Take 2: Recording at different bit depths/sample rates

[Ty Roberts]

I'm starting a new thread about this because I apparently didn't make my
question clear enough the first time.

I want to experiment with different bit depths and sample rates,
to get a feeling for how the sound qualities vary as the bit depths
and sample rates get down into the low end (below 16 bit sound).
I have a computer/studio setup with a Firewire 410 interface. I can do
recordings into my sequencer, but it defaults to 24 bit and I think the
only other choice is 16 bit. I want to do lower bit depths than that.

I took a look at Audacity, but it seems to only support 32 bit, 24 bit,
and 16 bit.

Bit depths higher than 16 bits are not interesting to me at this point.
I already know what they sound like. I'm only interested in going down
from 16 to lower bit depths and slower sample rates, the range that
would be considered unsuitable for recording. I'm really interested in
getting a feeling for the qualities of the sound in this lower range.

As you might know, 12 bit samplers like the SP-1200 and some of the
older Akai machines, etc, are still very popular with Hip Hop musicians.
This vintage gear in many cases sells for more than current gear with
"better specs." They don't want better specs. They want the grungy
12-bit sound of an SP-1200. Some people writing for severely constrained
gadgets are playing with the lower limits of bit depth and sample
rate. In many cases, they're going below 8 bit sound. I'm gearing up
for a project with an engineer to do some stuff along these lines.

So that's where I'm coming from. I want to get a feeling for the sounds
down at the low end. I'm not concerned with whether anyone thinks they
can hear the difference between 24 and 32 bit recording.

[Laurence Payne]

On Wed, 02 Jan 2008 04:00:16 -0700, Ty Roberts > wrote:

>I want to experiment with different bit depths and sample rates,
>to get a feeling for how the sound qualities vary as the bit depths
>and sample rates get down into the low end (below 16 bit sound).
>I have a computer/studio setup with a Firewire 410 interface. I can do
>recordings into my sequencer, but it defaults to 24 bit and I think the
>only other choice is 16 bit. I want to do lower bit depths than that.

OK. Well, Audacity is great, considering it's free! But it isn't
stuffed with functions that most users wouldn't consider useful :-)

Soundforge offers only 8 bit, and 4 fixed sample rates down to 8KHz in
the "sub-prime" range. I seem to remember Wavelab offers a greater
selection, but it isn't handy so I can check.

[Peter Larsen[_2_]]

Ty Roberts wrote:

> I want to experiment with different bit depths and sample rates,
[snip]
> I took a look at Audacity, but it seems to only support 32 bit, 24
> bit, and 16 bit.

Audition supports a lot of output formats .... get the demo!

> So that's where I'm coming from. I want to get a feeling for the
> sounds down at the low end. I'm not concerned with whether anyone
> thinks they can hear the difference between 24 and 32 bit recording.

That difference can be very obvious .... O;-) ... assuming of course that
steps are taken to prevent overload of the playback hardware.


Kind regards

Peter Larsen

[Anahata]

Ty Roberts wrote:
> I want to experiment with different bit depths and sample rates,
> to get a feeling for how the sound qualities vary as the bit depths
> and sample rates get down into the low end (below 16 bit sound).
>
> I took a look at Audacity, but it seems to only support 32 bit, 24 bit,
> and 16 bit.

Have a look at sox. It's a commmand line "Swiss army knife" tool for
converting between different audio file formats. You should be able to
convert at least to 8 bit depth, with or without dither (which you
should certainly investigate if you want to learn something useful) and
will convert between different sampling rates.

You can also get bit depth reduction by reducing the level, storing
e.g. as .WAV, reloading and increasing the level back to original. You
lose 1 bit per 6 dB level reduction. you can do that with Audacity using
the "amplification" tool.

Anahata

[philicorda[_5_]]

On Wed, 02 Jan 2008 04:00:16 -0700, Ty Roberts wrote:

> I'm starting a new thread about this because I apparently didn't make my
> question clear enough the first time.
>
> I want to experiment with different bit depths and sample rates, to get
> a feeling for how the sound qualities vary as the bit depths and sample
> rates get down into the low end (below 16 bit sound). I have a
> computer/studio setup with a Firewire 410 interface. I can do recordings
> into my sequencer, but it defaults to 24 bit and I think the only other
> choice is 16 bit. I want to do lower bit depths than that.
>
> I took a look at Audacity, but it seems to only support 32 bit, 24 bit,
> and 16 bit.
>
> Bit depths higher than 16 bits are not interesting to me at this point.
> I already know what they sound like. I'm only interested in going down
> from 16 to lower bit depths and slower sample rates, the range that
> would be considered unsuitable for recording. I'm really interested in
> getting a feeling for the qualities of the sound in this lower range.

