How to make OpenGL preview viewport faster in DAZ?

mwasielewski1990

So... Is there a way (besides hiding complicated objects like hair and other high-poly stuff) to speed up the opengl preview in DAZ? I often find that with 4-5 figures + some props the viewport becomes sluggish to the point that posing or moving objects is very difficult. Constantly hiding/showing stuff is kind of a workflow killer for me....

I'm on i7 8700 and GTX 1080ti, 32gb ram. Is this too slow to handle 5 figures in the viewport?

 

EDIT: I found some options in daz's preferences tab. Image included. My question is - does fiddling with for example, thetexture parameter will affect the final render quality?

Richard Haseltine

Those are just viewport settings, they won't affect Iray or 3Delight (I'm not sure about the OpenGL render modes).

Robinson

mwasielewski1990 said:

So... Is there a way (besides hiding complicated objects like hair and other high-poly stuff) to speed up the opengl preview in DAZ? I often find that with 4-5 figures + some props the viewport becomes sluggish to the point that posing or moving objects is very difficult. Constantly hiding/showing stuff is kind of a workflow killer for me....

I'm on i7 8700 and GTX 1080ti, 32gb ram. Is this too slow to handle 5 figures in the viewport?

In my experience most of the cost of the viewport is when things like smoothing modifiers kick in on clothing and things like that.  What I usually do is pose the character out of scene, hiding the hair and clothes.  Sometimes I will also change the character's mesh resolution down to "base" as well (base is good for testing out animation keys or using ActivePose - the viewport internally reduces to base when you press the play button in any case).

btw, 5 figures is quite a lot... especially at high res geometry with clothing and hair...

p0rt

its more down to all the default viewports besides iray being CPU only, which go along the lines of being useless with floating point math operations, while cuda can perform 1000x more floating point math operation per second, give or take, which you can benchmark using some GPGPU benchmark tool

 

the performance increase using CUDA would be along the lines of marvelous designer

 

 

while also being able to view realtime animations using G8 models, in all of the default viewport modes, besides iray

 

while AMD user's don't have to feel left out, as you can just convert cuda to openCL using the bridge downleadable from http://chrec.cs.vt.edu/cu2cl/downloads.php

mwasielewski1990

Thanks guys for answers. Just a small update: The Display Optmization parameter was set to none by default. Setting it to "best" basically didn't change anything spectacular in terms of preview display but gave an enormous speed upgrade :) So i guess that will have to do for now..