OctaneRender™ V4 XB4

Forums: OctaneRender™ V4 XB4
A forum where development builds are posted for testing by the community.
Forum rules
NOTE: The software in this forum is not %100 reliable, they are development builds and are meant for testing by experienced octane users. If you are a new octane user, we recommend to use the current stable release from the 'Commercial Product News & Releases' forum.

Re: OctaneRender™ V4 XB4

Postby Zay » Fri Jul 27, 2018 4:53 am

Zay Fri Jul 27, 2018 4:53 am
Seems like some of the material nodes settings gets carried over/mixed when switching nodes.
I made a glass material in the Universal node. The HDRI environment made a lot of fireflies as it sometimes does. Then I switch to the Diffuse node but I still had fireflies inside my object.
To fix that I had to delete the node completely and add a new Diffuse to "clean up" the material settings. I guess that's a bug?
Win 11 Pro | i5 12600K | 32GB ram | 2x GTX 1080Ti + 3080Ti - studio driver 546.01| Modo/Blender/ZBrush/Daz/Poser
Zay
Licensed Customer
Licensed Customer
 
Posts: 1113
Joined: Sun Jan 17, 2010 2:53 am

Re: OctaneRender™ V4 XB4

Postby bartosh44 » Fri Jul 27, 2018 12:44 pm

bartosh44 Fri Jul 27, 2018 12:44 pm
I'm happy with progress of V4.

I would like to ask polite about edge rounded shader which you stated that should be after last version.

Have you stopped you work on improving this helpfull feature?

I hope that quality of future edge shader will be similar to Modo (looks nice also on closeups).
bartosh44
Licensed Customer
Licensed Customer
 
Posts: 26
Joined: Mon Oct 31, 2011 8:06 am

Re: OctaneRender™ V4 XB4

Postby Notiusweb » Fri Jul 27, 2018 9:17 pm

Notiusweb Fri Jul 27, 2018 9:17 pm
haze wrote:
Notiusweb wrote:Hi! Haze had mentioned all around speed increases to most scenes, through improvements to tonemapping delta time and quality of samples. He had mentioned as upcoming for XB4, but it didn't look like the release notes for XB4 mentioned this at all. Was a systemic speed increase part of XB4, or not yet.
Thx.


You should see a substantial improvement in AI Light (especially scenes that have a lot of small directional emitters), which is mentioned in the release notes. We have been inching upwards on general render speed, and there are some more experiments currently in progress, but no guarantees can be made right now. We're doing these changes incrementally, otherwise we would be affecting stability, which is our highest priority currently.


I have a scene of a Daz character, with an HDRI background, and 1 emitter, in Standalone.
@1080p with Denoiser for 55 s/px takes 6 seconds on Direct Lighting, no matter if AI lights are on/off.
I toggle all the settings as far as enable, disable, light ID#, Environment, Light ID specified emitter, etc.
Always 6 seconds...
Previous XB3 was, also, 6 seconds.

Now, I am not complaining...At all! I am pleased with this rendering power.
3.08.2 would take 2000+ s/px, over a minute render-time, to match the presentation.

But I am curious if I am missing something...
As far as a 'substantial improvement', say in XB4 vs XB3 - what, if any, with 1 emitter, could I really be expecting here....anything?
I mean, is 55 s/px just too small of a sample range for render time numbers to be decreased, or like, should I indeed be seeing 3, 4, or 5 second render, instead of a 6 second render?

Or, is it just that the number of emitters just isn't a factor here because there is just 1?
As in, if it was a 10 emitter scene, now in XB4 I would then be seeing faster render time than before in XB3, something like that?

Please keep in mind I have no idea what is involved here, I only praise and complain ignorantly.
Thx!

(PS - Also, my interest here is not in a single frame render, it is in the realm of animations. Otherwise, this focus on a couple second gains per image wouldn't be as sought after. )
Win 10 Pro 64, Xeon E5-2687W v2 (8x 3.40GHz), G.Skill 64 GB DDR3-2400, ASRock X79 Extreme 11
Mobo: 1 Titan RTX, 1 Titan Xp
External: 6 Titan X Pascal, 2 GTX Titan X
Plugs: Enterprise
User avatar
Notiusweb
Licensed Customer
Licensed Customer
 
Posts: 1285
Joined: Mon Nov 10, 2014 4:51 am

Re: OctaneRender™ V4 XB4

Postby haze » Sat Jul 28, 2018 3:43 am

haze Sat Jul 28, 2018 3:43 am
Notiusweb wrote:
haze wrote:
Notiusweb wrote:Hi! Haze had mentioned all around speed increases to most scenes, through improvements to tonemapping delta time and quality of samples. He had mentioned as upcoming for XB4, but it didn't look like the release notes for XB4 mentioned this at all. Was a systemic speed increase part of XB4, or not yet.
Thx.


