[morbid angel]

go Frances!

[SJ]

SJ, when I render your scene it's very dark and I have to raise the iso quite a lot. In the better render, did you have to do this as well, or was the emitter strength already good? Your fstop is set to 200??

The brightness was definately different, but I had to adjust it also. Don't know anymore and sadly I deleted the mxi. But I think I'll re-render the scenes again anyway. I'll post the values then.

Back to reality - it doesn't sound like this test is an "apples to apples" comparison afterall. The point is not in trying to make the two renders look similar. It is more beneficial to see how the two engines react to the same settings. Even if the settings have different effects between engines. Once a baseline is established, then go from there.
Identical in this case would mean: same camera settings; same lambertian mat with same rgb value; same emitter material; same gamma and burn.

First: that what has been posted is just a part of what I'm trying to test. So it's pointless to assume I'm concluding anything from that alone at the moment.
Second: I want to narrow down the real differences. We all know that a V1 rendering with Burn 1.0/Gamma 2.2. looks quite different from a Beta with same settings. When sticking with these settings we can only repeat that again and again. So I tried to make them look as similar as possible using means that are provided by the M~R-interface to recognize which differences are impossible to "tune away" cosmetically in the M~R image viewer.

Regarding the identical settings: camera settings were similar, materials too, Emitter material and intensity plays no role because light should behave proportional regardless of intensity. Physical light distribution is similar regardless of intensity of the light source. If you illuminate a room with 10 Watts or 1000 Watts is equal for the relations of brighnesses. They stay same, just intensity changes proportional.

If you have any ideas how to nail down the problem: post it.

[seco7]

Fantastic job SJ, thank you for taking this on and putting quantitative numbers to such a vague problem. I really look forward to your continued tests. If there is anyway I can help, please yell.

[Mihai]

Emitter material and intensity plays no role because light should behave proportional regardless of intensity. Physical light distribution is similar regardless of intensity of the light source. If you illuminate a room with 10 Watts or 1000 Watts is equal for the relations of brighnesses. They stay same, just intensity changes proportional.

But in the emitter settings you had it set to 500 000 lumens, what was the emitter setting for the beta render? Also why set fstop to 200, might play a part. What was the fstop in the beta render?

Because illuminating a room with 10watts and then raising the shutter to make it look brighter, will naturally not be the same as having a 100w emitter with the same brightness. It's not the same light energy distributed in the scene.

[Frances]

First: that what has been posted is just a part of what I'm trying to test. So it's pointless to assume I'm concluding anything from that alone at the moment.

Second: I want to narrow down the real differences. We all know that a V1 rendering with Burn 1.0/Gamma 2.2. looks quite different from a Beta with same settings. When sticking with these settings we can only repeat that again and again. So I tried to make them look as similar as possible using means that are provided by the M~R-interface to recognize which differences are impossible to "tune away" cosmetically in the M~R image viewer.

Regarding the identical settings: camera settings were similar, materials too, Emitter material and intensity plays no role because light should behave proportional regardless of intensity. Physical light distribution is similar regardless of intensity of the light source. If you illuminate a room with 10 Watts or 1000 Watts is equal for the relations of brighnesses. They stay same, just intensity changes proportional.

If you have any ideas how to nail down the problem: post it.

You establish a baseline and work from that point. The baseline being identical settings. That is how you discover the differences. From this thread I've learned that a) optimum gamma settings are different between the engines, and b) emitters are different (differences could very well be a simple matter of terminology in the controls).

I think there is a contrast quality issue - the relationship between light and dark is not as well gradated as in the beta version. That's all I've got. I thought gamma was purely a display issue, so if that was the key, then adjusting it at the console would solve it. You've established that the issue is not correctable from the console, so I don't have an answer.

[SJ]

new testing results updated - see first post

[mverta]

Excellent tests, SJ, truly solid.

_Mike

[Mihai]

Please post your camera and emitter settings for the beta scenes.

[SJ]

OK the settings are as follows:

Emitter intensity: 10 000 000 Watts
ISO: 100
Shutter: 1/200 s
fStop: 128

rendered from C4D through Plugin 0.6b

[Thomas An.]

Very rational ... well thought out. Thank you !


The true measure now is (as you mentioned) which one comes closer to reality (we need to do this in a lab environment somehow).

The suspicion is that beta might be the more true ... but ... there has to be some experimental verification.

[SJ]

Thanks for the feedback anyone.

The true measure now is (as you mentioned) which one comes closer to reality (we need to do this in a lab environment somehow).

Yes! Just one little nitpicking comment: one would have to compare Maxwells output with certain reproductions of reality, not with "reality itself" because that's impossible. It would be really nice to have some predefined settings that would make the output look like defined films like say, Fujichrome Velvia, Kodak Ektachrome or whatever. Should be not too difficult to code [thinks the naive photographer, hehe]
And: a histogram for the Image Viewer would be really really helpful.

[mverta]

Yes, SJ, your comment is at the heart of what's being tested here - it's not a rendering of reality, but a rendering of photographed reality.

_Mike

[Voidmonster]

Yes! Just one little nitpicking comment: one would have to compare Maxwells output with certain reproductions of reality, not with "reality itself" because that's impossible. It would be really nice to have some predefined settings that would make the output look like defined films like say, Fujichrome Velvia, Kodak Ektachrome or whatever. Should be not too difficult to code [thinks the naive photographer, hehe]
And: a histogram for the Image Viewer would be really really helpful.

Yes, yes, yes! I would love to have a histogram viewer and choice of film-response emulations.

[Thomas An.]

one would have to compare Maxwells output with certain reproductions of reality, not with "reality itself" because that's impossible

Well, I am not sure about this "film emulation" thing though. (You could be right nonetheless).

If we have:
1. a long corridor with black diffuse walls and
2. we layout gray cards at regular distances thoughout the length of the corridor and
3. we put a light (with a diffuser) at the end of the corridor so the the first card is bright and the last is black.
4. take a precise lux meter and measure the light reflected by each card

Is that "reality" or "photographic reality" ?

[daros]

yes, yes!!! and a since 5 years expired 25 asa Kodak Supra film.