Head-Fi.org › Forums › Help and Getting Started › Introductions, Help and Recommendations › can't hear the difference between mp3 and flac
New Posts  All Forums:Forum Nav:

can't hear the difference between mp3 and flac

post #1 of 13
Thread Starter 

I did some comparison and it seems like i can't hear the difference between 128 kbps mp3 and flac ( i can hear the difference between low bitrate like 32 kbps tho). However, in Jriver Media Center's analyzer, i can see that the higher frequency for the 128 kbps mp3 looks not smooth compare to the flac.

 

I also did some online test from different websites and it seems like the highest frequency i can hear in a blind test is 14-16 khz.

 

So my question is:

 

-Does your ability to hear wide ranges of frequencies affect your ability to tell the difference between high and low quality music file?

 

-Does this mean i should reconsidering getting my music in 320 kbps (or lower) mp3 instead of flac? 

 

the websites i used to test my hearing:

 

http://www.noiseaddicts.com/2009/03/can-you-hear-this-hearing-test/

http://www.audiocheck.net/audiotests_frequencycheckhigh.php

http://www.audiocheck.net/blindtests_frequency.php?frq=14

http://www.phys.unsw.edu.au/jw/hearing.html

 

 

BTW what's the highest you can hear? 


Edited by imeem - 2/15/14 at 8:42pm
post #2 of 13
Quote:
Originally Posted by imeem View Post
 

I did some comparison and it seems like i can't hear the difference between 128 kbps mp3 and flac. However, in Jriver Media Center's analyzer, i can see that the higher frequency for the 128 kbps mp3 looks not smooth compare to the flac.

 

I also did some online test from different websites and it seems like the highest frequency i can hear in a blind test is 14-16 khz.

 

So my question is:

 

-Does your ability to hear wide ranges of frequencies affect your ability to tell the difference between high and low quality music file?

 

-Does this mean i should reconsidering getting my music in 320 kbps mp3 instead of flac? 

 

the websites i used to test my hearing:

 

http://www.noiseaddicts.com/2009/03/can-you-hear-this-hearing-test/

http://www.audiocheck.net/audiotests_frequencycheckhigh.php

http://www.audiocheck.net/blindtests_frequency.php?frq=14

http://www.phys.unsw.edu.au/jw/hearing.html

 

 

BTW what's the highest you can hear? 

 

Better hearing is always helpful; better listening can be learned.

 

What equipment is delivering your music to you.  The problem is that you are limited by the weakest link(s) in your chain, as well as how each piece fits together in the making the whole.

post #3 of 13
Thread Starter 
Quote:
Originally Posted by KG Jag View Post
 

 

Better hearing is always helpful; better listening can be learned.

 

What equipment is delivering your music to you.  The problem is that you are limited by the weakest link(s) in your chain, as well as how each piece fits together in the making the whole.

DAC/amp: fiio e17 and L7 LOD

IEM: Vsonic VSD1s

speaker: Audioengine A2

 

have other headphones but didn't test them 

 

When i introduce my friend to loseless audio, he wasn't able to tell the difference at first. But he said that the loseless version had better bass than the lossy version. He compared it by putting his head and ears between the 2 speaker enclosure.  He was using onboard audio w/ itunes and a unknown JVC mini hi-fi system via AUX in.

 

What should i pay attention when doing these comparisons? I try to listen for better bass, clarity, and distortion. 


Edited by imeem - 2/15/14 at 8:30pm
post #4 of 13

I'm not familiar with your specific equipment other than your speakers.

 

While most of my serious listening is not computer based, I can tell you that standard MP3's are mediocre at best and even a slightly above-average system will display the difference between them and various superior file formats.  You're on the right track with regard to what to listen for when upgrading from MP3's.  I would add soundstage, separation, air, treble extension and imaging.

 

Of course the quality of the master recording and its transfer are critical.  I have a few modern CD's from major artists that sound like compressed and crappy MP3's.

post #5 of 13
Thread Starter 
Quote:
Originally Posted by KG Jag View Post
 

I'm not familiar with your specific equipment other than your speakers.

 

While most of my serious listening is not computer based, I can tell you that standard MP3's are mediocre at best and even a slightly above-average system will display the difference between them and various superior file formats.  You're on the right track with regard to what to listen for when upgrading from MP3's.  I would add soundstage, separation, air, treble extension and imaging.

