So...general thoughts and experiences with SI2013.

[ActionArt]

Are your color management / pass gamma correction settings the same? Just checking because I already made that mistake and was frustrated for a bit.

[rray]

Yes, they're the same, tried a lot of things but found no way to get the details back. I'll probably drop a post at the mental ray forum. Shadowmap and detail shadowmaps shouldn't produce that different results without using transparent shadows.

[TwinSnakes007]

As a hobbyist, is MR attractive because of the level of integration into SI, versus other render engine alternatives?

You know, shader support, pass support, the whole render pipeline support that MR has, that other render engines may not.

[ActionArt]

Just personal preference I suppose and it has a lot to do with what you want to accomplish. I've tried several other alternatives and it looks great at first but I always run into a wall somewhere unexpected that destroys the experience. I would recommend trying others like 3Delight just to see how they work for you. It's not hard to try it.

Like it or hate it, MR can do just about anything and although people say it's slow in comparison to others, my experience is the opposite. For what I try to do day to day it's faster than anything else I can get and has unmatched flexibility.

You just need to spend a bit of time really getting to understand it, and it should be fine.

I'm really liking this version so far but just started testing.

I highly recommend this book to get started. It's great reading for anyone rendering, not just MR: http://www.amazon.ca/mental-ray-Maya-3d ... 73&sr=8-10

Only a few settings have really changed since this was written so it still applies.

[Daniel Brassard]

Those three guides helped as well

http://www.mentalimages.com/products/me ... rials.html

And this book although expensive (and old Mr3.4)

http://www.springer.com/computer/image+ ... 11-22875-3

And The book Writing Mental Ray Shader (if you're in for a challenge)

http://www.amazon.com/Writing-mental-ra ... 3211489649

[Mathaeus]

As a hobbyist, is MR attractive because of the level of integration into SI, versus other render engine alternatives?

You know, shader support, pass support, the whole render pipeline support that MR has, that other render engines may not.

I'd believe the SI render pass system is independent of particular renderer - at least, even in old Affogato to Renderman translator (which has no any support for Render Tree), it was possible to have different shaders in partitions, through custom properties.

Today, 3delight supports a lot of Render Tree nodes, also it does it's own optimizations when compiling. So no big deal to use the mix of three or four hair shaders in Render Tree, passed through final constant shader for transparency - or something like that.
Support for "direct" load of Renderman sdl is still here - so no problem to use the sl code written for another 3d apps.

Another question are habits, it takes some time to "feel" the different render engine.

Regarding MR, for me the last attractive feature belongs to 32-bit age, when MR was able to render what, for example, Lightwave renderer ( from LW 7.6 or so), wasn't - on the price of standard MR bad performance. Today.. actually I don't know... even MR has something attractive, simply I can't find anyone who want to use it. Where I am, everyone is Vray master, a few LW-Modo-and-so guys, that's it.

[rray]

(Just a quick follow up regarding the shadowmap issue I had with 3.10. Increasing bias from 0 to 0.0001 resolved this for me. Thanks for the tip from gunter on the mental images forum)

[gustavoeb]

[Hirazi Blue]

So what kind of hardware (esp. graphics card) was used to make this video?

[gustavoeb]

old geforce gtx 275...

ps: nice signature haha

[iamVFX]

I like new random number generator feature

[EricTRocks]

Did you log this as a bug? It won't get fixed otherwise.

[xsisupport]

I checked this out.

What's happening is that the array of locations changes every time the tree is evaluated.
So, every time you select in array, you are selecting from an array with a different order.

That's the issue and I'll log it.

iamvfx, I apologize if you feel I gave you a hard time in the xsibase bug thread when I asked for repros.

[iamVFX]

I checked this out.

What's happening is that the array of locations changes every time the tree is evaluated.
So, every time you select in array, you are selecting from an array with a different order.

That's the issue and I'll log it.

Thank you! Because of that I can't finish this compound (sneak peak, sort of ):

iamvfx, I apologize if you feel I gave you a hard time in the xsibase bug thread when I asked for repros.

You don't have to apologize, Steve, you do a great job helping us out.

I have no desire to report this bug because array of locations feature should be working in current release already.

Shame on beta testers.

[Tekano]

oh! I just found out something neat the other day iamVFX the get neighbouring particles compound allow you to search by cone (vector) so a quick modification of that allows you to do some neat viewport particle lighting effects! I imagine you should be able to get your closest point by vector in similar way to this.

also, don't be so hard on the beta testers, having done it for Softimage before it is not easy work, you get regular builds every week or so and really only have the time to test what you are currently working on or a new feature thats part of that build. - thats if your company allows you to run the beta version in the first place! there are thousands of ICE nodes - and millions of lines of code - do you really expect everyone to check every aspect of these for every build? its impossible.

[iamVFX]

Has anyone tried any of the python specific methods? Is it a joke or what?