> Date: Mon, 15 Jan 2001 23:33:02 -0500
> From: Anthony Lalande <[EMAIL PROTECTED]>
> While I'm not sure what exactly AGC means, I have heard somewhere
> that ears
> are able to focus and lose focus of sounds and noise in a dynamic
> way.
> Does anyone know enough about the working of the human ear to validate or
> deny this?
This is exactly what "psycho-acoustic" *should* mean. It is not a
physiological process, but a psychological/cognitive one. Your ears receive
a full spectrum of sound, your subconscious mind filters it a bit, and your
conscious mind filters it down to what you're interested in listening to.
Total aside: What we call "perceptual coding" and "psycho-acoustics" is
completely misnamed; the masking phenomena being exploited here are
physiological/neurological, not psychological/cognitive. Again with a visual
example: if you are standing in broad daylight and turn on a flashlight, and
aim it at something that is already fully illuminated, you won't see any
effect from the flashlight. This is a purely physiological effect: even
though you are fully aware that the flashlight is turned on and aimed at the
surface, no amount of cognitive effort will reveal its beam pattern to you,
because your nervous system is physically incapable of resolving that
detail. Whereas, with less than full daylight, or at night time, it becomes
perfectly obvious. Call the flashlight a 2-bit light signal and full
daylight a 24-bit signal. When the ambient light amplitude is small, you can
resolve a 2-bit signal just fine. When the ambient light amplitude is large,
the size of the smallest signal you can resolve also increases, and that
2-bit flashlight signal just disappears; it's truncated away by your optical
nerve. So, the final point: "masking" is a physiological phenomenon, not a
psychological one.
> Date: Tue, 16 Jan 2001 00:04:09 -0500
> From: Timothy Stockman <[EMAIL PROTECTED]>
> One of the problems of bit-for-bit comparison of a signal
> recorded by MD: I wonder if the problem could have to do with the
> sample-rate-converter in the MD? I've often had the nagging
> suspicion that the sample-rate-converter is not defeatable.
> My theory is that even when the input signal is 44.1, the sample
> rate converter resamples the signal at the MD's (slightly
> different) 44.1 clock. The theory is that it *never* tries to
> lock the MD to the clock recovered from the S/PDIF stream.
> Anyone know for sure? If this is true, an alternate way to get a
> known signal onto an MD might be to burn the test signal
> onto a CDR and then transfer it using one of the combo MD-CD
> decks. I bet that copying from the internal CD player
> *doesn't* go through the sample-rate-converter.
This sounds like a good experiment for someone with the gear.
(I don't have it.)
> Date: Tue, 16 Jan 2001 13:13:05
> From: "Crak Therapy" <[EMAIL PROTECTED]>
> Subject: Re: MD: Wanted: Usable MD/PC device
> i dont know whether this has been mentioned on the list before
> but there is
> a new 'Compact PC compatible MD deck'(with LP) from sony, the MDSPC3.
> it uses USB and a linkup kit is provided. it also has digital input and
> output.
Unfortunately, this is only a USB-Audio device. It's not a data drive, and
in particular the USB-Audio spec is deficient in regards to sending the
auxiliary
data from a digital (S/PDIF) audio stream. (I.e., it doesn't transmit the
subcode containing track marks and title information.)
> Date: Tue, 16 Jan 2001 15:48:18 -0500
> From: Stainless Steel Rat <[EMAIL PROTECTED]>
> Subject: Re: MD: ATRAC lossiness
> * "Howard Chu" <[EMAIL PROTECTED]> on Mon, 15 Jan 2001
> | about 4000 samples. So, it's been pretty much impossible for me to do a
> | direct bit-for-bit comparison of the ATRAC audio with the
> original WAV data.
> And useless, by the way.
> If you make a recording of a cricket chirping and a jumbo jet flies
> overhead, you cannot hear the cricket over the sound of the jet. ATRAC
> removes the sounds the cricket makes. When you do the bitwise comparison
> of the original vs. the reduced version, you can hear the cricket because
> that is what was removed, assuming that the fidelity of the original
> recording is sufficient to have picked it up in the first place. But that
> tells you ABSOLUTELY NOTHING ABOUT WHAT YOU CAN REALLY HEAR. The
> objective
> comparison is fundamentally flawed because hearing is subjective.
Yes, hearing is subjective, but there is a lot of objective data that feeds
in before you arrive at your subjective result. In a competent, high-quality
encoder that always manages to preserve only the "important" signals, the
bitwise comparison may be useless. But not every encoder is equally
competent,
and it's not always possible (bitrate limitations) to preserve every
important
signal. As an example, as you lower the bitrate on MP3, you lose more and
more
high-frequency content, and the loss is perfectly obvious. We're not just
talking
about comparing masking results, but also about outright signal loss, along
with any other (expected or unexpected) artifacts.
-- Howard Chu
Chief Architect, Symas Corp. Director, Highland Sun
http://www.symas.com http://highlandsun.com/hyc
-----------------------------------------------------------------
To stop getting this list send a message containing just the word
"unsubscribe" to [EMAIL PROTECTED]