 

Of course the quality of the master recording and its transfer are critical.  I have a few modern CD's from major artists that sound like compressed and crappy MP3's.

do u have any suggestions on which piece of music is good for comparison with mp3 and flac? I'm thinking classical music? 

post #6 of 13

320 MP3 and FLAC are not that different. The Audioengine A2 and VSD1S may not be the best gear to test FLAC vs MP3. The difference I found is mainly that the bass is stronger in FLAC. 

post #7 of 13
Quote:
Originally Posted by imeem View Post
 
Quote:
Originally Posted by KG Jag View Post
 

I'm not familiar with your specific equipment other than your speakers.

 

While most of my serious listening is not computer based, I can tell you that standard MP3's are mediocre at best and even a slightly above-average system will display the difference between them and various superior file formats.  You're on the right track with regard to what to listen for when upgrading from MP3's.  I would add soundstage, separation, air, treble extension and imaging.

 

Of course the quality of the master recording and its transfer are critical.  I have a few modern CD's from major artists that sound like compressed and crappy MP3's.

do u have any suggestions on which piece of music is good for comparison with mp3 and flac? I'm thinking classical music? 


Sorry--not my primary formats.  You could rip an audiophile/excellent quality CD.

post #8 of 13
Quote:
Originally Posted by imeem View Post
 

I did some comparison and it seems like i can't hear the difference between 128 kbps mp3 and flac ( i can hear the difference between low bitrate like 32 kbps tho). However, in Jriver Media Center's analyzer, i can see that the higher frequency for the 128 kbps mp3 looks not smooth compare to the flac.

 

Did you read up on how file compression in general and MP3 work? And what music are you listening to? Some music can be more susceptible to audible differences when compressed below 320kbps depending on what's in the recording. A mainly vocal track with a violin on even an audiophile CD has a relatively narrower range of frequencies and can be affected less by lossy compression than, say, classical music with a full symphony, or a metal track with lead vocals, a choir, two guitars and a bass guitar, a synth, drums, and a string quartet (if not a full symphony). Excerpts below with my notes:

 

-------------

 

Lossy audio compression algorithms provide higher compression at the cost of fidelity and are used in numerous audio applications. These algorithms almost all rely on psychoacoustics to eliminate less audible or meaningful sounds, thereby reducing the space required to store or transmit them.

 

It uses psychoacoustic models to discard or reduce precision of components less audible to human hearing, and then records the remaining information in an efficient manner.

 

The innovation of lossy audio compression was to use psychoacoustics to recognize that not all data in an audio stream can be perceived by the human auditory system. Most lossy compression reduces perceptual redundancy by first identifying perceptually irrelevant sounds, that is, sounds that are very hard to hear. Typical examples include high frequencies or sounds that occur at the same time as louder sounds. Those sounds are coded with decreased accuracy or not at all.

 

Some audio is hard to compress because of its randomness and sharp attacks. When this type of audio is compressed, artifacts such as ringing or pre-echo are usually heard.

 

Time resolution can be too low for highly transient signals and may cause smearing of percussive sounds

 

There is no scale factor band for frequencies above 15.5/15.8

 

-----------

 

Notes:

 

I. Psychoacoustic presumption

The algorithm for compression assumes that your brain only pays attention to some sounds, the same way that groupies flock to the band members generally in this order: vocalist, lead guitarist, rhythm guitarist, drummer, and then maybe the bassist too. Maybe. So basically unless you have a band that puts an emphasis on the loudness of the bass, like the Red Hot Chili Peppers,* Greyhoundz (local band here),* Mudvayne, etc., then most likely the algorithm will assume that since it's recorded too low, you wouldn't notice if it's gone. Kind of like when a band that doesn't have rad bass lines has its bassist getting sick or something and they get one of their crew to play on stage with them, and unless you're missing Flea jumping around on stage, you probably wouldn't notice. On some other types of rock and metal, the synth would be the next to go, if not ahead of the bass guitar (first example above was the more typical band format). It's like not hearing the triangle in a symphony unless they specifically put it "up front" in terms of loudness by putting a microphone next to it.

