This is an alternative weblog for the existing dave2002 weblog hosted at dave2002.pitas.com. This features material on weblogs, music, art, theatre, travel, news and software and hardware, amongst other things

My recent furls

Search this site or the web powered by FreeFind

Site search Web search

Site Meter

Saturday, July 30, 2005  

Compressed audio files and streaming audio

I've been revisiting audio files in the last few weeks, following the BBC's Beethoven Experience during which they made all the symphonies available in MP3 format at 128 kbps encoding. I, like many others, downloaded all of these, but I found them a little infuriating. Some of the performances are very good, but I just found the audio quality a bit tiring. This prompted me to check out compression formats again.

There is evidence that apart from lossless compression methods, that it is often possible for experienced musicians to detect the difference between an original source and a compressed copy. Less trained listeners will often accept lower quality, perhaps because they don't know what to expect. Some classical music can sound satisfactory at rates as low as 96 kbps (MP3, but more complex sounds will need higher data rates. Some music works moderately well at 128 kbps (MP3) but this is not always the case, and higher rates will be needed for good results. Most listeners will find it hard to detect much difference between MP3 compressed audio at 192 kbps or higer and the original source.

There has been discussion about which compression formats are best. This is not by any means certain, with some users preferring WMA, and others preferring Ogg Vorbis or MP3 - there may even be people who prefer Real Audio formats. Many writers have criticised MP3, but their justification for doing so seems uncertain. At high bit rates MP3 should give good results, as should most of the other methods. At lower bit rates MP3 may indeed be less good than some other methods, but all methods will give poorer results at low bit rates. Here I am not too interested in low bit rate compression, as I really don't like the sounds produced by low bit rate compression. It may be necessary sometimes, but should not be used for sounds where audio quality is a major part of the listening experience.

The quality of an MP3 encoding does depend on the encoder used. A few are apparently significantly better, though most do well at higher rates (say 192 kbps or above). THe better ones will also do reasonably well at lower rates, and here there can be significant differences between different encoders. Some of the better ones are those which use the LAME encoders, though these are not the only ones which give good results.

One problem with encoding is that some newer encoders use VBR (Variable Bit Rate) encoding - this may not be compatible with all MP3 players, though may give better results with compatible players.

Most MP3 decoders are very similar, though again there are slight differences. Partly this is because, apparently, the specification for decoding is tighter than for encoding, and there isn't too much to go wrong when implementors develop these. Most good MP3 decoders should sound very similar.

Winamp is an MP3 player which represents a reasonable reference standard, both for features and also for sound quality.

What I didn't realise is that most MP3 players now use at least 24 bit processing internally, and this can result in a 24 bit output which is applied to the DACs. If it's required to generate CDs from and MP3 source, this requires conversion back to 16 bit PCM. Some software simply truncates bits, which can give some slight extra distortion. Although this should not be too significant, better software allows dithering, which can subjectively give a closer approximation to the original 24 bit values. When generating WAV files, or burning audio CDs (not MP3 audio) from MP3s, for preference dithering should be applied to give better results. There are some converter tools which provide this facility.

I did try some tests myself on MP3 encoding, and I found that up to 192 kbps there were sections of a test CD which I could distinguish between low and high bit rate encodings - higher being preferable. Above 192 kbps personally I found it hard to be sure of any significant difference, though for preference I would always prefer the original source.

I also found some streaming audio radio stations, some of which are now putting out high quality audio streams. One is Radio Bartok from Hungary, which has an MP3 stream at 320 kbps. NRK Alltid Klassisk from Norway also puts out a high quality stream.

I am going to revise the posting shortly, to provide some references to justify the claims made here. I will also enhance the list of high bit rate streaming audio stations.

posted by David | 8:29 pm
I read over your blog, and i found it inquisitive, you may find My Blog interesting. My blog is just about my day to day life, as a park ranger. So please Click Here To Read My Blog
I read over your blog, and i found it inquisitive, you may find My Blog interesting. So please Click Here To Read My Blog

Get any Desired College Degree, In less then 2 weeks.

Call this number now 24 hours a day 7 days a week (413) 208-3069

Get these Degrees NOW!!!

"BA", "BSc", "MA", "MSc", "MBA", "PHD",

Get everything within 2 weeks.
100% verifiable, this is a real deal

Act now you owe it to your future.

(413) 208-3069 call now 24 hours a day, 7 days a week.
Post a Comment