Is there a reason not to use H265 and VP9 over their predecessors other than compatibility?
>>52338221
h265 is shit
Speed for h265 and doesn't look any better
>>52338247
Why?
>>52338350
Have you used it?
>>52338371
30% cpu usage
>>52338404
I'm guessing the post was meant for >>52338350
What are you trying to say?
>>52338423
nope, I though your argument for it's shittiness was high CPU usage
>>52338459
Why would that be a bad thing? I would think being able to utilize your entire CPU would be a good thing. If it's only using 30% of my CPu that means it's not going as fast as it can.
Slow encoder and lack of wider adoption in browsers. Mainly the slow encoders.
>>52338221
Massive encoding times
Higher decoding complexity
Inefficient in comparison to h264
libvpx is still slow as fuck
H265 is full of patent cancer, remember the two licensing pools thinge?
Try encoding a film in H265. The encode time is like 2 or 3 times higher, and the result doesn't even look very good when you reduce the bitrate enough for it to be worth it.
I thought H265 was supposed to compress far better than H264, while still being faster than VP9?
>>52338221
Only if you need speed. If you're encoding a video once for a website, the slower encode time of VP9 is worth it.
No web browser is going to support H.265, though.
>>52338600
libvpx and x265 encode at similar speeds and produce similar looking results.
Currently VP9 decodes much faster than H.265 because much more effort has been spent on VP9 decoders than H.265 ones.
>>52338488
hang yourself memer
>>52338718
Quality post.
>>52338488
100% CPU is good when encoding, but bad when decoding.
>>52338772
I see where I goofed then, yeah decoding should be lower cpu usage.
Thanks for clearing it up instead of being like this edgy faggot >>52338718
I read this thread and what a disappointment. It's only the typical /g/ meme because of the license or 265 is really shit?
>>52338850
Why don't you use it and see results for yourself seriously?
>>52338221
Same video I encode at 10FPS in VP8\Vorbis and 0.033FPS in VP9\Opus, and the result is compressed 15% better. Not worth it, I think.
>>52338221
Takes more computing power to encode and decode.
>>52338488
Not sure if parodying memory-hog apologists. Aside from the 100% scale, 30% usually means a poorly-threaded task is maxing out one thread/core, and the OS isn't always smart enough route other tasks to idle cores.
>>52338978
*enough to route
>>52338863
I just did, and it definitely looks sharper with x265 at the same bitrate.
Not that they are relevant on 4chan anyway, not even VP9 for whatever reason.