Film Negative Invert and Processing in L* Gamma FAQ

[rodeo_joe1]

Ooops. Missed a vital part off that screenshot.

[digitaldog]

Then somebody needs to tell the writer of the Wikipedia ProphotoRGB page.

[ATTACH=full]1385106[/ATTACH]

Screen shot from Wikipedia.

Are you now suggesting ProPhoto RGB doesn't have a gamma 1.8 encoding by design?

And do tell the person who believes the color space is named "ProphotoRGB". Both are wrong concepts.

 

Go to a color scientist and see the facts:

Welcome to Bruce Lindbloom's Web Site

Note: The gamma of sRGB is not exactly 2.2, but rather, is a grafting together of two different functions, that when viewed together, may be approximated by a simple 2.2 gamma curve. When using a simple gamma function, Photoshop calls this "Simplified sRGB." All calculators, spreadsheets and reference tables found on my entire site use the proper functions, not the simplified versions. The proper sRGB functions may be found here and here.

The gamma curve of ProPhoto RGB is 1.8. No extraneous discussions, just the spec's for a lot of RGB Working Spaces based on actual color science, by color scientists like Bruce.

And the basics of gamma and linearity:

http://digitaldog.net/files/LinearityandGamma.pdf

Edited April 22, 2021 by digitaldog

[digitaldog]

This is really useful info.

By the way, i just compared those "flat" exported files processed in Lightroom/ACR with Process Version 2003 or 2010 and with all adjustments turned off, and they still look not the same "flat" as real "flat" untouched exports from other raw editors.

And there's no reason why they should.

Overall i don't use Lightroom/ACR, so i may not understand it as pure as i understand other tools.

We are in violent agreement. ;)

[shijan]

Nope, you simply haven't learned how to render output referred** data from there!

http://www.color.org/scene-referred.xalter

** http://www.color.org/ICC_white_paper_20_Digital_photography_color_management_basics.pdf

I read deeper, and seems workflow described in that article is rather useless today. It was designed for Photoshop CS3 (2007) so it assume ACR use Process Version 2003 or 2010. Also as i remember in those days ACR was able to export only to sRGB, AdobeRGB and ProPhotoRGB profiles. So they create that "hack" to convert in Photoshop exported image form ProPhotoRGB gamma 1.8 back to ProPhotoRGB gamma 1.0.

Current versions of ACR can export directly to any installed ICC profile, so you can just switch to Process Version 2003 or 2010, reset all adjustments to 0 and export to custom-made ICC profile. For example to "native" scene-referred ProPhotoRGB gamma 1.0 or to ProPhotoRGB L*gamma if it required by your workflow.

 

To be honest all those Lightroom/ACR hacks are very depressive and i can't see any point to use it when there are a lot of other apps that can do proper "flat" scene-referred export in way more simpler and user friendly way.

It is real mystery why it is so complicated for Adobe simply add option in ACR to remove build-in contrast curve same as other raw editors normally can do:

[digitaldog]

I read deeper, and seems workflow described in that article is rather useless today. It was designed for Photoshop CS3 (2007) so it assume ACR use Process Version 2003 or 2010.

That assumption is wrong**. Those who desire scene referred rendering don't find setting ACR for scene referred useless today. NOTHING whatsoever has changed with respect to this since CS3.

Next question.

To be honest all those Lightroom/ACR hacks are very depressive and i can't see any point to use it when there are a lot of other apps that can do proper "flat" scene-referred export in way more simpler and user friendly way.

To be honest, that's rubbish. Again, some users want to produce scene referred rendering and that's why both Adobe and the ICC have provided a way to do so. IF you don't wish to follow that, fine.

It is real mystery why it is so complicated for Adobe simply add option in ACR to remove build-in contrast curve same as other raw editors normally can do:

I completely understand and accept after all these pages, when the discussion comes to ACR/LR/producing scene referred rendering or not, this is all a real mystery to you.

Next question. Or in terms of resetting curve points in ACR/LR; RTFM :p

 

**“Most of our assumptions have outlived their uselessness“.

-Marshall McLuhan

 

Edited April 23, 2021 by digitaldog

[shijan]

From https://www.color.org/scene-referred.xalter

- "The linear_RIMM-RGB_v4.icc profile contains a linear CCTF. It should be used when a linear scene-referred encoding is desired, but only with 16 (or more) bits per channel."

