Comparison table

[pciccone]

Hello.

There are several threads, here and in other forums, related to how LuxRender compares against other engines. Some comments that I saw regarding Cycles, for example, show a vast amount of ignorance about what Lux or Cycles can do. Some of us have been active in explaining the differences but that effort is just too much, too scattered and too prone to fail.

People who arrive at the LuxRender site generally want to know what this program does and how it's different from others. We know how many rendering engines are out there, we need to help people understand what is unique about Lux.

I suggest to make the task of "figuring things out" a lot simpler, and in the process create a tool to propagate more accurate information about LuxRender. This will result in a unified message and a lot less bickering and arguing in the forums.

We need a comparison table that shows, objectively, the features present in LuxRender and how they compare against other renderers. This table should be easy to be found from the top page of this site.
At the beginning we need three renderers to be compared against Lux:

- Cycles
- Octane
- Renderman

The table should be organized with each row describing a feature (e.g: Volumetric Light absorption) while each column would represent a renderer: LuxRender | Cycles | Octane | Renderman
The intersection of rows and columns would have a simple "Y" (yes) or "N" (no) or a small blurb to describe partial support or that sort of things.

I strongly believe that this effort will go a long way toward clarifying what LuxRender does and what it can provide of interest for 3D artists. In addition, whenever there is a discussion in some forum about "the future of such and such" or "XYZ can do this that ABC can't", we can simply point people to the table and let them learn the facts.

What is your opinion?

[J the Ninja]

Linking this prominently on the main site would be a good start:

http://www.luxrender.net/wiki/Features


As for the comparison table, I'm a bit hesitant about including commercial renderers on there, the companies in charge may not like it.

[pciccone]

That is why we need to stick to the absolute objective facts. This is not an attack or a "campaign", it's just informing the public about the power of an Open Source project. Linux has done it for many years, with comparisons against Windows and Mac OS, Blender does the same when talking about Maya. It's all about informing the public.
Most of my non-geeky friends still today don't know what OSS is. The need for information is very, very high.

When people read that Octane is physics-based they wonder how that stacks up against LuxRender, for example. After all, the reasoning goes, if physics is what slow down the rendering process, how comes that Octane is so much faster than Lux? A table can bring clarity. We are not going to go "against" anything, we simply need to explain what Lux is all about and provide choices. The same is for Cycles. I read some forums where people summarized the difference between Cycles and Lux with the fact that Lux has the Bloom effect. Clearly there is more to it but people don't know. A comparison table would help greatly.

The wiki page that you linked is great, unfortunately it requires a lot of reading, something that might discourage some people. Tables are great because they summarize, in a concise way, the relevant pieces of data. The wiki page should definitely be linked, I will actually post it on my Facebook page and other places. Neverthless I believe that the table would bring a lot of benefit in clarifying the uniqueness of LuxRender.

Let me just add that commercial products and technologies are often referenced in white papers and studies as the basis for expanding other technologies. As long as we are accurate and stick to the facts there is absolutely nothing wrong about this. We would be providing useful information. The vendors of the commercial products can publish their own material, if they want.

Don't underestimate the need to inform the public. Many OSS projects fade away because the developers fail to communicate with the public. This is a very important step and an easy one to perform. I anticipate a lot of positive outcome from publishing this piece of information.

Cheers.

[B.Y.O.B.]

I don't know if this is worth the effort... some people are simply lazy about proper research and at the same time think they know everything and teach everyone about it.
I think the people who can get the most out of Lux, create beautiful art with it, are usually so open-minded and careful that the do not belong to those riots.
What does it help to tell someone who thinks Cycles is the first renderer ever to support global illumination and real glass (and even AO - woohoo ) about the reality? Either they get some facts wrong and continue flaming, or they do not even read such a summary, as it will get rather long (I think).
The people who would read a long site like the current features page are the careful and loving artists I mentioned above.

[Pilchard123]

I don't know what it should be, but I do think there should be something.

...I feel really awkward saying that when all I've ever contributed is a fuzzy test render.

[pciccone]

I don't know if this is worth the effort... some people are simply lazy about proper research and at the same time think they know everything and teach everyone about it.
I think the people who can get the most out of Lux, create beautiful art with it, are usually so open-minded and careful that the do not belong to those riots..

This is reverse logic. A comparison table, or any information tool, is designed to provide information to people before they become familiar with a program. If we don't educate people then we can't complain that the public at large doesn't know about Lux. People make their opinion based on what they know and they get to know something if there is information about that something.

Lack of proper, user-friendly documentation is the most prominent problem of almost all OSS projects. That is why so many fail and why so many people don't know about better OSS alternatives to commercial programs.

Think about it. Why commercial programs flourish? Because of communication. It's all about informing the public. Some times the information is deceiving and that's why people use inferior products. Put your favorite beer example here. Seriously, that is why information is so important. It can actually stir the public away from great products. There is no doubt that commercial products invest money in informing the public. There is no doubt that Cycles is getting its own little "PR machine". Look what happened in just a few months of its development. Lot of attention has been diverted from Lux and SLG, which could have been easily integrated in Blender, in favor of Cycles. If we seat here hoping for "open minded people" we could see our beloved rendering engine being eclipsed.

It's a matter of information. If we inform the public then we will get the results.

In another thread here there was a call for suggestions on how to make Lux more popular.

This is how: inform the public.

[SATtva]

I've made a table mockup, feel free to update/extend it. I wasn't sure about Octane features as i don't use it -- please replace "?" with a proper mark.

lux-comparison.zip

Open Document Spreadsheet

(15.04 KiB) Downloaded 11 times

[B.Y.O.B.]

Cycles can use the CPU or the GPU (with CUDA and OpenCL, although OpenCL is lacking many features at the time speaking).
btw. here's the features page of Octane render: http://render.otoy.com/features.html

[pciccone]

Thank you Satva for starting this. We will need to flesh it out more but this is a good draft.

Cheers.

[B.Y.O.B.]

Maybe a compact list like SATtva's, but restricted to LuxRender would be better?
It would avoid this situation:

As for the comparison table, I'm a bit hesitant about including commercial renderers on there, the companies in charge may not like it.

And most commercial renderer systems do also have only their own features described.
It also removes the burden to update the list everytime another renderer gets a new feature.

I'm thinking of something like the feature's page, but in a short table.