I recorded an animation at 2k or 1440p using the H.264 default settings. But the final result i was not impressed with as it looked pretty much the same as in 1080 resolution.
Is there any encoding settings or even another codec that can bring out even more quality?
Im thinking of lowering it back down to 1080 but i want to push the encoding quality as high as possible, i just dont know if i should clock lossless output or increase the bitrate or change any of the other settings.
What are you going to do with the result ?
But the final result i was not impressed with as it looked pretty much the same as in 1080 resolution.
In what way ?
First, render your animation render as a lossless image sequence. NEVER render straight to a video format unless you knwo of a particular reason to do so. Render to image sequence and convert to a movie format afterwards
If you’re going to upload to a video hosting site use their recommended codec and output settings
You will most likely get a better result if you convert the video sequence to a video format with an external application.
I rendered straight to video codec, and seemed to get compression artifacts on the fast movement.
Is it for definite that you should render to image first? It will take more time and work that way…
Imagine you have a 250,000 frame animation.
You render straight to a video codec and a hiccup occurs at frame 249,995. Result: The video file is compromised and does not play back. Days (weeks perhaps) of render time for naught.
Now you render the same animation to image files. The same hiccup occurs at frame 249,995. Result: Just render the missing five frames again and compile your video from the image sequence.
That’s why you never render straight to video.
Or just take your case: You accidentally selected a suboptimal video codec.
Result: You have to render everything all over again. If you had rendered to an image sequence, you could just compile the video again with different codec settings in mere seconds.
Check your bitrate settings in the codec. There min and max settings. Many times the compressor tries to fit the bitrate to the min level.
About making image sequence. It’s much wiser to make images first, because after that you can experiment optimal video compression levels. No need to render from scene again and again.
It is good for non destructive workflow. Every stage is finished (perfect), no need to walk back and forward. Saves the time rendering it again & again. Just edit, composite & encode.
^hahaha…
Speaking of codecs… Does anyone know the optimal codec settings for a youtube video?
More specifically: in every step, render to MultiLayer OpenEXR. (In the “final print” you can omit MultiLayer if you want to.)
This frame-per-file format is specifically designed (by Industrial Light & Magic, no less … with a little help from Blender) to store data. Yeah, the files are big, but who cares: they capture the data, losslessly, in a high-resolution floating point format. (MultiLayer stores many different corresponding datasets … “layers” … in the same file.)
Yeah, your project’s output directory contains a gaggle of individual directories with maybe thousands of big files each. Your computer doesn’t mind … and multi-terabyte external drives are cheap(!) now.
The “final print,” too, is a directory full of OpenEXR files, carefully write-protected.
Now you concern yourself with making the video-file deliverables. These blend-files read the “final print,” and create an appropriate video file from it, doing whatever adjustments may be called for. This is the only time when information is being thrown away … and of course, it is not actually being “thrown away” at all since the source data remains unchanged. If you need several different video formats, frame sizes and so forth, you simply create a blend-file to produce each one.