- "Open the raw file in ACR. Note that the color rendering controls should be left at zero or linear in order to produce a scene-referred image."

- "Set the Workflow Options Space to ProPhoto RGB and the Depth to 16 Bits/Channel."

- "Open the converted (to scene-referred) raw file into Photoshop"

-"Edit -> Convert to Profile and CONVERT to the linear_RIMM-RGB_v4.icc profile using the relative colorimetric rendering intent."

 

As for me this is all looks the same as simple export to custom-made ProPhoto RGB with Linear gamma ICC profile directly from ACR :)

[shijan]

Also seems every raw processor use some unique detection and trimming of "empty" data range in debayered raw file. So the same "flat" scene-referred export from different raw processors usually have slightly different brightness.

[digitaldog]

As for me this is all looks the same as simple export to custom-made ProPhoto RGB with Linear gamma ICC profile directly from ACR :)

It isn't. They don't look the same, I could calculate the exact deltaE difference of all three and, I only know of one that has specific means to export Scene Referred rendering (although I also own Iridient Developer).

Also seems every raw processor use some unique detection and trimming of "empty" data range in debayered raw file.

Trimming empty data range?

Every raw converter is akin to your color negs; they differ based on a slew of factors. You're just NOW figuring out this fact or Trimming empty data range is another assumption?

This is what raw looks like:

 

http://www.digitaldog.net/files/ThisIsRaw.jpg

http://www.digitaldog.net/files/raw.jpg

 

The rest is created uniquely in each raw converter.

[digitaldog]

It is real mystery why it is so complicated for Adobe simply add option in ACR to remove build-in contrast curve same as other raw editors normally can do:

Found how how yet? There's at least 3 ways to do this IF you know how.....

[shijan]

Trimming empty data range?

I don't know how to name that correctly. But looking how it works in PhotoLine i assume that there is always some sort of reserved empty space in debayered source image. And it is usually trimmed under the hood. PhotoLine operates with imported raw file at very low level so you can see some things the that is impossible to see in other apps.

 

For example, here is how processed raw file looks when you open it in PhotoLine.

 

Now we can turn off that auto generated contrast curve:

 

And next we can go even deeper and bring back all trimmed empty data. This is totally useless empty data, so i guess other raw converters usually autodetect and trim it by default somehow.

[shijan]

I know how unprocessed RAW file looks, PhotoLine also can show image in raw pattern mode like this. And in reality those black and white pixels. They are usually shown as RGB pattern only for more human friendly preview:

Edited April 23, 2021 by dmitry_shijan

[digitaldog]

I know how unprocessed RAW file looks, PhotoLine also can show image in raw pattern mode like this. And in reality those black and white pixels. They are usually shown as RGB pattern only for more human friendly preview:

 

If you know what you say you know, you would know that this isn't front page news and never was:

The rest is created uniquely in each raw converter.

What you don't appear to know or can explain despite being asked is:

Also seems every raw processor use some unique detection and trimming of "empty" data range in debayered raw file.

Say what?

 

BTW, the terminology you're looking for is demosaicing (also de-mosaicing, demosaicking or debayering).

Demosaicing - Wikipedia

Edited April 23, 2021 by digitaldog

[digitaldog]

I don't know how to name that correctly.

So you made up a word for an attribute you can't explain?

But looking how it works in PhotoLine i assume

Yes, you assume. Please get facts then post**.

PhotoLine operates with imported raw file at very low level so you can see some things the that is impossible to see in other apps.

Assumption or factual data point you can prove?

Do tell us how many other raw converters, exactly the names of each app you compared this to, to come up with the assumptive text:

....impossible to see in other apps.

This is totally useless empty data, so i guess other raw converters usually autodetect and trim it by default somehow.

Yes, you're guessing again. Assuming, making up terms etc. It's getting pointless to continue because this is a photo, color, technical discussion, not a fiction fictional discussion based on assumptions.

Meanwhile, your time may be better spent learning how to reset curves in ACR unless you never care to use it which is fine with me.

**"Get your facts first, then you can distort them as you please." --Mark Twain

[shijan]

What is correct terminology for this "empty space" in debayered image :) ?

[shijan]

I don't use ACR/Lightroom and never will be, so i have no any interest to fit it to this workflow and search how to reset curves in it. I already post gif animation with possible option here Film Negative Invert and Processing in L* Gamma FAQ

If you know how to do it please provide info or links that may be useful for other people.

