My take is that Gladwell is post-peak and he knows it.
Give him a little bit of credit, at least try to address his arguments.
They're both rather accomplished bullshit-pop-sociology writers, but the real disagreement has a lot to do with their style. Anderson is like Tom Friedman and Ray Kurtzweil, in that he is a messianic This Is The Future pop philosopher-type, and tends to construct his argument around absolute, theoretical propositions, and asserts his case as if it were inescapable physical law. He writes like an Austrian Economist or a Straussian. He is a structuralist.
Gladwell is the skeptic. All of his books have mainly focused on picking-apart the assumptions of structuralists; you think entrepreneurs are the movers in an economy, he puts up 10 reasons why it isn't so simple. You think hockey teams always select the best roster of players? Outliers is about how social institutions are highly irrational in identifying the successful. You think people make rational decisions at all? He wrote a book called Blink where he calls it all into question.
I'm not saying he's right, it's just his MO. When he sees a book like Free, that makes Big Important Statements about How All Of Us Order Our Lives, it's bait to someone like him.
Without a quantitative analysis, or even an explanation of why stringed instruments don't sound the same, I'm not really prepared to agree. You have to control for mike selection and placement, instrument, performer, performance space, and preamp, and then the entire reproduction chain. A lot of consumer digital gear in this era has very very cheap signal paths compared to old analogue gear. The voltage rails on an iPhone's line amp are blah, compared to even an old Walkman, and the D/As are poor compared to an Apogee.
I'm not really sure what you're trying to say, do strings sound "better" A versus D, or do they sound more "correct"? Because a lot of subjective listeners say that X or Y sounds better in analogue versus digital, when the analogue recording was done by John Eargle with the LSO and the digital recording was done by Pete's Audio and Wedding Videos with the Bakersfield Light Opera Society Orch.
For the record, I personally think Electronic music suffers more in the digital domain, mainly because the DAs on most gear can't reproduce the proper low end. I was listening to Thriller recently, for no particular reason, and I swear that sounded better on vinyl (though the vinyl, in point of fact, might not have been an accurate reproduction of the master). A Moog is able to make noises that are very hard to record with any kind of equipment.
Re:Would you let it die already?
on
PHP 5.3 Released
·
· Score: 4, Insightful
I think if the functions, calling conventions and naming were regularized
...and they removed the sigils, and removed the silent type coercion, and eliminated the "@" function prefix, and removed the "global" keyword, and removed the weird "list" lvalue function that looks-like-functional-pattern-matching-but-really-just-cosmetically, and fixed the pass-by-reference semantic...
But if you did all of these things, it really wouldn't be PHP anymore.
I'm pretty sure the ENB for a noise floor (lets just say s/n ratio, to keep it simple) of 50db would be about 7.7, but even that is pretty optimistic for a compact cassette.
The reason that a compact cassette cosumes so many more (theoretical) bits is because, unlike an MP3, it isn't perceptually encoded, and the technology in fact is capable of carrying other kinds of non-acoustic data, like in the aforementioned C64 or PET computer drives. These, in fact, did about 300 baud and could only hold about 100KB per side, per the wiki.
That's basically what we went about proving wrong with a blind test using young music students. We succeeded. In my experience, humans are often far more amazing and capable of far more than we imagine...
It's important to make the point that people being able to tell the difference between analogue and digital is different from saying that analogue records a signal better than digital, or has greater fidelity to an input signal. In fact, people often prefer a recording of inferior fidelity if it "sounds good" in certain subjective ways (has a bigger bass, has sharp transients, etc.). The point we're trying to make is that you attribution of an analogue/digital discrimination based on the frequency passband of the respective media is not supported in principle. There may be any number of "tells" that clue a listener in to the format, and none of these have anything to do with the respective fidelity of the media.