 

Now as much as I critique it in practice, there is a reason why it was a viable theory. For example in any given moment you have dozens of sound sources but your brain sorts them out an pays attention to only a few sound sources, kind of like how the NSA is monitoring even this thread or that time I called my brother but they aren't really paying attention until the computer flags a few keywords that I wouldn't even type here, then an agent tunes in to see whether it's actually sleeper agents talking or Democrats and Republicans arguing in the comments section (or I was referring to my brother's classmate's gross zit instead of anything you can park a van filled with fertilizer on). Step into the world's quietest room, and suddenly you can hear your own heartbeat. It's not like your ears couldn't pick up the sound otherwise, considering it's that close, but your brain wisely enough would pay more attention to other people and vehicles to keep you alive because the only time you really need to listen to your heart is when you have a freaking stethoscope. And most likely that will be somebody else - a doctor - doing that for you. Deprive the brain of other stimuli and you'll finally hear it, and it might even drive you insane, kind of like the heart of that guy with the weird eye.

 

 

II. Music Genre

 

On top of the examples already mentioned above regarding bass, synth, vocals and violins, refer to one of the highlighted portions. When you have highly dynamic notes coming in you get a few troubles with compression. What I've noticed with my own music is that the bass drum and the other percussion tend to lose either loudness or lower frequencies, if not both; that's on top of getting even less of the bass guitar. Fast double pedal action can sound a little hollow, more like a school band drummer tapping on a snare with a pillow in it than someone working two pedals. Again, rip an audiophile-standard mastering vocal track, or a relatively simple pop track with drums and a synthesizer backing up Britney, and chances are both won't be noticeably affected.

 

 

 

*Coincidentally, people actually know the name of RHCP's bassist, Flea; and over here the Greyhoundz' bassist is popular with the ladies (not sure about groupies but my friends think he's the cutest bassist ever); now imagine Jason Newstead leaving after Load, Reload and Garage Inc, and the first thing you noticed about his replacement was his dreadlocks.

 

 

Quote:
Originally Posted by imeem View Post
 

I also did some online test from different websites and it seems like the highest frequency i can hear in a blind test is 14-16 khz.

....

BTW what's the highest you can hear? 

 

Assuming the setting of the gain and volume knobs anywhere in the chain remain constant, yes it is harder to hear the extremes of the frequency spectrum. You can pause that when you stop hearing anything, then boost the volume. Chances are you might hear a little bit more.

 

On my HD600 I can hear fine all the way up to 22khz, but of course that sounds a lot lower than 2khz; still that's an additional 2khz over where people say I shouldn't be able to hear anything ergo some get to say ribbon tweeters are useless. On my Aurisonics ASG-1 I can hear sound only up to 12khz, volume dropping at 10khz, using my regular listening level (obviously this one has a design compromise). If I pause and increase the volume, I can up to 16khz. Take note the Sennheiser is rated for 39khz and Aurisonics is rated for 25khz.

 

Quote:
Originally Posted by imeem View Post
 

-Does your ability to hear wide ranges of frequencies affect your ability to tell the difference between high and low quality music file?

 

It can, but that can vary with each genre. Some notes tend to have a very wide range of frequencies in it, and if the lossy algorithm for example essentially disregards say above 16khz, then the same violin, guitar, piano, or cymbal strike may not sound the same as on another. Of course, the headphone/IEM or speaker may not even be playing those frequencies at an audible level to begin with (as in the example above). There is a difference between a manufacturer claiming that their product goes from 20hz to 25khz and one that actually says, more precisely, "-3db at 25khz." The first just says that their microphone picked up sound from it up to 25khz, but for all you know it starts dropping sharply at half of that; the latter specifically says it starts dropping sharply at 25khz.

 

Quote:
Originally Posted by imeem View Post
 

-Does this mean i should reconsidering getting my music in 320 kbps (or lower) mp3 instead of flac? 

 

If it was just not noticing the difference on music, ie psychoaboustics, then I would say give FLAC a chance. However, if after trying the pause then increase volume test you can't hear anything, I'd say try keeping 320kbps if you have music that still has a lot of instruments in them.

 

 

Quote:

Originally Posted by imeem View Post
 

DAC/amp: fiio e17 and L7 LOD

IEM: Vsonic VSD1s

speaker: Audioengine A2

 

have other headphones but didn't test them 

 

For starters, the A2 is rated at 65hz to 20khz, -2db. However just because the bass drops sharply at 65hz doesn't mean it is very flat above that. In terms of psychoacoustics, that either tricks you into thinking it has full bodied bass all the time with successive single pedal drum hits, or that you'd be distracted by how it sounds like a mudslide on Pantera and Mudvayne.

