DAZ Studio Pro BETA - version 4.8.0.4! IRAY

[t_3]

To be honest with you, my kind of smell a rat, if daz was secretly, doing this in the background.
All the time while T3 was developing. The plug-in. (To be honest, I can see them making his life quite difficult)

Withholding source code etc (would certainly explain why he's been having trouble)
Because let's face it daz would not exactly want a superior render engine to the own.

"of course" they didn't inform me of their plans. interestingly though, if i now look back: when discussing some of the problems that i needed to solve with daz devs, there were a few sentences about similar problems they had to solve internally, that make much more sense now. my interpretation on this is, they were secretly working on iray integration since last summer the least.

my biggest concern though is the future of the studio sdk. there are a few things that could be added to the sdk, which would allow me to make ocds work better and faster in some particular tasks (that i have told them already). on the other hand they are constantly adding features to studio that is not even properly (or at all) reflected in the sdk - simply because the available interface is still 70% from v3, 20% from v4, and only 10% v4.5, with no updates to 4.6, 4.7, and now 4.8. i also know that other plugin devs had asked for additions in the past, apparently with no much luck (much in contrary to i.e. paul and smith micro).

on the other hand there is something good about iray also: it's pretty clear that content creators will jump on it as fast as possible. now the iray material system is obviously much more close to octane than the rigid 3delight "fake" material structure. so, if at all possible, it might open some door to auto create similar octane materials from iray materials that come with new products... lots of new work and investigation needed, but i will of course get on that and also try to get help from daz with that. so, if this would be possible, it would also make working with octane much more easy for ocds users, given octane is in other aspects still superior to iray.

a more general concern is the influence of this move to not only my plugin but other plugins and octane as well; studio is relatively easy to handle and now provides unbiased gpu/cpu rendering for free. even if not on par with octane, you can now simply use it just like the octane standalone (or even other unbiased renderers), means to pipe out render jobs also from other hosts. and i bet daz is going to push that aspect pretty soon, since more studio installs to them only means more potential content sales.

even nvidia must be clear about the fact that there is a notable difference if they partner with autodesk to have iray in max, or if they allow to ship a free product with almost complete iray integration - but of course they just like to sell more high-end priced cards

so this:

QFT, but not just because of this I haven't trusted them much ever. Their business model and what they do to maintain it...ugh. I'll just leave it alone

is also how i feel about the current situation.

[Spectralis]

DAZ has always changed it's sales strategy without much concern about customer input despite claiming otherwise. But what's in their favour are their very lively forums and their relentless approach to releasing content and expanding the DAZ universe. They're very good at creating a buzz and responding to problems with the content or software. Reallusion is pretty good at that too but I think DAZ tops the lot from a customer point of view.

As I do a lot of animation and have a lot of their animation products they sent me a survey about future animation developments so perhaps they're planning quite a few significant updates in a variety of areas in the future? The integration of an unbiased renderer being one of them.

[linvanchene]

obsolete post edited and removed by user

[Spectralis]

Whenever you change a setting the DS Iray viewport switches to a grey default rendering mode - it does some calculations and then displays the updated image.

I asked this question earlier but somehow it's disappeared. If iRAY take minutes to recompile a scene after any changes then animation must take forever. How does it update each frame without spending ages recompiling scene changes during animation rendering?

[Spectralis]

even nvidia must be clear about the fact that there is a notable difference if they partner with autodesk to have iray in max, or if they allow to ship a free product with almost complete iray integration - but of course they just like to sell more high-end priced cards

I am curious to see what the exact plans of nvidia are.
But for once I will refrain from public speculation.
I do not want to open up another can of worms right now.

But at the end of the day it could be worse.
What if DAZ would secretly have worked together with AMD?

Based on some reactions on the DAZ forum I get the impression that now a lot of people are considering to buy new nvidia graphic cards.

From that point of view having purchased a new video card can be considered as "sunk costs".

compare:
http://en.wikipedia.org/wiki/Sunk_costs

Now the upgrade price to OctaneRender for a lot of people is not anymore "New Videocards & OctaneRender license" it is now just a question of motivating the other DAZ users to purchase an OctaneRender license.

So basically it is now up to Otoy to convince all those new Nvidia card users that they have something to gain by upgrading to OctaneRender.

I think this will prove impossible unless the price of entry for Octane Render lowers significantly. Ok, iRay is not yet close to OR DAZ v2 plugin in features but it'll probably be out of beta before OR is. iRay can either use CPU or a NVidia GPU and it's FREE! No need to buy an expensive license to gain entry. The $500 cost of the OTOY license can be spent on a GPU instead. I can't imagine how OTOY can top this.

The only way I think OTOY can make their plugin more appealing is if it can create much faster and better renders. Even a 30% increase in rendering speed at much higher levels of quality would make the plugin more appealing but I'm not sure that's possible based on the difficulties t_3 has experienced getting up to date documentation from DAZ. I assume there's also the limits to what is possible with unbiased renderers so maybe it's not possible to increase rendering speeds significantly.

