Camera "scanning" with a fluorescent light box.

JWMster · Jun 2, 2020

Reference point for "Corvette Red" shouldn't be too hard to come by given Corvette owners clubs, Chevy still making the car, etc. I would use a shot of a colorbyte test card, too. Problem with the car test color is that paint fades. My guess would be that if these are collector cars as they look to be, they probably don't sit out in the sun that much and fade.
I've been happy with Vuescan colors on a Nikon LS8000, but not with the amount of time it takes to make a scan. Negative Supply equipment is my first choice these days, but I've been working with a lot of B&W lately. The Colorbyte will be my go-to when I get back to color.

Tom Kershaw · Jun 2, 2020

Here is a comparison of Pro 400H with and without the Nikon profile.
https://tomkershaw.com/apug/NikonProfileFuji400H_comparison.pdf

Tom Kershaw · Jun 2, 2020

JWMster said:
I've been happy with Vuescan colors on a Nikon LS8000, but not with the amount of time it takes to make a scan. Negative Supply equipment is my first choice these days, but I've been working with a lot of B&W lately. The Colorbyte will be my go-to when I get back to color.

From my limited experience with photographing negatives I've been surprised by the quality of the camera-as-scanner approach, although resolution isn't up to much compared to using the Nikon scanner with a 24mp APS-C camera.

PhilBurton · Jun 2, 2020

Tom Kershaw said:
Here is a comparison of Pro 400H with and without the Nikon profile.
https://tomkershaw.com/apug/NikonProfileFuji400H_comparison.pdf

Tom,
Thanks. The difference is striking. Which scan do you think is more accurate?

Tom Kershaw · Jun 2, 2020

PhilBurton said:
Tom,
Thanks. The difference is striking. Which scan do you think is more accurate?

Hi Phil,

The.one with the Nikon profile. However significant differences do seem evident based on apparently fairly small changes in computer settings. I should probably read up more on the operation of ColorPerfect. On another note, I suffered an unscheduled IT failure this afternoon so it may take a while before I can get get up and.running again with exploring this topic fully.

Tom

Les Sarile · Jun 2, 2020

Tom Kershaw said:
From my limited experience with photographing negatives I've been surprised by the quality of the camera-as-scanner approach, although resolution isn't up to much compared to using the Nikon scanner with a 24mp APS-C camera.

That's ok because the fact that a 36MP Nikon D800 applies more pixels on a 35mm film then that of the 4000dpi from a Coolscan doesn't mean it is extracting more info either.

On the right-bottom shows the test target used and taken with 35mm Kodak Techpan shot at ISO25 and processed in Technidol.
Above it is 100% crop of the center portion in red scanned with a 14.6MP Pentax K20D..
Above that is 100% crop of the center portion in red scanned with Coolscan 4000dpi.
Above it is 100% crop of the center portion in red scanned with 36MP Nikon D800.

Full res version -> http://www.fototime.com/8372250EA44CB06/orig.jpg

You will notice the 100% crop with the D800 has more pixels then the Coolscan at 4000dpi more vertical resolution achieved by the Coolscan and nearly identical horizontal res.
The big 100% crop on the right is the center section in red optically magnified about 4.5X which clearly shows there is much more detail on the film unresolved by these methods. It is therefore not the limiting factor for these scans.

jtk · Jun 2, 2020

Tom Kershaw said:
VueScan with the 9000 although I can scan with the Nikon Scan 4 software on an old Power Book G4. The 'off the scanner' file does look different - colour-wise. - I made a slight correction as the tiff had a slight blue / green bias.
View attachment 244136

Color appears to be crossed-over...not just unbalanced. Typically would result from bad processing, not a scanning issue.

Tom Kershaw · Jun 3, 2020

jtk said:
Color appears to be crossed-over...not just unbalanced. Typically would result from bad processing, not a scanning issue.

This negative was processed 10 years ago, I'm certainly more careful with processing today.

Tom Kershaw · Jun 3, 2020

I note from looking through the ColorPerfect Flickr discussion group that the author recommends keeping all profiles switched off in Photoshop / PhotoLine before use of CP which I've just tried and does seem to produce a more predictable result.

Tom Kershaw · Jun 3, 2020

Here is an example with no profiles applied apart from within ColorPerfect and conversion to sRGB on the jpeg.

Tom Kershaw · Jun 4, 2020

Spotted this thread on the pl32.com forums so I'm not alone then... https://www.pl32.com/forum3/viewtopic.php?f=3&t=5596&p=41994&hilit=colorperfect#p41994

Tom Kershaw · Jun 4, 2020

Here is an example of the negative using the workflow detailed in the discussion I've linked to above.

Lachlan Young · Jun 6, 2020

@Tom Kershaw - don't suppose you'd happen to have the uninverted negative of the Zephyr on hand? If it has a little neg rebate, it should be pretty easy to run it through PS and see if it does or doesn't have crossed curves.

Tom Kershaw · Jun 6, 2020

Lachlan Young said:
@Tom Kershaw - don't suppose you'd happen to have the uninverted negative of the Zephyr on hand? If it has a little neg rebate, it should be pretty easy to run it through PS and see if it does or doesn't have crossed curves.

Hi Lachlan,

I've not had a chance to look at this again since Thursday, I'll take another look.

Tom

Tom Kershaw · Jun 6, 2020

Ah, I've just checked and the scan I have from Thursday was cropped to just the image area, so I'd need to go back to the film.

Lachlan Young · Jun 6, 2020

Tom Kershaw said:
Ah, I've just checked and the scan I have from Thursday was cropped to just the image area, so I'd need to go back to the film.

No worries - I wouldn't want to make any assumptions about curve crossover on the film until I've seen it uninverted, with enough rebate to sample - on the basis of current evidence I'd be inclined to say that it's the software, not the film that's causing the problems.

Tom Kershaw · Jun 7, 2020

Lachlan Young said:
No worries - I wouldn't want to make any assumptions about curve crossover on the film until I've seen it uninverted, with enough rebate to sample - on the basis of current evidence I'd be inclined to say that it's the software, not the film that's causing the problems.

I suspect this is the case as well.

Camera "scanning" with a fluorescent light box.

Privacy & Transparency

Privacy & Transparency