New Nikon ES-2 Film Digitizing Adapter & Slide Scanner Announced

[Corran]

Sorry but that's exactly what it is...just like manipulating the color channels on your enlarger head. There is no "black magic" involved. Scanning with a high-end flatbed as a positive and adjusting the color channels + color curves is how I scan my film, but the same would apply to a DSLR "scan."

The problem with "profiles" or anything that purports to remove the orange mask is that it's imperfect - due to film age, development, etc. Also, there are issues with regard to density, DMax and range, etc. with regard to scanners (or DSLRs) that can greatly affect the results of scans.

By the way, there are some negative films available without an orange mask. I really like Rollei's CN200 film.

I'm with you that each film has a "look" to it. Obviously that's true, otherwise we wouldn't have the variety of CN films available as they would all be identical. But the color balance is still subjective in that you can change it to your whim and it's not "visible" on the light table, as compared to slides.

This is a 6x6 Fuji 160S scan from my Screen Cezanne, scanned as positive, inverted, and balanced with individual RGB channel settings and curves:

 

[Huss]

I'm sorry, didn't want to mock your image. As photographers we are inclined to have our own interpretation. ..

i'd like to see skin tones, reds and real greens from dslr neg scan

We're cool. 🙂

Here's some skin tones, and I'm pretty sure you can also find an assortment of colours to critique!
😉

Same 65 year old lens, same Fuji C200 that cost me $1.49 a roll

GonzaloS-1 by desmolicious, on Flickr

GonzaloTileS-4 by desmolicious, on Flickr

 

[Betin]

Sorry but that's exactly what it is...just like manipulating the color channels on your enlarger head. There is no "black magic" involved. Scanning with a high-end flatbed as a positive and adjusting the color channels + color curves is how I scan my film, but the same would apply to a DSLR "scan."

Sorry to disappoint , but it is not true. C41 color film is a sandwich of 5 layers - cyan, magenta and yellow with two masking layers in between. Each color layer has different density and sensitivity. Masking layers are non-uniform in density. Their density is dependent on upper layer density in shadows and lights. There is no straightforward approach to properly remove mask and adjust density.

Try the same scan in proper negative mode and compare.

 

[Corran]

The scanner's light doesn't change depending on the mode. And my scanner does not have profiles for newer films and older profiles look like garbage. I get excellent scans using positive mode and inverting, and setting W/B points as appropriate to balance the color (as shown).

Regardless of the uniformity of the density, even your enlarger head can't compensate for that at the negative level. The light is still uniform coming out of the head. You can argue that the color balance is done with the light as opposed to a plain whitish light from a scanner, but you are making the argument that a scanner and DSLR (with light source) is different. They are not. The light is uniform either way - though if you want to get into the nitty-gritty stuff you could debate the efficacy of Bayer filters here though. Doesn't matter to the overall point.

The idea that a scanner is fundamentally different than a DSLR with a good light source is incorrect (except for drum scanners, but that's a different topic).

 

[Betin]

The scanner's light doesn't change depending on the mode. And my scanner does not have profiles for newer films and older profiles look like garbage. I get excellent scans using positive mode and inverting, and setting W/B points as appropriate to balance the color (as shown).

Regardless of the uniformity of the density, even your enlarger head can't compensate for that at the negative level. The light is still uniform coming out of the head. You can argue that the color balance is done with the light as opposed to a plain whitish light from a scanner, but you are making the argument that a scanner and DSLR (with light source) is different. They are not. The light is uniform either way - though if you want to get into the nitty-gritty stuff you could debate the efficacy of Bayer filters here though. Doesn't matter to the overall point.

The idea that a scanner is fundamentally different than a DSLR with a good light source is incorrect (except for drum scanners, but that's a different topic).

It is not about a light source which can be compensated for. In fact color printing is done with yellow and magenta filters in between light source and a negative. Y/M filter density is how one controls color cast.

It is not that DSLR is fundamentally worse than a scanner (or enlarger per se) what matters is what you do with a projection.

RA-4 color paper itself is optimized for negative layer density.

Epson flatbeds are not bad in terms of color. I have one myself. Generic color profile is usually good for everything.