 

Quote:
Originally Posted by imeem View Post
 

When i introduce my friend to loseless audio, he wasn't able to tell the difference at first. But he said that the loseless version had better bass than the lossy version. He compared it by putting his head and ears between the 2 speaker enclosure.  He was using onboard audio w/ itunes and a unknown JVC mini hi-fi system via AUX in.

 

What should i pay attention when doing these comparisons? I try to listen for better bass, clarity, and distortion. 

 

I'll give you a few sounds cues to listen for and some artists who I can think off the bat would be best for listening critically for each, along with the others I posted above. A few caveats though:

- I won't type down what the sounds should sound like as in Adam West Batman visualized sound effects (like THUD! KAPOW!) as I think it can probably just add to the confusion; although it was easier doing that talking to my friend right in front of me

- I chose these among music I listened to recently, so they aren't necessary the best, just the ones I thought of quickly

- You might not have access to most of these past YouTube, but I'll try anyway

- These are primarily on bass and percussion, which despite my 22khz hearing, are what I easily pick out between gear and file compression level/format - I was a vocalist (although I play guitar and bass) although briefly but I used the beat as cues for when to start singing (again), metal or not (save for a few ballads with just a piano or violin)


 

1. Tone of the bass drum - does it sound deep, ie, with more lower frequencies? does it have some sense of reverberation (easier with subwoofers and large speakers)? 

ex. Deftones' Digital Bath and Passenger

 

2. Tone and detail, location of the other percussions - when they do a drum roll do they pan properly, ie, if the sound is not equal the ones in the center should only be slightly louder? do they sound deep but the sound of each hit doesn't blend into the next?

ex. Nightwish, particularly Sacrament of Wilderness from Oceanborn (the studio recording)

 

3. Tone and detail on bass drums with double pedals - do they sound exactly the same? does one sound deeper/fuller, each hit more detailed? is one audio format louder than the other?

ex. Mudvayne, particularly on tracks from The End of All Things to Come (mostly because that's the one I listen to most often); Pantera; Dream Theater, particularly Metropolis Pt.I;

 

4. Audibility and detail on bass guitar and other instruments - do you hear the bass guitar while the guitar(s) is/are playing loudly? can you even hear the bass guitar on quiet passages? what about the synth in the background - are they the same loudness relative to the other instruments?
ex. Nightwish, particularly Sacrament of Wilderness (the studio recording, too many of this live), Dark Chest of Wonders, I Wish I Had An Angel

 

5. General lower frequency - do the notes that carry the beat sound deep and detailed? do the notes fade faster on one file format?
ex. Feist - One Evening


Edited by ProtegeManiac - 2/16/14 at 11:46am
post #9 of 13
Thread Starter 

i only tested 2 songs so far : one of them was Japanese rock and another was home by Michael Bublé. Will do some more testing. 

post #10 of 13

The best test is to sample a wide range of songs. In my opinion, the more complex the song the better. For example, listening to Epica's 'Design Your Universe' which melds classical orchestral with metal, the difference is obvious. Lossless, the female vocals come across more clear and the male grunting sounds more deep and textured. The horns and drums in the background are less cluttered and have greater transparency. Lossy, both AAC 256kb/s and MP3 320kb/s, the female vocals become more veiled and sibilant, male grunting sounds muddled. Horns lose texture and drums sound more flat.

 

Let us not forget that not all ripping programs are created equal. Quality can vastly differ from one program to the next. Just because a file is FLAC or 320kb/s is meaningless if not done right. Programs such as iTunes actually license the MPEG technology, and I can say from experience make all the difference. From my own ripping experience, a 128kb/s song ripped in iTunes can sound better that a 256/320kb/s done in a freeware ripper. Where are you getting your lossless files from? Are you doing them yourself or downloading via torrent? What's to say that the FLAC file is first rip? What I mean is the FLAC could have come from a lesser quality file such as a 320kb/s MP3, and not from the master source (i.e.. CD).

