larryboy writes:
Farmerjoe was a popular, open-source, Blender render farm software written for Blender 3D versions 2.4x in 2006 by Mitch Hughes. It stopped working in Blender 2.5x and beyond because Blender changed, the Blender python API changed and because Python itself changed.
This new version for Blender 2.8x was born out of frustration that there weren't any simple, small-scale render farm solutions available and also my memory of how good Farmerjoe was when I used it for my master's films in 2010/11. With help from Mitch I have got it going again, fixed some bugs, added some features and written a Blender add-on which submits jobs through the Blender interface.
Features include:
- Rendering frames (any image format blender supports)
- Using project subdirectories for textures or baked information etc.
- Rendering the resulting frames to an AVI JPEG, AVI RAW or FFMPEG file.
- Rendering a single large image by splitting it into parts, rendering on separate computers and recombining them.
- Specifying a directory in the root of the Farmerjoe share where AVIs or composite images will be rendered. Different people can have different directories.
You can use it on a single computer to queue jobs overnight or on a multicomputer system.
The software is available on Github and I have started a thread on BlenderArtists.
Hope people find it useful :)
(More images available in the Github 'ForReadme' folder)
Thanks Laurence
6 Comments
Cool stuff - thx a lot !!
Sounds promising.
Could this be used on Azure/Amazon ECS or the serverless solution of Google... in an easy and easy to scale way?
Crowdrender has a cloud rendering mode with quite good pricing
I have no experience of Azure/Amazon ECS etc and unfortunately it is really outside the scope of this project. There are other cloudy render farms solutions out there.
Nice to see a revival of old plug ins. Would be nice to see BlenderCam return as a plug in (this time), now Fusion 360 is closing more and more.
Could this be used with a farm of low cost computers like the Raspberry Pi 4? Would there be any mileage in it or would it be cheaper to use more powerful PCs for the project?