Finalrender stage 1 v Vray

[1SHot]

Thanks GonçaloP!

what are your AA settings?

VR:

FR:

[marina]

Try using adaptative subdivision for some speed improvement (I think)

Regards

[GoncaloP]

You should change the Vray AA to Adaptative subdivision 0,2. Thats why your test with 3 subdivisions isnt good.

Gonçalo

[1SHot]

It is true!



Thanks

[GoncaloP]

This is my test:
A)Skylight
Vray = Mult = 1 | mult = 0,5
FR = Bounces = 1 | Mult = 1 | mult = 1

B)Skylight
Vray = Mult = 2 | mult = 1
FR = Bounces = 4 | Mult = 2 | mult = 1

C)Skylight + directlight
Vray = Mult = 1 | mult = 0,5 + VrayShadow
FR = Bounces = 1 | Mult = 1 | mult = 1 + FRRaytraceShadow

D) SubSurfScattering

Can you do this tests again with adaptative subdivisions 0,2?

Thanks, keep posting your tests please.

Gonçalo

[TRGraphics]

No offense to 1 Shot but it is obvious that he does not know vray very well and to be comparing the two programs makes no since to me at all.

Another point, I don't no about you guys, but I have never sold a render of a tea pot! What good is such a simple scene doing to show the power of each renderer.

The difference in the programs is going to be seen when they are rendering a complicated scene with lots of shadows and reflections and bouncing light.

I payed a visit to the F/R site last night for the first time and one thing I noticed in almost all the renders was that F/R has a problem with believable shadows close to an object. This is an important issue if you do architectural renders or product renders up close.

It seems to me that this issue alone puts F/R out of the running for production use, at least for me.

I have seen some amazing renders here with vray and I expected to see equally amazing renders at their site, but I didn't.

Thats all the information i needed to make my decision!

If the full time users of a program can't make high quality renders, why should I even try!!!

[1SHot]

No offense to 1 Shot but it is obvious that he does not know vray very well and to be comparing the two programs makes no since to me at all.

It's true, i'm a Maya user,but i try for statical render to use Max + vray/fr.I would want to learn to use one of the two well but I do not know which to choose.

[1SHot]

For GonçaloP:

A)

B)

C)


I don't find the SSS, i remake it.

[photonC]

next one...

FR




Vray

Where can i get this scene and the other scene you use here for tests?

[GoncaloP]

Thanks for posting the results.

I would like to look at that scene, can you email me to gprospero@netcabo.pt?

Regards,

Gonçalo

[1SHot]

I heave problem with mail go here:
http://oneshot.freemail.supereva.it/test2.zip

[photonC]

Thanx now I only need the pendzel's room file.

[rhino]

Set up an image with GI in fr to have good rendertime. Then zoom in on an object in the room for example and hit render again. Now it sucks, because you have to adjust sample distribution again. Vray doesn't have this problem and that's one of the greatest things of it imo.

Also to have detailed and good GI on detailed geometry or underneath beveled objects standing on the floor for example, vray is waaaaaay faster. Just a simple beveled cube on a plane will prove this. Just take a look at these two images in the fr forum, notice the bad GI on the floor just underneath the blocks...
http://www.finalrender.com/forums/fR...9579666;start=

The Absolute check box changes the sample distribution to absolute world units, the solution will be the same regardless of your zoom extent or the size of your scene. This was a major wish from 0 that has been fixed.

In my opinion Vray is superior at capturing tight shadow detail, it does an amazing job even at low settings. Finalrender is capable of the same detail it just takes slightly longer to render but IMO times are still acceptable.

There are a lot of uninformed posts about how the new GI is blurry etc. There are a number of controls for the sample distribution, you have a lot of control over how samples are placed. Understanding these parameters are essential to using FR, please spend some time learning them because you can achieve numerous looks. Complaints of flat blurry GI mainly come from uninformed users who haven't spent the time learning the program. It's simply not a one button solution program never has been.
If you reduce the filter size, this parameter controls how the solution is filtered (GI Options) you will see the GI is almost the same as 0, the only real difference is better light distribution and better edge detection. For example using a low filter size, tight distribution with edge detection (adds another pass to detect edges) you can get extremely "punchy" detailed GI.

Both Vray and FR seem to have problems with Soren Larssons room file. The single sided walls cause numerous problems, light leaks with FR1 with high multipliers and white spots with Vray. (though I don't know Vray well at all, a Vray expert probably knows what's going on) If someone could remake the walls two sided it would be a much better test.

[flipside]

The Absolute check box changes the sample distribution to absolute world units, the solution will be the same regardless of your zoom extent or the size of your scene. This was a major wish from 0 that has been fixed.

In my opinion Vray is superior at capturing tight shadow detail, it does an amazing job even at low settings. Finalrender is capable of the same detail it just takes slightly longer to render but IMO times are still acceptable.

There are a lot of uninformed posts about how the new GI is blurry etc. There are a number of controls for the sample distribution, you have a lot of control over how samples are placed. Understanding these parameters are essential to using FR, please spend some time learning them because you can achieve numerous looks. Complaints of flat blurry GI mainly come from uninformed users who haven't spent the time learning the program. It's simply not a one button solution program never has been.
If you reduce the filter size, this parameter controls how the solution is filtered (GI Options) you will see the GI is almost the same as 0, the only real difference is better light distribution and better edge detection. For example using a low filter size, tight distribution with edge detection (adds another pass to detect edges) you can get extremely "punchy" detailed GI.

Both Vray and FR seem to have problems with Soren Larssons room file. The single sided walls cause numerous problems, light leaks with FR1 with high multipliers and white spots with Vray. (though I don't know Vray well at all, a Vray expert probably knows what's going on) If someone could remake the walls two sided it would be a much better test.

That absolute button is a very good option then! Thanks for clearing this up.

About the GI, I know stage0, and I agree with you that you have to understand the way the samples are placed and it's effect on the GI look. In the beginning I really liked this way of controlling GI detail, but after using Vray, I prefer that way of adjusting it. For me it's just easier to set up, much less scene dependant etc. The tweaking time is smaller with Vray for me.

The tight shadow detail is one of the most important aspects of the GI calculation imo, so fr deinately needs some impovement there.

What is the soren larssons file?

thanks for your insights,

regards,

flipside

[GoncaloP]

I dont know about FR but Vray doesnt like thin one face walls. I usualy do the model with real dimensions, expecialy if the scene is lighted with exterior illumination, it helps to avoid those problems mentioned by Richard.

Gonçalo