. As film size goes up, so does the number of exposures that need to be combined. With 4x5 at 1x, scanning is only practical with an automated positioning system. As Ted has pointed out, what's really important is MTF. My feeling is that the dslr systems did a better job than the consumer flatbeds in that regard, but I don't have objective data. Currently, my dslr-scanner is disasembled, and so no updates will happen in the near future. My guess is that newer cameras, ones with higher resolution and no-vibration shutters, would produce even better results.

[sanking]

I spent a bit of time on dslr scanner development. In the threads on this in the DIY section, there are scans of USAF1951 targets and Stouffer step wedges. These were done many years ago now, and I don't remember all of the particulars. For instance, some sharpening might've been used. Anyway, with a D600 with a Rodagon D 75mm @f/4 and 1x magnification about 3000 dpi were achieved. With a D810 and a Magnagon, about 4000 dpi was achieved. Density tests showed about a one-stop advantage over a then current Epson flatbed. Caveats. At 1x, scanning 35mm film is easy, with no combining of exposures involved. As film size goes up, so does the number of exposures that need to be combined. With 4x5 at 1x, scanning is only practical with an automated positioning system. As Ted has pointed out, what's really important is MTF. My feeling is that the dslr systems did a better job than the consumer flatbeds in that regard, but I don't have objective data. Currently, my dslr-scanner is disasembled, and so no updates will happen in the near future. My guess is that newer cameras, ones with higher resolution and no-vibration shutters, would produce even better results.

I agree with Peter in that a dedicated high resolution digital camera set-up is capable of giving far better results than the Epson flatbeds in both resolution and dynamic range, and in fact better resolution than many high end flatbed and drum scanners. I set up a system with a Sony a7r 11, using the same Rodagon D 75mm f/4 lens that Peter mentions, and got about 4500 ppi in my tests of the system with the lens set to f/8. For the tests I used a chrome on glass target of the USAF1951 target, and focused with a Noritsu Koku Focus Test Chart.

Sandy

[Pali K]

Thank you Ropel, Peter and Sandy. And the additional benefit of a DSLR setup is that it will only get better as technology continues to improve. I don't shoot any digital and my last DSLR was a 5D MKII which I don't think would be much helpful to include. I don't even have a proper lens for it to do justice even for a 10 year old camera.

Pali