Wednesday, 23 May 2007

f - week 9 - construction/deconstruction

Another week another forum.

First up, the magic formula for a hit song... as Freddy read from Michael Stavrou, get a big pile of songs, compare and work out what they all have compared to non-hits... magic ...
Hit songs my arse, I don't write songs so I don't care (I compose pieces :).

Next up, Dragos played some tunes and pointed out various techniques of construction. Techno - start with a sound, add another etc.... at some point pull some out, then end - or something like that.

Thirdly, Matt Mazone presented an add he'd done sound/music for and how he'd constructed it. The most interesting of the lot. And that's all I've got to say about it.

I think I'm getting quite bored of forum.. it's really quite uninteresting. ho hum.
oh well, I'm sure there's worse things to do on a thursday.

One thing I'm getting quite annoyed at, is low quality mp3's being played at us - what's the story ?
Although I do enjoy myself listening to the fluttering around the hi hats and other constant trebly sounds - one day I'll work out what the encoding is doing and then I'll be happy :)

[1] “Music Technology Forum – Week 9 – Construction/Deconstuction”. University of Adelaide, Schultz building. 10 May 2007.
presented by Freddy May, Dragos Nastasi, Matt Mazone.


Freddie said...


John Delany said...

QUOTE: "One thing I'm getting quite annoyed at, is low quality mp3's being played at us - what's the story ?"

You'll be pleased to know I presented my audio samples this week (Week 11) as uncompressed WAV files in all their 16-bit, 44.1K glory :-) I don't mind mp3 too much, but if I can avoid it, I will.

Freddie said...

Whether it's an MP3, a CD or DVD it's all compressed. It's just a matter of how much. Besides, the dynamic range to begin with determines how "low quality" it is. Try writing some songs that utilize the full dynamic range of human hearing and then start complaining. All mainstream songs really only have about 6dB of DR anyway. Also the conversion from analogue to digital has a huge contributing factor of the quality. My samples were from the internet off some random site. If I could actually get those songs from iTunes, it would be a different story. Some people have no idea how to utilise the digital medium and think it's just a case of pressing a button. Some people have no idea between constant and variable compression. If I was the one that converted those songs to MP3 I would challenge you to be able to tell the difference.

John Delany said...
This comment has been removed by the author.
giant said...

it's the "poor" quality mp3's i object, noticeable unpleasant artifacts that git in the way of my listnn pleasure.

makes me wonder when random downloading of illegal music will be classified as plagiarism :P

John Delany said...

Any digital audio is of course an approximation.... and when referring to compression, we usually mean lossy data compression (ie. mp3, aac), as opposed to straight digitisation.

You mention creating recordings that utilise the full dynamic range of human hearing before complaining... Bit Depth doesn't only determine theoretical DR, (particularly when you utilise quality dithering algorithms and so forth) but also has a bearing on the amplitude resolution. Why then would 24-bit audio (with a theoretical DR of 144dB) be invented, if this is beyond our hearing range? I think it would be difficult to create a system with accurate playback DR of a full 144dB.

Secondly, you are totally right, an analogue to digital conversion will always create a theoretical compromise until we have infinite Sample Rate and BD.

Thirdly, I have spent way too much time over the years experimenting with audio and video codecs, and a wide range of applications, being the geek that I am. In listening tests on my Sennheisers, I find 128kbps MP3 to exhibit a level of phase-issues and spatial-issues which are eradicated at 160. I find 192 to provide a very close approximation to the original "uncompressed" digital audio for a lossy product. Anything above that however, is of a very marginal difference. I normally use the LAME mp3 encoder (with CDex), as opposed to Xing for example or RealNetworks codecs. I think VBR is not necessary for audio, more-so for video, where you are seeking the best trade-off between filesize vs. quality. Having said all that, if I can use uncompressed audio, I will, even if only to give me warm fuzzies. I'm sure you would do the same right?

Freddie said...

plagiarism is defined as taking someone elses work and claiming it as your own, which no one has done with any music thus far. Someone would look pretty stupid if they tried to take credit for writing a Metallica or a Beatles song.

Freddie said...

My comment regarding utilising full dynamic range was in referrence to say, classical music or music that is wildy dynamic. The tops are the first thing to get ruined, but some MP3's sound good and it is difficult to tell them apart from CD audio.
Nobody seems to care about MP3's sounding crap. The price (usually free) is more appealing. I think this is a complete shame as I love listening to good sounding recordings and until the general public feels the same way, we will continue to suffer MP3's (or MP4's or whatever is on the horizon). With all the 'hysteria' over MP3's it's a miracle we will even get BluRay music discs. Did you know that record companies are now making specific iPod mixes of their albums for sale. MP3's are unfortunately here to stay.

Of course BD doesn't just affect DR, but the BD does get affected by using up the msb with effects, automation etc. We certainly shouldn't be recording with 2 bits just because most music doesn't go more than 12db DR. Yes, I know about the thermal noise and the restrictions on 24 bit and the DR, and until that gets sorted we will continue to listen to compressed audio. This has all become a little off track, but, meh. *shrugs*

Ben said...

I find the step down from 192 to 160kBps MP3s to be noticeable, but anything above 192 has such a small difference it is negligible during general listening. Regarding 128, I liken the 'shitty' sound to badly compressed jpeg files. Now there's and idea for a sound installation- low quality jpeg pictures with equivalently low quality MP3s accompanying them. Sounds academic, no?

I have never done a BD comparison, and I wonder if it would even be worthwhile. I (erroneously?) doubt the theoretical dynamic range/resolution 'artifacts' caused by a 16bit BD would be perceivable when compared to a 24bit BD.

While we're talking like nerds, let's drop the acronyms- "One Hundred and Twenty Eight kilobytes per second Moving Picture Experts Group One Layer Three".