Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2026 Daz Productions Inc. All Rights Reserved.
Comments
They all are not fast when using CPU , that is the advantage here to use GPU other way make not sense and a single card for $250 will already kill any CPU rendering performance and it is all about , if 3DL used GPU for rendering there would be not much issues but if you have to wait hours for each renders and you need to do 100 of them this is killing the sprit , PAs need fast tools to do the job and 3DL is not what can deliver it that fast with less good shaders and old crappy Uber environment , you as customer can take hours for rendering it is your hobby you can take as much time you want, not if this is your job where time is money . 3DL is a great engine than can deliver the top quality for professionals but not in DS , not the way it is now and not with single CPU and it will never again as it is on the way of extinction already and as fast as the technology move in next 2 years you will have nobody that want to support it as nobody will do products for it anymore . My prediction
...Unfortunately my CPU does not have the routine which supports the Speed Boost in Lux 1.5 so I am still dealing with glacial render times.
Also R4 was buggy as all get out messing with assigning render cameras and not converting all materials, particularly from scenes created for an older version (including older patches of R4).
...Cath, you are the first to say what I have been sensing all along, 3DL is on the the road to becoming obsolete and given the speed of the tech curve it won;t take long. Wouldn't be surprised to find Iray being the only render engine in Daz5.
So looks like this "free" programme has just become more expensive to support as well as proprietary to one brand of hardware now, sad. So much for Carrara as well as it takes a lot more to convert Iray shaders to use there.
Makes part of me want to just pull the plug and uninstall it all.
Gonna' go out for a beer and think this over.
I doubt it will happen for a few years, given all the content with only 3Delight shaders, or shaders for both 3DL and Poser. And, as I said before, the Iray Realtime mode (which actually uses OpenGL and CUDA, if available) should enable folks that currently dislike the focus on Iray only content to use it with their existing hardware more readily than they currently can.
+9000 For me the key to being a "serious artist" is expending time, effort, and practicing to learn and improve. No matter what the medium.
(and personally learning new methods is half the fun)
@kyotokid I have said it before but this free program "requires" absolutely no expensive hardware to support. I'm on a 4 year old inexpensive laptop. I model in a program that is free. Sure If i spent a bunch of money some stuff might be easier and faster, But would that somehow magically make me a better artist? Don't think so.
Sorry but you're mixing stuff up. Maybe this is where your confusion stems from. For instance, modeling apps have nothing to do with materials.
In studio terms, to set up a material you as a "lookdev" (the artist who, well, sets up materials) only need two things: a shader from your "TD" ("technical director" who writes the shader code) and textures from the texturing department.
Your "TD" in the 3Delight/DS case would be Omnifreaker who wrote the UberSurface2 shader that you can buy in the store here. It goes on sale every now and then, and it's affordable (hint hint: I'm Russian, not a rich one, so when _I_ say "affordable", it truly is).
Your texturing department are the PAs.
You don't need their Iray materials. You just need to know where the maps go in _your_ shader and how to finetune parameters to get a specific look.
That's all. This goes for every renderer.
Unfortunately I see this happening as well, and a masterful stroke by Nvidia to force those who want to do 3D Art and to have fast renders but do not own one of their cards to buy one of their video cards.. This is where Otoy with Octane could pull a coup de grâce because it caters for all if it lowered its price by a lot, or released a version that only cost $50 to $100..
But yes the proprietry nature that seems to be slowly happening with Studio does not bode well for some of us..
You mean nobody in the DS world? Yeah, could be.
But the whole point is that the users who want to render in 3Delight or any other software, if they take some time to learn how their preferred software works, they don't need to have their vendors explicitly supply materials for those renderers.
Right?
Definitely! =D
Well no surprise I am past the limits of a 16GB 5 year old i5 / intel HD Graphics 3000 in iRay. I wonder how many GB RAM I need to render 12 different Genesis 3 characters on screen at the same time.
