This forum uses cookies
This forum makes use of cookies to store your login information if you are registered, and your last visit if you are not. Cookies are small text documents stored on your computer; the cookies set by this forum can only be used on this website and pose no security risk. Cookies on this forum also track the specific topics you have read and when you last read them. Please confirm whether you accept or reject these cookies being set.

A cookie will be stored in your browser regardless of choice to prevent you being asked this question again. You will be able to change your cookie settings at any time using the link in the footer.

[HELP] Question about SSIM and PSNR results
#1
Hi !

I did some SSIM tests using FFMPEG,and something is not clear now.

The movies/clips are encoded using Hybrid/x265.
SSIM results are :

0.97xxxx84  (AQ Mode 3 ,size 35 MB)

0.97xxxx15 (AQ Mode 4 ,size 20 MB)

Both are visually the same.
But why is the big difference in size and only 0.00xxxx69 in score?
Who can explain this?

Btw i know about different quality measurments.

Oh,before i forget ,another question:
What is a good value in PSNR if the movie/clip is a 1080p encode?

Thanks in advance!
Reply
#2
I'm by no means a friend or expert on quality measurement methods like PSNR or SSIM and think that unless they differ significantly they often don't really help.

Quote:But why is the big difference in size and only 0.00xxxx69 in score?
Who can explain this?
You look at it the wrong way, you should ask yourself:

a. What does SSIM measure ? (https://en.wikipedia.org/wiki/Structural_similarity)
The names of it structural similarity index measure (SSIM) basically says it, but you want to read up a bit to understand what that means when looking at images and video.
b. What are the difference between the aq modes and what does aq do?
see: https://x265.readthedocs.io/en/master/cl...on-aq-mode
(I assume you undestand what quantization and adaptive quantization does.)

c. What does it mean that I get this result (sizes shrinks when using aq mode 4, but ssim values stays the same) for my current source?

With that you might come to the conculsion that your source has tons of edges which can be more effectively compressed by adjusting the quantization of the blocks around them without really lowering what SSIM measures.

Whether this makes sense for your source is hard to tell without seeing the source. Smile (no I'm not interessted at looking at the source)

If you want a deep analysis of what happens if you use the differen aq modes on your source, you need to look at tools for codec analysis like:
Zond 265, HARP, Elecard HEVC analyser, or similar you can find through an internet search. Smile

Cu Selur
----
Dev versions are in the 'experimental'-folder of my GoogleDrive, which is linked on the download page.
Reply
#3
Ohh,your answer more than enough!Smile
And no,i dont want to send samples!Big Grin

Thanks!
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)