FLAC vs. 320 Mp3 - Page 19  

Quote:

Originally Posted by ForShure 


I explained psychoacoustics in my last post. I understand how it works. Everybody on head-fi is an anonymous source in some ways. Just trying to make my argument. I don't see you adding any new stuff to the debate, instead you choose to criticize the arguments that don't agree with your view without having support to back it up. But continue to enjoy your flat sounding mp3's believing that they sound as good as lossless.

I assume you haven't actually read this thread. Or any other thread regarding this subject in the sound science forum.

Every time somone claims easily discernible differences between 320kbps and lossless, they either refuse to do any proper ABX testing, or fail a proper ABX test. So, would you like to give it a shot?

@ForShure

Here - will make it easy for you.  Read this thread (http://www.head-fi.org/t/655879/setting-up-an-abx-test-simple-guide-to-ripping-tagging-transcoding) and it'll give you an idea of how to easily set up an abx at home.  From your comments, I'm willing to bet you've never done one.

As we're actually in the sound science section - it's reasonable to ask you to back up your claims.

Set up the test - using the guide.  You choose the music.  If you can, make the two tracks available here so we can try it as well.  Volume match and use the abx tool.  Use normal listening levels  Run for 15 iterations.  Post the log.  Be completely honest [the system can still be cheated - but there is no real point - it's not about ego, but instead about discovery].

The latest advances in LAME encoding for mp3 (and aac) are designed to essentially be transparent.  We're not supposed to be able to tell a difference (human hearing limitations).  Rather than fighting it (ego - my ears must be better etc), find out the truth and celebrate it.  Simply means that for our portable set-up you can carry more music.

And if you don't have a decent converter, I can convert for you, or just provide the files. Keep in mind that the highest quality encoding is not the default setting of the LAME encoder.

Quote:

Originally Posted by ForShure 

I explained psychoacoustics in my last post. I understand how it works. Everybody on head-fi is an anonymous source in some ways. Just trying to make my argument. I don't see you adding any new stuff to the debate, instead you choose to criticize the arguments that don't agree with your view without having support to back it up. But continue to enjoy your flat sounding mp3's believing that they sound as good as lossless.

If you understand the matter at hand, then explain your reasoning behind writing this: "If there was no difference then lossless files wouldn't be around because the need for them wouldn't exist."

With authority I was not referring to you, but the mentioned professor... look up the term.

I don't think I'm obligated to add (for you) new stuff to the discussion, but yeah I am criticizing generalizations/claims/... that are wrong. It has nothing to do with my views. There is evidence for transparency of MP3 with lower bitrate than 320 kbps, try Google.

I do enjoy "flat sounding" (wrong again) mp3/aac/vorbis/... files, do you enjoy ignorance? They say it's bliss after all.

Quote:

Originally Posted by chewy4 

Every time somone claims easily discernible differences between 320kbps and lossless, they either refuse to do any proper ABX testing, or fail a proper ABX test. So, would you like to give it a shot?

Hah, certainly not easy. Just last night I did your 128 kbps, 320 kbps, lossless test (didn't notice it till it recently got bumped) and successfully identified groups 1 and 3. I'll admit I ended up spending about an hour though... didn't realize it until I looked at the clock. The 128 kbps were relatively easy, the lossless vs 320 kbps sounded virtually identical.

Recently I have been doing my own ABX testing between LAME V2, V0, and lossless. My conclusion is that both V2 and V0 are virtually transparent; some music was transparent at V2, but for example Pearl Jam - Black, I could still identify an extremely minute difference with V0.

So yeah, now I know I'll be happy with V0 as it's virtually transparent (to me anyway) and saving the extra space over lossless is worth it (especially on a portable player).

Quote:

Originally Posted by chewy4 

 

The words lossy and lossless were intended to refer to types of compression algorithms, not the end result vs the real life performance before the recording.

 

And that 50% of space isn't wasted on a CD because it could have been compressed. It allows for universal compatibility and mechanical timing.

Sure. I'm just saying every recording process involves lossy 'compression' of one sort or another. It helps put a perspective to remember that. Not sure why you think CD is a universal format... or why that is of any importance now a days. There are probably far more computers than CD players in the world today! Nor what you think the significance of 'mechanical timing' is. You can easily get the same timing precision from all formats, especially with buffered readout, which is pretty standard.