[digitaldog]

I don't use ACR/Lightroom and never will be, so i have no any interest to fit it to this workflow and search how to reset curves in it. .

Fine, then we can jot this comment of yours down as rubbish (because due to your misunderstanding of how ACR works, that's exactly what it is):

It is real mystery why it is so complicated for Adobe simply add option in ACR to remove build-in contrast curve same as other raw editors normally can do.

Yes, you do seem to find items you don't understand but comment about, a mystery.

"Learning is not attained by chance. It must be sought for with ardor and attended to with diligence. "-Abigail Adams

Your lack of diligence is now clear.

What is correct terminology for this "empty space" in debayered image

There is nothing empty there. Another of your assumptions.

I believe you need a primer in understanding Histograms which I don't expect you'll study:

Video tutorial: Everything you thought you wanted to know about Histograms

http://digitaldog.net/files/Histogram_Video.mov

And (look, empty areas and more data to learn):

Beware the Histogram

Edited April 23, 2021 by digitaldog

[shijan]

Nothing new for me in that basic video about histograms.

All i wanted to explain here is that different RAW converters produce slightly different "native" scene-referred image look.

[digitaldog]

Nothing new for me in that basic video about histograms.

All i wanted to explain here is that different RAW converters produce slightly different "native" scene-referred image look.

It WAS explained and in these posts.

You assume (again) each converter is producing a scene-referred rendering. And they should match. Stop assuming.

 

Got Photoshop?

Make a document and fill it with 50% gray.

Examine the Histogram. Still confused by the “empty” areas or you are able to understand them even with a video primer on Histograms??? Fill the image with White or Black only. Understand the difference between each and WHY the Histogram shows what it shows?

 

Got nothing to do with raw per se. It's got nothing to do with Scene Referred or not per se. But if and when you understand how Histograms are drawn AND accept all raw converters are proprietary and differ based on the SAME raw, differ in what and how they construct that Histogram, maybe you'll connect the dots about “empty” areas of the Histogram.

Digital imaging 101 really. Nothing new or unexpected for some.....

Edited April 23, 2021 by digitaldog

[shijan]

Assuming, comparing and making conclusions is the core of everything that i do and learn.

I didn't talk about how different histogram redrawn or previewed in different raw converters

I exported images from different raw converters with same color space and gamma, put them to PhotoLine and compare their histograms from there.

I agree that different converters produce a different scene-referred rendering (or whatever you name that "flat" look). And i didn't talk that they should match. I already illustrate and confirm that difference in side by side examples.

I understand why the Histogram (as well as other more complicated monitoring tools) shows what it shows.

I can't see nothing bad if i name that "gap" in histogram as "empty space in histogram". But is that terminology really so matter for this particular case? Why cling to words?

[digitaldog]

Assuming, comparing and making conclusions is the core of everything that i do and learn.

Lots of evidence of assuming. Not so much elsewhere. Sorry.

You have made up terms that don't exist.

You have complained about an Adobe product you don't understand how to use.

You have assumed a lot about raw data, Histograms and more; much based on assumptions.

My paid grade here isn't sufficient to continue trying, sometimes, often in vain, to explain how this stuff actually works with of course, outside references.

I exported images from different raw converters with same color space and gamma, put them to PhotoLine and compare their histograms from there.

So what? They differ. Nothing at all unexpected from this fact! I don’t know if you are purposely trying not to understand this, or if you are really struggling with it. IF the former, let's move on. If the later, maybe I'll attempt to explain again, all I've already explained about Histograms, resetting curves, how all raw converters differ in rendering (AND HISTOGRAM). My patience with you is limited.

I agree that different converters produce a different scene-referred rendering (or whatever you name that "flat" look).

Again, you seem to have massive issues reading and comprehending. Perhaps it's a difference in native language? I told you this has NOTHING to do per se with scene referred rendering and you either didn't read or understand this as illustrated in the quote above. Clearly your idea of 'flat' being or having anything to do with scene referred shows you didn't read or understand the ICC white paper I co-authored and posted here. I can't help you any further if you refuse to slow dow, read and attempt to comprehend the facts!

I understand why the Histogram (as well as other more complicated monitoring tools) shows what it shows.

Great, IF so (if) my job is done here (despite your confusion in the last post, quoted here as well).

