development update Tweet

[Garrick]

I have not read through this whole thread but I am curious if and how far off displacement maps support will be?

[Refracty]

thanks for the detailed update Marcus !

[Rikk The Gaijin]

I have not read through this whole thread but I am curious if and how far off displacement maps support will be?

It won't be in 1.5, so maybe next year.

[cfrank78]

Hi Markus!

I probably didn't get your question, but you can already shift HDRI textures in the texture environment using the "rotation" pin, which is actually a shift of the lat/long coordinates.

I know the controls currently possible of an HDRI. Rotate x y, gamma and scale x and y. But the scaling is not right. Or at least not as powerful as on other engines. We had a detailed discussion with ahmet in the c4d plugin about that and he agrees! Maybe you should talk to him, cause he knows the problem quite well, but he said he cant do anything about it, because of your SDK. We made a clear comparison how other engines handle hdri, especially scaling and mapping (cubic, spherical,......) and how they handle scaling. They scale different and with more possibilities, to adjust the hdri, so it can be better adjusted to the szene you have, especially if you rotate in the y axis. In other engines for example you can take a sphere as an hdri and move it like a ball in all directions. This is one of many examples, that can not be done with the current HDRI handling of Octane. Also it is not possible to geht the full lightpower out of an HDRI. Ok you can rise or lower the gamma, but that doesnt help to bring all the information out of 32 bit hdri. Imagine a szene with a sun at about 8 am summertime. not on the horizont and not fully up. strong but not too strong. If you make a picture with a DSLR you can make a picture from an object with light surrounded, good shadows and still see some of the sun, even with 16bit. this is currently not possible if you use HDRI in octone, only with your sun system. But i prefer to work with HDRI and IBL, cause there i can get the full power of importance sampling, what is not possible with the sun system and i can avoid hotpixels and have a much faster rendering. In Vray for example you can adjust the HDRI in the environment, so you can "compress" the lights and the shadows, so it fits your szene, and make it look ultra realistic.

Hope you understand now what i mean!
Kind regards Chris!

Kind regards Chris

[dinf]

... It's definitely not a full blown render passes system, which is hard to do in a path tracer anyway, since most effects are just side effects. Everything is just reflections and refractions and emissions.

Hey Marcus, shouldn't it be possible to implement a thing like iray's LightPathExpressions (LPE) in Octane also?
http://blog.irayrender.com/post/3726005 ... al-effects

[cfrank78]

I read this whole thread...........and other threads too.
Please don´t get me wrong now, but i really have to ask this question. I read VERY often: we cant implement this n that right now, cause that would slow down Octane.

The main reason of selling Octane is, that its fast. I get that. As long as you have enough GPU power. But i really must ask myself, how fast octane would be, if it had all the functionalitys like other renderengines allready have! I know and HAVE TO work with some others too, because of my clients. Arion, Thea, Iray, just to name a fiew. They all have all the functions, that users here are begging for, for years now. For me it sounds, when i read all the commends of the coders, that you have 2 options:

Option 1: fast but with not all the functionalitys, like other engines!
Option 2: slow like other engines, but more functionalitys!

Tell me if i am wrong.....but after reading all of this, it really sounds like that!
I really don´t wanna cause any trouble here, but I have to ask myself if this is so, because it sounds more and more like that.

Kind regards Chris!

[NVN]

I just see it as cfrank78

[resmas]

Hope 1.21 bring some of those features, but please take in consideration some others.

like Undo and region Rendering

Please remember, that version 1.5 won't be the last one for Octane

Would be nice if for some "simple features" that we don't have yet you guys could make a statement, for example:

Undo - Not possible for this and that reason
Region Rendering - possible but still need this and that to be available.
etc

Both features are on our TODO list and they have moved up a bit. The node system rewrite was a requirement for the undo system. Region rendering is not a small feature, but do-able and in the not too distant future we will have a look at it.[/quote]


Thanks Abstrax for your reply.

keep up with the good work.

cheers
resmas

[abstrax]

I read this whole thread...........and other threads too.
Please don´t get me wrong now, but i really have to ask this question. I read VERY often: we cant implement this n that right now, cause that would slow down Octane.

The main reason of selling Octane is, that its fast. I get that. As long as you have enough GPU power. But i really must ask myself, how fast octane would be, if it had all the functionalitys like other renderengines allready have! I know and HAVE TO work with some others too, because of my clients. Arion, Thea, Iray, just to name a fiew. They all have all the functions, that users here are begging for, for years now. For me it sounds, when i read all the commends of the coders, that you have 2 options:

Option 1: fast but with not all the functionalitys, like other engines!
Option 2: slow like other engines, but more functionalitys!

Tell me if i am wrong.....but after reading all of this, it really sounds like that!
I really don´t wanna cause any trouble here, but I have to ask myself if this is so, because it sounds more and more like that.

Kind regards Chris!

In general it is true that if you add more stuff things slow down on the GPU. A lot. What we then usually do, is optimizing the heck out of the added piece of code and the corresponding data structures, plus some more optimizations in other places, to compensate for the loss. It's certainly a lot harder than adding stuff to say a CPU render engine. It's basically means, that we have to make compromizes all the time.

