H.264 vs H.265...

[Emanuel]

From '30s Weissmuller's Tarzan on nowadays 1080p B&W copy:

 

8-bit H.264

 

10-bit H.265

 

Similar bitrate. Different jagged edges though ; ) Out of curiosity, any other samples and takers... out there? ;- )

[leslie]

10 hours ago, Emanuel said:

From '30s Weissmuller's Tarzan on nowadays 1080p B&W copy:

 

8-bit H.264

 

10-bit H.265

 

Similar bitrate. Different jagged edges though ; ) Out of curiosity, any other samples and takers... out there? ;- )

i'm not much of a pixel peeper. So with my usual and completely unscientific approach,  an older gopro and a new nine. I did notice a difference of about a hundred megs of difference between .264 and .265 file sizes.  The .265 files always being about a hundred megs smaller than .264.

I had the two gopros mounted on a cheap steadycam and a bell near the handle. I'd start the gopros and then ding the bell for start and end points to help line things up in an editor.

Image wise i'd give the win to the nine. but then you'd expect the 9 to win as its the latest and greatest. Of course my test is only relevant for my gopros, but i would expect .265 file sizes to be smaller than  .264 files across the board regardless of camera. Whether file sizes and storage  is a problem for you i don't know. However you would gain a bit of space using .265. If you like the images coming from .265 

For your test I like the .265 looks a touch nicer (maybe sharper) to my eyes. Can you replicate the test with an another codec ? As i feel that might give another interesting point of reference to compare with. 

[Emanuel]

6 hours ago, leslie said:

i'm not much of a pixel peeper. So with my usual and completely unscientific approach,  an older gopro and a new nine. I did notice a difference of about a hundred megs of difference between .264 and .265 file sizes.  The .265 files always being about a hundred megs smaller than .264.

I had the two gopros mounted on a cheap steadycam and a bell near the handle. I'd start the gopros and then ding the bell for start and end points to help line things up in an editor.

Image wise i'd give the win to the nine. but then you'd expect the 9 to win as its the latest and greatest. Of course my test is only relevant for my gopros, but i would expect .265 file sizes to be smaller than  .264 files across the board regardless of camera. Whether file sizes and storage  is a problem for you i don't know. However you would gain a bit of space using .265. If you like the images coming from .265 

For your test I like the .265 looks a touch nicer (maybe sharper) to my eyes. Can you replicate the test with an another codec ? As i feel that might give another interesting point of reference to compare with. 

Thank you for telling us about your findings : ) Actually, I cannot now anymore once this test was already produced a few time ago when I was trying to understand the advantage as well as its limits on the newer H.265, B&W included from prior recorded footage. I concur with you, I always try to compare the variables following identical parameters despite my unscientific disclaimer at any rate : )

Earlier I had found good results on human faces for H.265, not so much details regarding the background. These gave me a much sharper outcome there too and jaggies free going beyond my own expectations then, comparing with my previous experience with. I believe because I ended with the same/similar bitrate this time.

[Emanuel]

As far as I could learn from my own testing, I guess H.265 actually serves not only for storage savings.

Seems 10-bit has a determinant role to play on IQ as well. Luminance-wise other than colour.

E :- )

[leslie]

Actually i tell a lie. That comparison i did was between a gopro 4 and a gopro 6.  Not the gopro 9 as i stated earlier. I recently acquired the gopro 9, so i think another test is in order, for completeness.