Comments on: Why mp3s suck, and how to hear it

By: Ian Shepherd

Ian Shepherd — Wed, 16 Jun 2010 21:41:01 +0000

Hey Tim,

Thanks for commenting. And, I’m sorry you think the article is a grossly oversimplified rant, I think that’s an exaggeration – the clips are there so people can judge for themselves. But hey – as I’ve said in a comment above, it’s not meant to be a scientific analysis.

The study you link to doesn’t show that mp3 outperforms AAC at all – the iTunes encodes used were still mp3 encodes. Additionally, the study clearly shows that all the encoders tested were statistically equal in quality – but over very small samples, in general. Given that the study has no control over people’s listening software or hardware, I would also hesitate to call it scientific at all.

Regardless, my real point is: go lossless. I’m not interested in the open-source debate, I just want encoders that work and sound great with no tweaking.

My experience of mp3s from Amazon is… not good. And as I repeatedly say in the article, most people use 128kbps still, this is the widely miss-named “CD quality” standard, and that’s what I want to debunk. I see so many more articles ignoring all the problems of lossy encoding completely, than ones criticising lossy compression, and that’s what I want to redress.

So again, sorry the article annoyed you, but as someone who’s already fully engaged with the issues, you’re not really the audience I’m trying to reach.

Cheers !

Ian

By: Tim

Tim — Wed, 16 Jun 2010 10:41:06 +0000

I guess I’m really tired of reading such grossly oversimplifying, pointless rants which seem to based on some snobbish attitude rather than on facts.
Despite of being theoretically(!) inferior to AAC, the design deficits of MP3 are compensated by more advanced psychoaccoustic models used by modern MP3 encoders. There have been many listening tests prooving that modern MP3 encoders compete well with AAC and outperform Ogg Vorbis at bitrates > 128kbps.
In fact the last scientifically conducted listening test showed that all(!!) modern MP3 encoders outperformed iTunes AAC at 128kbp average bitrate (http://listening-tests.hydrogenaudio.org/sebastian/mp3-128-1/results.htm). The often claimed inferiorty of MP3 is a myth based on the ugly performance of some MP3 encoders 10 years ago! The incredible development MP3 encoders have taken since was due to the competition between the open source Lame encoder and the commercial encoders. AAC remains unattractive for me as long as there’s no open source equivalent to Lame. Of course lessless compression should be considered an important alternative to MP3, as it is free from the drawbacks of lossy formats and disk space is continously increasing. Again, the most mature and widely supported standard is open source FLAC – for the same good reason that Lame became the most widespread MP3 encoder. The ~256kbps Lame mp3s offered these days by Amazon usually offer superior sound quality to all of the AAC files you’ll get from iTunes.

By: Liko Kine

Liko Kine — Thu, 13 May 2010 15:49:24 +0000

I am not an expert on any of this but is there any digital music that replaces the clipped audio to make it a digital signal?

By: Ian Shepherd

Ian Shepherd — Tue, 11 May 2010 12:01:47 +0000

Hi Suneel,

It depends ! There’s no point in using 24/96 on anything that came from CD, for example. Up-sampling won’t increase the quality and might even harm it.

If you have 24/96 sources, go for it ! Personally I would suggest FLAC at the source resolution – eg. 16/44 for CDs, 16/48 for DATs if you have them, etc.

Ian

By: Suneel

Suneel — Sat, 08 May 2010 05:30:18 +0000

Hi Ian,
I’ve started my music collection anew with Flac in 24-bit/96kHz, played through Rockbox.
Is it overkill or am I actually getting better sound?

Cheers
Suneel

By: Ian Shepherd

Ian Shepherd — Mon, 26 Apr 2010 22:48:57 +0000

@ Norseman

Realised I never replied to DampeS8N – fwiw I can hear mp3 artefacts in the car, on my iPod earbuds and over analogue FM radio – no comparison needed. Yes, people use mp3s for convenience (even me!) but we’re losing out as a result…

@ John – I have ABX-ed myself on all kinds of issues, including mp3 vs WAV etc – enough to know the differences are as plain as the nose on my face. Is the argument here tendentious ? Maybe, but so are all the posts out there that claim 128kbps mp3 is “CD Quality”…

I’m happy to stir it up a bit at the expense of some scientific rigour

By: Norseman

Norseman — Mon, 26 Apr 2010 16:53:36 +0000

@Ian Shepherd

Well, of course, skipped my mind
But I dont think convenience should come before quality

I use my iPod alot at work, but I’m not too picky on quality when rattling around in the forklift.

By: Ian Shepherd

Ian Shepherd — Sun, 25 Apr 2010 21:50:10 +0000

Hi Doedolf,

The mp3s were made from the WAVs, so I’m surprised you say they are different levels. This is absolutely an issue I appreciate – I’ve blogged about it before.

But I have to admit I didn’t check, in this case – how are you comparing their levels ? Perhaps your mp3 decoder is clever enough to pad the output to prevent clipping during reconstruction ?

Ian

By: Ian Shepherd

Ian Shepherd — Sun, 25 Apr 2010 20:50:39 +0000

@ Norseman – very good question ! Although, I do this myself, and the answer for me is convenience. If I want to listen hard and well, I fetch the CD. But if I just want to stick some music on, I listen to my iPod…

By: Norseman

Norseman — Sat, 24 Apr 2010 11:18:49 +0000

Love to read things like this. Im on my 10k$ stereo with high-end dedicated stereo soundcard in my computer. Ill never listen to an mp3 again, unless it the only way I can listen to that one great song (which is doubtful)

I knew mp3 had failed before I read this, but after reading, things stick out a bit more.

Tell me please why people use expensive stereosystems, mini-jack to motherbord soundcard, and listen to mp3?

Is there any reasonable answer to this… Probably not.