Cycles Standalone Engine

General Forums Latest News
61

The problem is, I’m sorry to say with Blender only that will never happen, for the reasons I stated in my post. Blender is a community project and while it has some production experience under it’s belt it’s far from the production experience that other software in this already niche market have. Ofcourse the BF will always throw out the qualifier that they aren’t trying to compete with commercial packages but that to me is just an excuse to not have to be held to the standards of production level software.

On top of that, while we have amazing work coming out of the community the level of actual professional production experience is still low in comparison to other software.

Cycles can only benefit from being ported to other packages, and maybe it can find it’s niche in an already crowded market. Right now folks think it’s amazing only because it is better than what was there before for the most part in-terms of OSS renderers and it’s already integrated into a package they have and can freely download. But there are thing Cycles can learn from other packages that can benefit Blender in the long run as well. Right now Blender’s hair, fluids, particles, etc. are fairly rudimentary at best. Let’s see what happens when you render something more complex from Maya, Houdini, Softimage, or Max. Does that break the renderer? If so let’s address that issue and maybe raise Blender itself to that level as well. Speed and resources are still issues. How can Cycles address that in a movie where you are rendering hundreds or thousands or elements (from foliage to an attacking army, with smoke and effects to boot). A lot of effects studios are now starting to do more “in-camera” work, can Cycles handle that kind data and scale well? These are things we aren’t going to find out from Blender alone because Blender doesn’t have the capability or functionality at this point.

Hamstringing Cycles because people want improvements to Blender is short thinking, not long term thinking. The added benefits to Blender are nebulous at most and as we have seen Cycles has been steadily adding features and fixes while Blender itself has remained stagnant in significant areas. There is a better chance that Cycles could become a world class renderer than Blender becoming a world class DCC app. Sorry if that sounds harsh but so far I’ve seen nothing indicating I’m wrong.

62

I really wish there was a like button on this forum. I really agree what your saying here.

63

'Cycles can only benefit from being ported to other packages, and maybe it can find it’s niche in an already crowded market. ’
It can benefit, but for me it seems that instead of fixing our car what has a flat tire, weak battery and it is structurally corroded like hell we just want to make the neighbor’s car compatible with our CD player, because we can. And there are 30 other manufacturers who already earns money from producing CD players for that specific car our neighbor has.
It makes no sense.
Let me to be honest: I do not care if Cycles would break under Max. For Max I have Corona and VRay what will not break and offers me everything what Cycles cannot.
What I do care is the production environment with Blender (and its renderers).
From archviz perpective Blender have VRay, Corona. The first one is supported officially and does not cost a lot. Good (but not OSS/free) renderer is checked.

But I just read a few days ago the ‘18 most anticipated’ stuff for Blender development and I’m always wondering: do we really need stereo viewport while it is slow like hell in ‘standard’ mode?
Would not make sense to improve layer management (this 20 layers limit is a joke) instead of projects like this?

There is a huge difference between Blender and Cycles. Blender as DCC has no alternatives for budget artists, for poor guys in developing countries, etc. Cycles has.

Some other thoughts:

  1. Talented people could make miracles if they use their talent in the way they should.
    The lack of long term thinking and proper directing in the development of Blender causes anomalies I wrote formerly.
    Could Blender be a better tool easily? Yes. But it needs pro leading, pro thinking.

  2. I do not think that Cycles could have any ‘market’ besides Blender. As I wrote, it is not outstanding at all, not in any feature or speed or quality or anything. When I first tried VRay or Corona, I had the feeling that ‘these will be successful’. And for the rest I didn’t have that feeling. And VRay became standard during the years, people started to migrate to Corona from VRay slowly.

  3. In both VRay and Corona a very-very small team made the miracles. Very-very small team would be able to make miracles with Blender itself. With focusing Cycles instead of its host the free/OSS 3d tool battle has lost.

64

But that’s a chicken and egg thing. Is Cycles not outstanding because it’s tied to Blender (which has no way to tap into what would make a renderer outstanding anyway and no real way to push the renderer to production level limits) or is Blender hamstrung by Cycles. Due to the level of advancement in just the few short years that Cycles has seen I think it’s the former.

Think about it this way. No one thought much of KHTML when it was first released. It was just another KDE project and people questioned why KDE didn’t just go with Gecko. Apple took the project removed all the KDE specific bits and now we have Webkit. Cycles has the opportunity, even more so since a lot the Blender specific features have already been decoupled.