Hopefully you found a copy of Photoshop and found out how a Histogram can be reported with flat (what you incorrectly call 'empty') areas. There is absolutely nothing unusual about this or that each raw converter shows this differently.

You wrote: "i assume that there is always some sort of reserved empty space in debayered source image."

Your assumption is just incorrect as well as the terminology . It shows an utter lack of understanding of what a Histogram plots. I've tried to explain it, I've provided a video AND article you say you saw/read but the evidence you understand the assumption above is wrong isn't yet clear to me and others I suspect. That's fine, assume and believe the earth is flat if you must.

I can't see nothing bad if i name that "gap" in histogram as "empty space in histogram"

You don't despite the fact it isn't a gap or empty space.

But is that terminology really so matter for this particular case? Why cling to words?

As the Chinese proverb says: "The first step towards genius is calling things by their proper name."

But if you prefer to make up terms and attempt to confuse yourself and worse, your readers, go for it. It's the opposite of genius.

[shijan]

First of all dude, we are not in your "online class" or whatever, so don't act like you are some sort of professor in kindergarten. If you came to participate in my thread please stay calm or go away.

Don't forget that this thread is all about real life film negative processing workflow, but not about some pure theoretical geek color science tech stuff and terminology.

 

A lot of that mess and misunderstanding was started after you introduce that idea of "scene referred rendering" described pdf and article on color.org. I never hear about that scene referred rendering thing before and it is essential that i can do mistakes at the first steps of understanding.

Anyway, i still have no idea how it may help with film scans processing. There is no clean technical explanation in that article and pdf about what actually those ICC profile do when you convert to them. Only some abstract concepts and the fact that i need to convert image to that profile. (Maybe i need to re-read it few more times).

I never convert something to something without knowing what is actually going on. So i checked those scene referred ICC profiles with different ICC inspectors and tools and i see now that they do some transformations. I can't explain for myself the purpose of those transformations yet, so i will stay away of those scene referred ICC profiles yet.

So after these deeper tests now i see that "scene referred rendering" is not the same as uncorrected "flat" export with removed contrast curve in raw editor.

 

At least you are correct that i am not a native English speaker. I am from Ukraine but my skills are enough to read/write tech articles in English natively.

You can see some of my works and contributions to other apps and projects during last years:

Blackmagic Forum • View topic - Final Explanation of Gamma and Color Shift Problems

VapourSynth + QTGMC Deinterlace + Hybrid UI FAQ for macOS

[Legacy] Large List of Suggestions and Requests for Photoline UI and Tools - PhotoLine Forum

PhotoLine UI Icons Customization Project by shijan on DeviantArt

Facebook

[digitaldog]

A lot of the mess was your inability to understand seen referred rendering histograms,color spaces, color accuracy, ACR, AND MORE. We’re done here. What we have is a failure to communicate. More and more likely what we have is a failure for you to be able to read and comprehend, colorimetric facts!

 

I agree completely with your last point:

“Anyway, i still have no idea how it may help with film scans processing.”

I have provided plenty of other areas you equally have no idea.

 

As to your chops (so called “experience”) if or not as sloppy as your understanding of our discussion, I'm not at all impressed by your multiple misunderstandings locked forever in this post.

 

You are lost in color space as the 8 pages of comments have illustrated towards your post here. Happy no one here took you seriously? :(

Try posting this nonsense in the color management forums on Luminous Landscape and you'll get ever harsher treatment from “your peers” if I can be that generous to you.

Edited April 24, 2021 by digitaldog

[frans_waterlander]

And that, Ladies and Gentlemen, is how digitaldog, aka Andrew Rodney, makes himself feel good.

[shijan]

Everything i wanted to share for now, i shared here in the first page of this thread.

Same discussion goes on other forums and people share ideas and suggestions with respect there.

But here the 7 pages of "discussion" consists of formal spamming by rodeo_joe|1's who seems barely understand difference between color space and gamma, and of digitaldog's nervous insolent posts that seems makes him feel "too smart", but have low practical real life usability and in most cases follow to areas not related to this topic.

[digitaldog]

And that, Ladies and Gentlemen, is how digitaldog, aka Andrew Rodney, makes himself feel good.

Ah, Frans, equally confused by colorimetric facts drops in to troll off topic....again.

Frans may invent an explanation for “empty spaces” in Histogams which should be a hilarious read.

Upload any image you have ever “made” to your gallery here yet Frans?

Edited April 24, 2021 by digitaldog