You should see a substantial improvement in AI Light (especially scenes that have a lot of small directional emitters), which is mentioned in the release notes. We have been inching upwards on general render speed, and there are some more experiments currently in progress, but no guarantees can be made right now. We're doing these changes incrementally, otherwise we would be affecting stability, which is our highest priority currently.


I have a scene of a Daz character, with an HDRI background, and 1 emitter, in Standalone.
@1080p with Denoiser for 55 s/px takes 6 seconds on Direct Lighting, no matter if AI lights are on/off.
I toggle all the settings as far as enable, disable, light ID#, Environment, Light ID specified emitter, etc.
Always 6 seconds...
Previous XB3 was, also, 6 seconds.

Now, I am not complaining...At all! I am pleased with this rendering power.
3.08.2 would take 2000+ s/px, over a minute render-time, to match the presentation.

But I am curious if I am missing something...
As far as a 'substantial improvement', say in XB4 vs XB3 - what, if any, with 1 emitter, could I really be expecting here....anything?
I mean, is 55 s/px just too small of a sample range for render time numbers to be decreased, or like, should I indeed be seeing 3, 4, or 5 second render, instead of a 6 second render?

Or, is it just that the number of emitters just isn't a factor here because there is just 1?
As in, if it was a 10 emitter scene, now in XB4 I would then be seeing faster render time than before in XB3, something like that?

Please keep in mind I have no idea what is involved here, I only praise and complain ignorantly.
Thx!

(PS - Also, my interest here is not in a single frame render, it is in the realm of animations. Otherwise, this focus on a couple second gains per image wouldn't be as sought after. )


I'm sorry if I didn't explain this well enough last time - there are several factors involved, and sometimes I use some of the terms interchangeably. Most importantly, when I think of render speed, I always think in terms of how long it takes to render a scene to the point where I can no longer see noise (time to clean image). We have been improving along several angles to make this happen faster:

reducing the amount of time it takes to:
1. load the scene from disk
2. process the geometry in the scene
3. getting samples from the scene

and another way to make the clean image happen faster:
4. improve the quality of the samples, for instance, using AI to do light sampling for us.

So as you can see, the Ms/sec render speed is only part of the picture. If we massively reduce the quality of the samples we get from the scene, you'll get a really, really huge Ms/sec render speed, but it will take forever to get to the clean image.

We're still making improvements across all of 1-4 above, but incrementally.
User avatar
haze
OctaneRender Team
OctaneRender Team
 
Posts: 961
Joined: Sun Feb 08, 2015 8:57 pm

Re: OctaneRender™ V4 XB4

Postby bartosh44 » Sat Jul 28, 2018 10:39 am

bartosh44 Sat Jul 28, 2018 10:39 am
What is wrong with my last question about edge shader? It`s a wrong place? Why it`s omitted again? I know also few licensed customers which asks same question. If it will not be implemented then I won`t ask again about that.

Best regards,
Marcin
bartosh44
Licensed Customer
Licensed Customer
 
Posts: 26
Joined: Mon Oct 31, 2011 8:06 am

Re: OctaneRender™ V4 XB4

Postby Notiusweb » Sat Jul 28, 2018 4:31 pm

Notiusweb Sat Jul 28, 2018 4:31 pm
haze wrote:
I'm sorry if I didn't explain this well enough last time - there are several factors involved, and sometimes I use some of the terms interchangeably. Most importantly, when I think of render speed, I always think in terms of how long it takes to render a scene to the point where I can no longer see noise (time to clean image). We have been improving along several angles to make this happen faster:

reducing the amount of time it takes to:
1. load the scene from disk
2. process the geometry in the scene
3. getting samples from the scene

and another way to make the clean image happen faster:
4. improve the quality of the samples, for instance, using AI to do light sampling for us.

So as you can see, the Ms/sec render speed is only part of the picture. If we massively reduce the quality of the samples we get from the scene, you'll get a really, really huge Ms/sec render speed, but it will take forever to get to the clean image.

We're still making improvements across all of 1-4 above, but incrementally.


Okay, thanks for clarifying!
Win 10 Pro 64, Xeon E5-2687W v2 (8x 3.40GHz), G.Skill 64 GB DDR3-2400, ASRock X79 Extreme 11
Mobo: 1 Titan RTX, 1 Titan Xp
External: 6 Titan X Pascal, 2 GTX Titan X
Plugs: Enterprise
User avatar
Notiusweb
Licensed Customer
Licensed Customer
 
Posts: 1285
Joined: Mon Nov 10, 2014 4:51 am

Re: OctaneRender™ V4 XB4

Postby haze » Sat Jul 28, 2018 9:28 pm

haze Sat Jul 28, 2018 9:28 pm
bartosh44 wrote:I'm happy with progress of V4.

