George Massenburg lecture in Sydney 2009

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

deuce42

Well-known member
Joined
Dec 31, 2008
Messages
645
Location
Sydney, Australia
This talk by George M made me really think alot. He is a truly an insightful guy and a doyen on the recording industry. Doesn't even need to be said does it:)

http://www.youtube.com/watch?v=wzJbjHc6bRE
 
A bit Ironic that it's totally distorted and very badly edited !!  ( or is that the point ?? )

I would have liked to have watched a better quality and better edited version, what's left just
makes George sound like a very negative and grumpy old man, who's stuck in 1982.


MM.
 
deuce42 said:
This talk by George M made me really think alot. He is a truly an insightful guy and a doyen on the recording industry. Doesn't even need to be said does it:)

http://www.youtube.com/watch?v=wzJbjHc6bRE

Not to mention, with the bad editing, you don't get to hear what the iTunes artifacts sounded like.  I was curious, so here are the differences in 44k/32bit WAV format, no dithering.  Just the straight difference.  (I slid the reference track a sample at a time to attain maximum phase cancellation)

I used a decidedly less acoustic track for my test than he did though.  Since he was saving iTunes for last, I thought it would be a lot worse.  In my tests, MP3 and ITunes scored similarly on the "suck meter".

iTunes v7.6.1.9 256kbps artifacts  (25 meg)

and for comparison:

MP3 320kbps (Sony 2.0 encoder) artifacts  (25 meg)

It's sad when you can still recognize the song by what's been thrown away.







 
did he say anything about FLAC ?

thansk for posting that [ shocking and disgusting ]
haven't that pat lenord / gilbert stuff for ahwile
 
In fairness to George the recording and edit of the lecture was just done by the people who ran the event. He probably had no say in how it was shot or edited.

I personally feel the salient points were that the quality of artist themselves is often being overlooked and may have been a priority in yesteryear. Maybe he is stuck in 1982 but sometimes its important to listen to people raise objections to make us question. Made me think anyway:)
 
Oh, don't get me wrong, he is probably "bang on" but the edit's in mid sentence are just a bit much
and so much is missing from his lecture.
I hate mp3's even though I've had an ipod since the generation 1 10gig !
BUT  .... so many people use it that we just can't get away from it ...... progress !!!

The only compression rate worth using is 320 kbps but I prefer a full 16/44.1 wav / aif file anyday.


Marty.
 
I don't think he's grumpy and negative.

He is passionate and opinionated because he really cares about the subject.

And he's spot on about Ms Spears. I had the displeasure of attending a show on her 2004 tour and it was a memorable event  :eek:

But hey, most people can't be bothered so I guess we have the audio quality that the vast majority of listeners think is "fine and dandy".


 
I'm grumpy and negative because mp3s sound terrible and hurt my ears. I feel like we've gone 10 years in the wrong direction. With very small hand held devices holding 30GB of data, I just want to be able to sling 24/48k on there and listen away. Think about that... 30GB is probably 30 albums worth of music. I'd rather have 30 albums of good quality over 500 albums of distortion.

Perhaps the processors aren't fast enough; I'm not really sure. But I don't feel like there's enough push from the consumer to demand higher quality audio.

The first time I saw a $0.99 digital song purchase I was stoked... until I realized it was an mp3 and almost puked at the idea of purchasing an mp3.
 
You can use the apple lossless codec in all iPods, so at least you can convert all your albums to that. Too bad they don't actually sell music in that format.

For non-iPod people, some of the better mpr3 players support FLAC.

It's not like anyone is forcing you to listen to mp3. Not yet anyway. Too bad more than 99% of all people don't hear or know the difference, and the situation is not going to get any better. Consumers voted and high resolution DVD-audio and SACD are dead already.
 
I'm not sure that showing that there are differences between an original signal and one which has been encoded is proving anything. From my understanding the goal of perceptual coding is to throw out information that is masked by other sounds. If there were no difference between an encoded and unprocessed signal then there would be something to worry about - an algorithm that isn't actually doing anything to your signal!

Maybe some believe that the goal of encoding is simply to save storage space but I don't believe that is the case.
 
mnats said:
I'm not sure that showing that there are differences between an original signal and one which has been encoded is proving anything. From my understanding the goal of perceptual coding is to throw out information that is masked by other sounds. If there were no difference between an encoded and unprocessed signal then there would be something to worry about - an algorithm that isn't actually doing anything to your signal!

Maybe some believe that the goal of encoding is simply to save storage space but I don't believe that is the case.

I agree, it's also a bit easier to email a client a 5 MB file than a 40 MB file !!
hmm .... where does that 35MB go ??

MM.
 
mnats said:
Maybe some believe that the goal of encoding is simply to save storage space but I don't believe that is the case.

Whoah! This isn't a question of belief! It really is the fundamental goal of these algorithms to conserve storage space or data bandwidth (same thing). It is also an equally important goal to keep this process as transparent as possible.

The data isn't thrown away anywhere, not even in algorithms that take into account frequency masking. All that "35mb" is still contained within an algorithm, as a set of mathematical rules. The best case scenario is that a very short string of rules can be expanded to a perfect copy of lots of data (takes a lot more storage space than the simple string) that was once compressed into this set of rules. The great thing about lossy coding is that we can relax the amount (depth or accuracy) of these rules to conserve even more bandwidth. That's of course where the artifacts (bad sound) set in.
 

Latest posts

Back
Top