"Sigma SD1 Merrill not good for scanning negatives/slides"

"Sigma SD1 Merrill is not good for scanning negatives/slides"

... that's what somehwere, someone said.

"A 36 megapixel full-frame (Bayer-) camera is sufficient to capture all the details of a medium format slide"

... that's what someone else, somewhere, said.

I tend to disagree. Take a look:

http://drtebi.com/dump/dslr-scans/occupied-house-berlin-2x3-dslr-scan-foveon.jpg

http://drtebi.com/dump/dslr-scans/sf-victorian-2x2-dslr-scan-foveon.jpg

These are stitched scans, 2×3, and 2×2 respectively; taken with the Sigma SD1 Merrill and the Sigma 70mm Macro lens.

These are just some early test results. My workflow is not yet optimized, colors could be better etc.

What do you think?

I scanned all my B&W 35mm negs with the sdQH and the Olympus 80mm macro lens. That lens was designed for use in slide duplication.

Results seem fine to me. I saved DNG files to speed up processing.

When I tried a few 35mm colour slides, I wasn't so happy with the colour. The Sony NEX-5N seemed to give better colour, and it has the great advantage of a built-in HDR mode. 35mm slides are very high in contrast, and you really need HDR.

I haven't tried stitching any medium format negs. Something to try later.

I think the're both a bit small (low resolution), and if you scanned them with a good flatbed scanner you'd probably get better results than that.

I think these are very good, detailed and I like the colour in the first one. The main reason I can think of that some would say the SD1M isn't good for scanning slides is ease of operation - focussing and no live view etc.

DrTebi wrote:

digi2ap wrote:

I think these are very good, detailed and I like the colour in the first one. The main reason I can think of that some would say the SD1M isn't good for scanning slides is ease of operation - focussing and no live view etc.

I think I must have confused everyone with the subject line of this post. If you read the body of my post, you will see that I tend to disagree...

In other words, I think the Sigma SD1 Merrill is ideal for DSLR scanning. At least on a technical level—the Foveon sensor really brings out the best of film.

I too was afraid that focusing and not having live view could be an issue. Now, after a number of test scans (including some stitched ones), I am happy to report that the 70mm Macro lens auto focus is spot on. So if that works... live view is not really needed anymore.

The SD-1 is slow, and It's a bit cumbersome to convert the files with Sigma Photo Pro, switching to another editor for further post processing etc., but in the end I like the results much better than what I got with the Nikon D810 or Pentax K-1, or even my Minolta Dimage Multi Pro film scanner.

That seems likely.

But putting twenty thousand or more files through SPP is simply not practical. One reason I went for the sdQH was the DNG output, which is fine for monochrome. Also the luminance resolution is probably as high as that of any Bayer FF camera, except perhaps the Canon 50 Megapixel model.

And the sdQH is within my budget. The alternative was the Sony A7r2, at more than twice the price. I've seen some excellent results posted by people using that camera with the Sony 90mm macro lens. That lens again is too expensive.

But I wouldn't use the sdQH for 35mm colour slides, as you have to shoot in SFD mode to deal with the contrast. This would be OK for perhaps a couple of hundred slides, but not for thousands.

Negatives have a lower contrast and single exposures are fine.

Resolution-wise, as I mentioned, it depends on your goals. If every last bit of detail is your goal, then stitching is necessary with the SD-1. The Pentax and Nikon have a little advantage here due to the higher resolution, but these will not give the same clarity as the SD-1 with its Foveon sensor.

Are you stitching 35mm? I couldn't see any improvement over single images with the sdQH. Medium format film is a different matter and certainly needs stitching to get crisp grain on silver negs.

Your mileage may vary, but I am quite satisfied with this setup, even if it takes a bit of effort to get great results.

FDecker wrote:

D Cox wrote:

Negatives have a lower contrast (...than slides) and single exposures are fine.

Isn't that the other way round? (note: I added the text in bold to give the complete context after snipping away most of the text).

Negatives have a max. DR of 12 (density 3.6) and slides are around 8 (density 2.4).

Maybe I am wrong....

Slide films are designed to have greater contrast (a steeper characteristic curve) than negatives or larger format positives. This is to counteract ambient light when they are projected.

You may be confusing contrast with density range.

https://www.kodak.com/uploadedfiles/motion/US_plugins_acrobat_en_motion_newsletters_filmEss_06_Characteristics_of_Film.pdf

DrTebi wrote:

Are you stitching 35mm? I couldn't see any improvement over single images with the sdQH. Medium format film is a different matter and certainly needs stitching to get crisp grain on silver negs.

I have tried some stitching of 35mm with the Pentax K-1, but couldn't really see any improvement. This may have been due to the fact that I had to use extension tubes, and mine don't have electronic contacts—hence I had to focus manually, which is difficult even with the live view.

I still need to try some 35mm stitching with the Sigma SD-1M.

By the way, the "Pixel Shift" feature of the Pentax K-1 can also come in handy for scanning negatives or slides. I could definitely observe more detail in a single-shot medium format slide, but it also introduced some weird artifacts, which I could not see in the same image when stitching.

By the way, what strategy have you used for converting negatives to positives? Normally "ColorPerfect" gives me some quite acceptable results, but for some reason it does not with images from the SD-1.

I loaded a few hundred B&W negs (DNG files) at a time into Adobe Camera Raw, selected all, and inverted them with the Curves tool.

I have only a handful of colour negs, none of much importance, so I did them on an Epson V750 scanner, in order to use the conversion software provided. However, I have also done it "by hand" using the Photoshop Curves tool -- "Auto" in this fixes the orange cast and gets you somewhere near the right colours.

I wouldn't use the scanner for important images, at least not 35mm.

atom14 wrote:

This could be the start of something old- like the old pixels vs DPI controversy.

I don't know if the Epson V750 scanner is any good, but on this page are its resolution specs: https://epson.com/Clearance-Center/Scanners/Epson-Perfection-V750-Pro/p/B11B178061

It does attempt to marry up pixels with DPI somewhere there.

The shots look "good enough to me", though I couldn't say how they compare with MF film detail. - It doesn't matter much as long as you can see the detail you want.

atom14.

Quick quote:

The Epson Perfection V800 Photo yields an effective resolution of 2300 ppi if the template is scanned with 4800 ppi. Using the professional scanning software SilverFast Ai Studio, this value is increased to an effective 2600 ppi.

from https://www.filmscanner.info/en/EpsonPerfectionV800Photo.html

Effect of Sony's HDR mode.

(SFD would be similar.)

Scottelly wrote:

I think the're both a bit small (low resolution), and if you scanned them with a good flatbed scanner you'd probably get better results than that.

I beg to differ. These are quite high-resolution "scans," even if you factor in the fact that the images were shot on medium-format film. And to be honest I have never seen anything from a flatbed scanner - with the possible exception of a Screen Cezanne - that came close to this.

DrTebi wrote:

What do you think?

Excellent camera, excellent work!

I enjoyed the imacon scanner we had at the university.

i tried a Nikon D800e some years ago and was disappointed. The setup wasn't good.
Did you tried much more, like stiched from 6 oder 10 shots? Want to see the single grain and color dyes

~32mb, just for the grain. should be iso 400 bw negative.

100%, scanned @ 3000pdi i think