I've cut the dimension size of my test renders way down in size. I've adjusted the render settings, is there anything else I can do? Any help would be welcome.

Below are my render settings. I'm just trying to get a basic idea of how the lighting and characters look so am not after the finest details in my test renders.

I would uncheck displacement and reduce the Pixel Sampling to 3.  I would also recommend using Dimension3D's render settings, there are far more parameters available to fine tune your renders.  The script can be found under Python Scripts>Partners>D3D>Render Firefly.

Set irradiance cache to 10 or less

Pixels samples to 2

Post filtes size 1

Post filter type to box

Cast Shadows Yes

Ray Tracing Yes

Ray Trace Bounces 1

Irridence Caching 15

Indirect Light Yes

Indirect Light Quality 10

Pixel Samples 1

Min Shading Rate 1.2

Use Displacement Maps YES

Post filtering adds very little processing time, but I set that to 1.  If you are only test rendering pixel samples doesnt need to be higher than 1,

You can set your render dimensions to 200x200 to speed things up also.

it will wash out the textures but who cares, its a preview - you dont need pore detail.

attached are my usual preview render settings.

as for the post filter - dont even touch that #$@%.  all it does is increase rendertime and wash out detail - the same can be accomplished interactively with much more control and cleaner results in an image editor in seconds.

Quote - as for the post filter - dont even touch that #$@%.  all it does is increase rendertime and wash out detail - the same can be accomplished interactively with much more control and cleaner results in an image editor in seconds.

I have proof that is not so. In fact, I wrote a very long thread with extensive demonstrations, but I don't know where the link is at the moment and would have to search for it and I'm kind of busy.

Do I need to show it, or will you take my word for it and just retract that statement?

Just a friendly challenge based on my evidentiary experience, no insult intended.

Found the thread easily when I remembered that I used the abbreviations PFS and PFT.

http://www.renderosity.com/mod/forumpro/showthread.php?thread_id=2787755

I designed a special prop to allow us to see the variation in handling tiny sub-pixel bits of geometry. Be sure to look at the images in full size, otherwise the whole exercise is pointless.

In that thread, scroll down till you see this picture:

When you look at this full size, you should observe the tapering red coil. The thinnest parts will disappear or smear if you don't use the right filtering. Postwork won't recover it.

When using Indirect Light I find it best to do so in a closed environment, if im not mistaken bagginsbill has a big sphere availabe.

In the thread linked, there is no mention of render times on the examples.

Render quality and speed are two different things.

No offense. Those are great example of the settings, but without render times for each, they can't be apply to render speed by anyone else.

Other tricks to speed up test renders is to ask yourself what really needs to be in the render. And what nodes really need to be there.

If you are looking at things like skin color, there is no need to render the props/hair/relections/etc unless they are light emitters in an idl scene. If they are light emitters, hide them from the camera, so they will still do the lighting. Carefully choosing what emits light and what doesn't, will change the render speed as well.

If just about everything in the scene is a light emitter, that pass will obviously take a while. Chances are good that can turn off light emitter on a lot of things and it will have very little change to the render quality, but improve the speed.

Closing unneeded programs running in the background can help as well, to a point anyway. Nothing worse than running a render and have the op system start a scheduled task.

Which way you render it changes the time it takes as well, internal, external, send it to queue.

Quote - In the thread linked, there is no mention of render times on the examples.

Render quality and speed are two different things.

No offense. Those are great example of the settings, but without render times for each, they can't be apply to render speed by anyone else.

Are you referring to the thread I linked to about Post filter validity?

My reason for posting it had nothing to do with the thread and its topic of IDL render speed improvement. In fact, it has nothing to do with IDL at all, as do many of the responses by others. It was in response to the notion introduced by BH that post filtering is borked and should be done outside Poser. It's not borked, and the correct value to use and why is explained in the thread.

If you want speed values, be my guest and try them all. I noted in the thread that Post Filter Size affects speed, without giving specifics. In general, such specifics from 2009 would be only directionally correct anyway. New Poser, new render engine, new computers - all change the answer.

I used to say, for example, that Reflect softness past 1 looked bad and increased render time by a factor not worth pursuing. That's totally incorrect now, and I am regularly rendering with softness = 4 or even higher, and I am slowly eliminating the use of specular nodes and lights from my repertoire. In three more years, rendering will be so much faster that this will be the norm, instead of for special renders.

^^^^^^^^^^^

And all this has what to do with the OP's question?

I was addressing the question in the first post, so were the others.

Feel free to post more irrelivant info. lol.....

BB:  on a technical level youre most likely right, and im probably just not using the right settings...

that said, i render organic figures and not thin geometric objects. your above image may be the ideal scenario to showcase filtering but is something that i have not rendered in a decade of using poser.  meanwhile, i have yet to render a single image where the filters have not blurred out/destroyed detail.  to reduce antialiasing on 'overnight' renders ive had better results setting pixel samples (AA) to higher settings (<15) than i have with my (albeit much more limited than yours) experiments with filters - which simply tend to blur the aliasing on the edges as opposed to smooth out the curves. kindof like trying to get rid of aliasing by applying a low quality version of gaussian blur.

that render you posted does look pretty damned crisp though. i wonder how it would compare to one with no filtering and just high AA settings - in both results and rendertimes.

im not averse to learning new things, however, and ive learned a lot from you in terms of lighting and shaders - so if you prove me wrong and demonstrate a faster/better way to get smooth images ill be grateful.

Quote - but is something that i have not rendered in a decade of using poser.

Really? You don't render hair?

^not really, i handpaint hair most of the time.  if i do render hair, 90% of the time its just there as a basis for postwork.

i am still waiting for someone (most likely william the bloody or carodan, the way things are going) to refine dynamic hair to a point that makes me want to use it.

what points about dynamic hair are you looking for improvements on?  a long time ago in a community far far away I used to spend all my time on Poser Dynamics, maybe I can help!

Attached Link: Geforce Fermi hair demo

the video is nvidia dynamic hair thats been rendering in realtime on geforce cards since 2009.

i dont even want realtime - hell if it took one hour to render one frame id be happy. just gimme hair that looks like that :)

(psst.. ill even render it with an x4 sinc shader if thats what it takes :P)