LWF questions

[markf]

I have just read a good description of implimenting LWF in the book Vray The complete Guide by Francesco Legrenzi.

 

I had a scene rendered previously with all of the default Max>Preferences.Gamma and LUT values and using Linear Color Mapping with the Bright and Dark Multipliers set at 1.0.

 

I followed the examples and in Max>Preferences.Gamma and LUT section I enabled Gamma/LUT correction, set Gamma to 2.2, enabled Affect Color Selector, enabled Affect Material Editor and in the Bitmap files section I set Input Gamma to 2.2.

 

When I render all of the texture maps are very dark, procedural materials are changed to a much less degree. If I open the texture map from the MAterial Editor and compare it side by side with the same map open in Photoshop (I have dual monitors) they look the same. In the rendered frame buffer, they are way to dark.

 

I think that I'd like to use a LWF but I seem to be missing something. Am I doing something wrong or how should I compensate for this?

 

TIA for any help with this.

[BrianKitts]

in the max gamma settings set the input value to 2.2, but leave the output value at 1.0

 

and then in vray color mapping be sure to either set the gamma value to 2.2 or leave it at 1.0 and check the vrayframe buffer's sRGB button.

[markf]

Brian,

 

Thanks for your reply. I didn't have the sRGB button selected in the frame buffer. That makes a huge difference. I re-read the section on LWF and it does mention that in one method the Vray frame Buffer sRGB button should be selected. If I understand correctly this option does not save the rendered image with the sRGB gamma corrections. One would need to save as HDRI (or maybe EXR?) and this would include the gamma settings. Or one would use the Gamma Overide option in the save file dialog and set it to (2.2 ?)

 

Another method is described that gives me pretty much the same results as my previous render when using the sRGB button in the Vray frame buffer. Instead of Linear Color Mapping one uses Gamma Correction Color Mapping. The Inverse Gamma is set to 0.454 which is the inverse of gamma 1/2.2. I must say that allot of this this is beyond my level of technical understanding. I do like the consistency of the appearance between the material Editor, Rendering Frame Buffer and Photoshop.

 

My texture maps look good but all of the procedural materials need to be tweaked to look the same as I had them before making the changes to the Max>Preferences> Gamma and LUT settings and the new color mapping.

 

I probably should not do this in the middle of a project but there's no time like the present.... I've got a feeling that all of the procedural and non texture mapped materials I've saved in my own Library are going to need to be re-worked.

 

Thanks again for your reply!

[BrianKitts]

Thanks for your reply. I didn't have the sRGB button selected in the frame buffer. That makes a huge difference. I re-read the section on LWF and it does mention that in one method the Vray frame Buffer sRGB button should be selected. If I understand correctly this option does not save the rendered image with the sRGB gamma corrections. One would need to save as HDRI (or maybe EXR?) and this would include the gamma settings. Or one would use the Gamma Overide option in the save file dialog and set it to (2.2 ?)

 

That's correct, that's my typical workflow.

 

Although sometimes I leave the gamma in vray set to 2.2 if I'm doing an animation writing out to jpg files so that the gamma is embedded.

 

If you ever render an image from same setup (gamma burned into the image) and save as an EXR if you open it in photoshop you'll have double applied the correction....and it will looked washed out. So you just need undo one of the corrections by doing an image> adjust>exposure and apply a gamma correction of .4545 to fix the image.

 

If you've got that down your pretty much 80% of the way there, the more you work with it the more it will make sense. Once you get it down then you might start experimenting with it in other ways to see what works best for you. For instance I remember someone only using LWF for animations and not stills, or I use a gamma of 2.2 for all exteriors, and 1.8 for interiors. I follow the opinion that the science drives the concept, but at the end of the day we are artists and what gives you the best result is the way to go.

Edited February 17, 2009 by BrianKitts

[markf]

Although sometimes I leave the gamma in vray set to 2.2 if I'm doing an animation writing out to jpg files so that the gamma is embedded.

 

...........use a gamma of 2.2 for all exteriors, and 1.8 for interiors. I follow the opinion that the science drives the concept, but at the end of the day we are artists and what gives you the best result is the way to go.

 

Do you set the Gamma at 2.2 in the Gamma Correction Color Mapping roll out?

