android
  #1  
Old 03-16-2011, 03:37 AM
Ravnefar Ravnefar is offline
Junior Member
 
Join Date: Feb 2011
Location: Eureka, California
Posts: 15
Default FLAC vs. MP3 bitrate 320

It would appear that there would be a significant gain in sound quality by using the FLAC format over the highest mp3 settings considering how many people on this forum use a lossless format. Especially since FLAC files take up twice as much space as mp3s at bitrate 320. So I made the following “experiment”:
Same digitally recorded track (pipe organ) ripped at mp3 bitrate 230, 320 and FLAC, which amounts to 13.8 MB, 19.1 MB, and 41.0 MB respectively.
I could hear some gain between the 230 and 320 tracks, but almost none between 320 and FLAC (“almost” might even be an illusion, because this was not a blind test after all).
My collection of classical music is fast approaching 60 GB. If I chose carefully, there might be room for an upgrade to bitrate 320 (most files are 230 or even lower), but updating to FLAC would be a different story entirely. Of course these files are for my ears alone, and I am well into the second half of my life. Still, I am intrigued: can other people really hear a “significant” difference? Should I perhaps have tested a wider range of sound tracks, because some recordings are more sensitive to the compressing process than others? Will be grateful for your input! (Cowon J3 and Monster Turbine Copper edition).
Ravnefar
Reply With Quote

Advertisement [Remove Advertisement]

  #2  
Old 03-16-2011, 05:21 AM
el maco's Avatar
el maco el maco is offline
Junior Member
 
Join Date: Nov 2010
Location: Finland
Posts: 57
Default

Quote:
Originally Posted by Ravnefar View Post
I could hear some gain between the 230 and 320 tracks, but almost none between 320 and FLAC (“almost” might even be an illusion, because this was not a blind test after all).
I made a casual blind listening test in which I failed to hear any significant difference between lossy Vorbis and Flac (using my portable gear with decent Shure SE530 IEMs). Before that I could easily "hear" differences i.e. the lossless track always sounded more detailed, lifelike, you name it. However, after that experience the difference was gone. (And no, there is nothing wrong with my hearing - quite the opposite.)

I guess many people grossly underestimate the power of self suggestion and overestimate the ability of the human ear. Our brain interprets the hearing sensation to match our expectations. If someone says there is a "huge" difference between Flac and Mp3 @ 320 kbit I think to myself: that guy hasn't made his homework properly (after all it never stops to amaze me how many people claim to hear similarly huge differences between hi-fi interconnect cables etc.) Surely I can accept that someone who knows what anomaly to listen to might be able to statistically differentiate between lossless and lossy audio in some extreme cases. Does that mean that we should waste three times the storage space just to be safe? I don't think so.
Reply With Quote

  #3  
Old 03-16-2011, 06:16 AM
esanthosh's Avatar
esanthosh esanthosh is offline
Member
 
Join Date: Oct 2009
Location: India
Posts: 207
Default

I usually rip my CDs to FLAC, but I mostly restrict FLAC to the PC. With portable players, it makes more sense to go with 320kbps files as you can fill up more music for listening on the go.

I have two Clip+s. I used the same song in both, the only difference being one had the 320kbps and the other had FLAC. I could not find any difference between them (same player, both rockboxed, both without EQ, same volume level, same IEM). If I am supposed to hear any difference, I have lost my hearing; so don't trust anything I say .

Last edited by esanthosh; 03-16-2011 at 12:07 PM.
Reply With Quote

  #4  
Old 03-16-2011, 07:26 AM
Satellite_6's Avatar
Satellite_6 Satellite_6 is offline
Senior Member
 
Join Date: Feb 2009
Location: VA, USA
Posts: 871
Default

Use high bit rate mp3, the difference between FLAC and mp3 on a DAP will be minimal at best and certainly not worth the memory. Hopefully you are using the LAME encoder (it sounds like you are). I don't think there is anything wrong with using 320 but others will tell you to use -v0 (which is ~230) because it is supposed to be just as good.
__________________
satelliteaudioblog.blogspot.com
Reply With Quote

  #5  
Old 03-16-2011, 11:05 AM
Syndrome's Avatar
Syndrome Syndrome is offline
Redneck Moderator
 
Join Date: Dec 2007
Posts: 1,278
Default

As its said here, on your portable mp3 is all you need. I've tried a few times to hear the difference coming from my old D2 and I couldn't. Coming from my PC sound card seemed to make a little more difference, and even then the FLAC had a slightly better sound stage, and that was about it.
__________________

Ruger M77 , Howa 1500 , Czech Republic 452-2E
Ruger .22 Single Six, Springfield Armory XD-9 , Springfield Armory XD45ACP
2007 Elite Archery Synergy

Gamo CFX, Gamo PT80
Large Rock


Need your headphones fixed or something soldered? Syndrome's Soldering Service
Reply With Quote

  #6  
Old 03-16-2011, 12:44 PM
Ravnefar Ravnefar is offline
Junior Member
 
Join Date: Feb 2011
Location: Eureka, California
Posts: 15
Default 230 vs. 320 (?)

