Should I wait for H.265?

[KyleMadrid]

I plan on buying some blu-ray movies and TV series over the next couple of months, and I like to rip all my DVDs/Blu-rays to my computer and then encode them usually to an h.264 MKV. I'm not one to worry too much about file size, because I much prefer to keep good quality. However, since I will be ripping quite a lot of high quality content, it is obviously going to take up a large amount of space on my hard drive.

I've read a little bit about the new H.265 codec and how it can produce much higher quality videos at lower bitrates, so I thought that maybe I should wait until it is available before I start encoding any more DVD/Blu-ray rips. But I'm just wondering if anyone knows when it is going to be available to use with free video encoding software such as Handbrake. I've read that it will be available this year, but does anyone have a more specific date, perhaps a month? Or am I asking too much? Basically, if it is going to be available by December this year at the latest, then I'd be happy to wait, but if it's any longer, then I think I'll just go ahead and encode to h.264.

Also, I've read that H.265 will require a lot more CPU power than H.264, so will many computers available today struggle to cope with it?

And I'd also like to know what some of you more experienced video encoders think of this new codec. Is it worth the wait? Is it really that much better than h.264?

Thanks in advance

[Cornucopia]

Don't hold your breath. SERIOUSLY!

First iterations of codecs are never that great qualitywise (not even getting into the much greater computational requirements). H.264 wasn't all that great for the first 3 or 4 years.. Remember, it is hitting ~10 years now.
Then you have to wait for coders to get past all the IP hassles and fully reverse engineer the reference model into something workable for consumers that is not $$$payware.

Either rip and leave as is un-recompressed, or rip and compress using the usual current methods (h.264). Don't believe all the rumor. Use h.265 when a good opensource implementation has been tested and proven to be an improvement.

Scott

[KyleMadrid]

Great, thanks Scott. That was exactly the sort of reply I was hoping for.

[deadrats]

i have a counter point, the divx/main concept/rovi people are almost ready to release their h265 codec/sdk, these guys are experts and their software is licensed by all the major software vendors. additionally the free divx converter will have a h265 profile and it's due to be released in a couple of months, you lose nothing waiting a little while and seeing for yourself.

[KyleMadrid]

You make a good point, deadrats. I guess you're right. You say it's due to be released in a couple of months, and that's probably when I will be ready to start ripping/encoding, so I might as well try it and see for myself like you said.

Thanks for the reply

[deadrats]

http://blog.divx.com/2013/01/25/wtf-is-hevc-2/#more-1137

http://www.mainconcept.com/hevc.html

they've had this up since January, they're waiting to get divx h265 devices certified and as you can see they already have a beta sdk available to ISV's under strict NDA.

let's assume that the first implementation of h265 is only 20% better than x264 (the truth is that even the reference h265 encoder is capable of beating it in most tests), it would still behoove you to give it a shot.

another codec to keep your eyes open for it vp9, a royalty free open source competitor to h265 being developed by google.

[jagabo]

I wouldn't bother with h.265 until there's widespread support. There are no playback devices except computers at this point in time. All MPEG family codecs can give good results given enough bitrate. So the only thing you'll be missing out on is a little higher compression ratio.

[Atak_Snajpera]

Do not forget that encoding time will significantly increase. h.265/HEVC gives up to 30% better quality/smaller size than h.264 specification. In order to achieve that 30% encoding speed will be few times slower than x264 running in Placebo preset (more complex calculations). So you will waste alot more electricity than before.

[jman98]

I have to agree with jagabo on this. Live on the bleeding edge, die on the bleeding edge.

KyleMadrid - If you REALLY care about quality as you claim, then you shouldn't do ANY conversions at all. Period. Whether you realize it or not, ALL of you conversions are actually lowing the quality. You may not be able to notice it, but it's there.

I do not advise anyone at present to spend a lot of time with vp9. I have a feeling that in the end it's going to wind up in the same graveyard as Theora and DRAC.

[Hoser Rob]

I'd definitely agree about not holding your breath. h.264 has been around for quite a while now and I don't think software support for it is really all that great now.

[Cornucopia]

But deadrats, like I mentioned before, you're not going to be getting a high-quality, efficient, workable "x.265" in FOSS anytime in the next ~3 years!. Sure, divxlabs might have something usable by the fall. But, heck man, it's owned by ROVI (remember: Macrovision!?!). Anything they put out now is not only not going to be a free lunch, it is going to be $$$ and probably phone home, to boot.

The level of computation necessary for even decoding h.265 needs to be at least another level of magnitude. That includes the chips in any devices you intend to use to decode (none of which have been manufactured yet). pdr, jman98, Atak, & jagobo are right. It'll be good in due time, but it's just a mirage right now. These things take time.

[fritzi93]

Do not forget that encoding time will significantly increase. h.265/HEVC gives up to 30% better quality/smaller size than h.264 specification. In order to achieve that 30% encoding speed will be few times slower than x264 running in Placebo preset (more complex calculations). So you will waste alot more electricity than before.

