Author Topic: Time to ditch sRGB/Linear as default (?)  (Read 33375 times)

2018-11-15, 00:00:41
Reply #165

steyin

  • Active Users
  • **
  • Posts: 313
  • BALLS
    • View Profile
Is TIFF preferred over PNG? Or any advantages of it versus the other?

2018-11-15, 11:10:50
Reply #166

Juraj Talcik

  • Active Users
  • **
  • Posts: 3518
  • Tinkering away
    • View Profile
    • studio website
Might be personal preference. I prefer Tiff for the following:

- more commonly accepted in all of its versions (16bit&32bit Tiff, whereas 16bit PNG (PNG-48) is less).
- more compression options
- more flexible alpha use (lot of applications read png with baked alpha, with Tiff they always give interpretation options)
- can save layers, good support for color profiles

It's good format to save from Corona, good format to store layers, good format to send to clients,/print.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2019-01-20, 17:52:16
Reply #167

Xntric

  • Active Users
  • **
  • Posts: 22
    • View Profile
Guys, i'm confused.

Corona can only render in sRGB color space, right?

No way to apply ACES colormanagement?

2019-02-07, 16:55:47
Reply #168

jpjapers

  • Active Users
  • **
  • Posts: 1122
    • View Profile
I'm really trying to get my head around ACES and tonemapping and general colour management. I've been using the settings dubcat posted a while ago but i'm only now getting into the details about what it does and why. I'm very confused about LUTs and ACES workflow right now. Could anyone perhaps help me understand it a little more as some of you seem to have a good grasp of things when it comes to colour and corona.

I have a couple of questions about which LUTs can be used with the ACES tonemapping settings to simulate various camera responses etc rather than to just produce a 'look'. I'd like to have a couple of LUTs to use to bring the data back into what a real world camera might produce right in the VFB. Or am I completely misunderstanding the workflow and I'm supposed tonot use the LUT then do my grading in lightroom? Ive tried using dubcats colour chart in aces scenes without a lut then matching in 3DLUT creator which gives a great result. Would i then save out this LUT and use it in the VFB for other images in this light/camera setup?

There's not much information out there for beginners on this but im finding the reading incredibly interesting.
That blender video only confused me more but it helped me to take an interest in colour space and tonemapping.
« Last Edit: 2019-02-07, 17:05:44 by jpjapers »

2019-02-07, 18:11:12
Reply #169

lupaz

  • Active Users
  • **
  • Posts: 282
    • View Profile
I'm really trying to get my head around ACES and tonemapping and general colour management. I've been using the settings dubcat posted a while ago but i'm only now getting into the details about what it does and why. I'm very confused about LUTs and ACES workflow right now. Could anyone perhaps help me understand it a little more as some of you seem to have a good grasp of things when it comes to colour and corona.

I have a couple of questions about which LUTs can be used with the ACES tonemapping settings to simulate various camera responses etc rather than to just produce a 'look'. I'd like to have a couple of LUTs to use to bring the data back into what a real world camera might produce right in the VFB. Or am I completely misunderstanding the workflow and I'm supposed tonot use the LUT then do my grading in lightroom? Ive tried using dubcats colour chart in aces scenes without a lut then matching in 3DLUT creator which gives a great result. Would i then save out this LUT and use it in the VFB for other images in this light/camera setup?

There's not much information out there for beginners on this but im finding the reading incredibly interesting.
That blender video only confused me more but it helped me to take an interest in colour space and tonemapping.

Why go through all that trouble?
Would the result be THAT much better?
Isn't it better to wait for the developers to work on the new built-in tonemapping?

2019-02-08, 10:23:32
Reply #170

jpjapers

  • Active Users
  • **
  • Posts: 1122
    • View Profile
I'm really trying to get my head around ACES and tonemapping and general colour management. I've been using the settings dubcat posted a while ago but i'm only now getting into the details about what it does and why. I'm very confused about LUTs and ACES workflow right now. Could anyone perhaps help me understand it a little more as some of you seem to have a good grasp of things when it comes to colour and corona.

I have a couple of questions about which LUTs can be used with the ACES tonemapping settings to simulate various camera responses etc rather than to just produce a 'look'. I'd like to have a couple of LUTs to use to bring the data back into what a real world camera might produce right in the VFB. Or am I completely misunderstanding the workflow and I'm supposed tonot use the LUT then do my grading in lightroom? Ive tried using dubcats colour chart in aces scenes without a lut then matching in 3DLUT creator which gives a great result. Would i then save out this LUT and use it in the VFB for other images in this light/camera setup?