I agree, I think you're failing at Nyquist, because the HF response of analogue cassette was nowhere near what a CD can do, and by the sampling theorem, a waveform under half the sampling freq is reproduced perfectly -- I hope you don't labor under the impression that the signal comes out of the speakers steppy, because it doesn't. I think the best high-end you could get out of a player was a 2 inch machine running at 30 ips, and even then you're only going to be able to squeeze 30-40 kHz bandwidth out of it, most of which is going to die in your amp and monitors. I will grant that induced dither has the effect of mushing up the high-end transients, particularly on earlier recordings that didn't used noise shaping, and this doesn't particularly cover strong-3rd-harmonic sustained-envelope sounds, like strings.
But in theory your issue could be addressed if you were to listen to some masters at 96k or 192k, or a 3 Mhz DSD record, and some people report these recordings having being "airy-er" or having "more space in between the instruments," and being able to describe the space in terms of which instruments are sitting in front of others, etc. But, a lot of this is more dependent on miking, and again, nobody has ever been able to get the "warmth" that a good phono or tape recording had. I think the consensus at this point is that it was induced distortion, and didn't actually reflect the signal inputting to the system.
Just curious, you didn't write up your experiment, did you? My old prof would probably be interested in reading.
Also, are you sure that the CDs were "worse" at reproduction on absolute terms, or that the analogue recordings simply induced distortions that you found pleasant, like tube-induced second harmonic distortion? It's almost impossible to do double-blind audio analysis with analogue v digital, because analogue always gives itself away with noise, and I've read that subjective listeners often cannot tell the difference between analogue and digital for most program material if the digital is noised up, or if needle pops are added, or if programs like string-heavy orchestral programs are given even-harmonic distortion.
I honestly wouldn't know how you'd get off saying a cassette has higher fidelity than a 128kbps MP3. The noise floor on a compact cassette was something like 40 dB below ref, only perhaps 10 dB at most of headroom, and the HF response over 8K was terrible. You'd need Dolby C or Dolby S to offset cassettes to the point where their noise floor character was equivalent to an MP3. Even though MP3s have more noticeable artifacting, what's an artifact on the cymbals compared to a Niagara falls-scale tape hiss?
The cassette doesn't have artifacts on sibilants, but then again it doesn't really have any HF at all, so if, for example, you were listening to a compact cassette recording of an MP3, the HF artifacts would probably go away, because the cassette simply doesn't have the response to capture those frequencies.
It is however possible, on both formats, for a loud sound like an kick drum hit to appear immediately before or after it actually is supposed to be heard, because the tape layers on the spool print through onto layers above them. When I used to do gun recordings with a Nagra 4-S you would always store the tapes "tails out" or FFwded to the end, so that any print through would sound after the actual sound, and would sound like an echo, rather than preceding the sound and ruining the attack.
I suppose I could make some joke about how Soviet Russia fell because they tried to pacify the public with a Sergei Eisenstein marathon and a special once-in-a-lifetime uninterrupted screening of Dziga Vertov's Tchelovek Skinoapparatom. But what are we to do when totalitarians have access to high-quality Hollywood content!
They've probably re-dubbed the movie in Farsi to make Sauron the good guy. This has happened before; when Mr. Smith Goes to Washington was screened in Eastern Europe, it was only in a dubbed version that made Taylor and Senator Paine into capitalist oligarchs, with Smith fighting for his worker's soviet back home.
I have no gripe with the SRBs, I simply found the rather sterile language in the vein of "working out bugs" and "real operating parameters" seemed to paper over the practical consequences of certain engineering decisions.
It wouldn't need an XLR connector, just something with three conductors. Balanced wiring on the chassis is sorta minimum standard requirements for professional gear. It's not my biggest complaint though, the lack of SPDIF or AES input or timecode are the dealbreakers.
No, but they can tell you who to shoot, and have. Information technology is really the most acute munition of war imaginable, particularly in internal warfare between a government and its people.
With regard to striping: This has actually come up in an experiment, and mpeg-compressed audio simply doesn't capture SMPTE timecode. It smears out the data words and you just end up with a tonal mush. Also, you can't insert edit on these cameras, so you can't pre-stripe or do record-run; you can only do free run.
