Studio oject render

[opifex63]

Corona studio renders.
And comparision to Vray.

At the first image i want to show the way Corona work with glass.  In real life, glass, under heavy light, usually light up it self in re-refraction.
So many years i try to force Vray to do that..)) But Corona, easily, did the same trick. And i like it a lot!
(render time approximately same, about 1 hr.)

Second render shows how re-reflection in gold changes color from light yellow to orange and gold starts to shine.
In this case vray renders metals slightly wrong. But i have to say, that this sconce in real life not so clean and so shiny....))

P.S. Best regards to developers team! Good luck guys! :)

Vladimir Urgalkin

[Ondra]

Nice pictures!

But isn't there a possibility to enable "reflect on backside" in vray? It does not help?

[Ludvik Koutny]

Nice pictures!

But isn't there a possibility to enable "reflect on backside" in vray? It does not help?

Yes, vray has "Reflect on backside"option, but for some reason it is disabled by default. That is probably the source of rumors that vray sucks for glass rendering.

[opifex63]

Nice pictures!

But isn't there a possibility to enable "reflect on backside" in vray? It does not help?

Thanks! ))

I just checked Vray scene, in fact, this option was enabled.
And, as you can see "Sub-pixel mapping" and "clamp output" are disable.
This scene was rendered with IM+LC scheme. But "brute force" do the same.
(At least in this scene)

So i can say it for sure, that the vray sucks for glass rendering. ))

[Ludvik Koutny]

In vray, go to global settings in render settings, enable refl/refr depth override and set it to 20, that should make a good base for comparison ;)

[opifex63]

In vray, go to global settings in render settings, enable refl/refr depth override and set it to 20, that should make a good base for comparison ;)

Let's see.. )

[opifex63]

In vray, go to global settings in render settings, enable refl/refr depth override and set it to 20, that should make a good base for comparison ;)

There is no big difference, as i can see... May be a little bit better, and little bit more saturation in glass edges.
But not even close to Corona.
And, as i expected, render time increase dramatically even in low resolution and with low DMC settings.
So... what else can i check out to prove my point? ;)

[maru]

To have realistic glass material in vray you should:
-enable "reflect on backside" in material
-set "reflection cutoff" and "refraction cutoff" to 0
-increase max ray depth in render settings
-remember about proper fresnel reflections and IOR, absorption etc depending on your material

There shouldn't be much difference in quality - it's simply raytracing...

[Ludvik Koutny]

It was not about proving a point... It was just about correct comparison. ;)

[opifex63]

It was not about proving a point... It was just about correct comparison. ;)

A correct comparison was not my intention..
I just wanted to show how Vray and Corona draws a glass in a normal situation.

But i was render one more test..
- reflect on backside: enable
- reflection cutoff: 0.00
- sub pixel mapping: disable
- Clamp output: disable
- Max depht: 20
- All material settings and IOR: same in Vray and in Corona.

As you can see, there is no big difference between first Vray image and last one..
But in real life in Vray i will never set-up my scene like that.. It is pointless, i think. :)

P.S. By the way, i use vray since its first stable beta. And i know what i'm talking about.
My personal imho - glass in vray sucks..))

[racoonart]

Nice test and a great model!
I'd be really interested where the difference happens. I mean, technically it's still raytracing, so casting rays on geometry with a specific refraction index, a number of bounces and so forth. If you look closely on the reflections and refractions it's behaving really different - not only the brightness.
So, is it just the brdf (Blinn is the Vray default afair) which is different to corona's or is it some other trickery.. ?

[Edit] I just noticed that the model (rotation) and maybe the lighting itself is not really the same in both pictures :-(

[opifex63]

It's strange.. I agree with you.
There must be no difference in raytracing process. But I can see this difference clearly.
Angle of this model in this renders and way it light itself really not the same. In vray there is photometric light inside bulbs, and in corona there is a emmiter material on bulbs. But I dont really care about it.
Because the way vray-glass looks don't depend on it.

[maru]

So many years i try to force Vray to do that..

i use vray since its first stable beta. And i know what i'm talking about.

No offence, but considering these two posts it's a bit strange you didn't know about "reflect on back side" and "refl/refr cutoff" for glass.

A correct comparison was not my intention..

I just wanted to show how Vray and Corona draws a glass in a normal situation.

If you want to show how Corona and Vray do the same thing, then why won't you want to use correct comparison? I want to see it because it's really interesting. Please make a comparison of EXACTLY the same scene - preferably with exactly the same lighting (use emitter geometry or area lights with physical units) and then we will talk. :)

I suspect there might be some differences in how filtering works in different rendering engines - how it treats areas smaller than one pixel.

[opifex63]

No offence, but considering these two posts it's a bit strange you didn't know about "reflect on back side" and "refl/refr cutoff" for glass.

What makes you think that I did not know about this? o_O ))
"reflect on back side" i enable always.
 

[maru]

I concluded this from others' comments. My bad.