So true, and there’s per say no chances these features/improvements will ever see the day light.
We would need 2-3 full time highly skilled devs to get Natron being suitable for real productions.
Too many things to be fixed/rewritten.
First we need to fix bugs. Having a stable core is more important than features (IMHO).
Also, some feature requests exists through third-party plugins (mostly commercial), Natron can’t include “all” the plugins 
We need developers willing to make the missing plugins or features. The easiest would be creating plugins, there are enough examples in openfx-*.
I will try to dedicate as much time as I can to Natron, but it will never be as much as it was (at least not at this moment), but my time will probably be used to maintain the build system and bugs. Features will be far between (and some features I just don’t have the ability to add (at least not without doing some research/reading etc)).
I’m not that hopeful in regards to new developers (please prove me wrong), Natron has been public since 2013 and there has been minimal contributions to the code from “outsiders” since that time.
I agree with everything you said.
Regarding possible new devs contributions, sadly, i have to agree too.
There was, and still is, a path for Natron to get attention. Not from devs in the first time, but from small/mid-size studios.
Of course, giving advices and critics after war is over is always easy.
I think really hard to find skilled developer without pay.
Natron is not in a good position on ‘marketplace’.
There are some cheap and expensive solutions on the market. Medium and larger companies choose old and proven software. Compositing is a special thing and a sensitive point at the end of vfx pipeline. Natron is not a content maker like Blender. You can make money quickly with Blender. Natron is in a much more difficult situation.
1 Like
Natron is definitely in a quirky position tool wise. Some of it’s nodes are honestly more advance than equivalent nodes in other software at this caliber, but some basic things people take for granted in other software, like grid warping and certain aspects of the text node require more complex alternative solutions in Natron without resorting to third party solutions. As an owner of Ignite pro I personally have access to it’s grid warp, but if I’m trying to demo for the average user I currently am mostly stuck with Warp fields or Magnetic fields. I’m 90% sure there might be a way to build a cage deformation using roto nodes and STMap nodes, involving inverting from a user defined value point, but I’ve not goten around to testing out my theory yet.
What do you need?
Maybe we should start talking about:
Why is it so hard to find new devs ? (supporting the few we currently have).
Is the code used very hard?
Is the code documentation not good enough?
Do we need higher bountys?
What else…?
I am not a coder, but I am willing to help and push if we figure out where the problem lies…
It’s more wants than needs, but there are silly things I get asked how to do quite often. Like uniform dashed outlines the aren’t limited to being static. Natron’s font outlines grow and shrink in length if they “march” around. Likewise we can’t really make dashed or patterned outlines around roto shapes without jumping through hoops. The fact most other comping software has those kind of things, but Natron would require far more complex steps to produce a similar result is just odd; however, there are very few other compositors (other than Blender really) that can open an SVG.
I can probably do something about the dashes, it was just a feature I added “for fun” many years ago, I never dedicated much time to the feature.
FYI, I have some major features planned for the Text node sometime this year.
Most OSS projects suffer from this. It’s nothing new.
1 Like
I look forward to everything you guys do. While I’ve only started to play with the BlackmagicRAW plug-in I’ve mentioned it to others with the cameras so hopefully there will be some more useful feed back on that soon as well. 
I’m a professional compositor, I wanted to use Natron earlier for some shots and promote Natron, but I can’t play a 4k footage in the Viewer and I found a lot of other problems. It would be good enough if the Viewer works well, other functionality is not so important yet, I think.
What do you mean by can’t play 4k footage?
I read a 4k exr scan, I push play forward button and Viewer is freeze, extremely slow or sometimes it stops after few frame and restarts the play. 64gb ram, RTX2080 super, Ryzen 9, ssd etc., With Nuke everything is okay.
I currently work on a project with multilayered EXR’s in 4k, I have no issues with playback. Make sure you have configured the cache settings.
What OS are you using? (I’m on CentOS7)
I test Natron under Gentoo and Windows10, I got the same problem. I check cache settings.
You might also need to adjust the threading settings. If you search the forum you can see that several people with threadrippers had similar issues.
I’m using an old i7 4771, 32GB of RAM and a 4GB Gforce GTX 970, and have no major issues working with most 4k footage in Natron given the age of my setup. So if one is having issues with 4k on a system with 64GB, a Ryzen 9 and a RTX 2080 then I’ve got to believe it’s either a system driver or something like OpenGL is turned off.
I think there is a problem with too many threads. I’m usually on older i7’s myself, no issues there (8 threads).
I have an “old” threadripper (1950 or something) that I never used, might set it up and see where Natron starts to run in to issues.
1 Like
See here: Natron crashes when rendering
Limiting threads to 8 helped me a lot!
Feel free to post more system specs and comp files so we can improve documentation on how to setup Naron for 4k exr handling…
I’m will be looking at hiring devs for my studio team for our own blender build and for Natron development as well. We havent started the hiring process yet. Waiting for COVID-19 situation to get better. Maybe we can ask these questions when hiring so we can get an idea of what the issue is.
2 Likes