Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: FLAC settings vs. decoder speed (Read 12006 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

FLAC settings vs. decoder speed

After a brief discussion with Polar on this thread, I decided to conduct a FLAC setting vs. decoding speed test. After about three weeks, I finally finished it.

The test and results page is here.

Boilerplate summary:
"0"-"8" correspond to "flac -V -0" through "flac -V -8". "ss" is "flac -V --super-secret-totally-impractical-compression-level".

[span style='font-size:8pt;line-height:100%']x axis is compression ratio, y axis is decoding speed. Note that neither axis goes all the way to 0.[/span]
  • "0" decodes faster than any other setting tested
  • "1", "2", and "3" decode faster than "4"-"8" and "ss"
  • "4"-"8" are faster than "ss"
  • "ss" is REALLY slow
You can see more detailed results, as well as the data (in Mathematica notebook form) on my page.

[edit: Changed thread link]
"We demand rigidly defined areas of doubt and uncertainty!" - Vroomfondel, H2G2

FLAC settings vs. decoder speed

Reply #1 –
What does ss do over --best? More work on the huffman coding?

FLAC settings vs. decoder speed

Reply #2 –
ss is equivalent to "--lax -P 4096 -b 4608 -m -l 32 -e -E -p -q 0 -r 0,16"(*), and --best (or -8) is the same as "-l 12 -b 4608 -m -e -r 6". So ss uses just about everything that's marked as "slow" or "expensive". I would recommend against using it. It takes me ~5 hours to encode one album with it, and it's not much better than --best.

(*) Although the settings in ss say it uses "-E", I have no idea what it does, and it seems not to be a valid switch. All the others you can find out about using "flac --explain".
"We demand rigidly defined areas of doubt and uncertainty!" - Vroomfondel, H2G2

FLAC settings vs. decoder speed

Reply #3 –
only the first graph image resolves for me.  here's a snippet of the html I get back:

Code: [Select]
<h3>Results</h3>
<p>The graphs are filesize ratio (x) and decoding speed (y).</p>
<p>Overall:<br />
  <img src="./main.png" width="576" height="355" /></p>
<p>Full:<br /><img src="file:///D|/Documents/AUDIO/FLAC%20test%202%20-%2020040814/site/mainnormal.png" width="576" height="355" /></p>
<p>With these graphs, one can say with 95% certainty, that:</p>


Josh

FLAC settings vs. decoder speed

Reply #4 –
Tee hee.  I guess that's what I get for making the HTML really fast in Dreamweaver. Try it now.
"We demand rigidly defined areas of doubt and uncertainty!" - Vroomfondel, H2G2

FLAC settings vs. decoder speed

Reply #5 –
Quote
After a brief discussion with Polar on this thread, I decided to conduct a FLAC setting vs. decoding speed test. After about three weeks, I finally finished it.

The test and results page is here.[{POST_SNAPBACK}][/a]
Although I realize I'm grossly late, I feel like I still owe you a word of gratitude for your terrific effort on this, Omion. In my view, your tests are just as authoritative and comprehensive as HansHeijden's and Speek's, even if yours only cover FLAC. I guess I'll have to change my opinion on the influence of compression level on decoding speed from 'nonexistent' to 'hardly any' . So once again: thanks!

One petty remark though: I'd refer to [a href="http://www.hydrogenaudio.org/forums/index.php?showtopic=24921&view=findpost&p=233988]this post of yours[/url] as the starting point of our discussion, in stead of to the whole thread like you do now.

FLAC settings vs. decoder speed

Reply #6 –
Quote
Although I realize I'm grossly late, I feel like I still owe you a word of gratitude for your terrific effort on this, Omion. In my view, your tests are just as authoritative and comprehensive as HansHeijden's and Speek's, even if yours only cover FLAC. I guess I'll have to change my opinion on the influence of compression level on decoding speed from 'nonexistent' to 'hardly any' . So once again: thanks!

One petty remark though: I'd refer to this post of yours as the starting point of our discussion, in stead of to the whole thread like you do now.
[a href="index.php?act=findpost&pid=242100"][{POST_SNAPBACK}][/a]

I was wondering where you were  Better late than never, I suppose!

Thank you for your feedback, and for egging me on to do the test. I guess I should have put you in the "special thanks" section, too! I'll change that when I get around to it.
"We demand rigidly defined areas of doubt and uncertainty!" - Vroomfondel, H2G2