MP3 frequency comparison

MP3 frequency comparison

Most of us at some time or other convert an audio recording to the more compact mp3 format for storage, or for play-back on mp3 players, or for sending audio files through web-space. Mp3 is very convenient for this but is what is known as a “lossy” format – which means the price you pay for a compressed file size is a loss of frequencies throughout the audio spectrum of the recording. You may not be aware of this frequency loss if you’re listening on low-end audio equipment, but it is very definitely there.

For my own interest I’ve done a little research into this effect. For my source I used a few seconds of a CD recording of the violinist Sophie Mutter playing part of Mozart’s A-major violin concerto. I chose this because the harmonic content of her playing on her Guarnerius violin on occasion extends beyond 20kHz (about the top limit for a quality CD), and her playing is accompanied by a full classical orchestra. I copied the extract onto my computer using standard audio properties - bit rate 1411kbps, audio sample rate 44kHz, audio sample size 16 bit, and 2-channel stereo. Using CoolEdit’s spectral frequency analyzer the frequency and harmonic content of the music was very obvious on screen.

I made several mp3 versions of the sample at different bit rates, using eRightSoft’s “SUPER ©” media converter, and compared the spectral frequencies of these mp3 files on screen using CoolEdit.

Here are some of the results (the column tabulation may not show as it should):

bit rate cut-off frequency compression
1411kbps >20kHz 1:1
320kbps 19.5kHz 1:4.4
192kbps 18kHz 1:7.3
160kbps 17kHz 1:8.8
128kbps 16kHz 1:11
96kbps 15kHz 1:14.7
64kbps 11kHz 1:22
32kbps 5kHz 1:44

I think you’d have to have an extremely good ear and absolutely top-of-the-range audio equipment to notice any significant difference between the source (1411kbps) and 320kbps – and possibly as far down as 160kbps (iTunes suggests 160 or 192 for most purposes). When you get down to 128kbps (the “near-CD” quality beloved of mp3 file purveyors on the internet) even my old ears can tell the clear difference between that and 192 and higher. 128 and 96Kbps may be just about bearable if you’re listening on low-end playback, but when you get down to 64-32kbps range you’re definitely in the voice (radio and telephony) recording range – but even that low range can have an application in recording tunes – of which more below.

When you do a mp3 conversion it’s not only the top end frequency that goes (although a top cut-off of 16 or 17kHz is not significant for most adults), but frequencies well down in the audio range are lost or attenuated, the effect of which is some level of distortion. This distortion, as I’ve said above, is apparent when you get down to 128kbps or below, but can be lived with at 192 and above, and certainly with 320kbps. I did a couple of further experiments to explore this distortion effect. I digitally subtracted the 320kbps file from the 1411kbps file (the original) to generate a difference file consisting of the lost or attenuated frequencies. These were at a very low dB level and in practical terms would be inaudible. This was certainly not so when I subtracted the 128 file from the 320 – the lost frequencies, although still at a relatively low dB level, were certainly audible on playback and are the cause of the audible distortion on 128kbps mp3.

I mentioned the 64-32kbps mp3 range. This can have an application for someone who records a tune and needs a little help in transcribing it from the recording. Convert the recording to 32kbps mp3 and you’ll have a top cut-off frequency of 5kHz (just above the top note on a piano). This means you lose a lot of the clutter of hiss and high harmonics; the tune notes therefore sound more distinct and, if you use the spectral frequency option in a sound editor, are very obvious on screen.

Summary: 192kbps and above gives best audio results; 128 is ok for web use and if you’re not too worried about a little bit of distortion, and 32kbps can have an unexpected use in transcribing tunes.

Re: MP3 frequency comparison

"Using CoolEdit’s spectral frequency"

Aha, another CoolEdit fan! Fantastic wave editor, too bad it’s gone :( I’m glad I burned my downloaded copy and license key before Adobe bought them out.

Interesting analysis, and worth thinking about since I suspect most people go with 128k or whatever the default setting is without really knowing the difference.

I’m not sure how practical this is, but is there a way for you to post your clips—without labeling the sample rates—and let us rank them according to how our ears hear them, and then post the results later? That would be an interesting experiment, and the results would add a lot to your analysis.

I may try it out with a field recorder myself and see what my ears make of things like 160k vs. 320k…but then, I’m lazy. Even lazier than lazyhound, apparently.

Re: MP3 frequency comparison

Posting the clips - an interesting thought but I doubt its practicality. I can’t post them here (as far as I know), so that leaves other sites like Sound Lantern. But there is the possibility that the other site might alter the sound files - putting them into another format, for example.
Also, the size of an mp3 file is directly related to its bit rate: the 32kbps file, for instance, is 40KB, while the 320kbps file is 388KB. That, and the information in the inbuilt mp3 tags, would be a dead give-away.
It would be easier for people to try the experiment for themselves, as long as they’ve got a sound editor with a spectral analyzer. eRightSoft’s “SUPER ©” media converter is one of the most useful freebies I use - I don’t know whether you know it, but it’s very comprehensive and has apparently been written by geeks for geeks.
"how our ears hear them" - it’s not just that, it will also depend in part on the audio equipment that people use, and there will be a large variation in that area.

