8-16-24 or 32 bits from Flac to Wav?, which option should i choose? Options

[blitzlash]

i want to convert a flac image to wav, and in the "others" tab, there is an option that says "Convert lossy and other formats to" 16 bit (default) but it's possible to choose between 8-16-24 or 32.. i don't want to lose quality so which option must i choose? and what's the difference between 16 and 24 or 32 bits?

saludos

[ghebert]

i want to convert a flac image to wav, and in the "others" tab, there is an option that says "Convert lossy and other formats to" 16 bit (default) but it's possible to choose between 8-16-24 or 32.. i don't want to lose quality so which option must i choose? and what's the difference between 16 and 24 or 32 bits?

saludos

Generally you would want to select the same bit depth as the source file. If you aren't sure, chances are it's 16-bit. Although I've never tried it myself, I would guess that changing to 24 or 32 bit would not improve the sound quality. It's like trying to add something that's not there...going from 16 bits per sample to 24 bits per sample...where are those additional bits coming from?

CD's are all recorded at 16 bit depth...if you downloaded files in FLAC rather than rip from CD then the person who encoded to FLAC may have used 24 bit encoding if ripping from vinyl or DVD-Audio in order to preserve as much of original recording's quality as possible since they are either higher resolution than CD and/or have more dynamic range.

This post has been edited by ghebert: Oct 8 2009, 18:52

[extrabigmehdi]

If you convert files from FLAC to wav, with the software xrecode, the default setting is to keep the same number of bit per sample, or same sample rate as the source.
It's probably true for other conversion tools, I mention xrecode because I know it better.

[[JAZ]]

That option of foobar2000 will do nothing for lossless sources. That's why it is separated.

So:
If you want to decode FLAC to wav, "convert lossless formats to" -> same as source
and "dither" -> never. (or at much for lossy sources only).

[DVDdoug]

...and what's the difference between 16 and 24 or 32 bits?

That's a hard question to answer unless you already understand something about digital audio... There is a quick introduction to digital audio on the Audacity Website.

It's the resolution (like dots-per-inch on a scanner). A bit is a binary "digit" that holds a zero or a one. With 8-bits you can "count" from 00000000 to 11111111 (binary) which is from 0 to 255 decimal.

8-bit WAV files hold values between 0 and 255, and 16-bit WAV files hold values between -32,768 and +32,767. (When you play an 8-bit WAV file it goes positive and negative, but for some reason it's saved as unsigned integers.)

This number represents the "height" of the "wave". At a sample rate of 44.1kHz, the height is recorded 44,100 times every second. 16, 24, and 32-bit waves are not "taller" or "louder", they just have more precision.

Another analogy would be the difference in measuring in inches instead of feet (or millimeters instead of meters). You can be more accurate if you can measure in millimeters or inches... These are integer formats, so you can't have fractions or a decimal point (except for 32-bit floating point).

This post has been edited by DVDdoug: Oct 9 2009, 00:11

[blitzlash]

thanks for the link and your very complete answer, it's very interesting

but i've got a question for you: cd's are recorded in 16 bits right? that means with a sample rate of 44.100.. so, which was the sample rate that used the cassettes, i want to know because cd's quality sound is better than tapes..

(sorry about my english)

This post has been edited by greynol: Oct 9 2009, 18:06

Reason for edit: Removed unnecessary full quotation of the previous post.

[pdq]

Cassettes are analog, they have no sample rate.

[[JAZ]]

the bits per sample and the sample rate are completely independent things. A CD is 16 bits 44.1Khz stereo, but could perfectly be 4bits 88.2Khz quad channel. (Not that it would be more useful this way..)

Analog media (cassetes, VHS, vinyl, reel-to-reel..) do not have a concept of sampling rate. At much they can have a concept of bandwidth.

Sampling rate is also explained in that audacity link.

[greynol]

...24 bit encoding if ripping from vinyl or DVD-Audio in order to preserve as much of original recording's quality as possible since they are either higher resolution than CD and/or have more dynamic range.

I'm surprised this wasn't addressed earlier. While vinyl is capable of delivering higher frequencies than CD, it cannot reproduce either the resolution or dynamic range that a CD can deliver.

This post has been edited by greynol: Oct 15 2009, 21:35