Your image has fundamental problems. It's not that it has uniform color cast. It is that this cast is not uniform. Skin tones need more red and less green. At the same time stones at the background need less red at the same density level. This is a result of improper negative mask removal and can only be fixed with creative masking techniques in photoshop. You may compensate for the model but everything else would fall apart.

(i'll take it down soon)

 

[Corran]

Again, you are bringing up darkroom printing...which has nothing to do with the product in question. You specifically postulated that a scanner scans CN differently (or better) than a DSLR with regard to color...which is generally incorrect.

I don't even do DSLR scanning...just pointing out the fallacy regarding scanner vs. DSLR. Darkroom printing is irrelevant. By the way, don't forget that Kodak says specifically that some of their film was reformulated for scanning. I think you are also somewhat mistaken that there's any difference in end result between darkroom and scanning (assuming a competent operator in both cases!) but that's a different topic and will not debate that here.

Realistically the biggest issue with CN scanning, via scanner or DSLR, is the one doing the scanning. I have read numerous posts all over the place about "blue cast on the film" because they have no idea how to scan. I myself tend to try to get a longer dynamic range rather than a "natural" image sometimes, which tends to mess with the color and look of the image when pushing down the highlights. I have to remind myself to let the highlights delicately go to white, and get a nice spread of tones in the higher zones.

One other thing I did think of - there could potentially be some issue with a DSLR scan if one were to use really bad light for it. A big topic these days in LED lights is the CRI spec. Really bad, cheap lights have bad color rendering. I imagine that this could make the DSLR scan troublesome - but most people tend to use their flash and diffusion. So maybe it's a non-issue. I am sure the scanner manufacturers have this figured out and use good lights. For my scanner, a pair of replacement bulbs costs as much as your typical Epson V-series scanner or more! If I did scan via DSLR I would personally be making a light table and rig for the camera, and would make sure to have high-end bulbs for the light.

Another scan, from Portra, of a difficult negative (overexposed sky, falloff from extreme wide angle):

 

[Corran]

Your image has fundamental problems. It's not that it has uniform color cast. It is that this cast is not uniform. Skin tones need more red and less green. At the same time stones at the background need less red at the same density level. This is a result of improper negative mask removal and can only be fixed with creative masking techniques in photoshop.
(i'll take it down soon)

What you are talking about is simply personal opinion. This girl had much lighter and less red skin. This is the difference between having seen the actual person and then making subjective opinions on the color of said skin. The edit you posted is extremely red tinted and looks very bad.

By the way, the image in question was shot on very expired film...and had color shift which was fixed in the scan.

Let me reiterate - there is NO CHANGE in the way my scanner works in positive and negative mode, except negative mode is limited to 8-bit (bad) and can only used fixed profiles which look like ass. You're quite simply wrong here.

At this point my belief is that you have a very poor monitor or one that is way off-target in terms of calibration. I would suggest you check that.

 

[Betin]

Again, you are bringing up darkroom printing...which has nothing to do with the product in question. You specifically postulated that a scanner scans CN differently (or better) than a DSLR with regard to color...which is generally incorrect.

Not sure where did you get it from. Nowhere I stated that DSLR is incapable of good scanning results. In fact I've entered this discussion only because I'm very excited Nikon brought this feature. Hope it's done properly and not just by "adjusting WB and some curves". Knowing Nikon's history with scanners my expectations are hopefully reasonable.

btw, I'm not so much relying on monitor but more on the densitometer in photoshop. Certain colors are subjective but there are other clues. Hair cannot be blue. Skin can be reddish or yellowish but not green. I suggest you take a look on a book on color correction by Dan Margulis.

 

[Corran]

I gave up on scanning color negs with DSLR. [...] All examples I found online how to deal with neg mask just can't deliver the same color fidelity even a cheap scanner is capable of.

Okay sure.

"Adjusting WB and some curves" is roughly what scanners are doing, as I've discussed. Again, it's not black magic. There's literally nothing else that is possible since the light source is a constant temperature and spectrum in a scanner, as would it be from a flash on the DSLR "scan."

I think it's pointless to discuss further. Check your monitor calibration.

 

[jsrockit]

And then waste your time with dust and scratch removal in PP ?

or just take care of your negative in the first place...like we do when we use(d) them in the darkroom.

 

[edge100]

850 camera has something called back illuminated sensor. Back light is the key for scanning.