It seems that 230 could sound just as good as 320 - hm! I am a half decent musician myself, and my ears used to above average. Maybe I am not even hearing such a great difference between the two as I think I am. It would be nice to not update and have 25 % more space. On the other hand I don't want to fool myself in the opposite direction. I want the best!

Thank you for your thoughts so far, everybody.

Ravnefar
Reply With Quote

  #7  
Old 03-16-2011, 01:24 PM
Satellite_6's Avatar
Satellite_6 Satellite_6 is offline
Senior Member
 
Join Date: Feb 2009
Location: VA, USA
Posts: 871
Default

I just though of something:

If you listen to a lot of classical music Vorbis may be a better choice than mp3.

Here is an old listening test: http://www.hydrogenaudio.org/forums/index.php?showtopic=36465 and I've seen others indicating the same thing. . .
__________________
satelliteaudioblog.blogspot.com
Reply With Quote

  #8  
Old 03-16-2011, 01:55 PM
berklon berklon is offline
Junior Member
 
Join Date: Oct 2010
Posts: 56
Default

I peronsally don't see the point of FLAC on a portable player - at least with the current limitations of space. 320 bitrate mp3 is a great balance between size and sound quality. To me the advantage of having a lot more music with me on the go supercedes the marginal difference in sound quality that FLAC provides on a portable system (even less of a difference when you factor outside noises interfering with my listening).
Reply With Quote

  #9  
Old 03-16-2011, 02:01 PM
sYlt's Avatar
sYlt sYlt is offline
Member
 
Join Date: Feb 2010
Location: Lisbon, Portugal
Posts: 291
Default

Quote:
Originally Posted by Satellite_6 View Post
I just though of something:

If you listen to a lot of classical music Vorbis may be a better choice than mp3.

Here is an old listening test: http://www.hydrogenaudio.org/forums/index.php?showtopic=36465 and I've seen others indicating the same thing. . .
Being that test almost 6 years old doesn't it mean things could be different now? Since I believe all encoders have been updated several times since. Either way that's an excellent test.

EDIT: Just did a quick test, a non blind one:
I picked one song - Plug in Baby by Muse, FLAC, then converted it to mp3 LAME on several kbps: 320, v0, 256, 192, 160, 128 (most common I believe). I used my IEM's, the Panas HJE900 and the result of my test was that in comparison with FLAC I could only detect small differences on 160 and bigger discrepancies on 128kbps. The song chosen wasn't probably the best one but I think it did the trick.

Personally I still like to have everything in my music library at least v0/256kbps and up, even if I can't really tell different from 192kbps for example. But I have a question, could be that listening at higher volumes it would be easier, if possible, to notice more differences between bit rates? That is, if speakers are used, no way I'm going louder than I usually do with IEM's.

PS: v0 on Plug in Baby was 250kbps.

Last edited by sYlt; 03-16-2011 at 02:33 PM.
Reply With Quote

  #10  
Old 03-16-2011, 02:53 PM
skip252 skip252 is offline
Administrator
 
Join Date: Dec 2007
Location: Chicago
Posts: 5,351
Default

I've said it before
Quote:
Originally Posted by skip252 View Post
Instead of guessing or taking the advice of someone that doesn't hear the same as me I'd ABX with foobar2000. That way you can determine exactly you can hear or not. If you can't reliably identify a FLAC file from a lower bitrate .mp3 why eat up disk space with a bunch of bits you can't hear?

ABXing will allow you determine the bitrate that is transparent to you. Once you determine that you can then target the bitrate that will work best for you. The difference in file size will then be determined by what bitrate you decide to use.
and I think it's still valid. Doing sighted listening test doesn't give the same results as double blind testing. As much as you may try not to, if you know which type of file is being tested your brain will use that information to determine which is better. ABX testing takes that information away and let's you determine at what level you can truly hear a difference from lossless.

