- Sauvik Biswas - https://sauvikbiswas.com -

Listening habits: Sound quality vs. convenience of portability

Long ago I had written why I prefer listening to whole albums [1]. My approach to listening hasn’t changed much. However, I am from a lesser group of the population. The majority consumes music in a very different way.

Convenience of portability

The idea of revisiting this topic stems from my listening to lectures by Andrew Scheps [2]. You can listen to one of his presentations here –

The talk is quite technical in nature. It talks about the various streaming services and the audio quality they deliver. It is surprising that most of our discoveries happen on 128 kbps compressed audio since YouTube, the largest online music repository, has 128 kbps audio stream for anything that’s not 720p or 1080p[1].

Let me offer a personal anecdote. A couple of years ago, a group decided to include a song from In Human‘s Voices [3][2]. I had engineered and mixed this album on really budget equipment but had put a lot of heart and soul into its quality. Sadly, the guy who mastered the album butchered the dynamics of the song. To top it all, the entire mix-tape was distributed via 128 kbps mp3 files. You can read more about it here [4].

The popularity of low quality, mp3 files can be traced back to the file sharing days of Kazaa and Napster, when bandwidth was a real issue. Nowadays, it’s not in most of the cases – especially the developed countries and urban population of developing countries. Yet it still baffles me as to why would anyone want to listen to these distorted streams of audio. An artist would never want a listener to listen to a badly reconstructed version of a song that he had put so much effort into.

A beautiful short documentary called The Distortion of SoundĀ  puts the artists’ viewpoints into perspective .

I call this the convenience of portability. To a person who wants to carry his entire library in a 64 GB memory card, such lossy compressions are godsend. To them it is necessary to carry as much as possible rather than carry as best as possible. Let me remind you that it is possible to carry about 150-180 full-length albums[3] a level 5 compressed FLAC (a lossless compression) in the same space, which I believe is more than enough music for carrying around, even for an extended period.

The term portability can also be applied to porting or copying of music from one place to another via a network. For early file sharing programs, this was important and that’s why lossy codecs like mp3 were developed in the first place. In a developing country like India, internet bandwith is costly. Not everyone can afford anything beyond 300 MB mobile 3G data for about 30 days. Most Indian online radio streaming services likeĀ  Saavn [5]offer 16 to 128 kbps streams. A competing service like Gaana [6]offers a HD version but that has to be paid for. Why would anyone listen to 16 kbps stream is beyond my comprehension. Even FM radio has better quality than that although the signal is compressed a lot prior to airing.

A small test on a sample clip

Since Mike Shinoda was one of the vocal opposer of this distortion, I decided to run some test on one of their songs[4].

128 kbps MP3 VBR

http://sauvikbiswas.com/wp-content/uploads/2015/01/LP-chunk-128.mp3 [7]

64 kbps MP3 VBR

http://sauvikbiswas.com/wp-content/uploads/2015/01/LP-chunk-64.mp3 [8]

16 kbps MP3 CBR

http://sauvikbiswas.com/wp-content/uploads/2015/01/LP-chunk-16.mp3 [9]

Now here’s the uncompressed source from audio CD. This one is about 3 MB in size as compared to the 200 kB file that the 128 kbps compression yielded.

http://sauvikbiswas.com/wp-content/uploads/2015/01/LP-chunk-unc.wav [10]

The first thing one will notice is the fact that the uncompressed signal has quite a bit of bottom and top that was stripped off in the 128 kbps encode. Psychoacoustically, it sounds a tad bit sharper due to absence of bottom end. On the other hand the 16 kbps encode doesn’t even sound like the original song.

My concluding 2 paise

I think it is as a reluctance to invest some thing in order to enjoy music. That investment may be monetary as in the form of equipment and better source. It may be time, the willingness to devote undivided attention to a piece of work rather than relegate it to background noise. Or it could be just an emotional investment – the willingness to accept that what one is listening to was not the intent of the artist, even though they might lack time or money.

I can’t blame a poor guy trying his best to listen to a streaming radio on a pair of weak, mobile mounted speakers. However, I can’t say the same for well to do people who prefer not to even pay for their music. I personally know a lot of people who fall into the latter category.

Notes:

  1. YouTube streams at 128 kbps AAC for 480p or less video and 384 kbps for 720p and higher res. video (read here [11]). 128 kbps AAC is roughly equivalent to 192 kbps mp3s sonically.
  2. The Reverbnation link streams 128 kbps mp3s.
  3. An audio CD can be compressed to about 350 MB of FLAC. 64*1024 MB / 350 MB/album ~ 187 albums
  4. Test specifications: The song was ripped from the original CD and was sliced in Reaper DAW. The encodes were direct renders from the DAW using the LAME mp3 encoder.
Listening habits: Finite resource for infinite music [12]
CN Remix: A good intent ruined by (possibly) the worst presentation [13]