I have found that the iPod badly distorts 320 kbps mp3. Most noticeable with classical music, such as piano concerts and opera.

The distortion is so bad on certain sequences that it is unbeleivable. I can best describe it as heavy wow and flutter as could be experienced on poor quality first generations cassette players.

In this test I have used EAC to extract a WAV file. Then proceded by converting the WAV file into mp3 320 kbps using different mp3 encoders as listed below.

I have tested two different iPods and the problem remain the same.

The track I tested was Hungarian Rhapsody No 2 performed by David Helfgott.

My findings were:

- MusicMatch 320kbps - Very High quality setting: No problem
- LAME 3.95 --preset insane -q0: Problem
- LAME 3.95 --preset insane -q1: Problem
- LAME 3.95 --preset insane: Problem
- LAME 3.93 --preset insane: Problem
- iTunes 320kbps: Problem

All mp3 files plays fine on iRiver IFP390T, Philips CD-8cm, a small no name stereo with MP3 capability and on the PC.

It seems that the problem is in the iPod decoder. Curiosly though that it can decode MusicMatch mp3's without distortion.

The problem is also present when using 256 kbps although less noticeable and even on 192 kbps.

Anyone knows if the iPod mini have a similar problem?

Does it happens with only some files, or does it occurs on all your music?

Btw, I would be interested by a sample where the IPod is adding distortion.

I ran a few tests awhile back and the iPod did seem odd:

http://www.audiosense.org/results/portable/iPod%20Mp3.htm

in comparison to the Karma:

http://www.audiosense.org/results/portable...Karma%20Mp3.htm

check the last graph on the iPod, the is something wierd with the Stereo decoding on high bitrate mp3s.

Oh yes, same iPod decoding a AIFF (Lossless, to rule out a hardware problem):

http://www.audiosense.org/results/portable/iPod.htm

Try "LAME 3.95 --preset insane -m s" and see if the problem still exists...

This happens with all my music, but is most noticable on music with quiet parts.

I also used another test track: The intro from the Opera Nabucco. I had to take off my headphones when I played it back on the iPod. It was sounding totally wrong, I initially thought I had broken the iPod.

The problem is that I have all my music in mp3 320 kbps, so I returned my iPod to the shop, and am currently unable to perform further tests.

When I get the chance I shall record the output of the iPod and send you a wave file sample.

It's been mentioned over at Head-Fi a couple times. Is your iPod the current gen? In a thread it seemed to be limited to 3G. I didn't have a problem with 1G.

My tests were done on a 1G

dman hope this is fixable with a new firmware

Thanks to BlessingX I have found the following threads on Head-Fi.org dealing with the same problem as described here on the iPod.

You can find the threads on Head-Fi by searching for:
- please AND verify AND ipod AND lame
- wav AND vs AND mp3 AND lame AND my AND experience

Users on Head-Fi report that the problem is not present in 1G iPods.

Still no word on the iPod mini.

I have verified this problem on a 3G 40GB iPod. Song: from Daphnis et Chloe, Suite No. 2, Lever du Jour, composed by Maurice Ravel, perf. by Cleveland Orch., BMG c. 1991.

Using EAC and LAME 3.90.3 I ripped (encoded) the song to WAV, APS (alt-preset standard), and API (alt-preset insane aka 320 kbps CBR). The beginning of the song is very quiet, with rapid flutes and bowed bass. The APS sounds quite similar to the WAV. The API (320 kbps) is extremely warbly/fluttery. This is not a subtle effect, it is quite unlistenable. Further into the track, when more of the orchestra is present and the volume is greater, the effect is not audible. The fluttering is audible both through the internal iPod amplifier and through the line out -> Yamaha Natural Sound Amp -> Sony MDR-7506 heaphones.

I've been trying to figure out the best setting/codec to use for my newly acquired iPod. I started out testing different iTunes settings and decided on iTunes VBR min. 224. During further listening I noticed skips in the audio, and found the highest freqencies to be unconvincing. I researched EAC and LAME. After 3 further grueling listening tests I settled for APS (alt-preset standard) and began to encode my collection. Upon further listening I wasn't happy with the high frequencies, esp. on cymbals. The way I hear it 14k-16k sounds really good, but uncompressed is more convincing in the upper reaches.