It’s not about whether it will be successful. That doesn’t really mean much in an open source context. Blender, The Gimp, even KDE, should be the poster child for that. Right now Cycles does a lot of things that someone who may be interested in rendering technology doesn’t have to rebuild or code themselves they can build on top of Cycles. Scratch their own itch as it were for specific features. That can be very enticing for smaller houses that may want have their own rendering solution but may not want to start from scratch. At the same time if someone is interested in porting Cycles to Max, Maya, or whatever, they will have to quickly get into use cases where production level features are needed. That can only help not hurt Blender in the long run.

How can Cycles be extraordinary if the software it’s tied to isn’t extraordinary as well. Arnold is a great renderer, and it’s been around for a while (via Project Messiah). It was always considered a pretty decent renderer but it wasn’t until Sony Pictures decided to use it in their production pipeline that it really took off. The developers ran into production level problems, with production level software/pipelines and had to meet the challenge. Right now there is no way for Cycles to have the same opportunity until it breaks away from Blender. Blender as software and Blender as a community isn’t up to the challenge at the moment.

65

I checked both VRay for Blender and Corona for Blender for a while for archviz. Both outperformed Cycles easily.

No one thought much of KHTML when it was first released. It was just another KDE project and people questioned why KDE didn’t just go with Gecko. Apple took the project removed all the KDE specific bits and now we have Webkit. Cycles has the opportunity, even more so since a lot the Blender specific features have already been decoupled.

Unfortunately I do not know anything about KHTML.

It’s not about whether it will be successful. That doesn’t really mean much in an open source context. Blender, The Gimp, even KDE, should be the poster child for that. Right now Cycles does a lot of things that someone who may be interested in rendering technology doesn’t have to rebuild or code themselves they can build on top of Cycles. Scratch their own itch as it were for specific features. That can be very enticing for smaller houses that may want have their own rendering solution but may not want to start from scratch. At the same time if someone is interested in porting Cycles to Max, Maya, or whatever, they will have to quickly get into use cases where production level features are needed. That can only help not hurt Blender in the long run.

I do not argue about the fact that if there are enough resources Blender could profit from Cycles ports. I see the resources and their usage differently; but if Cycles turn to be the best unbiased renderer on the world thanks to ports, still there will be no production-ready OSS/free biased renderer for Blender and its architecture will still remain outdated.

How can Cycles be extraordinary if the software it’s tied to isn’t extraordinary as well.

VRay, Corona are extraordinary within the same environment. The second one was developed by a single man, both Blender exporters were developed by one man team.[/QUOTE]

Blender as software and Blender as a community isn’t up to the challenge at the moment.

I’m trying to guess how could it be changed.
Being realistic I do not think if it ever could be changed; I miss the factor what makes a product outstanding.

You mentioned GIMP (God I Miss Photoshop;). PS and Adobe are ‘evil’ companies for most of the OSS users; I permanently hear - since decades - the complain about how non-innovative is PS as they do no development. And in the last 10 years the very innovative developers from both OSS and commercial side were not able to make a competitive product. That would not require any new ideas or revolutionary thinking. Just copying the features.
But nope. Still no competition.

I used to feel the same regards to Blender. I bought lots of addons on the Blendermarket to fulfill holes on the system. From my side Cycles cannot be improved significantly as it works well and it has known limits; new features could be added, etc.

But Blender itself desperately needs developers and yes, I would be more happy with focusing on the problematic parts.

66

I’m all for Cycles as a standalone engine (I also believe it could only benefit Blender), as long as someone else takes care of the development, that way the BF don’t have to worry about this project and concentrate on the software development.

67

I checked both VRay for Blender and Corona for Blender for a while for archviz. Both outperformed Cycles easily.

No one thought much of KHTML when it was first released. It was just another KDE project and people questioned why KDE didn’t just go with Gecko. Apple took the project removed all the KDE specific bits and now we have Webkit. Cycles has the opportunity, even more so since a lot the Blender specific features have already been decoupled.

Unfortunately I do not know anything about KHTML.

It’s not about whether it will be successful. That doesn’t really mean much in an open source context. Blender, The Gimp, even KDE, should be the poster child for that. Right now Cycles does a lot of things that someone who may be interested in rendering technology doesn’t have to rebuild or code themselves they can build on top of Cycles. Scratch their own itch as it were for specific features. That can be very enticing for smaller houses that may want have their own rendering solution but may not want to start from scratch. At the same time if someone is interested in porting Cycles to Max, Maya, or whatever, they will have to quickly get into use cases where production level features are needed. That can only help not hurt Blender in the long run.