[Hydra]

Physix is now free too...

NVidia is turning its software position into a beat stick against AMD. I Note that AMD isn't doing nearly as much as NVidia. I mean... buy a NVidia card, and use your CUDA cores OR OpenGL. Buy an AMD and you can do openGL only.

The choice is clear.

[t_3]

The $500 cost of the OTOY license can be spent on a GPU instead.

on a side note: octane is in euros, and since eur <> usd has nearly dropped to equality (which is probably going to last for a while), the entry point is already notable lower than $500.

I can't imagine how OTOY can top this.

apparently not by the price.

one point is that - like with 3delight - daz is also not developing iray and probably has no control over nvida's plans. so they depend on what nvidia is going to do with iray, or how fast they integrate new features. otoy is pushing everything into octane, it's future and to keep it on the bleeding edge of technology. because of that i.e. people use octane for max even if max also has iray.

most probably there will always be a significant difference in feature set and overall speed (apparently daz doesn't plan to implement realtime for example) that makes octane a worthwhile investment, even for some daz users. octane was a niche product in this area, and there is still room to grow with iray, and maybe even because of iray...

[t_3]

Even a 30% increase in rendering speed at much higher levels of quality would make the plugin more appealing but I'm not sure that's possible based on the difficulties t_3 has experienced getting up to date documentation from DAZ. I assume there's also the limits to what is possible with unbiased renderers so maybe it's not possible to increase rendering speeds significantly.

fortunately the render performance doesn't depend on problems with the studio sdk.

quite contrary: i took a route that allows me to devote best technical possible performance to internal tasks like i.e. mesh building - in exchange for lots of troubles with the api. had i used the standard render integration (like some wise body here told me already), it would have been a lot easier (and faster) to create a reliable integration, but there would be no way to provide the current level of interactivity and speed;

i.e. a fully collapsed, medium complex scene with 2m tris has a build time of less then 4 seconds on my machine (4-core haswell, plugin mesh transfer and octane mesh building together) - means this is the time needed to spend for frame changes within an animation (smoothing is probably going to last longer if you use it). also i don't reload textures for every frame, i do create material and geometry hashes to avoid costly mesh building or material updates for parts that didn't change, i even turn off post processing during a frame and apply it only once at the end to squeeze another 2-3% better speed out of octane. turning on coherent rendering during animation/final is pushing render speed by 30-50% alone. imo there is no doubt that octane (and ocds) is faster, but on the other hand it of course depends a lot on the materials you use, so an objective comparison is always difficult...

[Hydra]

IRay will have another side effect. Ever look at how crappy some Daz stuff is? If its Daz provided, you may have some issues trying to do a good render. Other artists provide much better models but its not always clear which ones do.

A lot of people are about to discover how poor their opacity and spec maps for hair are, and dang, prebaked hair shaders suck in this light. Who put the shadows in that scarf before rendering?

Any ways, I stand by my claims that it will be 6 months before Daz has this properly rolled out.

[Spectralis]

Even a 30% increase in rendering speed at much higher levels of quality would make the plugin more appealing but I'm not sure that's possible based on the difficulties t_3 has experienced getting up to date documentation from DAZ. I assume there's also the limits to what is possible with unbiased renderers so maybe it's not possible to increase rendering speeds significantly.

fortunately the render performance doesn't depend on problems with the studio sdk.

quite contrary: i took a route that allows me to devote best technical possible performance to internal tasks like i.e. mesh building - in exchange for lots of troubles with the api. had i used the standard render integration (like some wise body here told me already), it would have been a lot easier (and faster) to create a reliable integration, but there would be no way to provide the current level of interactivity and speed;

i.e. a fully collapsed, medium complex scene with 2m tris has a build time of less then 4 seconds on my machine (4-core haswell, plugin mesh transfer and octane mesh building together) - means this is the time needed to spend for frame changes within an animation (smoothing is probably going to last longer if you use it). also i don't reload textures for every frame, i do create material and geometry hashes to avoid costly mesh building or material updates for parts that didn't change, i even turn off post processing during a frame and apply it only once at the end to squeeze another 2-3% better speed out of octane. turning on coherent rendering during animation/final is pushing render speed by 30-50% alone. imo there is no doubt that octane (and ocds) is faster, but on the other hand it of course depends a lot on the materials you use, so an objective comparison is always difficult...

If the OR render speeds can be pushed further than iRay then that's great. One other advantage that might help OTOY is cross-compatibility. Now I don't know whether this is possible but if it was possible to render a project from one plugin in another - say DS to Lightwave - then that would be a massive advantage for those of us who use more than one 3D program. Set up a project in DAZ, add more elements in Lightwave and render using Octane. I have no idea if this could work but considering OTOY has a range of plugins based on the Octane Render architecture maybe a way could be found to integrate them much more? Neither iRay nor Indigo could compete with that.