How to Measure Video Quality / Compression Levels

By Ethan Ace, Published Sep 16, 2016, 07:44am EDT

Two cameras have the same resolution, frame rate and scene monitored but camera A consumes half the bandwidth than camera B. Is Camera A better? You cannot tell until you first check / compare compression levels.

It can be hard to determine which camera's video quality is "best" and whose bandwidth consumption is really lower, as manufacturer compression settings vary and are not comparable. However, you can accurately measure H.264 and even H.265 compression levels with a little skill and some available tools. In this tutorial, we explain how to do so.

A pre-requisite for this post is our video quality/compression tutorial.

H.264 *********** *** ************

*** ****** ** *********** is ************ ** ************ levels ** **** *.*** and *.*** (** **** as *** ***** ****-* codec). ************ ** ******** on * ***** **** 0-51 **** *** ***** the *****, *** **** compression *******, *.*., * camera **** *=** ***** have *** **** *********** than *=** *** ***** require *** **** *********.

***** ****** *** ********* not ******* ** ****** manufacturers, *** ******* *** adjusted ** ******* ******, such ** *-***, *-**, or ****** "***", "*******", and "****". ** *** be ******** ** ** compression, ****** *, ** (quantization *********) ** *** (detail ******* ******) ** various ******* ** *****. For *******, *** ***** below (**** *** ** ****** ************ *********** Comparison ******) ***** *** *********** used ** ****** *************:

Why ************ ******** *** *********

******* *** **** * cameras, **** **********, **** frame **** *** *** consumes ***/* *** *** other ***/* *** **** both **** ******* ** first ******. ** *** lower ********* ****** ******? Maybe *** *** *********.

*************, ************'* ******* ************ levels *** **** ************* with **** ************* ***** very ************, ******* * levels ** *** *** 20s *** ****** ******* them ** *** ***. See *** ** ****** ************ *********** Comparison ****** *** **** ******* and ******* ******.

********* *********** **** ********* differ *** ** **** scenes, ********** ****** **** with ***** ****, ** can ** **** ** notice *** ****** ******* difference. ******* ******** *** Q *****, *** ***** falsely ****** **** *** 2Mb/s ** ****** **** it ***** **** *** a **** ****** * level.

CBR *** *********** ********* *******

**** ******* ***** *********, either ******* **** *** using***** ******* **** **** a ********* *** (*** our*** ** *** ** MBR - ************ ********* ******). ***** ******* **** dynamically ****** *********** ****** and, ** *** ***** is *******, **** **** up *** ************ *****, often ******** ******* ***** quality ************* (*** ******** of******* *** *****). ******* ** ******* for ******** ********* ** the *****, ** *** be **** ** **** by **** *******. *******, it *** ** ******** **************.

Free *******: *********

*** **** **** *** analyzing ************ ***********, ***** ***** *** and **** *****, ********** multiple *******, **** ** bitrate, *********, ***** ****** of ******, *** **** importantly, ************ *********** ********* average *** *** ********** above *** ***** **** average. ** **** ******** graphs ******* ************ *** frame ***** **** ****. This ********** ***** *** various *** ***** ** the ********* *********:

** **** *****, *.*** ***** exported **** * *** may ** ********* ** AVInaptic, *** ** **** found * ****** ** cameras *****, **** ******** to ***, **** *** process. ** ***** *****,*** ***** ******, * ******* **** source *******, *** ** used ** **** *** camera's **** ****** ********, and ****** ** ** disk. **** ********** *** additional ************ ** *********, then ********* **** *** VMS ******, *** ******** files ***** *** **** likely ********** **** *********.

**** ********** ***** ************ the **** *******:

  • **** * ******'* **** stream.
  • **** *** ****** *** recording ** ***.
  • ******* **** ****** **** AVInaptic.
  • ******* ********* ******* ** compression ****** ** ********* differences ** ************ *****.