So, even if you can get the recording to work, you'll need about 5 seconds of leader before a slate at the head of the take for the synchronizer in the online bay to rechase, which is a finicky proposition. And now you need an online bay to resync your dailies, even for rough cut work, whereas with a real camera you can just load all the media from the camera and the Fostex (or Deva or whatever) into the Avid or FCP and it'll relink by timecodes automagically.
Touché. We don't allow citizens and companies to enjoy the right of selling arms to Iran.
Our government has, from time to time however, allowed arms to fall off the back of the truck if it got us something in return. These actions were very controversial at the time, were done completely in secret, are still (ridiculously enough) denied by the people who did it, had about zero popular support, and got a few people sent to jail.
Just throwing this in: a friend if mine was a DP on the last season of CSI and he'd demoed the 5dmk2 to see if it was usable for what they were doing. His opinion was that the image looked great, but it NEEDED a 24P mode (it only shoots 30fps nominal right now). Also, it only records MPEG-compressed movies, and there's no way to get a raw feed off it or at least something with less/more pro levels of compression.
I'm a sound guy, so my main complaints are that it doesn't have balanced wiring for the audio inputs, there is no digital audio input, and it's not clear how well the camera will hold sync if you were to do double-system recording, since the camera has no genlock/trilevel sync.
Now Apple could make updating Applications easier, as it's hidden as a small button in the iTunes Applications panel.
It's hard to miss when your third-party iPhone apps are ready to upgrade, the App Store app gets a big red badge telling you a new version of SuperFoo is available.
The memory point struck me too. BBs are the defacto standard for text messaging, but many of them only have 32 or 64 megs of RAM and maybe a Gig of onboard storage (though with an SD slot). Only the Storm begins to compare to something like the Pre or a 3G (non S) iPhone, and then marginally. I'm not saying BBs are bad, it's just clear how quickly the whole smartphone market has gotten out from under BB, and really how special-purpose their devices are compared to this next gen of "media" smartphones.
Slashdot Mirror
I wasn't aware Norman Mailer read slashdot.
My take is that Gladwell is post-peak and he knows it.
Give him a little bit of credit, at least try to address his arguments.
They're both rather accomplished bullshit-pop-sociology writers, but the real disagreement has a lot to do with their style. Anderson is like Tom Friedman and Ray Kurtzweil, in that he is a messianic This Is The Future pop philosopher-type, and tends to construct his argument around absolute, theoretical propositions, and asserts his case as if it were inescapable physical law. He writes like an Austrian Economist or a Straussian. He is a structuralist.
Gladwell is the skeptic. All of his books have mainly focused on picking-apart the assumptions of structuralists; you think entrepreneurs are the movers in an economy, he puts up 10 reasons why it isn't so simple. You think hockey teams always select the best roster of players? Outliers is about how social institutions are highly irrational in identifying the successful. You think people make rational decisions at all? He wrote a book called Blink where he calls it all into question.
I'm not saying he's right, it's just his MO. When he sees a book like Free, that makes Big Important Statements about How All Of Us Order Our Lives, it's bait to someone like him.
Maybe we can agree that analogue is always better for our favorite kind of music. But that's sortof the problem.
Without a quantitative analysis, or even an explanation of why stringed instruments don't sound the same, I'm not really prepared to agree. You have to control for mike selection and placement, instrument, performer, performance space, and preamp, and then the entire reproduction chain. A lot of consumer digital gear in this era has very very cheap signal paths compared to old analogue gear. The voltage rails on an iPhone's line amp are blah, compared to even an old Walkman, and the D/As are poor compared to an Apogee.
I'm not really sure what you're trying to say, do strings sound "better" A versus D, or do they sound more "correct"? Because a lot of subjective listeners say that X or Y sounds better in analogue versus digital, when the analogue recording was done by John Eargle with the LSO and the digital recording was done by Pete's Audio and Wedding Videos with the Bakersfield Light Opera Society Orch.
