Did Kodak color neg lose sharpness when it was reformulated to improve scanning?

[Pere Casals]

Moderator's note: tangential discussion extracted from "more advanced scanner" thread and moved here.

Main difference was the top coats

You are not well informed.

Show us the patents then.
sharpness enhancement in the emulsions was a long running project far pre-dating scanning needs. I know you are desperate for a magic solution that 'proves' high end scanners are reliant on magic software, but they aren't.

I explained you several times that small discrete color clouds generate aliasing, if the clouds don't overlap then a pixel can catch a cloud of a color or a cloud of another color, thus generating remarcable noise. Is the clouds overlap then you decrease color noise dramatically.

The re-enginering changes in the emulsion were to make clouds overlap, almost every CN film datasheet still mentiones that with weasel words, Prota 160 one says:

"Now the PORTRA Films have been reengineered to deliver significantly finer grain at all speeds for improved scanning performance and greater enlargement capability"

In fact they speak about finer grain "after scaning", because optic enlargement are practically extinct since very, very long ago, so when they say finer grain they mean less aliasing from overlaping clouds and adaptive sharpening to solve the side effects.

You probably know that velvia is noisy in a LS-5000, this is because velvia clouds were not touched.

I know you are desperate for a magic solution that 'proves' high end scanners are reliant on magic software, but they aren't.

Interneg, I showed you that Pro scanners perform a better adaptive sharpening than the Epson, but manually adusting the Eson scan you optain the same:

https://www.largeformatphotography.i...=1#post1479178


If you don't understand that from those samples, then sorry, I cannot do anymore, those samples speak on their own. That proof is clear like a glass.

[interneg]

I explained you several times that small discrete color clouds generate aliasing, if the clouds don't overlap then a pixel can catch a cloud of a color or a cloud of another color, thus generating remarcable noise. Is the clouds overlap then you decrease color noise dramatically.

The re-enginering changes in the emulsion were to make clouds overlap, almost every CN film datasheet still mentiones that with weasel words, Prota 160 one says:

"Now the PORTRA Films have been reengineered to deliver significantly finer grain at all speeds for improved scanning performance and greater enlargement capability"

In fact they speak about finer grain "after scaning", because optic enlargement are practically extinct since very, very long ago, so when they say finer grain they mean less aliasing from overlaping clouds and adaptive sharpening to solve the side effects.

Wrong, wrong and wrong again.

If you had any real knowledge of colour neg films from the last 30 years, you would see that the granularity gets finer, the dye clouds tighter, more densely packed, sharper and less cross contaminated. In anywhere apart from your imagination, this is how films are improved for both scanning and optical printing. And these improvements are not based on your fundamentally poor understanding of limited analytical aspects of the materials, but rather on a very extensive set of data covering the relationships of grain, sharpness, resolution etc and their interrelationships within a complex mathematical analysis of total image content. If your claims had any truth, current films would be natively a lot less sharp than, say, VPS when fully optically printed. They aren't. BTDT. If you knew anything of the use of DIAR & DIR couplers in C-41 films, let alone interlayer scavengers, you would know why colour neg films are so sharp relative to E-6 which is much more limited in the extent to which edge sharpness can be boosted. If you are not seeing this difference in your scans, it is because you are basing your conclusions off a scanner that either has such a low MTF response that it cannot represent the edge effects of the film, or that is so poorly operated or maintained that it cannot deliver its potential quality.

The surface finish apparently caused scanning issues in some cases, so it was altered to be smoother.

You probably know that velvia is noisy in a LS-5000, this is because velvia clouds were not touched.

No, it's because the light source in the Coolscan is extremely highly collimated like a point-source enlarger, quite possibly to try and compensate for flare in the optical system. Up to a certain enlargement size, the slightly more diffuse dye clouds in colour negative (compared to B&W neg grain) enable less apparent grain in optical printing. Going above 8-12x or scanning at decent resolutions with a more collimated light source and granularity will rapidly appear.

All you are testing is the lack of diffusion of the light source and the sensor response to that, not the film. Use a scanner with a better diffused light source before making these claims.

Pro scanners perform a better adaptive sharpening than the Epson, but manually adusting the Eson scan you optain the same:

These claims of yours have been repeatedly disproved by multiple sources (including those tests themselves) and the outright MTF performance of the Epson has been shown to be so poor that the level of sharpening needed to improve it even slightly produces an offensive level of aliasing and noise.

The sensor setup in Noritsu & Frontier scanners is so different from the 3xCCD or 3xPMT scanners that to try and draw comparisons is rather like stating the fundamental sharpness of a digital sensor without a Bayer array or an anti-aliasing filter is identical to one with both. Sharpening the one with the anti-aliasing will produce a degree of artifacts, especially if it's then being output as a compressed file format too.

[bob carnie]

I learn so much when you folks fight.. you are keeping it civil and I like it... I remember the days of Jorge, now he was a tough cookie to debate with and quite often wrong.

