Version 2.24.1-R3 (06.10.2015) - obsolete stable

[aoktar]

Hi Aoktar,

I have posted on August 07 the problem of hagout of C4D when drag a material on Live DB, and you have replied that you are investigating on the new SDK. For now the solution is to come back to 2.23 release .

I hope you will find soon a solution.
Many thanks for your precious support.

It's related to Sdk. I'm waiting newer version.

[alexmunteanu]

Hi Aoktar,
Am I doing something wrong or is there an issue with the new Geometry Engine/Control feature?
If I put it on "Resend only updated objects" the render time increases almost 4 times than if it was with "Resend all scene data" (from 43s per frame to 02:33 per frame).
And it doesn't matter if I choose Manual/All objects movable/Autodetect.
I tried to make a simple test scene (attached) to send it to you but in this one if I check "Resend all scene data" the emissive objects are not taken into consideration.
The render times are the same however and not slower on "Resend only updated objects" like in the main project which unfortunately I cannot send it over.
It seems I have to learn a bit more about this matter...
Can you please give me a few hints?
Thank you!

P.S.: the main project also has a closed interior, shaders with blackbody emission and IES lights like the sample project I attached

Edit: I forgot to check "surface brightness" in the blackbody emission and that fixed the lighting difference between "Resend all scene data" and "Resend only updated objects" in the project which I attached. And the render times are still the same and not like in the main project with such a big time increase.

[Silverwing]

Hi Oaktar,

I am having some weird results working with the C4Doctane save dialog.

When "Show Passes" is tuned off and saving through C4D Saving dialog C4D seems to only be able to save Multipass files in a multilayer file.
If you tick off Multilayer no Multipass is saved (Just the Mail pass)

When "Show Passes" is turned off and saving though Octane Saving dialog, Octane seems to only be able to save the Multipass files if there are not a multilayer file.

I do not know if you had a look at the problem that saving EXR´s with the Octane saving dialog the Compression slider seemed to be broken.

If you find the time it would be awesome if you could check that section of the code.
As always thanks for your hard work. Keep it up!

Cheers,
Raphael

[aoktar]

Hi Aoktar,
Am I doing something wrong or is there an issue with the new Geometry Engine/Control feature?
If I put it on "Resend only updated objects" the render time increases almost 4 times than if it was with "Resend all scene data" (from 43s per frame to 02:33 per frame).
And it doesn't matter if I choose Manual/All objects movable/Autodetect.
I tried to make a simple test scene (attached) to send it to you but in this one if I check "Resend all scene data" the emissive objects are not taken into consideration.
The render times are the same however and not slower on "Resend only updated objects" like in the main project which unfortunately I cannot send it over.
It seems I have to learn a bit more about this matter...
Can you please give me a few hints?
Thank you!

P.S.: the main project also has a closed interior, shaders with blackbody emission and IES lights like the sample project I attached

Edit: I forgot to check "surface brightness" in the blackbody emission and that fixed the lighting difference between "Resend all scene data" and "Resend only updated objects" in the project which I attached. And the render times are still the same and not like in the main project with such a big time increase.

Do you know the differences between two method? We explained lot of in forums and Inlifethrill's vimeo tutorials.

[alexmunteanu]

Hi Aoktar,
Am I doing something wrong or is there an issue with the new Geometry Engine/Control feature?
If I put it on "Resend only updated objects" the render time increases almost 4 times than if it was with "Resend all scene data" (from 43s per frame to 02:33 per frame).
And it doesn't matter if I choose Manual/All objects movable/Autodetect.
I tried to make a simple test scene (attached) to send it to you but in this one if I check "Resend all scene data" the emissive objects are not taken into consideration.
The render times are the same however and not slower on "Resend only updated objects" like in the main project which unfortunately I cannot send it over.
It seems I have to learn a bit more about this matter...
Can you please give me a few hints?
Thank you!

