[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Date Index][
Thread Index][
Author Index]
Re: A/B listening tests
Stefan Tiedje wrote:
> Bill Fox schrieb:
>> Speaking of A/B testing, according to what I've read, the difference
>> between 16 and 24 bit digital can be heard much more easily than the
>> difference between 44.1kHz and 192kHz.
> That means you read it and believe it?
Actually, in this particular case, I do and here's why. Recording at
the higher sampling frequencies eat up far more more memory than
increased bit depth and increases system costs. Then you have to down
sample to 44.1 in order to release product which requires expensive
software to do right. (Why not record at 176.4kHz and avoid the high
end software?) It behooves the hardware and software industry to
convince us that 16bit/192kHz is superior to 24bit/44.1kHz. (Of course
24bit/192kHz is even better!) Since the 24bit/44.1kHz supposedly sounds
better and costs me less, I'll go that way. I have also spoken to
industry people who confirm the increased bit depth is a far more
audible improvement than increased sample frequency. I have not heard
the other system nor do I have a room where I could hear the difference
so I can't speak from personal experience. However, the golden ears in
the industry who say that you need 192kHz all seem to have a stake in
its success. There is just as much physics at work when increasing and
linearizing dynamic range. As an engineer who believes in the Nyquist
theorem and as an over 50 year old male who (as I've read) can't hear
much above 12kHz, I see no reason to go to 192kHz, especially when most
listening environments are far less than ideal and most people are only
going to be listening to a crappy MP3 file anyway.
A higher sampling frequency gives you a wider frequency response that I
supposedly can't hear. Increased bit depth reduces the noise floor and
decreases distortion due to LSB errors. In any digital system,
distortion goes up as you approach the lowest volumes. These are things
that I *can* hear.
Cheers,
Bill