Re: Which edition?


D F Tweedie
 

David ...

Great discussion. For those that want to get deep in the weeds regarding sample rates and the differences between 16, 24 and 32 bit floating point I highly recommend this article:



While it is true that 24 bit audio provides finer discrimination for dynamic range than 16 bit, I do not agree that CD Audio truncates frequencies that humans can perceive. And with respect to frequency, I do not believe 44.1 kHz to be a compromise. Based on the Nyquest Theorem and the normal range of hearing of 20 Hz to 20 kHz, 44.1 kHz encapsulates all we can hear.

The issue of higher fidelity of 48, 96 or 192 sampling rates (or similar multiples of 44.1) has to do with artifacts (noise) introduced by the process of digital conversion (see article). In other words, recording at higher sample rate will result in 'cleaner' audio, but does not provide higher frequency content that can be heard.

So, while I now understand David's use of 'compromised,' I do not see that frequency range is truncated in any meaningful way ... so I would not consider CD Audio a compromise for frequency, although I do for dynamic range that could be used in a recording.

DF


From: "'David H. Bailey' dhbailey52@... [Band-in-a-Box]"
To: Band-in-a-Box@...
Sent: Saturday, December 10, 2016 1:45 PM
Subject: Re: [Band-in-a-Box] Which edition?

 
On 12/10/2016 3:28 PM, D F Tweedie bienpegaito@... [Band-in-a-Box]
wrote:
>
>
> David ...
>
> I think you mean 'compressed,' not 'compromised.'
>

Actually I did mean "compromised" -- the cd-audio format intentionally
cuts out very high frequency data. It's a compromise in order to get
more data into a fixed space. And a .wav file is simply a computer
version of the cd-audio spec. The audio is compromised. Not so that
most people would notice, which is what the engineers and the companies
that were in on the development were counting on. Convenience over
completeness.

> For those who haven't looked into this issue, calling files 'compressed'
> easily can lead to confusion.
>
> There are two meanings. There are compression algorithms that reduce the
> size of the audio file by removing parts of the audio content. These are
> called 'lossy' formats, meaning that once compressed, the lost part of
> the audio can never be recovered from the file. WMA and MP3 are the best
> known of these examples, both using psychoacoustic principles to
> determine what is removed and what is retained.
>
> There is a further consideration with this type of compression, in that
> compression can be performed at different levels of data reduction,
> ergo, "high quality MP3s," e.g., bit rates of 128, 196, 256, 320.
>
> For example, a bit rate of 128 renders an MP3 file with 1/10th of the
> original content of a CD-Audio file of 16 bit and 44.1kHz ... whereas a
> bit rate of 320 renders a file with 1/4th the original content.
>
> However, there is also file compression that has nothing to do with
> reducing or removing original audio data, it merely packs data more
> tightly for storage and then reverts it to its original size upon
> decompression. Examples are zip files or rar files.
>
> It is this latter type of compression that is used for the WAV files in
> the Audiophile edition.
>
> If you consider the old trick of using a copier and then copying the
> copy multiple times to see the degradation that occurs when you reduce
> data digitally, you will understand the value of the .WAV files for
> audio production in a DAW. Even though many DAWs operate at higher bits
> internally, each time an MP3 file is brought into a DAW it must be
> converted to a WAV or AIFF file, rendering a file absent at least 3/4ths
> of the original audio data. Now when this file is processed and then
> returned to a 16 bit, 44.1 kHz CD Audio file it is more likely to result
> in an inferior product than an originally uncompressed file.
>
> Which is why I say, that for audio production I value the Audiophile
> edition with WAV files that are available uncompressed with respect to
> reduction of data.

And that is exactly what you should do because it's the maximum amount
of audio data available in digital format. But 16-bit, 44.1K sampling
(which is the cd-audio standard) has admittedly been proven to be much
less complete than 24-bit or 96-bit sampling at 48K or higher. The
higher the sample rate, the higher the frequencies which can be
captured, thus the more complete sound. Which is why CD-audio files
sound better than mp3 files even at their highest bitrate, as D F
pointed out. But just as most people can't hear the degradation between
a full-spectrum vinyl LP recording and a CD-audio recording, so, too,
most people can't hear the difference between CD-audio and MP3 files at
320bps.

--
*****
David H. Bailey
dhbailey52@...
http://www.davidbaileymusicstudio.com


Join main@Band-In-A-Box.groups.io to automatically receive all group messages.