What scene is able to compare VRay and Maxwell ?
Smth, that could be easily noticed! No skylight or phys sky (for 1.47.03). Just pure light emmiters and materials! Typical scene full of boxes or smth complex ?
I think I don’t know !
Do you ?
P.S. I think it’s time to make available VRay 1.47.1x ! Really interesting less noise with QMC !
It very much depends on what you want to compare… in any case, do not expect identical results even for the same scene - for example, the maxwell image is always going to have vignetting, whereas the V-Ray one will not.
There are some simple scenes for which the global illumination can be computed by a formula, and V-Ray has been verified to produce correct results in these cases. I can post them here if you are interested.
Best regards,
Vlado
A big problem with the comparison is that maxwell requires you to use maxwell materials. And vray works best with vray materials (even though you can use standard materials, but not raytraced or the new “architectural”)
Until maxwell can render something in a reasonable amount of time, why bother. It’s to slow to use in production for most people. Unless you like grainy images that is. 
Tony
actually, with the latest build, its ALOT faster. still cant compare to vray, but its at least closer to brazil.
I mostly mean not rendertime, but quality of lightnig. If it said, that “maxwell - 100% physically correct result”, interesting to see the difference between VRay and Maxwell in identical lighting situation.
V-Ray can also produce physically correct lighting. Just because Maxwell is based on real-world physics, doesn’t mean all other render engines are not. Admittedly, V-Ray does not work with spectral calculations internally, but in (the majority of) cases where this can be ignored, V-Ray is physically correct.
Best regards,
Vlado
A little off-topic - but I’ve wanted to ask if there are any short-term plans to eventually incorporate this into Vray or if that is something which will take alot of re-programming?
A little off-topic - but I’ve wanted to ask if there are any short-term plans to eventually incorporate this into Vray or if that is something which will take alot of re-programming?
Bad idea - I want VRay 1.5 this year 
LOL - I was implying the 2.5 release.
Guys, infact I am at this moment sitting in the office on a sunday and doing a vray/maxwell/real world comparison. So, I will post my results in a while. My test focuses on comparing vray to real photgraph and maxwell to real photograph, as comparing vray to maxwell and vise versa is not correct. There for I made a stand point of photographs to mimic their lighting and conclude. My focus also is to determine the reasons for vray gi saturation and default calibration, gamma correction and etc.
I have made some interesting discoveries…
Me too…I’ve seen that most colour bleeding problems are history when applying a gamma curve. Even the combination gamma-curve and intensity exponential color mapping seems to be a very interesting combination.
I’m looking forward to seeing your results
Some images to illustrate it:
gamma-correction color mapping (g2.2):
As you can see it is essentially linear with the inevitable burnouts
intensity-exponential color mapping with gamma correction in post:
IMO right balance between color preservation and brightening of light areas:
Ha that attic model brings back old memories
Nice images, but I wonder what a real pic would look like in that lighting situation. Wouldn’t there be burned out areas too if you adjust your cameras exposure to leave the walls so bright?
Actually I am not quite interested how a camera image would look, I would be more interested how a HDR image of that situation would look.
But as Vlado has explained, Vray makes physically correct lighting calculations, so it only depends on materials and tonemapping if an image looks real or not. But what’s real? I would say if it looks real it is real. And even if it doesn’t look real, it can still be satisfying. In case of the image above, it is satisfying, because it maps a larger dynamic range into one picture.
If you look at professional photography, there are also many approaches to tonemapping, depending on the kind of film - or in case of digital cameras, depending on the software and how it is treated.
Look at this image below for example, it doesn’t necessarily look realistic, but it has a lot of color information a normal picture wouldn’t have, and certainly looks artistic
I agree Gijs, I just talked about comparing with a real picture, because you said ‘right’ balance, which is indeed very subjective and can differ a lot for various purposes.
The intensity color mapped image indeed looks nice, and will probably look balanced very well for most people (especially clients who usually hate burn outs)
The nice thing about intensity exponential is that keeps the saturation up to a certain level, whereas exponential washes out all colors and HSV keeps the saturation even in the hottests spots in the image.
Interesting tests…I’ve noticed that by lifting the gamma to 2.2 the whole image brightens, not respectivly to shadows. Unlike in maxwell, where the darkest areas are preserved.
I am interested to see some pictures to understand better what you mean with this.
---
Additionally, this is a picture with a red flooring. On top the rendering with the gamma corrected version below.
As you can see the color bleeding is acceptable and IMO quite natural.
Impressive Gijs, I definatly need that linear space workflow tutorial