3DL is not "on it's way out" any more than Pixar's RenderMan. It is not obsolete here in "DAZneyLand" nor is it obsolete in the film industry. I disagree with Mec4D on CPUs being useless. You're in hobby space, different world. I have machines with 48 and 64 Xeon CPU cores and RAM exceeding 256GB, some with RAM capabilities of 2TB INSIDE the machine (I can't afford 2TB of RAM). Blade servers make my machines look anemic on CPU cores and memory. Most professional level equipment (farms) have thousands of CPU cores and they do just fine. Another point is that these things also do not normally run in Windows, the overhead of the OS is too high. The vast majority of render farms are linux based as referenced by numerous studios in interviews. If you want to see some fast rendering under 3DL, then output a RIB and send it to a linux 3DL setup. The free standalone can now use up to 8 cores. It is also possible to offload processing to GPUs using custom code and the 3DL Sx Library. If anyone is actually using Sx to leverage GPUs they're not crowing about it.
GPU rendering is still not very trusted in the industry, and when it IS used, they're not using piddly little Titans plugged into PCI-e slots. They're using VCA's or 1 and 2U Tesla compute engines.
Kendall
So at a consumer level what? 64GB RAM and a quad core i7? Just for a DAZ render with say 16 Genesis 3 characters at FHD.
The point is that DAZ gave users access to PROFESSIONAL level rendering long ago, and for (mostly) free. To be used on consumer level hardware, but it could also be used on higher level machines. I remember when people were complaining that 3DL was "too hard" and required "unaffordable monster machines" and this was not long ago. Now people are complaining that Iray needs the "unaffordable hardware" and 3DL is now for the "consumer".
3DL is still used a lot in VFX and Film, DS is probably the only "hobby" platform in its portfolio. The hobby market is fickle, it flits from one thing to the next. The professional markets not so much.
Iray is neat, I like it a lot, but CPU rendering is not a dead process especially where one has access to a lot of them. GPU processing was used to get around specific problems when using CPUs, mainly OS licensing and the high-er electricity draw of CPUs at the time. Now, GPU cards/units pull huge amounts of power. Tesla units pull 1200W+ and need dedicated cooling systems, Titan-X can pull 375W. More than whole multi-CPU systems pull.
Don't forget that while CUDA can do a lot, it is still a limited platform. OpenCL is even less capable. These systems are for specialized types of processing, not for full general purpose computing.
There is something that does need to be mentioned. DNA Research, the company behind 3DL, isn't big on GPU processing. Whether they decided to leverage it at all is still only known to them. The new versions of 3DL have moved over to path tracing and to using OSL now, putting them on par with most other rendering engines (this was publicly mentioned through the 3DL for Katana release a week or two ago, I'm not breaking NDA). Whether these moves will push/allow the use of GPUs in 3DL is still unknown outside of DNAr.
Kendall
OSL has also been in the 3Delight for Maya for several months, as a "pre-release feature". It's awesome. I keep meaning to try and figure out if OSL totally requires the new Nodal Scene Interface API or if it can be used with "oldschool" RIB generation.
Indeed. However, I'm not taking any chances with NDA. It was never clear whether the "pre-release" information constituted a sanctioned "release" of information and released the NDA on that part. The Katana release is definitely "santioned" and therefore safe.
Kendall
I'm not on support with DNA Research (maybe one day, if 3D ever becomes a source of income), just a free user, and the OSL news and instructions how to enable the pre-release features were posted in the public 3Delight forums, so I'd say it has been legit to speak about since March =)
OK, thanks. I don't have nVidia so I am CPU rendering on a laptop. I found out though today after many iRay rendering out of memory shutdowns by Windows, changing my iRay rending in DAZ Studio to optimize for speed rather than memory stops my out of memory Windows closings of DAZ Studio.
The apps may be free, but the cost of a videocard if you were unlucky enough to have a system with a AMD/ATI card installed makes things a bit more interesting.. And well I decided to head on over to Otoy to see what system hardware was needed to run it.. Interestingly enough Version 3 of Octane Render is all Nvidia now, which I am pretty sure was not the case back in version 1..
...I do not have any of the Ubersurface Shaders as when they came out, I was still working in 32 bit and UE would crash the app almost every time. That was when IU stil was gainfully employed, nNow being on a a tight fixed income, I have to watch every purchase I make. I even had to let my PC+ membership expire early this year.