AVInaptic ***********

****** ** ** **** and **** ** ***, AVInaptic *** **** *** limitations ***** **** ***** to ***** *** **********.

  • ** *.*** *******:********* **** *** ******* H.265 ** ***, **** H.264 *** ****-*. *.*** encoded ***** ****** **** to *******.
  • ***** ***** ******:**** ********* ***** ******* using ***** ******, ********* may ******* ******* *******, including ****** ******* ******* and ************. ** *** be **** ** ***** dynamic * ***** ******** and **** ********** ***** sizes, *** ******* ****** be ********* **********.

Paid *******: ******* *********

*******'* ********* ********** ******* **** **** for ********* *********** *** other *******, **** * much **** ******** *******, including *****-**-***** ********, ********** display, *** ****. ************, Elecard ******** *.***, ** well ** ***** ******.

*******, ********* ** *** free, ****** *********, *** ***** for ********* ** ******* per *******, ****** ** unsuitable *** ****** ** hobby ***. *******, * feature ******* **** ***** version ** ******* ** available, ***** **** ****** users ** *** **** details *** *** ***** frame (** ******** ** overall ********). ******* ***** ** that ******* ** ****** manufacturers *** ********* ********** (though *** *** **** which). 

**** ***** ******* ********* use (**** ****.*** ** ******* ****):

****: **** **** *** originally ******** ** **** but *** ******* ** 2016 ** ***** ***** codecs *** *.***.

Comments (17)

Has VLC fixed their frame drop issues? I used to use it for just about everything but starting with around version 1.1, it became unstable with many file types.

VLC can also display some realtime decoding info while streaming a camera. Go to the Tools menu, select Media Information (Ctrl+I) or Codec Information (Ctrl+J), then click the Statistics tab.


Carl, VLC is up to 2.0.5 now... might be worth another try.

BTW Ethan, your AVInaptic link is broken (appears to point to a non-existent article). The AVInaptic homepage is here: Untitled Document - latest version appears to be from Dec. 2011.

Nice Post. I do some Video IP training and I have learned about Avinaptik and compression scales through the recording mode. I was only using global VLC statitics ...Thank you !

I agree with Karl :For example, VLC 1.05 tested on some Sony HD /H264 were working fine with 200 ms buffer when with the 2.05 version .. just doesn'twork under 500 ms and freezes. Why ? strange

Since H.265 (superior to H.264) standard introduced, I was wondering if there any manufacturer planning to use this compression standard and what experts thinking about it?

Regards

test

Agree
Disagree
Informative
Unhelpful
Funny

Gitl HEVC/H.265 Analyzer

This looks like a free utility which does HEVC analysis. :) 

Agree
Disagree
Informative: 2
Unhelpful
Funny

Truman, Thanks! We'll try it out and report back what we find.

Agree
Disagree
Informative
Unhelpful
Funny

Can I take a moment to say I miss Carl?

Agree: 2
Disagree
Informative
Unhelpful
Funny: 1

Is your aim getting any better?

Agree
Disagree
Informative
Unhelpful
Funny: 1

Carl, that was a compliment.  It was not intended as sarcasm or negative in any way.  I enjoyed your posts.

Agree
Disagree
Informative
Unhelpful
Funny

Sorry Undisclosed. That was my (admittedly poor) attempt at humor. I appreciate the compliment.

Agree
Disagree
Informative
Unhelpful
Funny

Just glad to have you back!

Agree
Disagree
Informative
Unhelpful
Funny

Thank you for the article.

Agree
Disagree
Informative
Unhelpful
Funny

An informative article.

Agree
Disagree
Informative
Unhelpful
Funny

Interesting to know about the compression levels.

Agree
Disagree
Informative
Unhelpful
Funny

Nice to know but I do not think it is something I would be using or when to use.

Agree
Disagree
Informative
Unhelpful
Funny

The applications are deep video inspection tools which I can't really see myself utilizing for my present or future functions however it is still good to be aware of for problem resolution.

Agree
Disagree
Informative
Unhelpful
Funny
Read this IPVM report for free.

This article is part of IPVM's 7,334 reports and 972 tests and is only available to subscribers. To get a one-time preview of our work, enter your work email to access the full article.

Already a subscriber? Login here | Join now
Loading Related Reports