 

It seems like if I were to use one set of gamma parameters for one scene and another set for another that would make re-using materials from a library problematic.

 

If I'm about 80% of the way there, what is the other 20%?

 

Last, Do you typically save out stills as HDRI or EXR? I have read about this and was thinking of trying it. I typically have the final image as a .psd file and I typically supply my clients with a highest quality .jpg as a final image for their use.

 

Thanks for your help!

[mi75]

I feel I'm about 79% of the way their personally, still searching for the last 21% of understanding.

 

If I'm rendering a still image I'll generally render to a .VRIMG or EXR file and apply the .4545 gamma correction in PS. Then I'll save a jpg for playing with in Lighteroom and PS. That way I have an original back-up HQ version for my own records. Not sure if it's the best way but it works for me.

 

So at the end of the day I have a EXR, PSD, Jpg.

 

I just ordered the same book and can't wait to read it.

[BrianKitts]

Do you set the Gamma at 2.2 in the Gamma Correction Color Mapping roll out?

only if I'm saving out to straight to jpg. If I'm saving to EXR then it's set at 1.0

 

It seems like if I were to use one set of gamma parameters for one scene and another set for another that would make re-using materials from a library problematic.

No, you only need one library.

 

the easy answer:

You're always applying the same gamma correction level, only if you do it one way it's A+B=C, if you do it the other way its B+A=C either way the end result will be the same.

 

the longer answer:

Gamma @ 2.2 in Vray saved as jpg > 2.2 correction applied in max = correct image

-OR-

Gamma @ 1 in Vray saved as EXR > photoshop recognizes the 32bit images and automatically applies the 2.2 correction to the image when it opens and you still have a correct image.

 

where many people go WRONG

If you set Vray to 2.2, save as an EXR, then photoshop when opening the file applies the correction to the 32bit image you end up with a washed out image, because the gamma got double applied = wrong answer.

 

If I'm about 80% of the way there, what is the other 20%?

80% = getting it setup right and getting a quality result that's usable

20% = understanding the process and why you do it, as well as how to fix it if you ever get an image that doesn't look correct.

 

Last, Do you typically save out stills as HDRI or EXR? I have read about this and was thinking of trying it. I typically have the final image as a .psd file and I typically supply my clients with a highest quality .jpg as a final image for their use.

Our workflow for stills is to render to 32bit exr. Open in photoshop and make any final adjustments. If it's overexposed or underexposed you can pull your image back into a workable color range using the exposure control in photoshop (the whole reason we work with floating point images) after the image is set, convert to 8 or 16 bit image and apply your photoshop work, keep your working file saved as a PSD. When all is said and done, save out final as a level 10 JPG. (I find that to be the best balance of quality and file size)

Edited February 18, 2009 by BrianKitts

[markf]

Brian,

 

Thanks for your reply. How do you make a reply with multiple quotes like you have done above?

 

>>Gamma @ 2.2 in Vray saved as jpg > 2.2 correction applied in max = correct image

-OR-

Gamma @ 1 in Vray saved as EXR > photoshop recognizes the 32bit images and automatically applies the 2.2 correction to the image when it opens and you still have a correct image.

 

I am using the Gamma Correction color mapping. I saved as a .tga with an alpha channel. I have Inverse Gamma = 0.4545 and Gamma = 1. Renders in the Vray Fran\me Buffer (without using the sRGB button) and opens the same in Photoshop.

 

I saved as .exr with all of the same settings except I deselected Clamp Output (from my reading I believe this needs to be deselected for creating a "floating point" image.) When I open it in Photoshop it is overexposed. I use the Image>Adjust>Exposure and change Gamma to 2.2 and it then looks identical to the other .tga file.

 

This seems to be different then what you are describing. I'm using the Inverese Gamma and you seem to be using the Gamma. This is confusing. Am I doing this "right"?

[BrianKitts]

I don't use the gamma correction color mapping, I think the need for that kind of become extinct when they added the gamma value into the other color mapping options. So, sorry I really can't answer your question directly as I've never used that one.