What it would mean is having to basically design my own shaders from the ground up. Not going to happen as I don't have PS.
I still own version 1 of Octane it is only Nvidia , it is about CUDA technology not just GPU
anyway that is for me for this thread .. have fun debating
...considering the fact most new content is coming out with Iray only shaders (which also do not translate to Carrara very well) you could fool me.
That'it's nice that you can afford such beefy systems, I cannot afford to build a dual 16 core Xeon system to speed up 3DL and UE (Crikey the server OS alone costs several thousand). I spent nearly 18 months scrimping, saving and purchasing components when I could afford them (on my meagre salary) for the system I currently have. To even build the base system for optimal performance for Iray would be around 6,500$ (based on 2 Titan-P GPUs with hydro cooling which I estimate will be around 1,400$ each as the Titan X is now obsolete). As I am on a fixed income that simply ain't going to happen unless I get a windfall big enough to cover the cost. My current 11 GB (I always subtract 1 GB for Windows) i7 system cost me 1,400$ (sans display) to build several years ago.
When I first got into this I could create really nice scenes on a 4 GB 32 bit notebook.
Today it seems to require a boatload of memory, fast multi core CPUs and beefy GPUs, making it get far out of hand price wise.
...that is disappointing as I thought they were going with OpenCL which would allow for the use of AMD GPUs as well. I have two HD 7950s that I was looking to use for Reality/Lux until Lux development pulled the plug on GPU and hybrid rendering.
It's not that I want 30 second renders, I just don't want them taking days.
I'd also liek to spend most of my time creating pics, not sweating nit picky details.
Kendall , You missed the point of my post , with 4 core CPU and 3DL you can do as much , it can't even use proper HDRI .. and you talking about film industry .. not with 3DL in Daz Studio .. and what Professionals do with render farms and 3DL engine in other professional programs have nothing to do with Daz Studio , and definitely not on a single CPU with half way features that is the reality
Btw Titan X use 150W while rendering in iray and max 170Watts when my cards are over clocked to the max clock speed .. .. even in games it don't get higher than 274Watts .. since it is the best power efficient card from all card bellow Titan X ah and it use 9 Watts while idle .. I know because I use it .
are you going to spend 12K dollars on a CPU's to have the speed of 3000 Cuda cores ? no you don't .. CPU rendering is about to die out soon , it is not the future
...and that is why I am pretty much feeling left behind especially when a new GPU (Titan P) costs almost as much as it did to build my entire system (when I was gainfully employed). Back then it was a beast, now it's a lamb. All I needed a GPU for was running the displays and working in OpenGL mode (which is what teh Viewport used). Back then 1 GB was more than sufficient. With Iray, if you want decent render performance, you need a lot of VRAM and cores which costs. 3DL only needed a decent hyperthreading thread CPU and physical memory.
Physical memory is relatively cheap these days. I can buy 128 GB for a little more than what 24 GB would have cost me when I built my system. What my system has now would cost less than my 1 TB HDD did. Crikey, I remember when 4 MB of RAM set you back a pretty.
GPU's on the other hand are expensive. Of the 6,000$ estimate to build what I consider is an optimal system for Iray (including the OS), 2,400$ of that is taken up by two Titan Ps (more if someone like EVGA comes out with a Hydro version). Not possible on my fixed income.
This is why I am becoming very discouraged
My system is less powerful than yours and only has an ATI graphics card, due to my finacial position a new computer is not likely to happen its way way down on the list, but I don't really let that discourage me I just find ways of working with what I have.
Um... UberSurface comes free with Studio.
As does Age of Armor's Subsurface Shader.
As does UberEnvironment...
You have quite an array of tools for free.
I was specifically disagreeing with your contention that CPU rendering is useless and GPU is the only way to go. On the whole, I agree with your points.
I cannot find the 375W reference URL now (didn't copy the orginal link) but it was a board that had a special cooling system to be overclocked.
