[Mihai]

A good idea is to place a neutral colour in your scene in a test render to use as a sample for the colour offset, and then use it to balance the scene in photoshop, but it's still not doing it with the full data range that maxwell see's in the mxi.

Why not load the MXI using the PS plugin? Or if you don't prefer that for some reason you can always save an exr from the MXI and work on that. Regarding the whitebalance I agree it would be a nice feature to have but it's not as crucial because in the real world you have no control over the temp of the light sources.

[zdeno]

Why not load the MXI using the PS plugin?

I know it sound like blasphemy but not everyone uses Photoshop or version which handles 32bit files.
and it would be great to deal with .mxi in maxwellrender window, it doesn,t sounds ultramega hard to do in my humble opinion.
BW request always appears in this forum throug all years, so many users have problems with it, specially with sun&sky outdoor scenes.

[Mihai]

Neither Photoshop or any other image editing application that can handle 32bit files? I'm sorry, but you shouldn't be doing CG then I mean I hope this doesn't sound arrogant, but it's like saying Win95 is good enough for me! There is really no excuse for not working in 32bit especially if you want to make these kinds of edits to your images.

As I said it would be a cool workflow feature to have "WB" in Maxwell, but I think some are missing the point here of why white balance is necessary in the real world, and what it does. It's sort of like the question of why you need two exposures to expose both for indoor and outdoor lighting. You have a WB setting, now what? Do you white balance in this case the yellow lights so they are more or less white? But then the sunlight will turn blue....so I hope people don't expect a "fix it all" setting, you still have to know what you're doing from the beginning. In this case you would be a bit screwed, WB setting or not, unless you had used color Multilight - in which case, set the lights to 6500K - you've "white balanced" the light from the emitter.

If you think from the beginning what the different light sources temp would be in the real world, but you just want a small difference in color cast, don't use 2600K lights together with physical sky for example, set the lights closer to the physical sky sun temp and you get less of a difference...or don't use yellow/blue lights at all if you don't want a color cast in your scene, you can have perfectly white emitters.... That's why I'm thinking how useful this would really be if you can already "WB" before hand (and afterwards with color ML which would be more flexible than an overall WB) and it's so easy.

Just saying, don't get too hung up on a feature name, and that it's "missing" when we are not dealing with the constraints of a real world situation.

[zdeno]

Neither Photoshop or any other image editing application that can handle 32bit files?

photoshop cs2 so for architectural purposes it is more than enough but 32bit possibilities are very rough

I'm sorry, but you shouldn't be doing CG then

aye Sir! I am an architect so this is not my primal target. CG is rather hobby

I mean I hope this doesn't sound arrogant, but it's like saying Win95 is good enough for me! There is really no excuse for not working in 32bit especially if you want to make these kinds of edits to your images.

with my canon that I bought oneday I get canon software that can magically sets BW in my RAW files with one click.

As I said it would be a cool workflow feature to have "WB" in Maxwell

I totally agree with You as always

, but I think some are missing the point here of why white balance is necessary in the real world, and what it does. It's sort of like the question of why you need two exposures to expose both for indoor and outdoor lighting. You have a WB setting, now what? Do you white balance in this case the yellow lights so they are more or less white? But then the sunlight will turn blue....so I hope people don't expect a "fix it all" setting, you still have to know what you're doing from the beginning.

It is mostly needed for getting rid allyellowishbubblegum of 5777K from sun. blue from sky is not so strong comparing to pure power of sun on clear sky

In this case you would be a bit screwed, WB setting or not, unless you had used color Multilight - in which case, set the lights to 6500K - you've "white balanced" the light from the emitter.

setting emitters to 6500K ? now it is blasphemy. maybe it works different in post-process multilight features, but doing something like that is fail wrong and even worst. scene lit with 2600 K with postprocess WB would looks completelly different than used 6500 K emmiters without using WB.

If they would looks the same in maxwellrender. then something is very wrong from beggining.

If you think from the beginning what the different light sources temp would be in the real world, but you just want a small difference in color cast, don't use 2600K lights together with physical sky for example, set the lights closer to the physical sky sun temp and you get less of a difference...or don't use yellow/blue lights at all if you don't want a color cast in your scene, you can have perfectly white emitters.... That's why I'm thinking how useful this would really be if you can already "WB" before hand (and afterwards with color ML which would be more flexible than an overall WB) and it's so easy.

Just saying, don't get too hung up on a feature name, and that it's "missing" when we are not dealing with the constraints of a real world situation.

translating 2600 K in wave equivalent is one of the most stunning features in maxwellrender. If someone tell me I should use MORE WHITE LIGHTS insted of physical correct one so there is something wrong.
when I render in biased render I use pure white to not fight with funny GI second bounce colors.
I hope maxwell treated this RGB-spectral conversion this way http://www.google.pl/url?sa=t&rct=j&q=a ... zQ&cad=rja

- so emmiters should be set on REAL color (as in real life)
- frambuffer in maxwellrender should looks little crappy at first look (as in real life RAW files without WB proper set)
- White Balance should be used in POSTPROCES (as in real life - processor in my canon does it for me, or canon software fixed it on my RAW fies)

-so if maxwellredner is "easy as taking a photo" so it would be nice to have few presets, and one place to set KELVINS, and one pickdrop to point white wall with one click telling maxwell "THIS SHOULD BE WHITE"
-as I said , it don't look so impossible to tallented coders in NL.

and last ...
this is Half Life's fault he was thinking WHY WISHLIST is so empty here You go Jason !