For the record, I personally think Electronic music suffers more in the digital domain, mainly because the DAs on most gear can't reproduce the proper low end. I was listening to Thriller recently, for no particular reason, and I swear that sounded better on vinyl (though the vinyl, in point of fact, might not have been an accurate reproduction of the master). A Moog is able to make noises that are very hard to record with any kind of equipment.
I think if the functions, calling conventions and naming were regularized
...and they removed the sigils, and removed the silent type coercion, and eliminated the "@" function prefix, and removed the "global" keyword, and removed the weird "list" lvalue function that looks-like-functional-pattern-matching-but-really-just-cosmetically, and fixed the pass-by-reference semantic...
But if you did all of these things, it really wouldn't be PHP anymore.
GIVE ME THE OVERTHRUSTER!
Yeah, the wheelchair can read minds, but can it flash a light to indicate "yes" or "no"?
I'm pretty sure the ENB for a noise floor (lets just say s/n ratio, to keep it simple) of 50db would be about 7.7, but even that is pretty optimistic for a compact cassette.
The reason that a compact cassette cosumes so many more (theoretical) bits is because, unlike an MP3, it isn't perceptually encoded, and the technology in fact is capable of carrying other kinds of non-acoustic data, like in the aforementioned C64 or PET computer drives. These, in fact, did about 300 baud and could only hold about 100KB per side, per the wiki.
That's basically what we went about proving wrong with a blind test using young music students. We succeeded. In my experience, humans are often far more amazing and capable of far more than we imagine...
It's important to make the point that people being able to tell the difference between analogue and digital is different from saying that analogue records a signal better than digital, or has greater fidelity to an input signal. In fact, people often prefer a recording of inferior fidelity if it "sounds good" in certain subjective ways (has a bigger bass, has sharp transients, etc.). The point we're trying to make is that you attribution of an analogue/digital discrimination based on the frequency passband of the respective media is not supported in principle. There may be any number of "tells" that clue a listener in to the format, and none of these have anything to do with the respective fidelity of the media.
I agree, I think you're failing at Nyquist, because the HF response of analogue cassette was nowhere near what a CD can do, and by the sampling theorem, a waveform under half the sampling freq is reproduced perfectly -- I hope you don't labor under the impression that the signal comes out of the speakers steppy, because it doesn't. I think the best high-end you could get out of a player was a 2 inch machine running at 30 ips, and even then you're only going to be able to squeeze 30-40 kHz bandwidth out of it, most of which is going to die in your amp and monitors. I will grant that induced dither has the effect of mushing up the high-end transients, particularly on earlier recordings that didn't used noise shaping, and this doesn't particularly cover strong-3rd-harmonic sustained-envelope sounds, like strings.
But in theory your issue could be addressed if you were to listen to some masters at 96k or 192k, or a 3 Mhz DSD record, and some people report these recordings having being "airy-er" or having "more space in between the instruments," and being able to describe the space in terms of which instruments are sitting in front of others, etc. But, a lot of this is more dependent on miking, and again, nobody has ever been able to get the "warmth" that a good phono or tape recording had. I think the consensus at this point is that it was induced distortion, and didn't actually reflect the signal inputting to the system.
Just curious, you didn't write up your experiment, did you? My old prof would probably be interested in reading.
Also, are you sure that the CDs were "worse" at reproduction on absolute terms, or that the analogue recordings simply induced distortions that you found pleasant, like tube-induced second harmonic distortion? It's almost impossible to do double-blind audio analysis with analogue v digital, because analogue always gives itself away with noise, and I've read that subjective listeners often cannot tell the difference between analogue and digital for most program material if the digital is noised up, or if needle pops are added, or if programs like string-heavy orchestral programs are given even-harmonic distortion.
The cassette doesn't have artifacts on sibilants, but then again it doesn't really have any HF at all, so if, for example, you were listening to a compact cassette recording of an MP3, the HF artifacts would probably go away, because the cassette simply doesn't have the response to capture those frequencies.
It is however possible, on both formats, for a loud sound like an kick drum hit to appear immediately before or after it actually is supposed to be heard, because the tape layers on the spool print through onto layers above them. When I used to do gun recordings with a Nagra 4-S you would always store the tapes "tails out" or FFwded to the end, so that any print through would sound after the actual sound, and would sound like an echo, rather than preceding the sound and ruining the attack.
