Serious question, are there cases where animators smuggled their fursonas into big budget movies?

>>4088 I'm sure they have. Several cartoon shows on CN, Disney, and Nic have had writer and creator self inserts, Owlhouse and nuPPG comes to mind.

>>4088 not a fursona per se, the kitsune from hotel transylvania looked so damned out of place, either someone in the crew with enough influence shoehorned her in, or someone paid serious bucks for that.

Spider has puss puss.

>>4091 Haven't watched the 2nd and 3rd movie yet, mind filling me in on who is this kitsune and from which movie she came?

>>4102 from 3rd movie https://hoteltransylvania.fandom.com/wiki/Kitsune

Here's the full movie if anyone wants to watch in 44 kb/s.

>>5439 If this doesn't go through I'll try again later.

>>5439 If this doesn't go through I'll try again later.

>>5447 It shows an attached file but opening it gives a 404 page. Is it a camrip or is the actual movie out already?

>>5448 It's been out for a couple weeks, but at these bitrates, it's pretty comparable to a camrip in quality.

>>5455 Finally, I probably should have given a couple more kb/s to the audio as there are a bunch of scenes with dialogue over music that really push opus to the limit of understanding.

>>5455 Nice encoding. What options did you use? Just a standard 2-pass encoding with a low target bitrate and a lowered resolution? >>5456 Yeah, probably. Even the Kung Pow webm has about a 2kb/s higher bitrate.

>>5461 It basically ended up being: ffmpeg -y -i "%%a" -c:v libaom-av1 -b:v 31k -passlogfile "%%~na" -pass 1 -deadline best -tile-columns 0 -frame-parallel 0 -row-mt 1 -auto-alt-ref 1 -lag-in-frames 25 -aq-mode 3 -an -f webm NUL ffmpeg -i "%%a" -c:v libaom-av1 -b:v 31k -passlogfile "%%~na" -pass 2 -deadline best -tile-columns 0 -frame-parallel 0 -row-mt 1 -auto-alt-ref 1 -lag-in-frames 25 -aq-mode 3 -c:a libopus -b:a 5.9k -f webm "newfiles\%%~na.webm" I originally used 8k for audio, but the video overran the target so much that I had to redo the audio at 5.9k to get it below 32MB. The source video was already 720x300 (1100 kb/s h264 + 256 kb/s aac).

>>5456 If I remember correctly opus is not great at low bitrates. >>5462 I admire your patience, that must have taken so many hours.

>>5462 >-lag-in-frames 25 As far as I can tell, the maximum should be 48 now. Other than that there's a few options I don't understand the point of: >-row-mt 1 With -tile-columns 0, doesn't that result in only one frame, and therefore make multithreading useless? >-deadline best Seems to be a libvpx-vp9 flag and it's noted in the documentation there that "best" may give worse results than "good". >-frame-parallel 0 Does this help improve the filesize? >-aq-mode 3 The places I've checked recommend not using -aq-mode 2 or 3. Another interesting flag for low bitrate AV1 encoding seems to be -aom-params deltaq-mode=1 From https://forum.doom9.org/showthread.php?t=183906 : >deltaq-mode=1 is spatio-temporal adaptive quantization based on objective metrics, working in tandem with temporal RDO (tpl-model) to get nice coding gains by deciding costs between inter and intra coding modes alongside temporal optimizations. Works well at low-mid bitrates, but at higher fidelity levels and especially grainy stuff, it can be a detriment to fidelity. >video overran the target so much that I had to redo the audio So you just reencoded the audio and muxed it with the old video track. Understandable I guess. These things take forever to encode.

>>5463 >If I remember correctly I don't think you do, opus is pretty damn great.

>>5455 I tried to transcode the whole movie at 576x240 to see if i could get better audio quality at the same size. It would take over 10 hours to finish. So here is Weird Al instead.

>>5455 >>5513 Through various tricks I managed to improve the audio massively without re-encoding the video.

>>5455 >>5513 Through various tricks I managed to improve the audio massively without re-encoding the video.

neat