I do not argue about the fact that if there are enough resources Blender could profit from Cycles ports. I see the resources and their usage differently; but if Cycles turn to be the best unbiased renderer on the world thanks to ports, still there will be no production-ready OSS/free biased renderer for Blender and its architecture will still remain outdated.

How can Cycles be extraordinary if the software it’s tied to isn’t extraordinary as well.

VRay, Corona are extraordinary within the same environment. The second one was developed by a single man, both Blender exporters were developed by one man team.[/QUOTE]

Blender as software and Blender as a community isn’t up to the challenge at the moment.

I’m trying to guess how could it be changed.
Being realistic I do not think if it ever could be changed; I miss the factor what makes a product outstanding (I mean the Blender community is outstanding when it comes to help each other - big respect for that - but somehow (I can’t really find the proper word) - ‘rigid’ when it comes to comparisons as a basis to developments.

You mentioned GIMP (God I Miss Photoshop;). PS/Adobe is an ‘evil’ company for most of the OSS users; I permanently hear - since decades - the complain about how non-innovative is PS as they do no development. And in the last 10 years the very innovative developers from both OSS and commercial side were not able to make a competitive product. That would not require any new ideas or revolutionary thinking. Just copying the features.
But nope. Still no competition.

I used to feel the same regards to Blender. I bought lots of addons on the Blendermarket to fulfill holes on the system. From my side Cycles cannot be improved significantly as it works well and it has known limits; new features could be added, etc.

But Blender itself desperately needs developers and yes, I would be more happy with focusing on the problematic parts.

68

Bear in mind that developers are people with specific interests and expertise. They’re not modular parts that can swapped in and out randomly. In short: just because Cycles is getting development energy, that doesn’t mean that other parts are proportionally suffering. If you told all of the developers working on Cycles (there aren’t that many) that they couldn’t work on Cycles anymore, that doesn’t mean they’d necessarily work on another part of Blender. More likely, they’d stop altogether.

69

Corona wasn’t technically written by “one man”, it piggybacks on Embree, a technology which Intel has been developing for years and which delivers outstanding CPU performance.

This “question” (or rather criticism) is pointless. Developer attention is not freely interchangeable. To develop Y, you most likely need a specialist in Y. On the other hand, for some reason somebody already implemented X, so even if you don’t think it is important, you can now have it, so be happy about that or just move along.

70

Dear Fweeb, BeerBaron,

partially I agree with both of you that coders are not ‘modules’, which is true. On the other way they technical skills are more universal, it is their goal what they need additional info/knowledge to work on. That is why I mentioned formerly that all the Blender development process should be based on putting focus on the important things. Correct me if I’m wrong that making stereo viewport for Blender does not really require any other skill than other - and more necessary - viewport improvements. I’m not discrediting anybodies work as I’m really grateful for all additions to Blender.

Anyway, I just hope that things will change for better usability in the future and someone will be payed by BF to make necessary changes, not just ‘optional’. I would be happy if Blender as DCC tool would reach the level of Cycles already reached.

71

Is that the level that you already discredited a number of times as not being production ready?

72

I cannot help you to understand what you do not want to understand.

73

You are wrong. Getting something on the screen and getting it on the screen efficiently are two related, but distinct sets of expertise. Also, people often complain about “viewport performance” when really they’re complaining about the performance of underlying systems that may have developed by someone external to Blender. Those systems may have merely been integrated into Blender by someone with little knowledge of the problem domain.

Furthermore, even if somebody may have the expertise to work on X, if they prefer to work on Y instead for whatever reason, they will most likely work on Y instead. In this decision process, it is not necessarily relevant what some users have voiced as important. Blender users aren’t really employers nor customers, so why should their opinions override what developers spend their days with? If you want to turn this dynamic around, you have hire developers for yourself.

Anyway, I just hope that things will change for better usability in the future and someone will be payed by BF to make necessary changes, not just ‘optional’. I would be happy if Blender as DCC tool would reach the level of Cycles already reached.

We’ve had tons of “usability” and “UI” debates here and if there’s any consensus to be found, it would be that the community can’t agree on anything.

74

This can turn in a flame war. :smiley:

What do I say last page?
Please get back on topic: Cycles standalone development.

Cheers, mib

75

Thanks for correcting.