Obviously, we don't know how good or bad other engines can be expanded, but I recently read a BOXX intervew with Chaosgroup and they said something similar about VRay RT.

What I really miss here in the forum, is the appreciation that Octane doesn't have that much development heritage: Arion had FryRender as a pre-cursor, Thea is the successor of Kerkythea, Iray was originally developed by Mental Images, which also wrote Mental Ray. And Vray RT is an add-on of Vray. You might ask, what has that to do with GPU rendering? Well, having written some CPU render engine certainly helps with writing GPU functionality, even if you have to restructure a lot and change algorithms.

And what is even more important: Octane is maybe 10% render engine. The reset is there to get data from the user/plugin, prepare it and get it into the render engine, and then to get the results out of the render engine and somehow present to the user/plugin. And all of that interactively. A little bit like a hybrid between a game engine and a render engine. It may not sound important, but those 90% need to be written, too. And this is where actually most of our resources go. And this where the pre-cursor engines of the other GPU render engines have an advantage.

Remember, I said a while ago, that one of the reasons why we don't have motion blur is the missing animation capability. That's still true. We are currently adding that capability (at least the ability to import, store and render animated data), but those things take time. Often more time than you were hoping for. And that's why for example the node system rewrite is so important, even though it doesn't immediately provide any benefits for rendering.

I hope that answers a few questions.

Cheers,
Marcus

[cfrank78]

In general it is true that if you add more stuff things slow down on the GPU. A lot. What we then usually do, is optimizing the heck out of the added piece of code and the corresponding data structures, plus some more optimizations in other places, to compensate for the loss. It's certainly a lot harder than adding stuff to say a CPU render engine. It's basically means, that we have to make compromizes all the time.

Obviously, we don't know how good or bad other engines can be expanded, but I recently read a BOXX intervew with Chaosgroup and they said something similar about VRay RT.

What I really miss here in the forum, is the appreciation that Octane doesn't have that much development heritage: Arion had FryRender as a pre-cursor, Thea is the successor of Kerkythea, Iray was originally developed by Mental Images, which also wrote Mental Ray. And Vray RT is an add-on of Vray. You might ask, what has that to do with GPU rendering? Well, having written some CPU render engine certainly helps with writing GPU functionality, even if you have to restructure a lot and change algorithms.

And what is even more important: Octane is maybe 10% render engine. The reset is there to get data from the user/plugin, prepare it and get it into the render engine, and then to get the results out of the render engine and somehow present to the user/plugin. And all of that interactively. A little bit like a hybrid between a game engine and a render engine. It may not sound important, but those 90% need to be written, too. And this is where actually most of our resources go. And this where the pre-cursor engines of the other GPU render engines have an advantage.

Remember, I said a while ago, that one of the reasons why we don't have motion blur is the missing animation capability. That's still true. We are currently adding that capability (at least the ability to import, store and render animated data), but those things take time. Often more time than you were hoping for. And that's why for example the node system rewrite is so important, even though it doesn't immediately provide any benefits for rendering.

I hope that answers a few questions.

Cheers,
Marcus

Hi Marcus!

Thanks for your very detailed answer. I am a C++ coder by myself. Ok....to be honest - i WAS. I am not that good, to write sutch complex codes and yeah i know, gpu coding is extremely complex. But i worked in "codergroups" too. I know that its hard and timeintensive work, especially if you are a small team. Don´t get me wrong, i understand all of that!

You on the one hand are a small team and work very hard. We know that and understand that and i really appreciate all the hard work you do, especially for sutch a low price. BUT.....you are a small team, with a HUGE usergoup.

Please allways remeber ONE thing. We decided ourself for Octane. Why did some or maybe most of us do that? Not ONLY because of its low price. We saw potential and capabilities in this amazing piece of software. Thats why this forum is so unique und so many users give their (little) sparetime, to help, to test, develop, help each others, bring ideas and much, much more. In all my web carrer i never saw a user forum like this! Its pretty unique, from my point of view.

But if after YEARS, still some of the most important basic features are still not implemented, the team should not be surprised that users here, that stood from the beginning on by your side, slowly get frustrated, especially if they read NOT POSSIBLE(right now) quite often in the past month. What do you think a users thinks, when he reads that. I thought: " So what, finally after years it MAY come and then......or not? Or...Yes. And IF Yes, it will slow down. What the .......*confused*"

Again, i really appreciate your hard work. I really do. But when i have a client, that gave me a "hard work job" and i failed, even if i worked 60 hours without sleep.......the client would be angry too. Understandably!

I am not angry, i am not frustrated, but i really have NO clue when Octane will have those features available and THAT makes me sad, cause i and sadly some person i know think, that i drive a ferrari in the 2nd gear only. We HAD TO bye and learn other Softwares, that made some of our projects possible and I HATED it, cause i like Octane very much. I really hated it, to learn other softwares, just to realise my clients Visions.

Thats all i wanted to say and describe why i think like that!

Kind regards Chris!