There's not much information out there for beginners on this but im finding the reading incredibly interesting.
That blender video only confused me more but it helped me to take an interest in colour space and tonemapping.

Why go through all that trouble?
Would the result be THAT much better?
Isn't it better to wait for the developers to work on the new built-in tonemapping?

Personally the results i have using a colour checker look alot more natural than without. Its probably adding maybe 5 minutes tops to my post processing and the colours look great.

2019-02-08, 10:36:21
Reply #171

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 5612
  • Let's move this topic, shall we?
    • View Profile
    • My Models
Personally the results i have using a colour checker look alot more natural than without. Its probably adding maybe 5 minutes tops to my post processing and the colours look great.

You mean , you add virtuall colour checker to your scene and that enables you to get more natural results? o_O Would love to see an example, if you will.
I'm not Corona Team member. Everything i say, is my personal opinion only.

2019-02-08, 12:05:12
Reply #172

jpjapers

  • Active Users
  • **
  • Posts: 1122
    • View Profile
Personally the results i have using a colour checker look alot more natural than without. Its probably adding maybe 5 minutes tops to my post processing and the colours look great.

You mean , you add virtuall colour checker to your scene and that enables you to get more natural results? o_O Would love to see an example, if you will.

Yeah so a while ago dubcat posted a virtual colour checker model with the correct sRGB values for the xrite colour passport but the thread didnt really get much attention.
I believe youve already commented in his hideout thread where i asked him about it. It seems to work sometimes but not all the time.
I was just curious and its possible that the scenes i tested on were just happy accidents.
« Last Edit: 2019-02-08, 13:04:51 by jpjapers »

2019-02-09, 20:05:05
Reply #173

James Vella

  • Active Users
  • **
  • Posts: 90
    • View Profile
Yes, I wrote that above, half-float .exr is 16bit internally. But to work with them properly, you need to be in linear environment, and Photoshop only offers this properly in their 32bit mode. And yeah...Photoshop doesn't give you all the tools in their 32bit mode, because Adobe doesn't care.

But Affinity does, I suggest you give it a try. It's not perfect yet but they have much better 32bit support.

Also, do you really need linear information for this kind of post-production ? You could simply stay in normal environment and use 'Screen' mode instead of 'Linear(Add)' which the name implies what it does :- ).

Wouldnt it be easier if you wanted to stay in photoshop to have your 32bit exr file as a smart object within your 16bit post file. This way you can always readjust the exposure/gamma without crush or burn within the smart object?

Im curious to how this workflow compares to affinity as i have not used it before
« Last Edit: 2019-02-09, 21:42:54 by James Vella »

2019-02-11, 09:57:12
Reply #174

Juraj Talcik

  • Active Users
  • **
  • Posts: 3518
  • Tinkering away
    • View Profile
    • studio website
Yes, I wrote that above, half-float .exr is 16bit internally. But to work with them properly, you need to be in linear environment, and Photoshop only offers this properly in their 32bit mode. And yeah...Photoshop doesn't give you all the tools in their 32bit mode, because Adobe doesn't care.

But Affinity does, I suggest you give it a try. It's not perfect yet but they have much better 32bit support.

Also, do you really need linear information for this kind of post-production ? You could simply stay in normal environment and use 'Screen' mode instead of 'Linear(Add)' which the name implies what it does :- ).

Wouldnt it be easier if you wanted to stay in photoshop to have your 32bit exr file as a smart object within your 16bit post file. This way you can always readjust the exposure/gamma without crush or burn within the smart object?

Im curious to how this workflow compares to affinity as i have not used it before

Yes, as long as I do all changes in smart layer container (but don't apply them via filters like CameraRaw directly in-file), that is viable choice indeed. As long as you are content of having a nested file.
You would have to be fully non-destructive and don't mix your color grading with any image manipulation, and I work somewhat dirty.

But over the past two years, Adobe has changed so much how 32bit mode applies, for example excluding direct access to CameraRaw as filter (can by bypassed by direct loading),etc.. that I just concluded to be futile using it at all.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2019-02-11, 10:16:36
Reply #175

James Vella

  • Active Users
  • **
  • Posts: 90
    • View Profile
But over the past two years, Adobe has changed so much how 32bit mode applies, for example excluding direct access to CameraRaw as filter (can by bypassed by direct loading),etc.. that I just concluded to be futile using it at all.

What version of Photoshop are you using? They removed Camera Raw filter?

As you mentioned I usually keep all 32bit render layers in a smart object for relighting and then in the 16bit master file (color grading file) put a camera raw filter on the smart object, you are saying this is removed now?