It's subtle, but as a practicing recording engineer I've often heard the difference between live, 16-24 track 2" tape, and 16-24 bit mastering formats. That's why I'm desperate to find the most transparent format possible for my iPod short of lossless, which would only allow me to store 80-90 albums. I have 250 and would like to have room to grow. I realize I haven't included AAC in my listening tests (my only other option for a compressed codec with the iPod - I may be forced to try it).

I decided to do another listening test with a larger variety of source material, all with very high production quality. I found the APS to sound pretty good, but the API (320 kbps) just sounded more convincing. It was in the "hard to put my finger on it but less annoying" category. The high freqs in some material with APS encoding can annoy me, and while the highest freqs in API aren't aren't totally comparable to CD quality (44.1khz PCM) it's doesn't tend to annoy me. However, my tests were mostly with limited dynamic range material like rock or highly processed music (read analog compressed and effected). The sound quality on classical music at low volume levels was awful and unlistenable.

I would estimate that the level of the problematic music was approx -20 to -16 dbFS. Higher levels seemed to mask the problem. As we know, the whole field of audio compression is based on masking of audio effects. An encoder that competes poorly at low bitrates may shine at highter bitrates, or as I've experience in this case, vice versa. As a practical matter I'm considering encoding my 'loud' type music with API, and high-dynamic range material with APS for use on my iPod.

I am surprised not to see more about this issue with the iPod. Apple is obviously not going to comment on this. The only mention of this issue I can find is on Hydrogen Audio. Are more people noticing this?

I would love to hear feedback on this issue.

Some questions: 1. Do you have problems with APS (alt-preset standard)? 2. Do you have warble problems with API (alt-preset insane aka 320 kbps) on low volume music? 3. What have you come up with as a practical approach for use on the iPod (3G)?

Peregrine Maese

This is a bit close to rule 8. Have you actually abxed APS to be inferior than API on such samples ?

I think APS is more than enough for ipods and you can always try --preset extreme. If you can do it, transcode from lossless to APS, APS -Y or APE. You will save heaps of space and there won't be audiable differences under non abx listening.

Your reply is general and doesn't address the problem of 320 kbps playback problem on iPod.

I'm currently in the process of adding AAC at various bitrates to my comparison. More on that when I've had time to conduct this new test.

It doesn't really matter whether I've ABXed APS and API because I don't appear to be able to use API as a 'one size fits all' encoding scheme that will work with my iPod (due to the distortion problem) and prevent me from having to re-encode my large record collection.

I reopened this thread primarily because I was shocked at how little documentation there is of the 3G iPod's incompatibility with 320 kbps mp3's.

Showdowking, in case of the iPod we have to bend rule 8 until some method of ABX becomes available on the iPod.

The files sound OK on a PC, 1G and 2G iPod's. This problem is only reported for playback on a 3G iPod. Three other users on Head Fidelity Headphones report similar problems. AFAIK nobody have reported the problem on the iPod mini either.

I cross my fingers for the iPod mini, as I will get one in two weeks time. First thing to test will be playback of Hungarian Rhapsody encoded by LAME at 320 kbps.

ABX on iPod

I don't know why ipods screw up 320k, they really should fix it.

I would go apple lossless if your ipod could handle it - even if it meant re-encoding. For people who aren't comfy with APS/APE/MPC go lossless at all cost.

If the ipod is also distorting APS / APE than its very serious.

WreakyHavoc,

To get quality sound on your 3G iPod you can use MusicMatch to encode your files.

MusicMatch uses an MP3 encoder from Fraunhofer.

These files sounds perfect on the 3G iPod I was testing.

However I don't know if the high quality FhG MP3 encoder is sold seperately as CLI.

Anyway I was lucky that I could return my iPod to the shop where I bought it as they had a 14 day money back, no questions asked policy.

Now I'm getting the iPod mini and hope that will not have this problem.

Thanks for your input Sehested. I will continue to test codecs. This iPod was a birthday gift. I don't have the option of using a different player. I am quite happy with the size of the disc (40 GB), as I intend to use it as a portable hard drive ferrying uncompressed audio files from the recording studio to home.

