My understanding is that HEVC (aka h265) can provide roughly the same video quality as h264 in roughly half the space, for 1080p+ footage. I have a large collection that I would like to re-encode to reduce my storage needs, and am trying to determine the "best" compression options to use. I understand that "best" is very subjective, but generally speaking, my goal is just to save as much space as I can while (ideally) not reducing visual quality by a noticeable / perceptible amount.

I'll be using Adobe Media Encoder (w /a GPU). My collection is recorded at 1080p, between 35 & 45MBps.

For instance: if HEVC can be expected to yield comparable quality in half the space, does that mean I could compress a 40MBps h264 video to 20Mbps h265 (50% bitrate), and expect visually similar results? And therefore if I wanted to really minimize the risk of noticeable degradation, I could set the encoder to, say, VBR target = 70% of the source bitrate (28MBps), with a max a bit above that?

Are there other things I should consider when setting up the encoder? Other relevant options: Quality (from highest/slowest to lowest/fastest), Use Maximum Render Quality (checkbox), Render at Maximum Depth (checkbox).

Any guidance would be appreciated :)


I've transcoded some of my AVC (h.264) 1080p movies to HEVC (h.265) with 2 pass encoding (using my GPU) and target bitrate set to 50% of original. The other streams (like audio and subtitles) were only copied into the new file.

Mostly I'm using Potplayer to play these files with Madshi renderer, and I can't see any difference on a 55" 4k HDR TV with naked eye. The only thing, that I've noticed at some movies, if there are dark scenes with big black backgrounds, the transition from one shade of black (dark grey) to another is built from much more bigger blocks in HEVC, like in AVC version. I say again, these blocks are visible both on AVC and HEVC but only in some of the movies and in HEVC they are a little bit more noticeable.

But I'm only a home user, I'm not making professional videos and the 15GB video file vs the 30GB worth it to me.


When you are perceptually coding a signal (eg Wave to mp3), the codec removes 80% of the original signal data. This is possible because the codec removes material that humans cannot easily perceive is missing. This is a one way process - you cannot get all of the original data back. If you convert from one codec to another (eg mp3 to AAC) this is called transcoding. The software would have to decode the MP3 back to a Wave and then perform the perceptual encoding again for the rules of the AAC codec. This would not sound very good.

  • I'm not really sure how to interpret that as it pertains to my question - do you mean that regardless of what options I choose, transcoding from h264 to h265 will always yield poor results...? (i.e. "this would not sound very good" -> "it would not look very good")...? – Metal450 Mar 29 '19 at 19:41
  • Yes. I'm just talking about it in principle. Anything that uses a perceptual (or lossy) codec (video or audio) will be severely degraded by converting it to another perceptual codec format. – goodie Apr 11 '19 at 13:07
  • By that logic, wouldn't essentially every edited video (i.e. on YouTube etc) look "severely degraded", since it's always re-encoded after editing? I've read many places (including here on SO) where people discuss recompressing h264 to h265 with minimal or no perceptual loss (in mathematical terms, lots of loss of course - but in terms of what the viewer can actually perceive, virtually none). I'm a bit surprised to hear the perspective that effectively it would be impossible to use h265 without a "SEVERELY degraded" result... – Metal450 Apr 12 '19 at 21:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.