Announcement

Collapse
No announcement yet.

Edius to X264 & X265 via TMPEGenc Video Mastering Works 6

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • GrassValley_SL
    replied
    This thread is moved from the main EDIUS area. While interesting it has very little to do with EDIUS. :)
    Last edited by GrassValley_SL; 10-26-2016, 11:26 AM.

    Leave a comment:


  • Liverpool TV
    replied
    Originally posted by GrassValley_PS View Post
    However if h.265 give equal quality at a lot smaller file size, it will be very valuable for transmitting files back from the field.
    That's right Pat, but these tests do not show much of an advantage, if any.

    There's a couple of things I can think of that may skew these results. Maybe TMPEGenc's implementation of X265 has some issues? Maybe X265 is designed more for UHD and above? Although it should really show the same bitrate/filesize advantages regardless of resolution.

    I'm doing another test at UHD but with Handbrake. I'll do another post shortly if there's obvious differences.

    Leave a comment:


  • GrassValley_PS
    replied
    However if h.265 give equal quality at a lot smaller file size, it will be very valuable for transmitting files back from the field.

    Leave a comment:


  • Bassman
    replied
    Imho, if h.265 is not A LOT better it just needs to stay away. The last thing we need is a format war/choice between h.264 & h.265 and all of the compatibility issues that might ensue. Not being able to reliably deliver your product makes one want to quit this business.

    Leave a comment:


  • David Clarke
    replied
    I think More4K is probably right and the problem is the encoder in TMPEG, not that H.265 is over-hyped. The information I have seen says that the H.265 encoding in Handbrake and TMPEG is not as good as it could be. I don't know about the quality of Adobe's encoder. I did some tests a while back but gave up because I could not see much difference but the encoding was 4-5 times longer.

    Leave a comment:


  • Bassman
    replied
    Yes, thanks Dave. Nothing beats primary information!

    Leave a comment:


  • Jerry
    replied
    Thanks for doing the tests Dave.

    Leave a comment:


  • Liverpool TV
    replied
    I was also very surprised with the apparent lack of obvious difference.

    Although I get the point of using movement, as Jerry has pointed out there is a lot of water movement in the dock. Plus a lot of very fine detail in most of the backgrounds.

    I had already done some progressively decreasing bitrates and these didn't show too much difference between the two codecs. The bitrate of 2Mbit is very low for a 1080 picture and is at the point where you start seeing encode artefacts, especially in the water. This is where I would have expect to see the obvious differences, but they both appear to be very similar with h.264 holding up surprisingly well against h.265 with the water, as far as the amount of artefacts.

    I used the very slow setting and 2 pass VBR. I can't remeber the actual encode times as I batched them and went for a drink. But on other tests, X265 takes a lot longer than X264 when matching their encode parameters as close as they can be.

    From what I've tested over the last couple of years, I doubt I'll move from h.264 as I don't see the obvious advantage to h.265, especially taking on board the much slower encode times and problematic playback compatibility. I've also tested UHD and seen very similar results. Maybe the advantage comes with high motion content? Although if that were the case, it would make sense that h.265 should have been obviously better with locked off shots at the very low bitrates.

    I've seen blurb on the web that suggests h.265 is twice as efficient at the same bitrate, I personally question this. As it stands, I can't see any obvious gain in efficiency in these examples, and at certain points on certain clips, while I do appreciate the obvious encode errors at these very low bitrates, it even appears that h.264 makes a slightly better attempt at the water.

    Leave a comment:


  • Bassman
    replied
    Yes, I agree, the h.265 is not looking too good from a forward progress point of view. The test would debunk the hype or maybe the TMPG implementation needs some attention?

    Leave a comment:


  • Jerry
    replied
    Tim, look at the water in the lower left side. Both display encode problems. Taking it lower would make it even worse.
    Plus, there is a time factor that we haven't mentioned. Encoding to H.265 will take longer.
    Dave didn't mention the time difference between h.264 and h.265 on the 1080 versions.

    Leave a comment:


  • Bassman
    replied
    Thanks Dave. I too do not see much difference. I agree with Jerry as what I have heard is that h.265 is supposed to yield way smaller file sizes while retaining image quality.

    Maybe a good test would be to create a 2Mbps h.264 file and make a series of h.265 encodes with decreasing bitrates from 2Mbps. Then compare to see how far you can get away from the h.264 file while retaining the same image quality. Some motion in the video might be a good test as well.

    Leave a comment:


  • Drbgaijin
    replied
    Thank you for taking time and effort to make the comparisons Dave.
    At a quick glance I am not seeing any noticable difference either.
    However I will take a closer look later in the week.
    Regards
    Douglas

    Leave a comment:


  • Jerry
    replied
    I may have misinterpreted what h.265 was supposed to accomplish. I thought is was supposed to give the ability to give the same quality as h.264 but at a lower bitrate and file size.

    The H.265 1080 file size was more than the h.264. I can't see any difference between the quality of the two 1080 files on a 30" 2560x1080 monitor. Both showed the same good and bad points.
    I played them in both the Windows 10 Movies and tv app and VLC.

    Leave a comment:


  • Liverpool TV
    replied
    Yes, it's an interesting comparison. Although they're matched as close as I can get them for bitrate etc. There isn't the huge difference that I would have expected, considering all the so called advantages of h.265, regardless of the very low bitrates. Maybe it shows how good h.264/X.264 really is?????

    Leave a comment:


  • T-Bone
    replied
    1 Mbit 264 better than 265 ?

    I don't have a very good eye for this, but on the 1 Mbit, the 264 looks better than the 265. Example: at 00:12, left center, the black edge at the water - more shaking of the black on the 265.

    Leave a comment:

Working...
X