Quote:

Originally Posted by xsal 

Sure. I'm just saying every recording process involves lossy 'compression' of one sort or another. It helps put a perspective to remember that. Not sure why you think CD is a universal format... or why that is of any importance now a days. There are probably far more computers than CD players in the world today! Nor what you think the significance of 'mechanical timing' is. You can easily get the same timing precision from all formats, especially with buffered readout, which is pretty standard.

I get what you're saying, but I just don't think lossy or compression should be used to describe flaws in the recording process. It just makes things confusing.

PCM data, the uncompressed data on a CD, is pretty much universal. FLAC, ALAC, or pretty much any other compressed format not so much. With raw PCM data you don't need some software in the middle, whereas with a compressed file you need something to convert it back to PCM.

Yeah, you're right that mechanical timing isn't relevant in terms of any performance increase/decrease. Probably not really relevant to anything nowadays, but it certainly made sense at the time the format was released.

Quote:

Originally Posted by xnor 

Get foobar2000, choose full on installation, add your 24/96 file, right click - convert - ...

 

Choose output: Add new

encoder: lame.exe (download it from rarewares.org)

extension: mp3

parameters: -S --noreplaygain -b 320 - %d

 

format is: lossy

highest bps supported: 32

OK

 

Choose processing: Add Resampler (SoX)

(download it form here and install it using File - Preference - Components - Install ...)

Configure the resampler for 44.1 kHz

 

Convert the file.

 

Now select the 24/96 file and mp3 and right click - ReplayGain - Scan per track gain.

Good to know! Thanks!

Quote:

Originally Posted by xnor 

Get foobar2000, choose full on installation, add your 24/96 file, right click - convert - ...

 

Choose output: Add new

encoder: lame.exe (download it from rarewares.org)

extension: mp3

parameters: -S --noreplaygain -b 320 - %d

 

format is: lossy

highest bps supported: 32

OK

 

Choose processing: Add Resampler (SoX)

(download it form here and install it using File - Preference - Components - Install ...)

Configure the resampler for 44.1 kHz

 

Convert the file.

 

Now select the 24/96 file and mp3 and right click - ReplayGain - Scan per track gain.

...

Technical data is all well and good but, as has been mentioned a fair few times, it is your ears that you should trust at the end of the day. If you can't hear the difference between formats, more power to you, you can save a whole lot of disc space and more "fully" utilise your portaplayers.

I am not what most people would call a "bona fide" audiophile. I lack the funds for high-end equipment, I think that my 175 Euro Sennheisser HD 25 II's are great (they are the most expensive earphones I have owned to date), am reasonably happy with my Skullcandy Aviators (though they are somewhat lacking in bass impact), and often enjoy the sound of my $99 Marshall's (who, at risk of sounding pretentious, have "personality" and make the music sound "come to life" to a certain extent). My stereo is, by today's standards, very dated, and my DAC is an onboard laptop sh*tbox. Furthermore, my higher frequency register is slightly impaired.

That being said, I have noticed that in many cases, (probably not all), the base sounds fuller, richer and has more impact in lossless formats. The fundamentals shine through. In chewy4's DBT I scored 2 out 2 (there are three but I skipped one) based almost solely on bass response. It's true that the difference isn't enormous between lossless and 320kpbs but is is there. And if you have higher quality sound sources than I do, which many of you probably do, then those differences should be more noticeable.

So if you, like me, a fan of electronic music you are probably into those lovely warm bassy sounds. They just sound better in lossless formats. Just my 2 cents.

Quote:

Originally Posted by MrLazyAnt 

Technical data is all well and good but, as has been mentioned a fair few times, it is your ears that you should trust at the end of the day. If you can't hear the difference between formats, more power to you, you can save a whole lot of disc space and more "fully" utilise your portaplayers.

 

I am not what most people would call a "bona fide" audiophile. I lack the funds for high-end equipment, I think that my 175 Euro Sennheisser HD 25 II's are great (they are the most expensive earphones I have owned to date), am reasonably happy with my Skullcandy Aviators (though they are somewhat lacking in bass impact), and often enjoy the sound of my $99 Marshall's (who, at risk of sounding pretentious, have "personality" and make the music sound "come to life" to a certain extent). My stereo is, by today's standards, very dated, and my DAC is an onboard laptop sh*tbox. Furthermore, my higher frequency register is slightly impaired.