The bitrate at which anyone else can hear a difference is really meaningless to you. People say they hear a difference between lossless and lossy files all the time. Others will chime in with what they use or what they believe is a good bitrate for you. Considering they don't have your ears or equipment I'd consider that well meant but useless. Once you establish what your transparency level is you can use that to encode to the level that's best for your personal use.

I'd take a good look at the LAME page at Hydrogenaudio. There's a lot more information there including a chart that gives guidelines as to which settings are recommended for what environment. I regard those as suggestions not rules. Very good suggestions but not locked in stone. Your ears and equipment may find that what's suggested is either overkill or not adequate for your use.

You ears should be the final judge as to what settings to use. ABX testing is considered the best way to allow them to make an informed decision. I'd give it a try. You may be surprised how difficult it is to reliably identify a difference between a relatively low -V quality setting and lossless.
Reply With Quote

  #11  
Old 03-16-2011, 03:10 PM
Satellite_6's Avatar
Satellite_6 Satellite_6 is offline
Senior Member
 
Join Date: Feb 2009
Location: VA, USA
Posts: 871
Default

Quote:
Originally Posted by sYlt View Post
Being that test almost 6 years old doesn't it mean things could be different now? Since I believe all encoders have been updated several times since. Either way that's an excellent test.

EDIT: Just did a quick test, a non blind one:
I picked one song - Plug in Baby by Muse, FLAC, then converted it to mp3 LAME on several kbps: 320, v0, 256, 192, 160, 128 (most common I believe). I used my IEM's, the Panas HJE900 and the result of my test was that in comparison with FLAC I could only detect small differences on 160 and bigger discrepancies on 128kbps. The song chosen wasn't probably the best one but I think it did the trick.

Personally I still like to have everything in my music library at least v0/256kbps and up, even if I can't really tell different from 192kbps for example. But I have a question, could be that listening at higher volumes it would be easier, if possible, to notice more differences between bit rates? That is, if speakers are used, no way I'm going louder than I usually do with IEM's.

PS: v0 on Plug in Baby was 250kbps.
It is old, but buggering the listening tests I still think that violins and stuff still sound better on Vorbis. It doesn't matter much for me, but I still think it warrants a recommendation for people who listen to classical.

I think it is easier to tell the difference at louder volumes. . . and with headphones that have a more sensible quantity of mids. . .
__________________
satelliteaudioblog.blogspot.com
Reply With Quote

  #12  
Old 03-16-2011, 03:45 PM
WalkGood's Avatar
WalkGood WalkGood is offline
Moderator
 
Join Date: Aug 2006
Location: FL
Posts: 12,579
Default

Just make a few rips and try some ABX'in on foobar2000, works great and it could be different for everyone. Personally I don't use anything larger than lame 192vbr rips on the go and it's all fine for me.
__________________
WalkGood,
Ramón

abi >> | Forum Rules | Glossary | Why Rockbox | FLAC or MP3? | irc
Reply With Quote

  #13  
Old 03-16-2011, 03:49 PM
sYlt's Avatar
sYlt sYlt is offline
Member
 
Join Date: Feb 2010
Location: Lisbon, Portugal
Posts: 291
Default

Just made mine:

20:28:14 : Test started.
20:30:06 : 01/01 50.0%
20:31:59 : 02/02 25.0%
20:36:33 : 03/03 12.5%
20:41:50 : 04/04 6.3%
20:46:17 : 04/05 18.8%
20:46:34 : Test finished.

----------
Total: 4/5 (18.8%)

Gotta give some rest to my ears now, it took some concentration to discern between both, FLAC vs 192kbps. So I guess it's safe to assume that for me anything above 192kbps is just fine, at least with this IEM's at this volume.
Reply With Quote

  #14  
Old 03-16-2011, 09:34 PM
Daanish Daanish is offline
Junior Member
 
Join Date: Oct 2010
Posts: 45
Default

Just did a test with VBR 320 Ogg and Flac, results 6/10 chance of guessing 37.7%.

