OctaneRender® 1.024 beta 2.48 TEST (win) [OBSOLETE]

[abstrax]

So does this means that now we can mix "very different" Geforce Cards whith different chip sets?

Yes, that's correct

Cheers,
Marcus

[gzavye]

I'm doing some tests about portals (I was waiting for that since long months )
I am a little bit confuse about results.
For me images are equals (maybe little bit accurate with portal)
Render is slower with portal cause I have to check alphashadows to see shadows in the room...
I perfectly understand portal concept (I used to use it for MR, vray, Indigo, Luxrender) but ATM it seems not very usefull.
I tested with a very simple scene; maybe is the cause You tell me

PT_alphashadows OFF_4000s/px_7mn54_2.6Ms/s

PT_alphashadows ON_4000s/px_10mn04_2.02Ms/s

[Ayertosco]

Maybe the scene is too simple? Don't know much about portals but if they increase the efficency of the lightning coming in from the outside, maybe you'll see some differences on larger scenes, with more openings\windows and objects.

[necko77]

PMC and PT dont have sunlight if alpfashadow are unchecked

Alphashadows slows down rendering

i cant believe that this scene is running at 0,26 at PMC with portal and 0,36 PT with portal -512x512 rendering size-
very slow if you ask me . . .
I need much faster GI

512x512, 200 S, PMC and PT ( after 200 semples) around 3 minutes

512x512 PMC after 2000 semples, around 30 minutes

PMC 2048x2048 around 30 minutes, scaled down

its better, so when you doing your rendering alwys do them in higher resolutions and scale them down to needed

octane stoped here...i was not possible to continue my rendering after few stops and restarts
PT eat all graphiccard resorces
I will finish now PT for 30 min just to see which one wins in this scene PT om PMC
after 30 minutes...here it is

blender image with portals and normals faceting inside interior

[Timmaigh]

Hi there.
Just tested the new release with my new GTX 590 and I dont know ,but multy GPU
worked for me better with the preview release , even if the multy GPU not worked
properly with release 2.47. I know the sample counter has been updated ,but regardles
I think 2,48 is slower then 2,47. you shouldn't have changed anything about the Fps ore Ms/sec.
because now we cant compare with 2.47
regards
Rico

Which kernel are you talking about? Again, if you are doing performance comparisons, please compare render times.

The reason why samples/second has changed is because it was incorrect before and now it's correct. Obviously that makes it hard to compare against older results, but broken things should be fixed, shouldn't they?

Cheers,
Marcus

Well to be fair, in the case of the incorrect Ms rates its bit more complicated. The basic line of thinking is, ofc, if its broken, it needs to be fixed, as you say. But when you think about this deeper, that Megasamples per second value says you basically nothing on its own. You need to compare it to something, to previous versions to know, if its faster or not. That is not possible anymore. If you kept it the way it was and now it would show for example 4 Ms/s instead of 2 Ms on the same scene as with 2.47, i would know its 2x as fast, even if the real values should have been 3 and 1,5 Ms. Thats all that matters IMHO. Now you may have correct value to be shown up, but as there is nothing, you can compare it to, its basically irrelevant. Really the only way now is to go with render times, but what is the point of this value then? I thought its there to let you know, how long render time can you expect.

Hopefully i made sense here Just wanted to point out, maybe it would be not so weird to keep incorrect value, as it does not influence the functionality of Octane in any way (i suppose), but its imho IMHO dependent on comparisons.

[vipvip]

when i adjust the point of depth of field and save the ocs, it doesn't keep it when re-rendering the picture ( the picture is very blurry with high aperture values ).
To correct this, i'have to uncheck 'export camera' and it works.
but the big problem is: no more camera animation possible...So that it's actually impossible to render animation anymore ( with blurry DOF effects) i've tried to check the focal distance, and even the autofocus option is not memorised...
This procedure was working before. ( both on maya and lightwave exporters )

Does anybody experiment this ?

Another thing: animation on light-emitive objects seems to be not functionning ( lightwave export)...
if you have informations about these things...

otherwise this test-beta seems to be promisefull.

thanx

[MaTtY631990]

I'm doing some tests about portals (I was waiting for that since long months )
I am a little bit confuse about results.
For me images are equals (maybe little bit accurate with portal)
Render is slower with portal cause I have to check alphashadows to see shadows in the room...
I perfectly understand portal concept (I used to use it for MR, vray, Indigo, Luxrender) but ATM it seems not very usefull.
I tested with a very simple scene; maybe is the cause You tell me

noportal.png

portal.png

At the moment you will see a bigger benefit with evenly light through windows, no sunlight, which I think will improve in the next release. Also make the window much smaller and try another comparison.

[lilbignik]

This my simple test on gtx 580; 1024x768; 700 s/px

octane 2.48, path, 4.42ms/s, 2min04

octane 2.48, pms, 3.07ms/s, 2min59

octane 2.48, path, 3.82ms/s, 2min24, with portal

octane 2.48, pmc, 2.73ms/s, 3min23, with portal

I don't understand, why render with portal are slower and so different from image without portal. This is normal???
I noticed that pmc in octane 2.48 was slower. this scene on 2.47 has 4.09ms/s, 2min31.

[colorlabs]

Hi, I just noticed I can't rotate HDRI environments. The rotation x/y sliders do nothing.

[abstrax]

when i adjust the point of depth of field and save the ocs, it doesn't keep it when re-rendering the picture ( the picture is very blurry with high aperture values ).
To correct this, i'have to uncheck 'export camera' and it works.
but the big problem is: no more camera animation possible...So that it's actually impossible to render animation anymore ( with blurry DOF effects) i've tried to check the focal distance, and even the autofocus option is not memorised...
This procedure was working before. ( both on maya and lightwave exporters )

Does anybody experiment this ?

The exporter always overwrites the focal depth of the scene file and if you want to to animate it, you have to animate it in your 3D application. That should not have changed since beta 2.46b. If the focal depth gets overwritten by the exporter, autofocus gets disabled. If you want, we can make Octane keep the settings from the saved OCS file if the exported focal depth is 0.

If you set the focus depth in Octane using the focus picker, the autofocus gets disabled and the focus depth in the camera node gets changed. If you enable autofocus the focus depth of the camera node does not get changed, but the depth/distance of the center pixrl is used for rendering. These settings should get saved with the OCS file and should be restored, if you open the file from Octane (not via the exporter, which overwrites those settings).

As far as I can see, the concept of a "focus point" never existed in Octane and focal depth was handled implicitely and never sored.

Another thing: animation on light-emitive objects seems to be not functionning ( lightwave export)...
if you have informations about these things...

otherwise this test-beta seems to be promisefull.

thanx

Sounds like a bug to me. Could you please explain that in more detail?

Thanks,
Marcus