Also, people often complain about “viewport performance” when really they’re complaining about the performance of underlying systems that may have developed by someone external to Blender. Those systems may have merely been integrated into Blender by someone with little knowledge of the problem domain.

Yep, but then how could this be solved? I mean I know that viewport performance is a problem in the most of the 3D apps (except of Max), but if I have to address weaknesses in Blender this is in the top 3.

Furthermore, even if somebody may have the expertise to work on X, if they prefer to work on Y instead for whatever reason, they will most likely work on Y instead. In this decision process, it is not necessarily relevant what some users have voiced as important.
Blender users aren’t really employers nor customers, so why should their opinions override what developers spend their days with? If you want to turn this dynamic around, you have hire developers for yourself.

Yep, I understand this, I just always wondering how non-effective way communities handle problems (not just related to Blender and not just software-related).

We’ve had tons of “usability” and “UI” debates here and if there’s any consensus to be found, it would be that the community can’t agree on anything.

Technically this is why I suggested to have some kind of ‘centralized’ development plan, guided by pros with experience in other tools and some vision for the future.
Most of the people (again, it is not Blender related, it is a generic problem in any society) simple does not have the skills to make worthy decisions on anything.

76

Sorry for being offtopic. Fortunately we are not in a ‘flame war’, just discussing politely.

77

It depends, may be some of it simply can’t realistically be solved, or maybe users have unrealistic expectations or other misconceptions. For example, sculpting in Zbrush works completely differently than in Blender, therefore performance cannot be compared. Some developers are already working on OpenGL performance, but putting more developers on the task isn’t necessarily going to make it go faster.

Technically this is why I suggested to have some kind of ‘centralized’ development plan, guided by pros with experience in other tools and some vision for the future.

Development is already being guided by professionals, but not necessarily the kind of professionals with long and broad experience in production with other packages. I agree having input from such people would be valuable, but then again why would they busy themselves with something like that?

Maybe it’s more reasonable to accept that Blender is going to stay in its own league and that this isn’t necessarily bad?

78

That’s a really defeatist statement. Even if the Blender Foundation doesn’t want to compete with the big guns in the industry it doesn’t mean we as users should be ok with having incomplete features or weak points that clearly need to be adressed.

Even freelancers and small studios need good tools, maybe we cannot pay the overpriced licences of AD, but we can surely donate some money for the Blender development, shouldn’t that count for something?

79

I don’t think it would be realistic to expect ZBrush-like performance and I would not expect that performance. I use Max and Blender and my primary focus was on archviz in the past; as far as I know no other app could compete with the viewport performance of Max with huge archviz scenes and I can accept it. I don’t expect something from an OSS app what commercial competitors could not deliver.
But when using Blender I always have the feeling that even the DOS version of the 3D studio performed faster (it is probably not true, but makes me feel like that:). It also has another aspect (I will write about it).

Development is already being guided by professionals, but not necessarily the kind of professionals with long and broad experience in production with other packages.

Generally speaking I don’t think that professionals just sit and guess the best implementations without knowing other packages.
I saw this kind of guiding in other areas, never worked out well.

I agree having input from such people would be valuable, but then again why would they busy themselves with something like that?

Why people bother with making OSS apps?:wink: Because it is good to make something for others. I was involved in non-commercial projects for years and I enjoyed it, although it cost money for me.

Maybe it’s more reasonable to accept that Blender is going to stay in its own league and that this isn’t necessarily bad?

As an example: see it from the side of poor guys who would like to learn 3D in developing countries. On the computers they usually have or get as a donation it is a pain in the ass to do anything with Blender. Cycles needs GPU for relatively fast results, viewport performance is weak, etc. (A few years ago I was thinking about making a media school somewhere in Africa.)

Or something else: the only reason I’m active on the forum at this level that someone wrote a post that the Blender-Cycles combo is production-ready for archviz, while it is - as you wrote - not in the same league. But I would like to see Blender in the same league as it is not something impossible, but it has clear rules how to get there.

For the things I do now Blender and Max serves me very well. But let me to be honest, I would like to drop Max in the future. And as Blender is a very good swissknife, it would deserve more focused development and let all the good things make people happy through that specific area we focus.

80

You guys really know how to ruin a thread. It was a thread about Cycles Standalone and you turn it into thread number 1000 about Blenders/Cycles place in the industry.

And people wonder why some devs avoid this forum nowadays…go figure. It’s tiresome to read the same story over and over again.