More gamma stuff and linear workflows

[GVC1123]

Right then,

I think i'm starting to understand how to go about setting up LWF but still not sure why i would?

So far i've calibrated my monitor to gamma 2.2, enabled gamma/lut in max, changed that to 2.2 as well, then ticked affect colour and affect materials. (is that right?)

I can see from a before and after render that there is more setail showing up in the shadow areas, but there also seems to be a lot more noise, is this right? Also, some glossy highlights have now been made more sublte than before?

Also, if you are going to render an animation over a netwrok using backburner, does every machine need to be setup for a LWF or can i get away with only having my workstation converted?

Personally, to avoid the noise in the shadow areas, I set VRay's gamma (in the color mapping) to 2.2. This way VRay will be set up to better calculate these areas.

If you want to run an animation, there is a "load enable state" checkbox which you should check on for every farm box. This will cause each box to adopt whichever gamma settings you send it.

[GVC1123]

Also, it's good to know which frame buffer we are using since Max frame buffer behaves different from the VRay frame buffer when you've set the gamma preferences. I prefer the VRay frame buffer because it has more viewing abilities.

[MikeeMax]

in method II, i understand the settings, but what number should i put in "Gamma" field of the "color mapping" section ? 1 or 2,2 ?

[DanielBrew]

You should leave it at 1. As Vlado said at the end of II,

(*) Do not apply gamma correction to your rendered result.

Dan

[plastic_]

this is my take on it:

FORGET ABOUT GAMMA 1.0!


i think vlado was a bit too theoretical about that one.
sure, it's an option to calibrate the display to gamma 1.0, but it's not practical.

i will use a screenshot to explain why:

it's perfectly possible to calibrate your display to gamma 1.0, with a spyder 3 elite, for example...i just tried that. the problem is, afterwards your windows will look like that: (see attatchment).

windows will put the new gamma settings over _everything_, so the vray vfb will look great, but everything else is wrong.

besides, at some point you will have to exchange files with other (uncalibrated) systems, it has to look (quite) right on those too.

gamma 2.2 is a standard. uncalibrated displays are close to gamma 2.2 (even macs by now)

so i don't think it makes sense to shift away from that, since there is no need to.
linear workflow works great with gamma 2.2, you just have to specify it in max and in the vray color mapping options.

[MikeeMax]

linear workflow works great with gamma 2.2, you just have to specify it in max and in the vray color mapping options.

Didn't Daniel Brew just said the opposite ?

You should leave it at 1. As Vlado said at the end of II,

so...." don't apply color correction to your image" means:
no color correction after render
, or
no gamma 2.2 in color mapping mode ?

[DanielBrew]

MikeeMax,

When I said leave your gamma under VRay's color mapping at 1, I was answering your very specific question relation to 'Workflow II' of Vlado's 3 methods of LWF.

I think what Marc is saying is that methods II and III are all very well in theory but if you use your computer for anything other than rendering and compositing then you are likely run into problems.

Dan

[MikeeMax]

Thanks DanielBrew. i'm starting to see the light at the end of the tunnel.
i got it almost completly. only thing left to understand for me is this sentence
"(*) If a bitmap texture is created with method I and has gamma baked into it, you need to specify this gamma when loading the bitmap in 3dsmax, so that the correction can be removed."
how can i check if a gamma is baked into an image ? and how can i remove this gamma correction if i come to the conclusion that there is a gamma baked into the image ?

[DanielBrew]

MikeeMax,

I don't know of a way to check if an image has gamma baked into it other than to look at it. Generally, most images taken with digital cameras or saved from websites already have gamma baked into them.

If you want to remove gamma correction from an image when it's loaded into Max there are two places you can do this.

1. 'Gamma and LUT' tab on the preferences box - if you want to globally remove gamma from all bitmaps then set your input gamma to 2.2.

2. The 'Select Bitmap Image File' dialogue box - if you want to specify the gamma for each bitmap individually you can use the override button in the gamma section.

Hope this helps,

Dan

[GiacomoGreenspace]

Hi to everybody,
I'm new to the forum and really pleased to join the vray community.

I have a question related to the LWF and I'm sure there's no better place to ask for info.

I switched to LWF since 2 year and I'm pretty happy with it, but still can't find the depth that an image in exponential color mapping (tweaking bright and dark multiplier) can give and sometimes think that LWF is washed out. Love and Hate! So I'm here to post my workflow and ask you if it is correct.

Under the Gamma and LUT panel I switch on the Gamma and LUT correction and set:

Gamma 2.2
Affect Color Selectors ON
Affect Material Editor ON

Bitmap Files
Input Gamma 2.2
Output Gamma 1

then in the vray color mapping panel I set linear with a gamma 2.2.
SubPixel Mapping and Clamp output are off
Affect Background and Don't Affect color ON (so I can save the EXR)

and finally I switch on the sRGB on the VFB.

That should be a correct LWF setting.I am pretty sure it is fine. Can you confirm it?

Now the doubts.

When I load a texture it si displayed correctly and it is not washed out.

When i want a color I have to use a colorcorrector to specify a 2.2 gamma to obtain the same color coming from photoshop or an RGB value. Is that normal? Why?

the fresnel effects and falloffs maps needs to be treathed with a color corrector or they are fine as they are loaded?

a bump or grayscale map is ok as it is loaded from the gamma and lut panel or does it needs a color corrector as well to be used as bump, displace, reflection?

I know it is a little bit tricky and confusing.

In few words what I want to know is how to treat grayscale maps and fresnel for reflection / bumps etc etc etc when I have a correct LWF settings. They are automatically loaded correctly?

Grazie.

