I'm making a feature animated film by myself and it renders 1 minute per frame. I have 300,000 frames. And this is AFTER I gutted everything to the bare minimum and upgraded my specs.
Before I kill myself, is there anything I can do besides go back in time two years and use Unreal instead of
[...]
?
You don't even have 300 frames, cris
Yes I do, go masturbate to chicken meat
Not Cris. Fuck outta my thread.
Cris, fuck outta my thread.
Chicken meat fucker, get out of my board
What's with this chicken thing? The unfathomable jealousy of someone actually making something.
I've been doing render tests since day one. Despite optimization, they've ballooned up to a minute because of the scope.
a minute per frame is a best case scenario.
you should consider moving to eevee next for improved raytracing.
you should consider using 2.39:1 @ 108p to cut down on pixels (1920*804).
assuming you're not a lying homosexual that should kill himself, by the time you've completed production (minimum 3-5 years) gpus should cut your render times in half or less.
Use EEVEE but use the screenspace raytracing branch of blender. And use cycles in scenes where it is necessary(refractions and other effects EEVEE struggles with).
Otherwise maybe try tweaking the settings of cycles. Lower samples with denoising, lower resolutions with upscaling in post, etc.
You could even import your scene into unreal, and only learn enough unreal to use it as a renderer.
Or rent a renderfarm
This is EEVEE... 1080p, 3 render layers. I cut it down from 8 at 1440p.
Where did I say it was ready to render? I'm only two years in, nearing the end of preproduction. I obviously make plenty of test renders which is the reason for this thread: a tiny ten second animation with just one character takes basically the whole night to render, and that's if it doesn't crash. It's fucking my workflow and I don't know if this is a viable state to enter production in. Once I do, I'm locked in.
Are you rendering your chicken tube meat Coomer simulation?
Stop lying, you are 2 years on a production and never made render tests?
You are full of shit or completely retarded, maybe both
Have you tried lowering the amount of samples in EEVEE? The default is 64. But going lower can decrease the amount of render time a lot. It only somewhat causes aliasing and might undersample some effects. But you can then raise resolution while still having shorter render time.
It also depends what your bottleneck is. 1 minute per frame sounds unreasonable for what is essentially a realtime render, unless your hardware is over a decade old.
If it's a draw call issue you might be able to run multiple instances of blender to use more cpu, but then you might run out of vram.
Maybe post some in blender screenshots and render. You might be doing something that costs a lot of performance without even realising it.
Use 3d render farm services like anybody that does not have 2 3090 or a 4090.
>EEVEE
It's been broken from 3.0
>I have ~166 minutes of animation ready to render and only now do I discover it takes 1 minute per frame.
Yeah you've animated an entire feature film by yourself without test rendering, sounds legit OP. I don't tell people to gtfo, but I mean come an GTFO.
Is it too late in the production timeline to consider using Unreal for rendering?
Unfortunately yes, the artstyle is wrapped up in too many Blender-specific shaders and exploits to migrate. It must be so nice to use Unreal.
Thanks for the suggestions. I don't work in the industry and I can't find much information about this, but from what I've gathered render target time should be under thirty seconds for an indie project like this. I've looked into render farms but a lot of them don't work with render layers.
sure, just buy a 7090 or w/e is out at the time and you'll rip through renders in under 30s
that's only a few days, a week at most. just deal with it impatient retard. kids these days and their instant gratification. fuck off with this shit
Are you stupid? It's almost an entire year of render time.
Op wants us to believe he's made a feature length film ready to be rendered while simultaneously not knowing how to rent time on a render-farm. Nobody renders CGI beyond test renders on their home computer you absolute poser.
You have some wild idea of making the thing you claim to have and pretend like a 1 minute render is the thing holding you back. GTFO.
Read the thread. I'm two years in, nearing the end of preproduction and finalizing all my shaders. At the moment I have an average 1 minute render time and that seems cripplingly high for a project like this. I want to fix it before I'm locked in.
>finalizing all my shaders
You are ~2 weeks in and prototyping all your shit-posts.
Nobody makes a finished ~90min film and hit render. That is not how this works.
You'd know nothing about how your scenes look in motion and how the light need to be tweaked etc.
Every few seconds worth of animation need to be tested re-rendered and tweaked and corrected many times before the final version
of a scene is assembled as what goes to the final cut. You're a massive LARP'er and a fraud.
I'm not wasting my time writing to someone who can't read.
Nah, not me. This is an original IP.
Are you the 'Berserk in style of Arcane' guy Op?
I think you are the Berserk in style of Arcane guy.
anon... what framerate is your animated feature at... lol
24 since I hate the 12 fps meme with all my soul. Yes, it will be a four hour movie.
can you appreciate why people might think this is a low-effort troll post?
Not really, no. There's nothing particularly unbelievable about it and I'm acutely aware of my technical, artistic, and practical limitations, hence this thread. It's not your run-of-the-mill two minute short film with ugly clay people or unrigged objects because that simply doesn't interest me. What does interest me is the anger some people display whenever I discuss it.
Sometimes I forget there is a certain class of person on the internet who gets legitimately upset whenever anyone dares to make anything or have ambition - I guess because it reminds them of everything they lack.
Believe me or don't and close the tab. Like I said, I'm two years in and all I'm interested in is optimization advice.
okay, friend; good luck with your endeavour.
i think the advice you've received here is sensible and without seeing the project i don't think anyone will be able to offer more specific insight.
Thanks very much. I'll be posting bits of it on the board after I go public with it, since it's under my real name.