I would like to ask polite about edge rounded shader which you stated that should be after last version.

Have you stopped you work on improving this helpfull feature?

I hope that quality of future edge shader will be similar to Modo (looks nice also on closeups).


I don't have any updates for you on this right now, but it is still on our list for V4.
User avatar
haze
OctaneRender Team
OctaneRender Team
 
Posts: 961
Joined: Sun Feb 08, 2015 8:57 pm

Re: OctaneRender™ V4 XB4

Postby Notiusweb » Sun Jul 29, 2018 4:09 am

Notiusweb Sun Jul 29, 2018 4:09 am
Had this happen in a scene so made this test scene to show.

Test Scene has:

1 Sphere Emitter (White)
1 Sphere Object (Purple)
1 Object Background (iPad)

GOAL - I want to have the background iPad receive no shadow from the Purple sphere blocking the White emitter sphere.

1) In Object Layer Node for iPad background, checking/unchecking the check box for Shadow Visibility does not have any effect....the iPad always gets a shadow from the Purple Sphere blocking the white emitter sphere.
I do not want this.
ShadowVis1.jpg


2) In Emission-Visibility for the emitter sphere, I can uncheck 'Cast Shadows', and this causes no shadow to be produced on the background iPad, which is what I want
ShadowVis2.jpg


So I find that Object Layer's Shadow Visibility checkbox does not do anything.
Possible bug?

Thx.
Win 10 Pro 64, Xeon E5-2687W v2 (8x 3.40GHz), G.Skill 64 GB DDR3-2400, ASRock X79 Extreme 11
Mobo: 1 Titan RTX, 1 Titan Xp
External: 6 Titan X Pascal, 2 GTX Titan X
Plugs: Enterprise
User avatar
Notiusweb
Licensed Customer
Licensed Customer
 
Posts: 1285
Joined: Mon Nov 10, 2014 4:51 am

Re: OctaneRender™ V4 XB4

Postby enricocerica » Sun Jul 29, 2018 7:45 am

enricocerica Sun Jul 29, 2018 7:45 am
roeland wrote:The format of the textures loaded by the image tiles node is the same as the image files. If an image is saved as grayscale, it will be loaded as a grayscale texture.

How to set up:
  • First you select one of the images, eg. C:\path\to\tile-01-01.png
  • Add placeholders, eg. C:\path\to\tile-%u-%v.png.
  • Then select the grid size.
  • Then set up the index for the first tile and the amount of digits, eg. if the file above is the first tile, you have 2 digits and start index 1.
  • Click preview to see which list of images will be loaded


Great feature, I thought that the result grid image would fill the UW map as if it was a single image but it seams that you have to play with UV transform to place the result on the mesh and there is no border mode pin so we cannot set any repeat of the tiles.
Did I missed something ?
Also, would be great to generate index only on %u value, I mean if I have let's say 4 images 001, 002, 003, 004 and set matrix with 2 columns and 2 lines it will generate incremental numbers with 001, 002 on first line and 003, 004 on second (hope this is clear).
Anyway great release and the light link is something I was waiting since a while, very useful ;)
Thanks
Modeling system : I7 32GB Windows 10 & Fujitsu Celsius H720
GPU : 1x Gigabyte GTX580 3GB + 1x MSI GTX780 3GB + 1x PALIT GTX780 6GB +1x Asus Stix GTX1070 8GB
http://www.myline.be
User avatar
enricocerica
Licensed Customer
Licensed Customer
 
Posts: 1012
Joined: Wed Nov 25, 2009 7:32 pm

Re: OctaneRender™ V4 XB4

Postby Zay » Sun Jul 29, 2018 8:50 am

Zay Sun Jul 29, 2018 8:50 am
Notiusweb wrote:GOAL - I want to have the background iPad receive no shadow from the Purple sphere blocking the White emitter sphere.


To kill the shadow coming from the purple sphere you need to add a "Object Layer Map" to it.
Win 11 Pro | i5 12600K | 32GB ram | 2x GTX 1080Ti + 3080Ti - studio driver 546.01| Modo/Blender/ZBrush/Daz/Poser
Zay
Licensed Customer
Licensed Customer
 
Posts: 1113
Joined: Sun Jan 17, 2010 2:53 am
PreviousNext

Return to Development Build Releases


Who is online

Users browsing this forum: No registered users and 7 guests

Fri Mar 29, 2024 8:10 am [ UTC ]