No.

(10 char)

 

[edge100]

I gave up on scanning color negs with DSLR. The biggest challenge negative orange mask is not uniform. In fact there are two separate masks - magenta and yellow. Their density varies depending on upper color layer density. All examples I found online how to deal with neg mask just can't deliver the same color fidelity even a cheap scanner is capable of. Your examples are just another proof, sorry.

This new feature of D850 is a HUGE selling point for me. I might even upgrade my DSRL finally. How good it is remains to be seeing. And of course it needs to be capable of dealing with 120 and large formats as well to be really valuable.

Sigh.

DSLR scanning works. On colour negatives, positives, and B&W negs (and sides, too).

Colour inversion simply isn't an issue. If you're having trouble, you're doing it wrong.

My tutorial: http://www.mfphotography.ca/michael...-guide-to-scanning-film-with-a-digital-camera

Every single image on my site from the last ~3 years is scanned with a D800. ANd I'm happy with the colour in every single one of them.

 

[edge100]

Okay sure.

"Adjusting WB and some curves" is roughly what scanners are doing, as I've discussed. Again, it's not black magic. There's literally nothing else that is possible since the light source is a constant temperature and spectrum in a scanner, as would it be from a flash on the DSLR "scan."

I think it's pointless to discuss further. Check your monitor calibration.

It is pointless to discuss, because the evidence that it works (and works well) is plentiful.

 

[edge100]

One other thing I did think of - there could potentially be some issue with a DSLR scan if one were to use really bad light for it. A big topic these days in LED lights is the CRI spec. Really bad, cheap lights have bad color rendering. I imagine that this could make the DSLR scan troublesome - but most people tend to use their flash and diffusion. So maybe it's a non-issue. I am sure the scanner manufacturers have this figured out and use good lights. For my scanner, a pair of replacement bulbs costs as much as your typical Epson V-series scanner or more! If I did scan via DSLR I would personally be making a light table and rig for the camera, and would make sure to have high-end bulbs for the light.

This is 100% correct. The accuracy of the light source will determine the colour composition of your scan. If the light is missing a bunch of frequencies, those will consequently be missing (or under-represented) in your scan.

 

[Betin]

It is pointless to discuss, because the evidence that it works (and works well) is plentiful.

Your page encouraged me to try this method once again. I must admit, it turned out pretty good. I will experiment further, maybe i don't need D850 after all...

 

[edge100]

Your page encouraged me to try this method once again. I must admit, it turned out pretty good. I will experiment further, maybe i don't need D850 after all...

Glad I could help! Looks great.

 

[splitimageview]

Sigh.

DSLR scanning works. On colour negatives, positives, and B&W negs (and sides, too).

Colour inversion simply isn't an issue. If you're having trouble, you're doing it wrong.

My tutorial: http://www.mfphotography.ca/michael...-guide-to-scanning-film-with-a-digital-camera

Every single image on my site from the last ~3 years is scanned with a D800. ANd I'm happy with the colour in every single one of them.

Nice tutorial! Thanks for the link.

 

[Betin]

Results are promising. I need to keep tuning into the process tho. DSLR scans come out super grainy and over saturated. Grain is exaggerated even from 120 format and probably may be tuned down a bit in lightroom which applies rather harsh sharpening by default. Somewhat better results are coming with inversion done using ColorNeg plugin (colorneg.com) than manually dragging curves. Optical prints are far from being that saturated. On the other hand, prints don't fit full range of brightness and clip. More tests and calibrations are needed to understand where discrepancy is.

The process is far from being straightforward for sure.

The film here is Kodak Colorplus 200. Scanner - Primefilm XA which i have no plans to retire. DSLR - Sony A850.

Scanner recovers a lot more detail from better film (Ektar) but I have 1:2 macro lens only which is a limiting factor for 35mm.

Light source - Artograph lightpad a920.

 

[Corran]

My hypothesis is grain aliasing caused by the film being "scanned" at too low of a resolution (24mp camera at 1:2 equates to around 2000-2400 DPI I think).

This is one area where the D850 will not have a problem, having well over 5500 DPI in the "scan."

 

[Huss]

You need to use a 1:1 macro lens. I have no grain or splotch issues with mine.