 

To get the full range of colors in a render you should be rendering in linear color mapping to start with. This way the exposure controls don't limit your render. Once you have that under control then you can start to play with the other modes. I typically use Reinhard for everything which is a blend between exponential and linear. Exponential clamps any over-bright colors, Linear allows them to burn out. If you use Reinhard you can use the burn value to adjust the blend between linear and exponential which basically allows you to control the clamping of the lighting.... want the sun to burn out more, raise the burn value...... burning too much, take it down. It's a great option.

 

And yes don't clamp your output, with the exception of if you are using a color mapping that will do it for you.

 

(to do multiple quotes I just copy a couple of times what's in the text box when you originally hit the quote button to reply, then just erase the extra text that doesn't apply to that part of the answer)

Edited February 18, 2009 by BrianKitts

[markf]

 

I was using Reinhard until I started down this LWF path. It sounds like what your saying is I can maintain the LWF with Reinhard, Linear, Exponential etc color mapping by setting the Gamma in the color mapping roll out to 2.2. If saving as an .EXR file I could leave it at 2.2 and re-adjust in Photoshop>Exposure, or Set Gamma to 1.0 and enable to sRGB button for an acccurate view nin the frame buffer and then no need for Exposure adjust in Photoshop.

 

And yes don't clamp your output, with the exception of if you are using a color mapping that will do it for you.

 

I'm not sure I am understanding this. All of the Color Mapping types allow for Clamp Output. My understanding is that Clamp Output would cut off all of the color values above 255,255,255. These colors can only be saved in a HDRI or EXR floating point file type. If one is saving as .jpg, .tga, etc it would be advisable to use Clamp Output. If one was saving as .EXR, Clamp Output should be disabled because it cuts off the higher range values that are the whole purpose behind saving as EXR. Does this sound correct?

 

Thanks for your help with this. It's kind of confusing and I have spent an entire day adjusting the materials in my scene to the new Gama settings and experimenting with this. I think I'm going to like this. I can see a noticeably closer match in how my maps appear in Photohsop and in the Material Editor as compared to the finished rendering. Same with the colors in the Material Editor as compared to the finished rendering.

[BrianKitts]

I was using Reinhard until I started down this LWF path. It sounds like what your saying is I can maintain the LWF with Reinhard, Linear, Exponential etc color mapping by setting the Gamma in the color mapping roll out to 2.2. If saving as an .EXR file I could leave it at 2.2 and re-adjust in Photoshop>Exposure, or Set Gamma to 1.0 and enable to sRGB button for an acccurate view nin the frame buffer and then no need for Exposure adjust in Photoshop.

 

Yup. To tell you the truth the more you'll probably start using exr and see the control it gives you in post you'll quit saving anything out of max as a jpg except for quick previews and drafts. Anything for production you'll save out to EXR

 

BTW you keep talking about saving out to HDR or EXR, EXR has better compression but is still a lossless format, so there's really no good reason that i know of to save to HDR if EXR is available..... saves server space

 

I'm not sure I am understanding this. All of the Color Mapping types allow for Clamp Output. My understanding is that Clamp Output would cut off all of the color values above 255,255,255. These colors can only be saved in a HDRI or EXR floating point file type. If one is saving as .jpg, .tga, etc it would be advisable to use Clamp Output. If one was saving as .EXR, Clamp Output should be disabled because it cuts off the higher range values that are the whole purpose behind saving as EXR. Does this sound correct?

 

If you save it as a jpg its going to clamp it off anyway because a jpg can't hold colors brighter than 255,255,255 anyway, so why do it before the fact. Leave it off in case you do end up saving to EXR that way you have the full range of the image. Personally I can't think of a scene where I've needed to clamp the output (specially using reinhard)

Edited February 19, 2009 by BrianKitts

[markf]

Yup. To tell you the truth the more you'll probably start using exr and see the control it gives you in post you'll quit saving anything out of max as a jpg except for quick previews and drafts. Anything for production you'll save out to EXR

 

What type of process besides exposure control does the EXR format allow for? Do you have a typical type of adjustment you use? I notice that most of the other filters and adjustments are ghosted out in Pshop for the EXR.

 

Thanks again for your help with all of this!

[BrianKitts]

What type of process besides exposure control does the EXR format allow for? Do you have a typical type of adjustment you use? I notice that most of the other filters and adjustments are ghosted out in Pshop for the EXR.

