Epson 4990 vs Nikon D850?

[Gabe]

Who's going to spend $6000+ for this camera plus lens just to scan with? You and maybe ten other people.

It won't be that much forever.

Besides, the A7RIV will give a not too dissimilar result for 1/3rd of the cost right now.

[Sal Santamaura]

Who's going to spend $6000+ for this camera plus lens just to scan with? You and maybe ten other people.

It's risky to extrapolate from one's own ability/attitude with respect to disbursement of financial resources, discretionary or otherwise, to others' positions in those respects. I'd wager your estimate is off by quite a few orders of magnitude.

[stephenmcateer]

I used to use a D810 and stitching for my digitisation, and got good results (~ 90 megapixels for 4x5). These days I use a Sony A7RIV and pixel-shift capture of the whole frame, which is faster and yields even better quality (~ 180 megapixels).

Have a look at my Flickr for examples. I upload the originals at quite a high resolution, so you can get a decent idea of what is possible:

https://flickr.com/photos/atomstitcher/

Here is a 51 megapixel resized version of a sheet of 4x5 TMX for example (captured with the A7RIV): https://live.staticflickr.com/65535/...8fa2b67f_o.jpg

Gabe — that example image is outstanding. It makes the results I get from my flatbed look archaic, which I suppose they are. I've ordered a Kaiser LED lightbox and am looking forward to using it. Pixel shift is something I have to look into. Thanks.

[Gabe]

Gabe — that example image is outstanding. It makes the results I get from my flatbed look archaic, which I suppose they are. I've ordered a Kaiser LED lightbox and am looking forward to using it. Pixel shift is something I have to look into. Thanks.

Thanks. I use a Kaiser Slimlite Plano too, and it does a decent job. Sure, a 99 CRI panel would be preferable for colour work, but to get one in that size with good uniformity gets very expensive. Best of luck with the new method anyway.

[sanking]

Who's going to spend $6000+ for this camera plus lens just to scan with? You and maybe ten other people.

Believe I made it clear that I bought the Fuji GFX 100s for general photography, and for scanning, "NOT JUST TO SCAN WITH."

I share my experience precisely because the GFX 100S is a very expensive camera, and not many people have the experience to make a valid comparison with other equipment.

Sandy

[Sert]

Who's going to spend $6000+ for this camera plus lens just to scan with? You and maybe ten other people.

I don't think anybody purchased GFX100s just for scanning.

[stephenmcateer]

UPDATE August 24th 2022: I got a Kaiser Plano a few days ago. I've only digitized one neg with it so far, from my linhof Technorama 612. See it here on Flickr if you fancy: https://flic.kr/p/2nF8QbQ.

The inital image was not 100% sharp, so I used Smart Sharpening in Photoshop which seems to have worked. [I made some notes below the image on Flickr.]

I'm reasonably happy with this one image so far. I think a copy stand would help reduce vibration but I don't want to spend the money at this point.

[Gabe]

UPDATE August 24th 2022: I got a Kaiser Plano a few days ago. I've only digitized one neg with it so far, from my linhof Technorama 612. See it here on Flickr if you fancy: https://flic.kr/p/2nF8QbQ.

The inital image was not 100% sharp, so I used Smart Sharpening in Photoshop which seems to have worked. [I made some notes below the image on Flickr.]

I'm reasonably happy with this one image so far. I think a copy stand would help reduce vibration but I don't want to spend the money at this point.

Congrats on getting the Kaiser. It's a good panel.

A few comments:

- Whilst evaluating your setup I would advise using images with plenty of fine detail (preferably across the entire frame), as well as a good tonal range, to more easily judge the results critically from a technical perspective.

- It looks to me like focus was off on this one, as the frame edge and hairs/fluff are not sharp. If your lens has autofocus, using that to focus on fine detail in the frame can help immensely in achieving critical focus (might be difficult on this image though).

- Parallelism of the film/sensor is absolutely critical for good results (especially with panoramic formats). One of the easiest and most accurate ways of doing this is to use a small mirror placed on top of the film holder you use (assuming it has a level surface), or on the light panel, then in live view with the lens stopped down you can check if the aperture is centered or not in a magnified live view image (you may have to bounce a light off the mirror into the lens in order to see this). Don't trust bubble levels for this purpose as they are generally not accurate enough.

- I would definitely advise getting a good copy stand for stability and ease of adjustment.

- You are using far too small an aperture, and the resulting diffraction will be robbing a lot of detail (and also introducing possible vibrations from long exposure times, as you noted). I would advise not going beyond f11, preferably f8, for best results. If your film/sensor are parallel you will still have sufficient DoF.

- Make sure you use a cable release or timer function to avoid vibrations (you may have done this already).

[koraks]

To get the best scans when using a digital camera, you'll need a solid copy stand (many use re-purposed enlargers), a light source with very high CRI rating (99 is good)

Given the discontinuous nature of both dye transmission spectra and CCD/CMOS element sensitivities, how did you arrive at the high CRI requirement?

[stephenmcateer]

Congrats on getting the Kaiser. It's a good panel.

A few comments:

- Whilst evaluating your setup I would advise using images with plenty of fine detail (preferably across the entire frame), as well as a good tonal range, to more easily judge the results critically from a technical perspective.

- It looks to me like focus was off on this one, as the frame edge and hairs/fluff are not sharp. If your lens has autofocus, using that to focus on fine detail in the frame can help immensely in achieving critical focus (might be difficult on this image though).

- Parallelism of the film/sensor is absolutely critical for good results (especially with panoramic formats). One of the easiest and most accurate ways of doing this is to use a small mirror placed on top of the film holder you use (assuming it has a level surface), or on the light panel, then in live view with the lens stopped down you can check if the aperture is centered or not in a magnified live view image (you may have to bounce a light off the mirror into the lens in order to see this). Don't trust bubble levels for this purpose as they are generally not accurate enough.

- I would definitely advise getting a good copy stand for stability and ease of adjustment.

- You are using far too small an aperture, and the resulting diffraction will be robbing a lot of detail (and also introducing possible vibrations from long exposure times, as you noted). I would advise not going beyond f11, preferably f8, for best results. If your film/sensor are parallel you will still have sufficient DoF.

- Make sure you use a cable release or timer function to avoid vibrations (you may have done this already).

Thanks Gabe. I've made a note of your comments and will apply them next time I'm digitising. I think, as you say, a copy stand would help a lot. Autofocus is probably a good idea too — I used manual focus as the YouTube tutorial I watched advised that. Anyway, I'll report back here with any improvements next time.