Why Steve Albini Still Prefers Analog Tape 440
CNET's Steve Guttenberg ("The Audiophiliac") profiles prolific audio engineer and general music industry do-it-all Steve Albini; Albini (who's worked on literally thousands of albums with musicians across a wide range of genres) has interesting things to say about compression, the rise of home-recording ("The majority of recordings will be crappy, low-quality recordings, but there will always be work for engineers who can do a good job, because there will always be people who appreciate good sound."), and why he still prefers to record to analog tape. (Note: Albini is justly famous not just for his production work, but in particular for his essay "The Problem with Music.")
Re:how can you not play an audio file? (Score:5, Interesting)
How will a regular musician know if the format or encoding is common enough to have decoders in the future?
Perhaps in the same way that VXA [mit.edu], for example, allows you to future-proof compressed archives?
Re:how can you not play an audio file? (Score:5, Interesting)
Analog master tapes normally have extensive printed notes on their label, about things like the speed used and which tracks are in what location. Digital files need similar documentation on things like format used. Studio masters being made by the musician shouldn't have any DRM silliness to deal with.
The main challenge for digital audio preservation is that all audio tracks need to be exported into simple PCM files. I would agree that some common studio digital formats will not be readable in the future. That means the musicians need to get .wav files instead of things like ProTools files. But saying properly exported and documented digital is fragile compared to analog tape is ridiculous. I expect to be able to read PCM files saved onto current CD and DVD media for at least another 50 years, while it's already hard to get good quality tape playback.
Re:No Analog is not better... (Score:5, Interesting)
This is so fraught with unsubstantiated nonsense it makes me ill. Which resolution are you referring to, sample rate or bit depth? Do you know what those actually do?
Sample rate: a higher sample rate allows for higher frequency representation. As in, if you have a sample rate of 48,000Hz, you can play back a frequency of 24,000Hz (already above the range of human perception). Higher sample rate = more high frequencies you can't hear.
Bit depth: higher bit depth increases dynamic range (think: ability to represent *quieter* sounds), and reduces quantization error (white noise). A 24bit CD has a 144dB dynamic range and 1/33,554,432th of the signal will be noise. Even 16 bit (CD) has 96dB range and 1/131,072th noise. Going higher won't make anything sound better.
It is MUCH more important that you have a guitar amp that doesn't buzz, a drum kit that's been tuned well, and a singer that can actually sing than it is to get 'super digital quality.' But please, continue to believe in nonsense numbers like 500MB per song.
Re:No Analog is not better... (Score:2, Interesting)
Anti-aliasing filters were a real problem when ADCs sampled at the intended data rate, such as 44.1 KSPS. You had to have something that passed everything up to 18 KHz or so, but was a brick wall at 22 KHz, with no ripples or phase distortion. That was a problem and bad recordings got made.
But ADCs don't work at the sample rate any more - they work many, many times faster which means the analog anti-aliasing filter is trivially simple. Then, before the data even leaves the chip, it goes through a digital low pass filter which constrains the bandwidth to what the output sample rate can support, with far fewer distortions than were possible in the analog realm. That filter is typically built right into the process of decimating the sample rate to the 44.1 or 96 or 192 KSPS output rate.
Re:Why Analogue? Stranded investment. (Score:5, Interesting)
I can credibly claim to be in the "golden ear" crowd as a former high-end audio reviewer [soundstage.com]. You need functional ears, but that's more about training than anything else. The better reviewers have spent years of their life carefully listening to different equipment and music, trying to become good at hearing small differences.
There are a few small tricks people usually fall for that good listeners try to get a handle on:
The last one is the most insidious, and I have an anecdote on how deep that goes. When 24/96K digital was first being released for studio use, I sat through a single-blind demo room at an AES show. They played an excellent analog master jazz recording, a version sampled at 24/96, and a version at 16/44.1 CD resolution downsampled via their equipment. I correctly graded the three from better to worse, seemed pretty obvious to me even though the high res digital was very close to the original.
As the presenter worked the room asking people which of his samples A, B, and C were, it was obvious mine was not the majority opinion. There were a few vocal people expressing their opinion that got things completely backwards. They thought the CD quality version was the "best", and therefore it had to be the original master. As this was an AES show, these were people who worked with audio all day, and their preference didn't match reality as I heard it at all.