Re: MP3 frequency comparison

Fab stuff, Lazyhound. And I agree about Cool Edit Pro - particularly that it keeps on playing when you go do other windows, unlike the Adobe version.

mp3s generally are being increasingly used and I’m sure will take over from CD if they haven’t already. I’m involved in a project where we’re recording a CD (and we’re putting up mp3s to get some user feedback / suggestions before we finalise the mixes). We were commissioned to "make a CD" but it’s aimed at young parents and our research now indicates that more or less everyone in that age group uses mp3s via ipods / phones rather than good old CD. So mp3s are likely to be increasingly important - and in our case the website will stay on as a source of free music / info / downloads for that particular project.

Just to add to the great technical info you’ve posted, if the mp3 comes via a flash player rather than a link (as it does on the site I’m building and on many other sites) you’re restricted to a very few sampling rates

Re: MP3 frequency comparison

…the sampling rates normally recommended when playing mp3s via flash are multiples of 11: ie 11kHz, 22kHz, 44kHz. This doesn’t apply when simply playing mp3s via a web link, only when the audio is called up by a flash player. I’ve no idea why!!

Re: MP3 frequency comparison

That hydrogenaudio link - now that’s how the professional does a listening test!

Re: MP3 frequency comparison

I’ve long hated to listen to mp3’s but for many uses, especially, it seems, for emailing ITM tunes from sessions or to learn (etc.), they seem to be the dominant form.

It’s always pained me to work hard on the audio quality of music mixes in the studio and then to hear so much missing from the lossy compression of mp3s. A lot of audio engineers I know now do specific mixes for mp3s, often while listening to average-to-crap-grade mp3 devices as converters. And the top mastering houses now have made sure that their mastering engineers are fluent in the characteristics of mp3 compression.

However, the conversion codecs seem to be getting significantly better, which will almost always will happen to whatver medium is dominantly popular at the time.

The main advantage of the mp3 compression format is that the result is -tiny- and easily emailed and shared. So if one doesn’t mind using a larger file to share, there are a lot better options, and as broadband connections become more common, it’s easier to share even full-bandwidth (CD versions) of music and audio files. Still, emailing a tiny mp3 will remain attractive to most folks.

There seems to be a (somewhat nonsensical) race between
shrinking file sizes and growing hard drive sizes at shrinking per-byte sizes… If one can carry a player with 120 gigabytes of storage, why not use full-bandwidth files? I don’t really know the answer to that… maybe someone will not be able to change the catalog within the device for two years and so needs the tiniest files possibe … ?

So anyway, if file size isn’t quite so much of a concer there are alternatives for encoding that shrink the files with a lot less damage to the original audio recording, and thus the music.

I have no ‘dog in this hunt’, that is, I don’t necessarily prefer nor advocate any of these, but I present these compression methods as alternatives to the mp3 losses that lazyhound measured and described.
Here are Wikipedia pages for some of them.

http://en.wikipedia.org/wiki/Comparison_of_audio_codecs

http://en.wikipedia.org/wiki/Free_Lossless_Audio_Codec

http://en.wikipedia.org/wiki/Apple_Lossless

Finally, the super-talented producer and musician T-Bone Burnett has developed some sort of hardware/software box that he says makes mp3 files that are really whole and still tiny and compressed. By most accounts the sound of those files is wonderful, full and true. But Burnett says that he has no plans to make the thing, nor its processes, generally (commercially) available to the public. grrr.

Happy listening,

stv

http://cdbaby.com/Culchies
http://cdbaby.com/Lopers2

Re: MP3 frequency comparison

I’ve read a lot about lossy vs. lossless, etc. I use .flac when transferring audios, especially cassette rips of live material particularly sensitive to frequency loss due to digital compression (often recorded on inferior audio equipment or already victimized by generational deterioration or the ravages of time and atmosphere on the magnetic tape).

But the real issue is beyond anyone’s control, unless you’re in the studio making the masters. Engineers and producers continue to master recordings that I can’t turn up without making everything louder at once, and this drives me crazy. Sure, it sounds magnificent at low or medium volume settings on headphones or small computer speakers, or those dreadful all-in-one CD-radio-tape combo boxes they make. But when I put it on a real stereo, my gain and volume controls is overridden by practically uniform levels boosted above midline. Acoustic guitars are not louder than saxophones. Why they should be made to be this way in the studio confounds me. What ever happened to headroom, or am I going crazy?

Re: MP3 frequency comparison

Ramiro, that’s a very informative article about loudness.