 

All things aside, the ONLY way you are going to hear any difference in sound quality is to take a CD you are familiar with, rip the disc in a wave file, AIFF, or FLAC, and then re-rip but in MP3 format. Compare the 2 yourself. Simply downloading 2 versions of a song has huge room for error as stated above. 

post #11 of 13

I think you need to learn more about digital audio. Hell, most of headfi could, but I digress.

One of the factors affected by mp3 (and most audio file-size compression) is the sampling rate (CDs are 44.1kHz). Dropping the sampling rate gives you a smaller file, but sampling rate is directly related to the highest frequency that can be sampled (and therefore stored in the file).

So those really low bit-rate mp3s are noticeable to you because they've probably halved the sampling rate (from my experience 64kbps would do this), meaning anything over ~11kHz is gone. That's extremely noticeable.

So yes, if you can only hear up to 14kHz that would certainly affect your ability to tell the difference.

post #12 of 13
Thread Starter 
Quote:
Originally Posted by Bansaku View Post
 

The best test is to sample a wide range of songs. In my opinion, the more complex the song the better. For example, listening to Epica's 'Design Your Universe' which melds classical orchestral with metal, the difference is obvious. Lossless, the female vocals come across more clear and the male grunting sounds more deep and textured. The horns and drums in the background are less cluttered and have greater transparency. Lossy, both AAC 256kb/s and MP3 320kb/s, the female vocals become more veiled and sibilant, male grunting sounds muddled. Horns lose texture and drums sound more flat.

 

Let us not forget that not all ripping programs are created equal. Quality can vastly differ from one program to the next. Just because a file is FLAC or 320kb/s is meaningless if not done right. Programs such as iTunes actually license the MPEG technology, and I can say from experience make all the difference. From my own ripping experience, a 128kb/s song ripped in iTunes can sound better that a 256/320kb/s done in a freeware ripper. Where are you getting your lossless files from? Are you doing them yourself or downloading via torrent? What's to say that the FLAC file is first rip? What I mean is the FLAC could have come from a lesser quality file such as a 320kb/s MP3, and not from the master source (i.e.. CD).

 

All things aside, the ONLY way you are going to hear any difference in sound quality is to take a CD you are familiar with, rip the disc in a wave file, AIFF, or FLAC, and then re-rip but in MP3 format. Compare the 2 yourself. Simply downloading 2 versions of a song has huge room for error as stated above. 

 

I used dBpoweramp to convert from flac to mp3. The 2 songs i tested so far was not ripped by me. I ripped 1 album using Jriver Media center (Eminem's The Marshall Mathers LP 2) but i didn't test those songs (yet). Mainly because for some reason, i find it lack detail and it's compressed/not dynamic (hope i'm using the right terminology). Maybe because Rick Rubin's the producer.  Overall the sound was not enjoyable and i'm not used to this sort of sound for a rap/hip hop album. 

Quote:
Originally Posted by kraken2109 View Post
 

I think you need to learn more about digital audio. Hell, most of headfi could, but I digress.

One of the factors affected by mp3 (and most audio file-size compression) is the sampling rate (CDs are 44.1kHz). Dropping the sampling rate gives you a smaller file, but sampling rate is directly related to the highest frequency that can be sampled (and therefore stored in the file).

So those really low bit-rate mp3s are noticeable to you because they've probably halved the sampling rate (from my experience 64kbps would do this), meaning anything over ~11kHz is gone. That's extremely noticeable.

So yes, if you can only hear up to 14kHz that would certainly affect your ability to tell the difference.

i was able to make one of those really low bitrates mp3 44 khz dBpoweramp 


Edited by imeem - 2/18/14 at 6:07am
post #13 of 13
Quote:
Originally Posted by imeem View Post
 

i was able to make one of those really low bitrates mp3 44 khz dBpoweramp 

If you analyse the mp3 you'll see that the sampling rate has been reduces, hence the cut-off in the high frequencies. Only 3 factors determine bit-rate, so the only way to reduce file size is to reduce sampling rate, bit depth or number of channels. Modern lossy (and lossless) encoders are very clever and do this intelligently.

 

Bit-rate (bits per second)  = sampling rate (Hz) * bit depth (bits) * channels

File size =  bit-rate * length (seconds)

 

e.g. Stereo CD bit rate: 44,100 * 16 * 2 = 1,411kbps

New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Help and Getting Started › Introductions, Help and Recommendations › can't hear the difference between mp3 and flac