 

EXRs benefits are mainly the exposure control, and the ability to contain multiple render elements all in one file. (zdepth, shadows, specular, etc)

 

Yes most filters can't operate on a 32bit image (this is really dumb on adobe's part) I try to get my lighting set in max so that I don't have to adjust the exposure of the EXR in post.... but it's a nice safeguard if you have to.

 

I convert the image to 8bit after the lighting is right, apply all my lighting changes as layer corrections in photoshop. That way if you have to re-render you can open the new EXR convert to 8bit and drop it into the PSD in place of the old image and all your corrections are identical to the previous image.

Edited February 19, 2009 by BrianKitts

[markf]

EXRs benefits are mainly the exposure control, and the ability to contain multiple render elements all in one file. (zdepth, shadows, specular, etc)

 

I have not used render elements, but I understand the idea. I'm going to try using it. The EXR format allows one to render elements, other file types do not allow for this?

 

 

I convert the image to 8bit after the lighting is right, apply all my lighting changes as layer corrections in photoshop. That way if you have to re-render you can open the new EXR convert to 8bit and drop it into the PSD in place of the old image and all your corrections are identical to the previous image.

 

I typically work with adjustment layers and etc. and dropping re-rendered scenes as you describe.

 

thanks very much!

[BrianKitts]

no there are other multi-layer formats, such as an .rpf but right now EXR is the most popular.

[Thablanch]

only if I'm saving out to straight to jpg. If I'm saving to EXR then it's set at 1.0

 

 

No, you only need one library.

 

the easy answer:

You're always applying the same gamma correction level, only if you do it one way it's A+B=C, if you do it the other way its B+A=C either way the end result will be the same.

 

the longer answer:

Gamma @ 2.2 in Vray saved as jpg > 2.2 correction applied in max = correct image

-OR-

Gamma @ 1 in Vray saved as EXR > photoshop recognizes the 32bit images and automatically applies the 2.2 correction to the image when it opens and you still have a correct image.

 

where many people go WRONG

If you set Vray to 2.2, save as an EXR, then photoshop when opening the file applies the correction to the 32bit image you end up with a washed out image, because the gamma got double applied = wrong answer.

 

It is not exactly the same..

When you work in gamma:1, the sampling is done on darker surfaces, and the correction applied after, when you work in 2.2, the calculations are made directly in the rendering, wich gives you a slightly faster rendering and a better quality of image...

 

The overall look is the same, but you will need higher settings to achieve the same result.

[BrianKitts]

It is not exactly the same..

 

That's true, I didn't mention the part about using the "don't affect colors" option which I do use when rendering in either 1.8 or 2.2 gamma. which aleviates the situation your mentioning.

 

Didn't want to make it any more difficult by bringing that part into it, but you're correct as I presented it.

[Thablanch]

http://www.spot3d.com/vray/help/150SP1/tools_vrimg2exr.htm

 

You can work in regular 2.2 mode without having the double correction in the export. So you keep the quality..

 

Also allows you to take whatever passes you need, as the .vrimg contains all the data.. a little like the rfp files, there is a few scripts out there to do that.. ( and I read that the next Vray buildt in converter will be able to export the separate passses as well)

 

So you can keep a full 2.2 workflow, either you save .jpg / tif or vrimg

[markf]

That's true, I didn't mention the part about using the "don't affect colors" option which I do use when rendering in either 1.8 or 2.2 gamma. which aleviates the situation your mentioning.

 

I'm not sure if your still watching this thread or not but I will give it a shot. I read about the option. I think I understand how it changes when the gamma is applied and why rendering at gamma 2.2 with "don't affect colors" results in a higher quality GI and faster render time.

 

I rendered my scene with Reinhard, Multiplier 1.75 and Burn 0.7. I rendered to an exr file using gamma 1.0. Next, rendered to a tga file with gamma 2.2. Those two renderings appear identical. Then I rendered exr with gamma 2.2 and "don't affect colors" enabled . The rendering is quite noticeably different from the previous two. It is significantly darker.

 

Am I mis understanding or using incorrectly the "don't affect colors" option? I would like to have the advantages of rendering with 2.2 gamma but the "don't affect colors" option seems to return an incorrect result.