Listening to their (incorrect) arguments for why they made their decisions, I realized they liked CD quality and its limitations. There was some compression to the CD version and a bit of a fuzzy/harsh roll off at the top end. But it was what they were used to. They thought recordings were supposed to sound that way, because most recordings they listened to did. You can see "familiar is better" in every generation of listener. People who grew up on vinyl like surface noise, early CD listeners are used to terrible aliasing filters, and people who grew up with low rate MP3s like their artifacts. And on the studio side, there are people who like the way analog tape sounds. To be fair, that was better than any digital available until very recently. Recent remasters of old analog recordings are still digging out details you couldn't hear in the earlier digital transfers. I think current generation 192/24 bit digital equipment is more than good enough to replace analog tape though; we passed that point a few years ago.
Re:how can you not play an audio file? (Score:4, Interesting)
its not like the old days when devices were dumb and we had new physical formats for every music generation.
The "having new physical formats" is a relatively recent thing. From 1894 for the next hundred years phonographs changed little, and it was always backwards-compatible. When it changed from 78 RPM to 33.3 and 45, newer players would still play the old 78s. When stereo was introduced the new stereo records would play on old monophonic players with both channels playing through its one speaker. The design was engineered that way. A monophonic record had the up and down motions translated to sound, while a stereo record had both channels in the up and down motion and a single channel in the sideways motion, which combined with the up and down signal filtered that channel out through destructive interference.
With cassettes and 8 tracks (I never had an 8 track, I was using cassettes before 8 tracks were popular) most people recorded the record the first time they played it so they could hear it in the car. My old '02 has both cassette and CD. It was probably 1995 before I had a CD player. And a turntable bought today will play records from 1894.
Re:how can you not play an audio file? (Score:4, Interesting)
It's quite hard to find a player for wax cylinders. And if you find an old one, it likely won't play celluloid cylinders as they've shrunken enough over time to be a problem.
The broadly successful digital recording formats of today will be easily playable by players in 100 years as well. The secret to a "future-proof" format is mass use of that format, not analog-vs-digital, or open-vs-closed.
Re:"Digital recordings will be unplayable" (Score:2, Interesting)
Albini has no idea what he's talking about.
Considering that he's an engineer with decades of experience, that seems unlikely. You remind me of a 19 year old classmate in college who questioned the professor's knowledge of the subject, who told the kid "Son, I've forgotten more than you ever learned."
He's obviously done the math. Can you tell a 15kHz sine wave from a 15kHz sawtooth wave? A CD can't, because there are only three samples per crest and almost every teenager can easily hear 15kHz.
I do fault more modern engineers, though. Why does the Led Zeppelin "Presence" LP have more dynamics than the CD does? CDs have a greater dynamic range so someone must have screwed up the remix. Yes, I was disappointed when my brand new CD didn't sound as good as the old LP of the same album. Boston's was so bad the musicians complained. Zappa refused to release a digital version of their Fillmore East album until his deathbed because the quality wasn't good enough for him.
don't trust Albini (Score:5, Interesting)
Albini is not a trustworthy opinion on this stuff...
He is a disciplined **studio engineer** but he only worked on *one* national release album after In Utero...because he's actually kind of a douchenozzle...
Here's a post from another thread that gets into detail [slashdot.org] that I wrote...
The point is, I don't trust technical opinions from people who can't see beyond their own expertise...
Here's a recent interview he gave: http://vishkhanna.com/2013/08/16/ep-24-steve-albini/ [vishkhanna.com]
He's the bad kind of luddite audiophile...the guy who understand waveforms and shit but really just likes to thrown around their expertise b/c it gives them social power...they always hear things that are 'obvious' that no one hears adn they love it...
Also, this caught my eye in your comment:
and *you* remind me of the old, lazy tenure Prof. who teaches a course on tech business but can't check his own email...
that scenario you present is a common trope of human behavior...just as often that 19 yr old college classmate drops out and starts their own company...
in my experience teaching HCI at WSU-Vancouver I never encountered a scenario like you describe...sure I had 'know it all's' who like to hear themselves be smart...but my job as an educator is to focus that into productive work...
a good prof doesn't need to bring out credentials to sit a sophomoric undergrad down