More goodies for subscription users! iray!


As I have said, I am all for partnering…there is real power in that for sure…what I mean by innovation is bringing in the use of new tech and better usages of new hardware…I am sure I am not the only max user that feels this way…I am really looking forward to what XBR reveals for us max users in the future. That being said, the implementations of the tools that exist already for max is fine, but they are already being developed and made available for purchase or free in some cases…a few of the plugins that were integrated I already had purchase licenses for ie…polyboost, CAT, because I needed them, it’s great that they are now integrated, but as others have said…these things get integrated and then development or new innovation within them slows to a crawl…I mean Pflow for example…what an amazing piece of tech that AD developed, but here we are years later after the implementation and it is still not multithreaded, and completely ridiculous to work with compared to other applications particle itegrations…Ie…ice, and maya’s particles, which are practically realtime in the viewport…then there is cloth…I was glad to see that a few new things were added to the last release, but every other package can run realtime sims of cloth in the viewport…we still wait for ever to sim these things…reactor, well that has been the same since it was implemented and good to see that new tech is being put in place there…hopefully it gets the attention to be developed and bring it up to speed…Hair was a great partner add on, but is slow in comparison to other plugins and other software package implementations…so even though the partnerships are a great thing, if that’s the strategy then why are most of the implementations outdated in comparison to what’s currently available in other packages? They need to be developed to bring max back up to at least what other apps have had for years…The bottom line is yes I am happy to have these new “partner” features, and I am a happy Max user, I just want to see it get the lovin it deserves, and the users are looking for. And yes Iray seems like a cool thing and I am sure that some people will benefit from it, but from all the demo videos I have seen on the area and at siggraph…it’s not any faster than current solutions available, in fact it seems a bit slower…anyway…I gone on enough…you hopefully can see where I am coming from, and know that there are plenty of other users that feel the same way…if not then we will just have to wait and see what the future brings and hopefully it’s bright for max.


one last thing I have to say…I realize that this is an extension release…and not a “point” release or a “full” release…so with that in mind I can understand the extensions being existing tech that is becoming integrated…and I do think they add value to the subscription plan…I can only hope that for the future full releases, we will see either dramatic improvements to the existing integrated elements, or a push toward new features that will make users workflows faster and better and are not just “currently available” plugins that users can already purchase if they need them…


They have been doing that. You could never buy the hair, cloth, ProOptimizer, Slate, or Graphite tools that are currently inside of 3ds Max. All were available as 3rd party tools, but the developers have worked with Autodesk to improve the tools and add new features to them that in some cases wouldn’t be possible if they were still 3rd party plugins.

Now on the topic of included versus 3rd party tools. Hairfarm exists to address needs that some users need in a hair system (same with Hairtrix), but not everyone needs it or every Max user would be buying it. Same reason why Shave and a Haircut (same technology that Hair & Fur is based on) is available for Maya. Different users need different featuresets.

That’s my 2 cents,


This is surely untrue, Eric, as it assumes that the 3rd party developers involved never intended to incorporate new tech/tools/innovation into the plugs in the state they existed when purchased by AD.
And as always advertised by AD, the original devs remain onboard as AD employees to continue development of their respective tools.

What most people seem to be irked about is the fact that newly incorporated 3rd party tools are most welcome, but should surely be improved upon with every iteration of Max. I mean, i still see threads posted here with people asking questions about the Physique modifier…
It has always seemed to me, as a Max user since R4, that whilst we’re extremely loyal Max advocates, we more often than not seem to be playing catchup in terms of bleeding edge innovations in opposition with Maya/Xsi/Houdini.

And considering a FREE app like Blender has come on leaps and bounds since it’s inception in terms of what can be achieved with it’s toolset, i think loyal,paying customers of Max deserve a little more love.

Would it be such a bad thing spending six months to bring some of the existing tools up to scratch before adding new ones.


IMHO the way Autodesk have developed Max by letting 3rd party plugins flourish and then acquiring them seems to be the totally wrong way to develop a cohesive application.

You have products with little or no interoperability, their own ideas of UI design, it’s no wonder that it’s such a complete mess when they’re brought into Max.

3 character systems? One of which the creator has left Autodesk, so who knows whether that is another stagnant feature?

Frankly, I’m not optimistic about Max - I feel it will just pootle along with mediocre new features and the odd gem that a talented third-party will create.


Actually it is very true (read my statement again) some of the stuff he has done with Graphite Tools never would have been possible without having access to Autodesk source, and being able to compile the code in C++ as part of the 3ds Max core. Same is probably true for Slate, as some of the access they would need to do what has been done has to be done at the source level. Also, most of the recent additions haven’t been purchases, but have been long term development deals to ensure, I hope, that past mistakes don’t repeat themselves.



Dude, if you want people to take you seriously and read your posts try to formulate them into coherent paragraphs and sentences. This wall of words and ellipses’ is hard on the eyes.


Yes, Philip did leave, but Stephen is now working with Autodesk. So there must be something in the contract that they can only have one Taylor brother on staff at a time or something like that.



