Hi everyone, I'm currently doing a series of tests with some renderers to compare the refraction of some glass materials. For that I made a basic model of a product, a simple glass bottle and in its inner part another glass and I tried to get closer to the maximum of the reference using Corona Render, and Fstorm. The purpose of this discussion is to understand a little more how the refraction difference works in the software calculation and understand which delivers better quality.
Reference: https://prnt.sc/IoAo48ru8qFq
We can see in the reference some effects generated on the slightly darker edge and we can notice a high brightness in the reflection.
Example - https://prnt.sc/A6Gsb4D6-p-O
Now let's compare the reference with the images generated by Corona Render
Refraction Corona Render - https://prnt.sc/8-c9y0tApEGn
In the corona render on a studio light, we can see low reflection compared to the reference, and we can also see that we don't have dark parts on the edges. As much as I place objects and or replace all reflections to an HDRI it still doesn't come close to the reference.
https://prnt.sc/d9Twxw7QL0_z
In the link above I tried to add a little absorption and dispersion, it gained a little volume but it was darker than the reference.
Now let's compare the refraction of the Fstorm renderer
https://prnt.sc/7FF1afcnWEA5
We can see that in the Fstorm rendering the refraction brought details closer to the reference, bringing the darker areas but a little stronger. but still it can be improved. so to bring a little more detail of the refraction and the light I added some imperfections such as fingerprints and scratches to see if the volumetry of the glass would be sharper and get even closer to the reference.
https://prnt.sc/r_NpyybHL_cz
The model was a little darker, more highlighted the bottle and brought a little more detail. but still it is not similar with the reference.
I would like to know about your analysis, if there is something that I am doing wrong, or to understand a little the behavior of the refraction calculation or maybe to do more tests with other renderers such as vray. let's go. I would like to know your opinion.
The text was a little big, but I thank those who contribute to this analysis.
Probably best to share this on the corona forums as they have quite a technically solid active community.
Its hard to investigate further without seeing your tone mapping or material setup but I can also see those dark edges you refer to in the corona render, however they appear mute/grey. I would assume this is due to your tone mapping, fstorm in my opinion has excellent tone mapping with default settings.
In order to post an answer, you need to sign in.