2019-02-11, 10:51:27
Reply #176

Juraj Talcik

  • Active Users
  • **
  • Posts: 3518
  • Tinkering away
    • View Profile
    • studio website
Direct access to 32bit file. In previous CC versions (<2018) you could have applied CameraRaw in 32bit mode. ( Of course, it would apply auto-leveling in its default mode, by you could switch to 2013 process and zero it out, and use CameraRaw on actual HDR data).
This can still be done by auto-loading into CameraRaw with "Open As" of 32bit file as .raw file. But no longer as filter.

Applying it in 16bit mode still works just fine, but there isn't much point to it without access to any dynamic range, so it's just different GUI.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2019-02-11, 11:08:31
Reply #177

James Vella

  • Active Users
  • **
  • Posts: 90
    • View Profile
Direct access to 32bit file. In previous CC versions (<2018) you could have applied CameraRaw in 32bit mode. ( Of course, it would apply auto-leveling in its default mode, by you could switch to 2013 process and zero it out, and use CameraRaw on actual HDR data).
This can still be done by auto-loading into CameraRaw with "Open As" of 32bit file as .raw file. But no longer as filter.

Applying it in 16bit mode still works just fine, but there isn't much point to it without access to any dynamic range, so it's just different GUI.

Fair enough, I suppose my camera raw method is just as dirty :) Im not doing full dynamic range stuff with it just a slight few adjustments which is easier than white balancing/vignette without extra plugins. I also like the sharpen in camera raw and having the ability to switch it on/off as a smart layer filter, depends on your workflow I suppose.

2019-02-11, 11:34:52
Reply #178

Juraj Talcik

  • Active Users
  • **
  • Posts: 3518
  • Tinkering away
    • View Profile
    • studio website
Yeah the whole thing (CamRaw) is crafty, which is why I lament every time why Adobe just couldn't make it fully functional. It would be blessing to have it work on linear files without any hassle as that would make the post-pro workflow absolutely identical to photography, one smooth process without unnecessary division between "big" changes in linear 32bit, and "small" changes after.

Now you originally mentioned the Affinity and I have to contend that it isn't as good as it appears.. we bought two licences but almost kinda regret it as while it is lot more ambitious (much better 32bit support, simultanenous layer adjustments (super good for textures),...) it's just not even stable. At the moment I am fully back at Photoshop for almost anything. It's not just habit...it's the same stuff as 3dsMax, might not be ideal, but still the best choice.
talcikdemovicova.com  Website and blog
be.net/jurajtalcik   Our studio Behance portfolio
Instagram   Our studio Instagram, managed by Veronika

2019-02-11, 12:52:11
Reply #179

James Vella

  • Active Users
  • **
  • Posts: 90
    • View Profile
Yeah the whole thing (CamRaw) is crafty, which is why I lament every time why Adobe just couldn't make it fully functional. It would be blessing to have it work on linear files without any hassle as that would make the post-pro workflow absolutely identical to photography, one smooth process without unnecessary division between "big" changes in linear 32bit, and "small" changes after.

Now you originally mentioned the Affinity and I have to contend that it isn't as good as it appears.. we bought two licences but almost kinda regret it as while it is lot more ambitious (much better 32bit support, simultanenous layer adjustments (super good for textures),...) it's just not even stable. At the moment I am fully back at Photoshop for almost anything. It's not just habit...it's the same stuff as 3dsMax, might not be ideal, but still the best choice.


Cheers for the info on Affinity, stability is a concern I might wait it out. Another thing that concerns me is I use lightroom for my photography so the Adobe suite suits my pipeline, as well as the terrific catagorizing/tagging etc and similarities between adobe tools.

Regarding what you said about not being ideal but the best choice, I see you did a test on the ACES workflow using Davinci with dubcat - out of curiosity did you implement this workflow? Are you working in a higher gamut space or just sRGB in post? I ask because I have read the tech papers, did some tests on my own, I also work in Adobe 98 space for my photography on a wide gamut monitor for print (output to sRGB for screen), however for 3D archviz I just dont see any major advantage currently - mostly because Photoshop is the main post production tool (with the limitations we spoke about), easy to swap PSD files with other studios and if its kept in sRGB (even if its a low gamut space - which still looks great even with photography) theres no confusion along the way for input, output, space conversion, color matching, etc. I wont go into print for now, just curious about your view on the topic currently.

 edit: Not to mention I wont be remastering any of my old work, which I doubt many people do in archviz.
« Last Edit: 2019-02-11, 12:58:46 by James Vella »