That alone deters me right there. It would be worse than going back to encoding BDs to BD9 on my old dual-core machine.

If encoding time increases that drastically for a mere 30%, well...it's hard to see much advantage. Though I've read predictions that the file size reduction may be more on the order of 50%. Eventually.

I'm very interested in test driving it though. I wonder how much CPU horsepower it will take to play it, and what software players will add support for H.265?

[deadrats]

But deadrats, like I mentioned before, you're not going to be getting a high-quality, efficient, workable "x.265" in FOSS anytime in the next ~3 years!. Sure, divxlabs might have something usable by the fall. But, heck man, it's owned by ROVI (remember: Macrovision!?!). Anything they put out now is not only not going to be a free lunch, it is going to be $$$ and probably phone home, to boot.

The level of computation necessary for even decoding h.265 needs to be at least another level of magnitude. That includes the chips in any devices you intend to use to decode (none of which have been manufactured yet). pdr, jman98, Atak, & jagobo are right. It'll be good in due time, but it's just a mirage right now. These things take time.

a few things to consider:

1) we don't need to wait for a FOSS implementation of h265, divx converter is already legally free and so will be the next version with the h265 profile.

2) main concept's sdk is licensed by ever major ISV out there, including sony (vegas), sorenson, pegasus (tmpg), elecard, ulead, cyberlink, the list goes on.

3) the supposed computational overhead required for decode and encode is greatly exaggerated, it's all based on tests with the unoptimized reference encoder, you can bet that the main concept people are experts in video compression and theirs will be optimized with assembler and simd. furthermore, h265 is designed from the ground up for scaling across multiple cores, it lends itself to very easy highly threaded implementations, like gpu acceleration.

lastly, the supposed lack of hardware support is shortsighted, as i already said divx certified devices that support the h265 profile will be out in less than a year and more importantly why would you not encode with a future proof codec? let's assume you're correct and it will take at least 3 years for h265 to be widely supported by devices, that means in a bit over 3 years h264 will be outdated technology like xvid is now, why would you choose a lame duck codec.

and let's remember this; the OP wants to back up commercial blu-rays to the high possible quality, yes not re-encoding them would be the highest possible quality but if you must re-encode you would want to use a codec that is more efficient than what is used the the blu-ray.

what does anyone have to lose by waiting a couple of months to see what h265 has to offer?

[vhelp]

i agree with everyone else here too. i mean, lets face it.. the level of (exhaustive) support and amount of time that we get with the x264, i don't think there will be one like it again, not another x265 encoder.

[KyleMadrid]

Thanks for all the replies everyone. After reading what everyone has said, I guess I should just go ahead and encode with H.264

[fritzi93]

Surely HEVC is not needed *unless* 4k (or 8K!) takes hold.

If 50% reduction can be achieved, you could just about get a 4k movie-only on a BD50. You'd need a new player, obviously. But will there ever be a physical media for 4k? If Sony is any indication, they seem to be leaning towards a media server and DRMed downloads.

[yoda313]

If Sony is any indication, they seem to be leaning towards a media server and DRMed downloads.

I sure hope not. The video streaming websites "recommend" 10mpbs internet speeds to watch 1080p. What would you need for 4k??? 20-30mbps?

Until true high speed internet is a right and not a privilege there will be some form of physical media. They can still do flash media for it. Granted it would be really pricey for a 60gb drmd flash card to sell movies on but it would beat only being able to stream or download it.

Although I suppose they could have some sort of download kiosk for people to put 4k movies on harddrives and stuff. Obviously drmd to the hilt but there has to be some alternative to streaming or downloading only. We just aren't there yet for streaming only to be viable.

Granted 4k is super edgy right now and only a fraction of 1% could use it but when 4k becomes a "standard" you still have to be able to sell the media to those without a fast enough internet connection.

And I don't know that I'd want to download a 50gb file even if i had the bandwidth or patience to do it.

[posdnya]

[QUOTE=deadrats;2236509]

2) main concept's sdk is licensed by ever major ISV out there, including ... elecard

It's incorrect a bit. In reality Elecard made codecs for Moonlight Cordless (Generation 2 codecs), then the same group of developers work for Mainconcept (G3 codecs) and now Elecard stay on G4, which are now licensed to "major ISV out there".
BTW. Couple TS streams with HEVC+AAC in different resolutions http://www.elecard.com/en/download/videos.html
In-depth HEVC analysis tool http://www.elecard.com/en/products/professional/analysis/hevc-analyzer.html
HEVC Player sample http://www.elecard.com/en/technology/researchlab/hevc-player.html

[Atak_Snajpera]

4K is just a marketing gimmick to "force" clients to buy new TVs. Many people still do not see difference between 1280x720 and 1920x1080. 4K vs 2K will be even less noticeable. To see difference you would need huge TV and very good source material. Also do not expect 4K TV stations to be very popular in Europe or USA in next 10 years. Japan and Korea will have native 4K TV stations a lot faster ofcourse.