LuxCoreRender Forums

CodeHD wrote: Wed Oct 21, 2020 9:45 am There is also the Discord, with about a similar number of people who are very active. (Driven also by a very lively off-topic chat going on there)

Yeah, I'm planning to spend more time there as well.

Once Luxcore gets more user attention, more donations and interested developers will follow.

It is this exact point. And this what i'm talking about we can't simply let time make this happen. Maybe people don't see it that much but it is extremelly hard to get a place alongside native blender renderer aka Cycles.

__Cycles have a super wide community
__ Supported by real animation studio
__ Used in the Film industry
__ Cycles Devs get paid for their work
__ Cycles is involved in Gsoc almost every years
__ Defacto integrated in Blender
__ Work straigth with Eevee realtime renderer
__ Via Blender cycles is supportes by the graphic industry (Epic / Nvidia /AMD / unity and many more )
__ Cycles is supported by most popular Blender community ( Blenderguru /Creativeshrimp / Reynante martinez ...)
__ Recently the Development increase due to enormous cash entrie

And For those who are not entirelly satisfied there is Cycles on Steroide E_Cycles wich is average 2 to 3X faster than normal build.
After that you have to fight for External engine battles wich include many paid renderer like :

Vray
Corona
Unity
Unreal Engine
Radeon Prorender
Octane render

But even with that there is a clear path and opportunity to get a solid place in this bloody battles.
For user to switch one engine to another there are some issue to be overcome :

Number 1 : Performance

For someone to move an entire Library and years of workflow being etablished in a given render engine it require an obvious advantage.And performance is the biggest factor followed by quality. If you take a look at my channel you will see that the most viewed video is the Cycles vs Luxcore Benchmark with a total of 13 115 view : https://youtu.be/euEtA3RQg0I
For interior rendering Luxcore is ahead of normal Cycles but with E-cycles this gap is not that obvious for someon to make the switch.
For Product visualisation and simple scene cycles is ahead E-cycles too.Volumetric rendering being the worst part for Luxcore (smoke/fog).


Number 2 : Quality

Few people among artist consider it but it is a very important one for very best artists. Luxcore is doing a great Job with the caustic feature and overhall quality is better than Cycles and Vray from my POV.
But your not going to make a large userbase with this factor. And it is even true when it come with a Performance penalty.


All that being said Performance should be our focus.If you're not convinced that performance is important just look people interest for realtime engine like Eevee Unity Unreal.
It is not Lost but the performance level required 3 years ago to be competitive has been reset.

Good points, I largely agree.

I regard Luxcore as the Rolls Royce of Blender renderers: it's not the fastest car, but it is the most polished, attractive car for rendering connaisseurs.

If I may direct the devs to an area that would benefit a lot from improvement:

LuxcoreRender SSS is great, but difficult to control (needs multiple nodes and studying of the many parameters), and it is very noisy / takes a long time to become noise-free.

Cycles performs better in that area, even when using the slower Random Walk SSS type. To me as a character / toy modeler, SSS is an important instrument for plastics and credible organic materials like skin and food. It would be great if Luxcore SSS would become both faster and more accessible.

I use an area light with very narrow light angle as light source.

Hello,

MetinSeven wrote: Wed Oct 21, 2020 11:54 amRolls Royce

I don't think that the parameters are to complicated. They in some cases just different. If you really like to remodel real world behaviour you'd probably need even more parameters.

The question is where does luxcore want to go to?
Does it like to be a cycles clone, or does it like to attract scientists and engineers.

Of course both would be possible too...

Since I don't know any open source full spectral renderer (maybe Mitsuba but V0.6 isn't developed any longer and V2 is not as complete as before). So supporting a full spectral work flow makes luxcore attractive and unique in one step.

How do we proceed to start the implementation ?

BR

The question is where does luxcore want to go to?
Does it like to be a cycles clone, or does it like to attract scientists and engineers.

Theses questions has been discussed right after the rebirth and feature poll vote.having Cycles as direct competitor doesn't mean we are or need to be cycles clone in terms features and global philosophy. And the community choose to make Luxcore a Fast yet efficient photorealistic render engine.
It has been done for thoses who need a better solution for high quality render compared to Cycles and other solution out there (Vray / Corona).
But all that should not be done by sacrifying Quality/Precision. The reason why Spectral rendering is still on the table.If even cycles is considering being spectral it isn't Lux to be RGB forever .

BUT the team wann't replicate past error where the project died because of :

__ Tiny User base
__ Lack of support from community to the Devs
__ Lack of clear roadmap and Goal
__ Lack of donation

Considering the current Devs team size and financial support you necessary have to make priorities. Old Lux main issue was speed not quality/precision.
So the engine Development start there. And we get PGI cache / Light tracing and more recently RTX optix Support. But all our competitor increase pace too.

We need a new feature poll guided by vote and bounty support.

lighting_freak wrote: Thu Oct 22, 2020 4:45 am Does it like to be a cycles clone, or does it like to attract scientists and engineers.

If one coniders scientists and engineers, you have to have a clear idea what features you would offer them (and advertise as such). The vertex-based Blender is not a replacement of CAD software for engineers, and the 32-bit workflow will be insufficient for many scientific users. By my experience, you can still use it nicely for serious purposes, but as I know my fellow scientists, you will have a hard time converting them

Anyways, like Sharly I don't think it is necessarily an either or. Sharing features with Cycles is a good thing because it makes it easier for users to switch or use both in paralell with a similar workflow. If you do nothing else than cloning, it will be so similar that there is no point in using both.

I imagine luxrender as a rendering engine which delivers state of the art rendering quality wen using bidir, spectral or not, but by sacrificing speed. however LuxCore offers alternaive engines which allow you a realtime experience at lower quality. That means that the production process is done using faster engines and the final renderings using bidir. To archive that all rendering engines should be slightly better alligned as they are which should not be difficult.

Some note about Project donations because I have the feeling people are sometime detached from the facts:

- 50% of our monthly donations come from https://www.vectary.com

- the other 50% from about 10 Blender users (pretty much always the same 10)

- LuxCoreRender v2.4 for Blender has been downloaded about 20,000 times

- recently, there has been a noticeable surge in LuxCoreRender success across Blender users

- the increase of Blender user hadn't lead to any noticeable change in the amount of donations (only more questions to answer on forums, etc.)

The rational conclusion is that a special project like Vectary is worth many thousands of Blender users from the donations point of view.

This is not a business and I'm not saying that this project should be driven by donation sources but the above facts are quite clear and worth some tough.

P.S. as proof of what I'm writing about "Special projects", it looks like we have recently received a $1000 donation from https://www.felixrender.com

- the increase of Blender user hadn't lead to any noticeable change in the amount of donations (only more questions to answer on forums, etc.)

This is normal as people are just starting with a new engine.Some are in the middle wondering where to go cycles or luxcore.Some already like The engine quality so they start learning and converting workflow and assets this move and learning process is by itself an investment. But the majority of downloader haven't completelly make the switching step.

- 50% of our monthly donations come from https://www.vectary.com
- the other 50% from about 10 Blender users (pretty much always the same 10)
The rational conclusion is that a special project like Vectary is worth many thousands of Blender users from the donations point of view.

The fact that Vectary contribution is equal to 10 regulars blender luxcore Users is proof that a good user base can make a difference.When the switch is done or you are sure to do it then you start support by :

Donating
Producing art with the engine
Report bug
Teaching
etc ...

What we want is converting simple casual engine tester into real regular Users.