There is a difference but I had to listen hard to make, it certainly wasn't glaringly apparent. But in general one had guitars where the crunch of distortion sounded more lifelike while the other was flatter.
__________________
Player: Cowon J3
IEM's Headphones: Panasonic HJE900, RE-262, Fischer Audio DBA-02, Fischer Audio -003
Amp: Ibasso T3
Past: Fischer Audio Silver Bullet, Ortofon e-q5
Demoed: Monster Miles Davis,
Reply With Quote

  #15  
Old 03-16-2011, 09:43 PM
Satellite_6's Avatar
Satellite_6 Satellite_6 is offline
Senior Member
 
Join Date: Feb 2009
Location: VA, USA
Posts: 871
Default

Well neither of those tests actually mean anything. . .
__________________
satelliteaudioblog.blogspot.com
Reply With Quote

  #16  
Old 03-16-2011, 11:31 PM
Daanish Daanish is offline
Junior Member
 
Join Date: Oct 2010
Posts: 45
Default

Quote:
Originally Posted by Satellite_6 View Post
Well neither of those tests actually mean anything. . .
How many would we have to do for it to mean something then
__________________
Player: Cowon J3
IEM's Headphones: Panasonic HJE900, RE-262, Fischer Audio DBA-02, Fischer Audio -003
Amp: Ibasso T3
Past: Fischer Audio Silver Bullet, Ortofon e-q5
Demoed: Monster Miles Davis,
Reply With Quote

  #17  
Old 03-17-2011, 07:33 AM
WalkGood's Avatar
WalkGood WalkGood is offline
Moderator
 
Join Date: Aug 2006
Location: FL
Posts: 12,579
Default

It’s going to be different for each individual so it’s only meaningful to you …
__________________
WalkGood,
Ramón

abi >> | Forum Rules | Glossary | Why Rockbox | FLAC or MP3? | irc
Reply With Quote

  #18  
Old 03-17-2011, 10:01 AM
sYlt's Avatar
sYlt sYlt is offline
Member
 
Join Date: Feb 2010
Location: Lisbon, Portugal
Posts: 291
Default

Quote:
Originally Posted by Satellite_6 View Post
Well neither of those tests actually mean anything. . .
Care to explain? =O
Reply With Quote

  #19  
Old 03-17-2011, 10:10 AM
Satellite_6's Avatar
Satellite_6 Satellite_6 is offline
Senior Member
 
Join Date: Feb 2009
Location: VA, USA
Posts: 871
Default

You need at least 10 trails with a low chance of guessing.
__________________
satelliteaudioblog.blogspot.com
Reply With Quote

  #20  
Old 03-17-2011, 10:27 AM
Thib Thib is offline
Junior Member
 
Join Date: Jul 2010
Posts: 60
Default

I'd would probably loose blind test, I know I can discern between VBR V2 and Flac or 192 and Flac, because at some point I did massive replacements and compared both of them, by the way it's not because the song isn't "proper" like electronic music with overwhelming highs etc that it won't benefit from sound quality upgrade, on the contrary I noticed a lot of difference with Trance music.
Anyways, Flac is good for me because :
- I don't have to wonder whether it's good enough...
- Computer storage is very cheap (2To = 80€), it may be a bit of a problem with DAPs though even if Memory Card are getting cheaper as well.
- Why the hell would I legally download a downgraded track from my favorite artist ??!! It's no more the original track to me...
- For the sake of it, yes, Flac is one of the greatest invention ever, if I'd see the inventor of Flac I'd go completely mad... It has its own tags, encodes fast and looses a lot of size especially for classical music, Ravnefar

The bad side is that I've gone extreme with that because I don't want to hear about mp3 anymore and of course I'm surrounded by "128 kbps ibuds robots" or "128 kbps Beats by dre robots" so I can't exchange Flac with anyone ...

People have their own opinion (with extreme stubborn cases sometimes by the way...) I don't blame them for that, but I blame people for not knowing that lossless exists and that share their music only in lossy quality. To me it's like a producer releasing a movie in divx... To get back to the topic, be sure you have the original files on your computer, we don't ever know what could happen !

EDIT : Rafnevar, I believe you could probably gain in sound quality getting rid of those monster turbine copper edition, I don't know how good they are but you should absolutely consider high-end IEMs, I was sceptical at some point too but the headphones quality make a lot more difference than the bitrate of your files.
__________________
Portable Rig :
White Cowon J3 32Go -> Sennheiser IE8 Custom Sleeves basses at the middle / Shure SE420 Custom Sleeves / M-AUDIO Transit USB + JBL Duet 2.0
Home Rig :ASUS Xonar STX -> IE8 / Harman Kardon Soundsticks 2.

Last edited by Thib; 03-17-2011 at 10:35 AM.
Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump



All times are GMT -5. The time now is 02:06 AM.