P.S.: the main project also has a closed interior, shaders with blackbody emission and IES lights like the sample project I attached

Edit: I forgot to check "surface brightness" in the blackbody emission and that fixed the lighting difference between "Resend all scene data" and "Resend only updated objects" in the project which I attached. And the render times are still the same and not like in the main project with such a big time increase.

Do you know the differences between two method? We explained lot of in forums and Inlifethrill's vimeo tutorials.

I do know what are their functionalities, but what I do not know is why I got a render time 4x slower in that specific project using "Resend only updated objects".
If I couldn't replicate the situation means there's something wrong with that project. Who knows...?
Thanks!

[aoktar]

I do know what are their functionalities, but what I do not know is why I got a render time 4x slower in that specific project using "Resend only updated objects".
If I couldn't replicate the situation means there's something wrong with that project. Who knows...?
Thanks!

Because, i think you have lot of indivual meshes. Renderer performance can decrease when you have many seperated meshes. "Send all data" collect all objects to a single mesh. So can be faster, bur can differ the result than LV's output if you use "Auto" mode. Propably i'll rename this as "All movable" avoid confusion.

If you check the manual you can read more info. Even in tutorials.

[alexmunteanu]

I do know what are their functionalities, but what I do not know is why I got a render time 4x slower in that specific project using "Resend only updated objects".
If I couldn't replicate the situation means there's something wrong with that project. Who knows...?
Thanks!

Because, i think you have lot of indivual meshes. Renderer performance can decrease when you have many seperated meshes. "Send all data" collect all objects to a single mesh. So can be faster, bur can differ the result than LV's output if you use "Auto" mode. Propably i'll rename this as "All movable" avoid confusion.

If you check the manual you can read more info. Even in tutorials.

Yes, indeed that is the case as there are quite a few objects in that scene.
So as a rule of thumb if you have a relative simple scene with a couple of meshes sending the geometry only in the first frame will decrease the overall rendering time while if your object count is high you'd get better results resending the geo for every frame?
Did I understand it right? Does the poly count affect any of the two methods?
Thank you again!

[cbmotion]

Hi ahmet

i tried to install your last 2.24. and the stand alone
but i get an error - i must use cuda 7 -
but in my osx the highest cuda is 6.5 i can get
mountain lion 10.8.5 -

i tried mavericks and Yosemite but then i have a GPU limitation of 2 Cards.
From my external 4 Titan Cards are only 2 shown.
i was in contact with apple and the Engineer from Netstor.
apple said it is not their problem - hihi - funny but it work before
they say - it is the problem of the 3th party hardware

with netstor i tried different kext tuning and terminal orders. but didn't got it to work.

so i went back to Mountain lion and i have my 4 cards back again.
i heard from a apple worker that apple don't want that my old mac pro is faster than the new mac pro.
So they limited it for only two cards. isn't this a shame for apple.

i have the latest Cuda and Nvidia driver i can get for my OSX.
can you help to get your last version run on my mac pro with Mountain lion?

cheers carsten

[sdanaher]

Hi ahmet

i tried to install your last 2.24. and the stand alone
but i get an error - i must use cuda 7 -
but in my osx the highest cuda is 6.5 i can get
mountain lion 10.8.5 -

i tried mavericks and Yosemite but then i have a GPU limitation of 2 Cards.
From my external 4 Titan Cards are only 2 shown.
i was in contact with apple and the Engineer from Netstor.
apple said it is not their problem - hihi - funny but it work before
they say - it is the problem of the 3th party hardware

with netstor i tried different kext tuning and terminal orders. but didn't got it to work.

so i went back to Mountain lion and i have my 4 cards back again.
i heard from a apple worker that apple don't want that my old mac pro is faster than the new mac pro.
So they limited it for only two cards. isn't this a shame for apple.

i have the latest Cuda and Nvidia driver i can get for my OSX.
can you help to get your last version run on my mac pro with Mountain lion?

cheers carsten

Oh, really? That's not good. I'm running Yosemite, so 2 cards is the best I can hope for? Would be nice if Otoy would allow some limited network rendering in the license to combat this somewhat.

[aoktar]

Are you sure that you don't run OctaneVR?