Though I have used the Epson, Fuji Frontier, Flextight , Creo and tested the drum scanners it is interesting to see some of the behind the scene stuff... my frontier was good up to about 16 x 20 and then the files fell apart.

[Pere Casals]

These claims of yours have been repeatedly disproved by multiple sources (including those tests themselves) and the outright MTF performance of the Epson has been shown to be so poor that the level of sharpening needed to improve it even slightly produces an offensive level of aliasing and noise.

hmmm, interneg, have you eyes in your face?

Where is it that V700 "offensive level of aliasing and noise" ?

Top left, sharpened by me is the V700 crop, at top-right you have Creo/Scitex Eversmart Supreme, not more, not less. Botom left: Creo/Scitex Eversmart Pro , Bottom-right: Scanmate 11000

Do you know those scanners ? its aprox price when new?

(The red rectangle was sharpened by me) here the original: https://www.largeformatphotography.i...=1#post1479176



https://www.largeformatphotography.i...=1#post1479178

Look, I see the same, please tell me if you see a single pixel that's different !!!

___________


Portra datasheet has the wording: "Now the PORTRA Films have been reengineered to deliver ... improved scanning performance ..."

What do you say? a change in the supercoating ? you are not well informed. Please get information, and then we'll debate


__________

No, it's because the light source in the Coolscan is extremely highly collimated like a point-source enlarger, quite possibly to try and compensate for flare in the optical system.

interneg... velvia was not touched because it was projection stuff, while CN color films were reengineered for larger clouds to be easier to scan, collimation apart, velvia behaves different than CN in the LS-5000, I was pointing that to make you understand what change CN film experimented, with the digital minilabs introduction.

Please... just inspect with a good microscope old and new CN, and slides, see the clouds, see with your own eyes what changed, as I did, then you'll know about what you speak. Supercoating ?

Look, I did it, do it, take the microscope and see.


From that toy will realize why all those scanners are discontinued, why the X1/X5 are discontinued, and why instead the V850 will stay and will be manufactured in the long term.

[Oren Grad]

If you had any real knowledge of colour neg films from the last 30 years...

hmmm, interneg, have you eyes in your face?

We are happy to have vigorous debate - at least until it starts going in circles - but per Forum guidelines we won't tolerate personal insults. No more of that, please.

It's fine that you both feel strongly about the topic. But it's easy to get carried away in the heat of the moment. So after you've drafted a reply but before you post it, wait a little bit and then make another pass through it to remove anything that could reasonably be construed as a personal attack.

[Pere Casals]

We are happy to have vigorous debate - at least until it starts going in circles - but per Forum guidelines we won't tolerate personal insults. No more of that, please.

It's fine that you both feel strongly about the topic. But it's easy to get carried away in the heat of the moment. So after you've drafted a reply but before you post it, wait a little bit and then make another pass through it to remove anything that could reasonably be construed as a personal attack.

ok

[interneg]

Portra datasheet has the wording: "Now the PORTRA Films have been reengineered to deliver ... improved scanning performance ..."

What do you say? a change in the supercoating ? you are not well informed. Please get information, and then we'll debate

I was paraphrasing several of Ron Mowrey's comments on both dye clouds and digital optimisation. Bigger dye clouds would be less sharp & current Portra 160/400 are clearly natively sharper than the previous generation(s). If you believe that you know more about this than the engineers who designed the materials, that's your prerogative, but you'll need more than basic microscopy to prove anything useful. X-ray microdensitometry and scanning electron microscopy for starters. What films did you look at anyway?

And as for 3D LUT's, I know what they are, I have made & used them for various jobs. They are not a panacea, but can be very handy.

[Pere Casals]

Bigger dye clouds would be less sharp & current Portra 160/400 are clearly natively sharper than the previous generation(s).

Let me reiterate, first take the microscope and inspect 1990 vs 2000 color film, later we can continue debating. Clouds are not well seen in the scans, you need a good microscope, you need to see it.

They are not a panacea, but can be very handy.

They are a panacea when the deformations you want in the color space cannot be made by operating with 2D LUTs or curves.

A 3D LUT can map any color in the original color space to any color in the destination color space, so this it TOTAL flexibility, if this is handy or not it depends on if you have the tools to build a 3D LUT that solves your problem.


____


By the way, please tell me how a "trash" scanner like the V700 can match that easy the crop from the ScanMate 11000 https://www.largeformatphotography.i...=1#post1509776 , or the Creo

No guess ?

[interneg]

Let me reiterate, first take the microscope and inspect 1990 vs 2000 color film,

Which specific films?

[Pere Casals]

Which specific films?

You can start with bare consumer films.

Later you can compare Vericolor with Portra that was introduced in 1998 to be the Vericolor VPS and VPL substitute. Probably the early Portra was changed soon.

Also you can compare equivalent Fuji films.

Then compare with Velvia.