I suppose I could make some joke about how Soviet Russia fell because they tried to pacify the public with a Sergei Eisenstein marathon and a special once-in-a-lifetime uninterrupted screening of Dziga Vertov's Tchelovek Skinoapparatom. But what are we to do when totalitarians have access to high-quality Hollywood content!
They've probably re-dubbed the movie in Farsi to make Sauron the good guy. This has happened before; when Mr. Smith Goes to Washington was screened in Eastern Europe, it was only in a dubbed version that made Taylor and Senator Paine into capitalist oligarchs, with Smith fighting for his worker's soviet back home.
An AC writing like an old codger. I think I bought my first Java book in 1995... By this standard of "fad," C was a fad when Windows NT came out.
I have no gripe with the SRBs, I simply found the rather sterile language in the vein of "working out bugs" and "real operating parameters" seemed to paper over the practical consequences of certain engineering decisions.
There's a VERY good reason to re-use the SRB's, they are a well tested design with the flaws worked out and the real operating parameters known.
We would be remiss if we did not note that the engineering kinks of the SRBs have been ironed out, by that they killed seven people in the process.
It wouldn't need an XLR connector, just something with three conductors. Balanced wiring on the chassis is sorta minimum standard requirements for professional gear. It's not my biggest complaint though, the lack of SPDIF or AES input or timecode are the dealbreakers.
No, but they can tell you who to shoot, and have. Information technology is really the most acute munition of war imaginable, particularly in internal warfare between a government and its people.
With regard to striping: This has actually come up in an experiment, and mpeg-compressed audio simply doesn't capture SMPTE timecode. It smears out the data words and you just end up with a tonal mush. Also, you can't insert edit on these cameras, so you can't pre-stripe or do record-run; you can only do free run.
So, even if you can get the recording to work, you'll need about 5 seconds of leader before a slate at the head of the take for the synchronizer in the online bay to rechase, which is a finicky proposition. And now you need an online bay to resync your dailies, even for rough cut work, whereas with a real camera you can just load all the media from the camera and the Fostex (or Deva or whatever) into the Avid or FCP and it'll relink by timecodes automagically.
Touché. We don't allow citizens and companies to enjoy the right of selling arms to Iran.
Our government has, from time to time however, allowed arms to fall off the back of the truck if it got us something in return. These actions were very controversial at the time, were done completely in secret, are still (ridiculously enough) denied by the people who did it, had about zero popular support, and got a few people sent to jail.
These companies sold network equipment.
Guns don't kill people, people kill people.
But we still don't sell guns to Iran.
...or any conception of timecode, for that matter.
Just throwing this in: a friend if mine was a DP on the last season of CSI and he'd demoed the 5dmk2 to see if it was usable for what they were doing. His opinion was that the image looked great, but it NEEDED a 24P mode (it only shoots 30fps nominal right now). Also, it only records MPEG-compressed movies, and there's no way to get a raw feed off it or at least something with less/more pro levels of compression.
I'm a sound guy, so my main complaints are that it doesn't have balanced wiring for the audio inputs, there is no digital audio input, and it's not clear how well the camera will hold sync if you were to do double-system recording, since the camera has no genlock/trilevel sync.
Now Apple could make updating Applications easier, as it's hidden as a small button in the iTunes Applications panel.
It's hard to miss when your third-party iPhone apps are ready to upgrade, the App Store app gets a big red badge telling you a new version of SuperFoo is available.
The memory point struck me too. BBs are the defacto standard for text messaging, but many of them only have 32 or 64 megs of RAM and maybe a Gig of onboard storage (though with an SD slot). Only the Storm begins to compare to something like the Pre or a 3G (non S) iPhone, and then marginally. I'm not saying BBs are bad, it's just clear how quickly the whole smartphone market has gotten out from under BB, and really how special-purpose their devices are compared to this next gen of "media" smartphones.