I'm not distressed about this issue. I would, of course, love to get right to ripping my entire CD collection (250+) and enjoying random, high quality playback, but I don't mind taking the time to find the best solution in the meantime.

I'm puzzled that so few have reported the 320 playback problem. I read your post during my research but never gave it much thought. I had tried 320 iTunes, but mostly with rock/metal/punk material. It wasn't until I had given up on iTunes' Fraunhofer codec in favor of LAME, and decided to add classical that I ran into this problem.

I was one disappointed customer, discovering that the "perfect" iPod, could have such a problem.

The only thing we can do to change this are to report the issue to Apple.

Please take the time and file a complaint.

iPod feedback

and while you are at it, may I suggest that you file one for lack of gapless audio playback as well. When you listen to concept Rock albums, such as Pink Floyd the Wall it is really annoying that the transition between the songs is not natural.

The warbling problem is there with more bitrates than just 320kbps. I tried a few test encodings using a problem sample I had uploaded (opening to Stravinsky's Rite of Spring; see upload section) to determine the point at which they would disappear, and found that --preset cbr 160, --preset cbr 192, et al until 320kbps exhibited blatant, clearly audible warbling/gurgling. The problem got better as the bitrate got lower, but even at 160kbps, wasn't corrected. Same goes for iTune's MP3 encoder; even VBR encodings showed this problem! However, --preset standard played back flawlessly, which is curious: if CBR frames are flawed, and iTunes VBR is flawed, why would --preset standard not be? In any case, all that's left of my iPod is a feedback rating on eBay After all I had heard about the iPod being the "prima donna" of MP3 players, it was a grossly underwhelming experience. Even a cheap $35 MP3-CD player from K-Mart can play back files without mangling them, so why not the iPod?

Any word on this from the new 4G users? I'm (perpetually) considering an ipod but I don't like the sound of this. I have quite a few 320's even though the bulk of my stuff is ~230 vbr.. but in any case I don't want anything but perfect playback for the price those suckers go for

4G users, comments?

Has anyone actually conducted any AAC tests on the 4G(or 3G for that matter) concerning this distortion issue, or is it only concerning the MP3 codec?

strangely enough i dont hear the warbling, distorted sound that is being reported, and i have plenty of music that ive encoded at --api (320k cbr)

It completely depends on the type of music you encode; I only noticed it with some classical pieces. The warbling effect is only noticeable when a single tone or instrument (like a bassoon) holds a note. Check out my sample in the uploads section for an idea of what I mean.

I have now tested my iPod mini for high bitrate MP3 distortion and am happy to say there is no trace of this problem.

Seems the problem is isolated to 3G iPod's.

After weeks of listening tests that I found to be intense and at time exasperating (I mean, there are some songs I never really want to hear again) I have decided on a practical solution. I'm encoding at 224 AAC using the iTunes application. I have been quite happy with the playback and have found no trace of the distortion problem. High frequencies are convincing enough that I don't spend all my time thinking about sound quality. I was having that problem with the LAME mp3's. I would hear an artifact and it would draw my attention, thereby destroying my immersion in the musical experience. This still happens at times with the AAC, but not as often.

I really wanted to use the highly tested and mostly open standard solution of EAC with LAME. I was not happy with having to use a proprietary format (Apple AAC). But it is nice to actually be using the iPod now, instead of doing listening tests. What I'd really like to do is find a way to make EAC and iTunes play well together, but the times I've tried to rip with one and encode with the other required so many extra steps and manual data entry it just wasn't worth it.

I've got over a hundred discs encoded so far and am really enjoying this little wonder.

The next step will be to get a giant HD (250 gb or more) and rip lossless for future transcoding. I'm happy for now, though.

Would you feel better if I said to you that AAC is from the same committee that developed MP3?



A nice point, the really important one.



Have you tried this whole thread? I've heard people there able to integrate EAC and Itunes AAC encoder from the first attempt flawlessly.

Yeah.. the other name for AAC in MPEG4 is MP4.. Compared to MP3, MP4 is a very advance encoder.. comprising many new tools which even Apple didn't implement in its ITune !!

AAC has much improved temporal and aliasing handling capabilities than MP3..