Re: MP3 frequency comparison

I understand that To Petty has been taking it on recently. The new Mudcrutch release was on vinyl, and came with a CD option of the LP master to eliminate dynamic compression.

Bob Dylan complains about the same thng, I don’t know if he’s doing similarly, haven’t bought a Dylan album in a while.

Re: MP3 frequency comparison

Thanks for bringing up the Loudness Wars. It’s really tiresome, that practice. It’s one reason that I’ve largely stopped listening to contemporary pop music.

There’s a vital backlash against it. Mastering engineers hate it, most serious mixing engineers hate it, but somehow the myth that
over-compressed, clipped music is somehow more successful in broadcasting persists.

It’s a mystery to me. Mainstream radio is still ‘playlisted,’ that is, they only play music that is on lists that they get from consultants that they hire, so it’s desperately hard for new music to break into the lists. The consultants and the broadcasters have a separate ‘backchannel’, fueled by major label $$ that determines what gets played, so why in the world would it matter to make the song louder? I have no idea.

Over-"normalized", clipped files sound terrible as mp3s, too.

All the best ways to get heard just now seem to favor music with real dynamic range…

I think we’ll see an end to it. In general, the further one gets from the mainstream of pop and rock music the better the recordings are. Of course there are notable exceptions among artists and producers.

Personally, I’m a pacifist in the Loudness Wars. I like natural dynamic range. And I have a certain allergy to music made or recorded in a spirit of competition.

stv

Re: MP3 frequency comparison

Hmmm. stv, if they didn’t maintain that backchannel, they’d lose any claim to being a participant in the "industry". After all the changes of the last few years, they continue to insist that they not only own the racecourse, but all the dogs as well, in spite of plenty of evidence to the contrary. I really think they’ve already been superceded by technology but the world hasn’t figured that out yet. And of course they don’t want us to. "Pay no attention to the man behind the curtain."

Turns out that for all those years listening to session recordings from Ireland and small-run cassettes from independent distributors, I was helping to smash the imperialist/capitalist recording industry!

What I really want to ask was:

1) Is there any real difference between Cool Edit and Adobe Audition? Other than a couple of minor GUI changes, I don’t see any, but I was going to hang on to a spare copy of AA if it has better converters or something. It seems to be pretty bloated…

2) Is there any reason why eRightSoft SUPER is better than any other modern converter? Other than it’s free? Because their whole modus operandi is unfriendly to the user, and I don’t want them messing with my computer.

Posted by .

Re: MP3 frequency comparison

Gzeg, if by eRightSoft’s unfriendly modus operandi you mean their predilection for getting the user to visit their website by devious means (such as suggesting the current product isn’t up to scratch - this happened a while ago), I agree and have foiled that dastardly behaviour by blocking them at firewall level. However, I’ve no complaints about the way SUPER itself operates; as far as I’m concerned it does a fine job.

I can’t comment on differences between CoolEdit and Audition since I’ve never used the latter, which would be pointless since CoolEdit does everything I ask of it (and Audtion is hideously expensive, anyway). Like others here, I’ve taken the precaution of backing up CoolEdit together with my licence details for whenever I need to transfer it to a new computer - which I’ve done twice over the last few years.

Re: MP3 frequency comparison

Cymbal crashes are always the telltale sign of compression distortion for me - even if everything else sounds good, even a little bit of MP3 compression will cause a cymbal crash to sound unnatural…

Re: MP3 frequency comparison

There’s nothing worse than a compressed cymbal crash to ruin a classic ITM recording of "pure-drop" musicians…

Re: MP3 frequency comparison

I’ve uploaded 7 vorbis files coded with different quality to more or less match the sizes of the mp3 files.
Listen and judge yourself.

Re: MP3 frequency comparison

Loudness War wikipedia article - really interesting.

While centered around recordings, they do briefly mention radio, where this sort of thing has been going on for decades. Optimod and other multiband compression systems were programmed to give variable compression during the day - so during "drivetime" (ie people going to / coming home from work), a station’s output was more compressed than at other times. And if a station sounds "louder" than others, people are more likely to favour that station (that’s the theory!).

I think compression is OK if an intended part of the sound, but obviously for recordings of acoustic instruments it doesn’t really sound "right" - at least to me. Often big broadcasting organisations end up having systems which introduce opportunities for unintended compression problems. For example, working in a studio, you would control levels carefully as if anything went out too high in level, the limiters further along in the system would kick in and mangle the sound. You can sometimes hear this happening during loud parts of live classical broadcasts - particularly on descants in choral works, which always seem to catch the engineers by surprise.

However, it isn’t normal practice to monitor "off air" during a live broadcast as this is the job of someone in another studio where the radio network is "assembled" and leaves for the transmission system. Therefore you would never know what sort of stuff was happening to your broadcast once it left your studio.