Just received the link to a render engine poll from master zap (ex mentalimages - now autodesk)…just vote for our favorite renderer 
http://strawpoll.me/2958/
gogogo
I can’t open that page…
hmmm…works still fine over here…sorry man…anyone else having trouble?
must be my connection from the office…worked fine from home. How can arnold have more votes then vray?
i think because this link maybe posted on the arnold forums as well vote boys! maybe someone can post it in the max section of this forum too…
Arnold 55% ?
I cannot believe.
yes i cannot believe this too…this poll is still very biased…don’t know why…will post in the max section now…
Sincerely doubt that many people have access to Arnold, unless a large number of them pirated it or something (or just voting based on hype). I wish they would just release a consumer version of it, I’ve only heard great things about it and can’t wait to try it myself.
Feels like people at Sony have been voting.
What’s interesting his this is an initiative from Master Z and Mental Ray is way behind.
Arnold is very popular in communities where custom coders sits next to artists. They simply can get anything coded very quickly and easily so I do understand why its so popular.
People that try Arnold often say it’s incredibly fast, even with high complexity scenes, and without to bother about prepasses.
But some stuff are still missing, for example it seems there is no light-linking, you have to use all lights in the scene.
Anyone can confirm it’s true ?
EDIT : please note that I have no proof of any of this, just heard about it. If someone has extensive knowledge of both renderers, please enlighten us.
EDIT2 : people that really tried Arnold and VRay in this forum say that instead of “incredibly fast”, Arnold is only 10% faster on Brute Force, and slower than VRay if you use interpolation in VRay.
This is false. I have tested an interior scene with Vray and Arnold.
Arnold is more fast than Vray in Brute force GI. But if i use IR, Arnold is dead.
Well, that’s what I said, people that don’t want to bother with prepass get better render times in Arnold.
They think they save time by *not* having to tweak prepasses over and over until it is good enough.
As I haven’t tested Arnold myself I will not argue 
In the future we will probably see independant comparisons appear in the web (with people that know both renderers very well, because if someone knows mostly one of the two he will have better render times as he knows tricks to get the best of it).
It would really be best if you make such statements after you have made the comparisons yourself For example, what I have heard people say, is that the speed of brute force tracing is roughly similar between the two. (Of course, brute forcing an architectural interior is mostly out of the question anyways, due to the required high number of secondary bounces). There is some stuff which Arnold seems to do better (like hair), but we’ll get to that too. There is also seems to be some stuff which V-Ray does and which you accept for granted, which are not present in Arnold.
Best regards,
Vlado
Well, I’m not stating anything, as I asked if someone could confirm (or not) what I have heard.
Most of the hype is about a software that we haven’t even tried.
And I also asked for independant comparisons, which will take time.
Anyway, I would not drop a renderer that I know is very good (and has excellent support), for a renderer that I heard was very good, and has very limited support (Solidangle clearly stated they don’t want to take too much time on the hotline).
I have added an “EDIT” line to my original message, to make even clearer that I ask for comments, and have no proof about stuff I heard from artists.
I have done a more or less in depth comparison and also involved one of the guys who is more of an advanced user in arnold, our result, nearly identical speed between vray and arnold.
Thank you Dmitry.
For the comparison, did you use Brute Force on VRay, or something else ?
I have licenses of both, and using both of them. Arnold brute force is indeed faster, in my tests, not significantly though. Vray looks like more mature product, so you have DR, more output passes etc. I am used more to Vray material system. I love progressive thing from Arnold. Also VrayRT is great. Arnold is more expensive. Vray offers brute force but also interpolation, Arnold does not, with short deadlines this could be really important. Vray is more widespread, easier to find learning materials. I mean it is just preference I think at the end, as this poll shows both are 2 best engines today.
To compare them u need to use Brute Force with Vray. What is point of comparison when you are comparing 2 different things than.
Thank you for you input.
To be fair, I think a comparison should use the fastest method available for each renderer, whatever method is used.
For a given result quality, if VRay is faster using passes, then VRay wins.
The main question is how much time is needed to setup the lighting and shading, and render everything.
The method does not matter.
If VRay render is faster using passes, and the setup of those passes is fast, then VRay wins.
If VRay render is faster using passes, but the setup of those passes takes more time than what you save compared to Brute Force, then it’s different.
And then you add the support part, which we know is awesome in VRay.
And about the price, I don’t know how much costs Arnold, their website is empty, is it some kind of big secret or something ?
You can not look it like that. You compare interpolated result with Brute Force. Interpolated means that Vray guessed a lot while computing, so results are not at that level of quality as with Brute Force, also interpolated are prone to flickering etc. It is not always about how fast you can get final image. For example for animations, if you are using interpolation you need to take care a lot, and work a lot, to have flicker free render. With Arnold or Vray BF you need to calm down noise and you do not need to think about flicker at all, that is great workflow.
Support is great on both sides. ChaosGroup did significant job to bring that on higher level that is for sure.