:hmm: i thought we had finished this discussion … i just have to frown.

we are right in the middle of a big renewal of 3ds Max which conists of 3 stages. If a not wron, stage one was completed with 3dsmax 2011. So there is a lot more to come…

I am not a plugin developer, but my common sense tells me, that he must be someone who has experience in 3D and experience in programming. But in a lot of cases these guys are no experts in
human or user interfaces. If a remember the original polyboost menu, it was a beast to handle.
It can’t be better, when these guys are integrated in the 3dsmax development team and get support when renewing or integrating their plugins.
And the speed max develops is absolutely right.
i can’t imagine, that since april 2010 everybody is 100% into slate, max composite or the 3dsmax material library.
i can speak only for myself, …i still have a lot to learn.

We are getting a new physics engine and a realtime raytrace…Iray is something i need right now. And the fact the Physx isn’t GPU supported is good, because other wise i would have to replace my brand new ATI card.

I wonder if Autodesk will support ATI or in the future AMD gpu’s?

As a resumee i must say, right now i am happy whith AD’s version policy. And like Ken said, the have learned there lessons.


I would say it would make more sense to get these guys onboard earlier, when an application is at alpha or beta and offer UI advice or support at an earlier stage, rather than wait until it’s finished or a couple of versions in, then decide to get them in, and having to spend time retooling it to fit.

As for a new dynamics engine… Well, not really - people have been able to download the PhysX engine for a while now, for free.

I’m not sure what the extent of the integration into Max is, but it’s not new.

As for the GPU thing, there’s always CPU support, and considering that nVidia own the physics engine, it doesn’t seem too unreasonable to expect it to be supported on their own graphics cards.


Nvidia is the one doing the mental ray/iray (they acquired mental images back in 2007) and physx stuff. They are basing their GPU acceleration on their CUDA technology, which is currently only supported on Nvidia hardware, AFAIK. So if you want to GPU accelerate Nvidia software you will need Nvidia hardware (preferably something based on the Fermi architecture). Otherwise you will probably be stuck with CPU only processing for that technology.



I know that … and agreed!
But ATI (in the Future the gpu will also be called AMD) has it’s own
stream technology.

(… you can find more in youtube)

perhaps there will be some kind of translator, which will give us perhaps not 100 % gpu support, but 90% gpu support.

AD would risk, to make itself dependent from NVidia. And if you remember at the beginning of 2010 there were no direct11 cards from nvidia available. That would be a very serious situation.


Using OpenCL will solve the multi vendor problem in theory. For nVidia OpenCL runs on CUDA, for Ati OpenCL runs on Stream.

So on paper an OpenCL app will run on both platforms, but from what I’ve read Ati has made a bit of a mess when it comes to implementation, documentation and development tools.

For example Vray RT GPU is made for OpenCL but doesnt run on Ati cards, the comments of the Vray developers speak volumes about why. Adobe went for a CUDA setup for there Mercury playback engine because again they found the OpenCL on Ati sub par, more examples on google.

So my bet is if you go do anything GPGPU go for nVidia, you get you get CUDA, OpenCL, PhysX and APEX and so on.


I think that ATI should surrender and develop some kind of bridge between Stream processors and CUDA technology, they should forget about their OpenCL implementation.
They win some battles against Nvidia, but this is one that they have lost, and they should accept that for the shake of their buyers.




it give no problems with iray and ati, because all mental images products work on both NVidia & ATI.

hot chip


Hot Chip…in GPU mode?

How can you achieve mental images products working over ATI in GPU mode, not CPU mode ( of course iRay will work over CPU mode also, but AFAIK iRay is only Nvidia compatible for GPU mode) ?



iray runs on CUDA, so it looks like it’s nVidia only for GPU mode.

“iray achieves its high level of performance by taking full advantage of the CUDA programming model, allowing interactive previewing on both single and multiple NVIDIA GPU platforms”


“The exact GPU hardware specifications vary with content and application requirements and CUDA-enabled GPUs will be required for high performance. While all CUDA-enabled GPUs are supported, iray’s memory requirements will make NVIDIA Quadro®, QuadroPlex® and Tesla® systems for workstations as well as Tesla server units the platforms of choice in most situations.”



yes you can read in all forums, nvidia/cuda + iray. That is marketing, because nvidia had buy mental images, but the developer team of mental images say:“We are programming for nvidia & ati”. Not all functions work on a ATI-Card, but all this new technics need time. Believe me, at the next year, comes ATI with a graphiccard, there works iray with GPU much better as now. At last, at this time give it only the Fermi-Cards from NVidia, that make sens to use iray.


Johnathan you missed the more important info from the FAQ.pdf:

15. Will products with iray run on NVIDIA GPUs only?
While iray requires NVIDIA GPUs for interaction and high-performance, products such as mental ray with iray and neuray with iray will continue to support all platforms that our customers require and for which their developers see a sufficient market.
16. Does iray also work without CUDA hardware?
When running on a system that lacks compatible NVIDIA GPU resources, iray operates as a CPU-based renderer, albeit significantly slower. iray can also operate in a “hybrid GPU/CPU mode” where it can utilize any collection of local GPU and CPU resources to compute a single image.
20. Does iray run on non-NVIDIA GPUs?
21. Will there be an OpenCL version of iray?
There currently are no plans for supporting OpenCL.
So Thorsten is incorrect in his statements. An ATI card WILL NOT accelerate iray rendering, according to the FAQ. I don’t have any ATI cards so I can’t test to verify this.



oh sorry, than was my knowledge not up to date.