As for the general Titan-X power use, here are some references:
http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-titan-x/specifications states 250W for this reference version
http://www.tomshardware.com/reviews/nvidia-geforce-gtx-titan-x-gm200-maxwell,4091-5.html shows 224W reached in their tests.
If one HAS the CPUs, then one doesn't necessarily need to spend the extra on the GPUs. If the image is rendered in seconds using the CPUs, why waste the extra time transferring the data to the GPUs?
What professionals do with 3DL doesn't affect DazStudio? Then the professionals coming up with using RiCurves and the Marschner shader for use in King Kong in film had no effect on DS? Two hair plugins would dispute that. Many things that start off in the high end uses of both 3DL and PRMan do work their way down to DS. With more things on the way. This is only possible due to the selection of a professional level engine being made available in the first place.
The posts that I was referring to were saying "3DL is dead, no one is supporting it." I am highlighting that 3DL is not dead, not in professional use and not in DAZ either. Just because specific people don't want to take the effort to convert materials from Iray to 3DL does not make the engine dead. Sure, us PAs like to support the thing that sells, and that is easiest to support. Guess what, once the version of 3DL that supports OSL makes it to DS, then the material conversion becomes a moot issue. Whether PAs go back to putting 3DL versions of their maps into products is yet to be seen. 3DL getting path tracing will make huge differences in speed. One can already increase 3DL rendering speed by getting out of Windows and using RIB on systems with lower overhead. But few do it even though the method is free.
Your statement about "can't even use proper HDRI" makes no sense. 3DL *has* to replicate the actual environment where the action occurred when being used for VFX to insert into pre-recorded film. Go look at the films that 3DL was used in, those items inserted by 3DL into the scenes (Wolverine's claw blades for instance) had to match the lighting and environments EXACTLY. This was done using HDRI among other methods.
"are you going to spend 12K dollars on a CPU's to have the speed of 3000 Cuda cores ? no you don't .. CPU rendering is about to die out soon , it is not the future"
If you are defining the course for the future by the hobbiest market, then maybe you have a point. But the problem IS NOT the cost of hardware, but the cost of Windows. Studios, businesses, and professional freelancers will spend $25000 on a VCA if they need GPU processing, not $1000 on a PCI-e card to stick into a slot. There are numerous reasons, many of which will be dismissed in these forums. There are reasons a Quadro M6000 (2048 CUDA and 8GB VRAM) costs $5000 and a Titan-X (3072 CUDA and 12GB VRAM) costs $1000 and lots of people happily pay the difference.
I have personally used DS to generate tens of thousands of frames of RIB to be sent to Linux 3DL to be rendered -- each scene containing many millions of polygons. I've also had DS under Windows Server 2008 render over 2000 frames of 1080p 3DL animation in one shot simply because I didn't want to reboot the machine. I've had rendering sessions run for over 3 weeks 24/7 with the systems getting zero time off. Would anyone be confident subjecting their systems to that sort of use? Could anyone trust consumer level GPU cards to run for weeks full-on at 100% without failing?
So long as DAZ continues to provide 3DL to users, then the access to the professional methods will be available. Also, don't discount that there are many more people using DS for high-level work than many in these forums would think.
Kendall
PS: For those who want to call "BS", I've attached a photo of my daughter preparing a rack of machines to start a processing job. The 1U unit below the top machine is a Tesla. Each of those large units weigh 97 pounds.
As far as I know, Octane never ran on anything other than an Nvidia GPU, and I've been using it since version 1. However, OpenCL (specifically AMD GPU) was planned for Version 3, but it looks like Otoy went to plan b, and came up with a method to port cuda to AMD, Intel, ARM GPU's (http://venturebeat.com/2016/03/09/otoy-breakthrough-lets-game-developers-run-the-best-graphics-software-across-platforms/). Other than the initial announcement back in March, things have been very quiet at Otoy, so it's hard to say when (or if - hard to say what Nvidias response to this might have been) we will actually see this new technology in Octane.
I agree with Will here. I just purchased a $1300 laptop to use during my travels for rendering images for my next app, it has 32GB of RAM and a GTX980M (4GB VRAM) and I have rendered some pretty complex scenes with it, always needing way less than 1 hour to complete.