Just a simple Divx HEVC / x264 Comparison

[HerpaDerp]

http://screenshotcomparison.com/comparison/58721

Non Hovered = Divx HEVC
Hovered = x264

HEVC: Converted using Divx - High Quality Preset, 10,000 average bitrate.
x264: Converted using XMedia Recode - Slower Preset, Blu-ray compatible checked, 10,000 bitrate

I realize it's way too soon to make any conclusions. x264 has had many years to mature.

However, if I had to make a statement here, it would be that the Divx HEVC encoder is (so far) disappointing. One would assume that HEVC would be able to at least compete with x264, but the results here say otherwise.

When a community supported HEVC encoder comes around (x265???) that will probably drastically improve things.

Edit: Screenshots taken using VLC.

[detmek]

It is not same frame and luminance is off. So, comparison can not be done.

[x265_Project]

When a community supported HEVC encoder comes around (x265???) that will probably drastically improve things.



Yes, it will.

[HerpaDerp]

It is not same frame and luminance is off. So, comparison can not be done.

I know it's not the same frame and I know the luminance is off. You can blame the Divx converter for this. Even with a luminance discrepancy, the difference in visual quality is quite extreme.

Yes, a comparison *can* be done. Just use your eyes.

[RedDwarf1]

That is like comparing an apple to an orange, a total waste of time IMO. If you cannot ensure the same frame then why bother doing this?

A small video would of given a better comparison rather than using an isolated frame.

[hello_hello]

I know it's not the same frame and I know the luminance is off. You can blame the Divx converter for this. Even with a luminance discrepancy, the difference in visual quality is quite extreme.

Yes, a comparison *can* be done. Just use your eyes.

The Divx HEVC version doesn't appear to have as much detail, but that's the encoding and not the fact the screenshots show different frames?
It'd be better to fix the luminance levels and show the same frame, or better still a small section of each video. I'd be interested in a quality based comparison. Does the Divx encoder have a quality based encoding mode?

I tried a few little x264 v x265 encodes the other day. With the default settings for each (CRF23), x264 was the clear winner, but it also used about 4x the bitrate, so I guess you can't directly compare CRF values. I thought x265 needed about CRF16 to match the quality of x264 at CRF23. By then the bitrates were pretty similar. Nobody seemed to want to tell me the bitrate of the x265 version, but the file sizes indicated it was around 5% less. (1400kbps for x264).
The difference in encoding speed was greater than I thought it'd be. At least for my fairly static test video and an old dual core. 12fps for x264 v 1fps for x265.

Keep in mind I wasn't out to prove anything and I was only having a rough look.