[Mihai]

Thanks for decomposing my post. So you work in RAW, why not find a free application than can read EXR files and do the point and click WB? Why not just set the sun temp a little higher, say 6000K?

Define "completely different". So if I take a photo of a color chart with a 2600K bulb, my camera being set to 6500K white balance, and I white balance that color chart, it's going to look completely different than if I took a photo of that chart with a 6500K bulb, without white balancing it? I can agree with extreme color temperatures you can't "get back" certain colors depending on the temperature, but I think you exaggerate with this "completely different" statement.

[Half Life]

this is Half Life's fault he was thinking WHY WISHLIST is so empty here You go Jason !

I say go for it -- I'm all for better post processing options inside Maxwell. Personally I really dig Motiva Real Camera and would love to see similar functionality built into Maxwell

http://www.motivacg.com/descargas_en.php

Also, even though I was an Adobe Photoshop Certified Expert I gave up on Photoshop (and Adobe overall) due to their ridiculous pricing and upgrade policy flip-flopping and MVP (minimum viable product) "upgrade" policy and release schedule.

I have been searching for the best alternatives, and have found some suitable stand-ins for most of what I used Adobe products for... quite literally somebody would have to pay me to use Adobe products again.

If that means I should be banned from CG then I'm sorry I don't belong to the overpaying suckers club

Best,
Jason.

[zdeno]

Thanks for decomposing my post.

Your welcome. It was easy. nothing to talk about

So you work in RAW, why not find a free application than can read EXR files and do the point and click WB? Why not just set the sun temp a little higher, say 6000K?

I don't say it is impossible. I said it would be good feature to have WhiteBalance in maxwellrender.
As I understood maxwell is operating on spectral data not pure RGB, so I assumed , RAW .mxi modyfication in maxwellrender would be more accurate than in 3rd party software. But looks like I overrated this spectral solution, and plain RBG computation (like levels in photoshop) would do the trick.

Define "completely different". So if I take a photo of a color chart with a 2600K bulb, my camera being set to 6500K white balance, and I white balance that color chart, it's going to look completely different than if I took a photo of that chart with a 6500K bulb, without white balancing it? I can agree with extreme color temperatures you can't "get back" certain colors depending on the temperature, but I think you exaggerate with this "completely different" statement.

I didn't run tests so it was my assumption, but You have right. phrase "completely different" is much too strong.
Difference would be more subtle (I think) but still even little glitch in HUE is easily read in our brains as FAKE and all photorealistic goes home to sleep.

try something like this. find in maxwellrender forum a picture with notreal colors, and desaturate it. It would look quite convincing and reallife. because our brains are easy to fool with brightness/darkness level , but not with hue or saturation level.

I don't know exactly how mulilight with color works inside, but I think it is only POSTPROCES merging different .32bit mxi files, and have nothing to do with GI solutin in scene.
so setting emitters at 2600 K in all red cornell box, would give us different results , than setting emitters at 9000K in the same red cornell box and AFTER render is done changing sliders of emitter color back to 2600K.
I assume red material bouncess off much more energy from RED TINTED light than from BLUE TINTED one. so all scene energy solution should be different or even completely different ergo colormultilight is just funny surogate of physical correct solution, but still very usefull.

and at the end. I am still at old version of maxwell so maybe I should not complain about new wishlist or features but still adding White Balance into maxwell ? +1

[Half Life]

and at the end. I am still at old version of maxwell so maybe I should not complain about new wishlist or features but still adding White Balance into maxwell ? +1

Sure, except this is the not the Wish List forum.

Best,
Jason.

[tom]

I don't know exactly how mulilight with color works inside, but I think it is only POSTPROCES merging different .32bit mxi files, and have nothing to do with GI solutin in scene.
so setting emitters at 2600 K in all red cornell box, would give us different results , than setting emitters at 9000K in the same red cornell box and AFTER render is done changing sliders of emitter color back to 2600K.
I assume red material bouncess off much more energy from RED TINTED light than from BLUE TINTED one. so all scene energy solution should be different or even completely different ergo colormultilight is just funny surogate of physical correct solution, but still very usefull.

Maxwell Color ML is not such a post trick. MXI stores spectral illumination/solution per each pixel and Color ML spectrally transforms the resulting emission accurately before ToneMapping.

[zdeno]

THX Tom for joining in

... MXI stores spectral illumination/solution per each pixel ...

It sound very good. so I assume again, doing White Balance tunning on pure .mxi in maxwell core would give us more accurate solution than rough RGB approach in .hdr files outside maxwell ? or am I thinking too much ?

[tom]

Right. But, it's still not a big loss until we add these controls. Because, currently doing it on 32-bit render isn't any worse than doing it on the RAW image from a digital camera.

[zdeno]

indeed.good point.

for me it is much more easier to whitebalance picture (RAW) in dedicated canon software than guessing which slider have to be moved to rebalance colors in photoshop. so still I think this feature would be very helpfull. or I have to run through some tutorials for photoshop how to deal with this issue. Maybe I just don't know tools to do it quickly and efficently.
I think this subject is clear now. THX for patience and eplanations