Giacomo Arteconi.

[pipjor]

Hi to everybody,
I'm new to the forum and really pleased to join the vray community.

I have a question related to the LWF and I'm sure there's no better place to ask for info.

I switched to LWF since 2 year and I'm pretty happy with it, but still can't find the depth that an image in exponential color mapping (tweaking bright and dark multiplier) can give and sometimes think that LWF is washed out. Love and Hate! So I'm here to post my workflow and ask you if it is correct.

Under the Gamma and LUT panel I switch on the Gamma and LUT correction and set:

Gamma 2.2
Affect Color Selectors ON
Affect Material Editor ON

Bitmap Files
Input Gamma 2.2
Output Gamma 1

then in the vray color mapping panel I set linear with a gamma 2.2.
SubPixel Mapping and Clamp output are off
Affect Background and Don't Affect color ON (so I can save the EXR)

and finally I switch on the sRGB on the VFB.

That should be a correct LWF setting.I am pretty sure it is fine. Can you confirm it?

Now the doubts.

When I load a texture it si displayed correctly and it is not washed out.

When i want a color I have to use a colorcorrector to specify a 2.2 gamma to obtain the same color coming from photoshop or an RGB value. Is that normal? Why?

the fresnel effects and falloffs maps needs to be treathed with a color corrector or they are fine as they are loaded?

a bump or grayscale map is ok as it is loaded from the gamma and lut panel or does it needs a color corrector as well to be used as bump, displace, reflection?

I know it is a little bit tricky and confusing.

In few words what I want to know is how to treat grayscale maps and fresnel for reflection / bumps etc etc etc when I have a correct LWF settings. They are automatically loaded correctly?

Grazie.

Giacomo Arteconi.

bump...i pretty much have the same questions

[jstrob]

Hi to everybody,
I switched to LWF since 2 year and I'm pretty happy with it, but still can't find the depth that an image in exponential color mapping (tweaking bright and dark multiplier) can give and sometimes think that LWF is washed out. Love and Hate! So I'm here to post my workflow and ask you if it is correct.

To understand the depth it can give you can think of a "linear image" as an image you can adjust the exposure. You can understand this better if you do some photography. If you take a picture with low exposure, it will be very different from taking a picture with high exposure and then lowering the levels in photoshop on a non linear image. But if you take your picture with a digital camera in RAW format (or high dynamic range format, what I call a "linear image") you will be able to get the same result in photoshop if your RAW format have enough depth. So you can play with exposure in the computer and also with white balance. And believe me once you start taking pictures in RAW format and adjusting them afterward, you don't want to go back to low dynamic range format like jpg in your digital camera. It is exactly the same thing when rendering with any 3d software. If you render using the linear workflow you will get something like RAW format. For example if you use 32 bits EXR, you will be able to adjust exposures and color in ways impossible before.

Under the Gamma and LUT panel I switch on the Gamma and LUT correction and set:
Gamma 2.2
Affect Color Selectors ON
Affect Material Editor ON
Bitmap Files
Input Gamma 2.2
Output Gamma 1
then in the vray color mapping panel I set linear with a gamma 2.2.
SubPixel Mapping and Clamp output are off
Affect Background and Don't Affect color ON (so I can save the EXR)
and finally I switch on the sRGB on the VFB.
That should be a correct LWF setting.I am pretty sure it is fine. Can you confirm it?

No it's not.
You should leave the input gamma to 1 instead of 2.2. When you import a hdr format like a 360 environment in 32 bits for image based lighting, you don't need to do anything. When you import old fashion jpg (or nay ldr format) for texturing you need to tell max to use 2.2 only for this picture (in the bmp dialog)

You should also put gamma to 1.0 in the vray color mapping panel.
You should leave "dont' affect color" OFF and "Linear Worklfow" OFF (for this one it does not make any difference since gamma is set to 1.0 in the vray color mapping section and this one I think is used if you put 2.2 in the gamma and want it reversed).

Then when you load those images in after effect, you have to set after effect for a linear workflow and your EXR will not be washed out and then you will see the real difference with a old fashion workflow.

Now the doubts.
When I load a texture it si displayed correctly and it is not washed out.
When i want a color I have to use a colorcorrector to specify a 2.2 gamma to obtain the same color coming from photoshop or an RGB value. Is that normal? Why?
the fresnel effects and falloffs maps needs to be treathed with a color corrector or they are fine as they are loaded?
a bump or grayscale map is ok as it is loaded from the gamma and lut panel or does it needs a color corrector as well to be used as bump, displace, reflection?
I know it is a little bit tricky and confusing.
In few words what I want to know is how to treat grayscale maps and fresnel for reflection / bumps etc etc etc when I have a correct LWF settings. They are automatically loaded correctly?

You will not need a color corrector when you choose a color in max. It will be different in photoshop, but photoshop is not working in a linear workflow. So yes it is normal.
If you load jpg for bump or any greyscale effect, you should just import them as is (leaving the gamma at 1.0).

[MoonDoggie]

The "Linear Workflow" checkbox under the color mapping settings are just for converting old images to LWF. If you use an entire setup from the beginning, don't check this.

[jstrob]

Also I just noticed that it's better to put dark multiplier: 1.0, bright multiplier 1.0, gamma 2.2 and "DON'T AFFECT COLORS ON".

Otherwise you will get a lot of noise in the darker area.

If I understand correclty, this way the image will be calculated axectly like if you put gamma 1.0 but only the sampler will try to remove the noise like if the gamma was 2.2...

[gunny263]

since the sticky post was created in 2006, is it even valid anymore? would be good to have a sticky post with screenshots that match current versions of the software