 

That being said, I have noticed that in many cases, (probably not all), the base sounds fuller, richer and has more impact in lossless formats. The fundamentals shine through. In chewy4's DBT I scored 2 out 2 (there are three but I skipped one) based almost solely on bass response. It's true that the difference isn't enormous between lossless and 320kpbs but is is there. And if you have higher quality sound sources than I do, which many of you probably do, then those differences should be more noticeable.

 

So if you, like me, a fan of electronic music you are probably into those lovely warm bassy sounds. They just sound better in lossless formats. Just my 2 cents.

Bass should be the last thing effected by lossy compression. The amount of data required to accurately reproduce bass is very tiny compared to the high treble, so compression algorithms don't really deal with it until you get to really low bitrates.

I'd suspect that your results were coincidental if that's what you were basing it off of. As suggested several times on here I would suggest using Foobar's ABX comparator;it is really the only way to know you can hear the difference. As stated before: if you are able to do this and get 15/15(or 11/11 at the very least), you would be the first I've seen.

Always keep probability in mind when doing this kind of test... If 500 people participate in a coin flipping competition to see who can get heads the most number of times, chances are at least one person is going to be able to get heads 8 or 9 times in a row. Doesn't mean that they're good at flipping coins.

Quote:

Originally Posted by chewy4 

 

Bass should be the last thing effected by lossy compression. The amount of data required to accurately reproduce bass is very tiny compared to the high treble, so compression algorithms don't really deal with it until you get to really low bitrates.

 

I'd suspect that your results were coincidental if that's what you were basing it off of. As suggested several times on here I would suggest using Foobar's ABX comparator;it is really the only way to know you can hear the difference. As stated before: if you are able to do this and get 15/15(or 11/11 at the very least), you would be the first I've seen.

 

Always keep probability in mind when doing this kind of test... If 500 people participate in a coin flipping competition to see who can get heads the most number of times, chances are at least one person is going to be able to get heads 8 or 9 times in a row. Doesn't mean that they're good at flipping coins.

I wasn't basing my assessment only on your DBT, this is a general impression whose accuracy I've grown more and more convinced of over time. And I probably won't be able to hit the nail on the head every single time, but I've listened to many a track on youtube, directly after listening to it lossless, just to check if it's the right track before sharing on facebook, and the bass always suffers. Maybe it is just my mind playing tricks on me, I don't know. When I can be bothered, I will eventually do the ABX test. But that won't be today, and probably not before I have upgraded my music system, and grown accustomed to it. And I did keep probability in mind, which is why I listened to each sample 6 times, once for reference and the rest for comparison. The "Shpongle" I got the last three sets right and the "Mighty River" I got sets 2, 4, 5 and 6 right. I was fairly happy with these results. Maybe I just think I am noticing the bass fundamentals when really I am listening for the harmonics, which play an important role. Either way, it sounds like the bass to me =)

From "Sound On Sound" Magazine

http://www.soundonsound.com/sos/apr12/articles/lost-in-translation.htm

"The MP3 format has a reputation for making bass and low-frequency content sound weak: that slammin’ bass line can easily lose its phatness! (See audio examples L and M referenced above.) Low frequencies are harder for DSP algorithms to analyse because their durations are long, and amplitude differences over the short analysis windows used by the encoders may only be slight — so the analysis system doesn’t get an entire cycle of a low frequency per analysis window. In some situations, the encoder will be presented with less than a half cycle of any frequency below 114Hz. The AAC format fares much better in bass resolution, and it is thus much more forgiving to the bass."

I hate to admit I get a disproportional amount of pleasure in being right about this.

Quote:

Originally Posted by MrLazyAnt 

And I probably won't be able to hit the nail on the head every single time, but I've listened to many a track on youtube, directly after listening to it lossless, just to check if it's the right track before sharing on facebook, and the bass always suffers.

Yeah but this is about 320 mp3s and not (most probably) transcoded from lossy to lossy audio in youtube videos..