If a 16bit file contains only 12 bits of information it will practically
be the same as a 12 bit file. This is what some of the earlier replies
about reducing the volume were getting at.

>
> As you might know, 12 bit samplers like the SP-1200 and some of the
> older Akai machines, etc, are still very popular with Hip Hop musicians.
> This vintage gear in many cases sells for more than current gear with
> "better specs." They don't want better specs. They want the grungy
> 12-bit sound of an SP-1200. Some people writing for severely constrained
> gadgets are playing with the lower limits of bit depth and sample rate.
> In many cases, they're going below 8 bit sound. I'm gearing up for a
> project with an engineer to do some stuff along these lines.

SP-1200 is 12bit 22KHz. You should be able to resample to 22KHz in
audacity I think.

What you might find is that a properly converted 16bit 44.1Khz to 12bit
22Khz does not sound as grungy as you might hope. It might just sound
noisy and muffled.

I bet EMU allowed some aliasing in the SP-1200 to get a bit of brightness
despite the low sampling rate. Many people disable the VCFs too to get
even more aliasing.

There are some bit and sample rate reduction plugins that do a really
nasty aliasy conversion. The Waldorf D-pole VST plugin is my normal tool
for this kind of thing. I'm sure there are many others. CMT Bitcrusher
looks interesting for free.

>
> So that's where I'm coming from. I want to get a feeling for the sounds
> down at the low end. I'm not concerned with whether anyone thinks they
> can hear the difference between 24 and 32 bit recording.

Don't worry too much about the actual file bit depth. If you want it as
an effect, keeping the file 16/44.1 and using bit reduction effect
plugins will be more versatile.

[Ty Ford]

On Wed, 2 Jan 2008 06:00:16 -0500, Ty Roberts wrote
(in article >):

> I'm starting a new thread about this because I apparently didn't make my
> question clear enough the first time.
>
> I want to experiment with different bit depths and sample rates,
> to get a feeling for how the sound qualities vary as the bit depths
> and sample rates get down into the low end (below 16 bit sound).
> I have a computer/studio setup with a Firewire 410 interface. I can do
> recordings into my sequencer, but it defaults to 24 bit and I think the
> only other choice is 16 bit. I want to do lower bit depths than that.
>
> I took a look at Audacity, but it seems to only support 32 bit, 24 bit,
> and 16 bit.
>
> Bit depths higher than 16 bits are not interesting to me at this point.
> I already know what they sound like. I'm only interested in going down
> from 16 to lower bit depths and slower sample rates, the range that
> would be considered unsuitable for recording. I'm really interested in
> getting a feeling for the qualities of the sound in this lower range.
>
> As you might know, 12 bit samplers like the SP-1200 and some of the
> older Akai machines, etc, are still very popular with Hip Hop musicians.
> This vintage gear in many cases sells for more than current gear with
> "better specs." They don't want better specs. They want the grungy
> 12-bit sound of an SP-1200. Some people writing for severely constrained
> gadgets are playing with the lower limits of bit depth and sample
> rate. In many cases, they're going below 8 bit sound. I'm gearing up
> for a project with an engineer to do some stuff along these lines.
>
> So that's where I'm coming from. I want to get a feeling for the sounds
> down at the low end. I'm not concerned with whether anyone thinks they
> can hear the difference between 24 and 32 bit recording.

In addition to what's been posted, there are also plugins; once called "bit
crusher" comes to mind. The challenge you face certainly has somewhat to do
with bit depth, but more importantly, perhaps, what chips and algorithms were
used. In other words, different mfgrs use different ingredients to bake
their cakes. If you want the "Akai Sound", simply going for the same bit
depth may not give it to you.

Regards,

Ty Ford


--Audio Equipment Reviews Audio Production Services
Acting and Voiceover Demos http://www.tyford.com
Guitar player?:http://www.youtube.com/watch?v=4RZJ9MptZmU

[Richard Crowley]

"Ty Roberts" wrote ...
> I'm starting a new thread about this because I apparently didn't make
> my
> question clear enough the first time.
>
> I want to experiment with different bit depths and sample rates,
> to get a feeling for how the sound qualities vary as the bit depths
> and sample rates get down into the low end (below 16 bit sound).

Others have already explained how you can reduce 16-bit audio to
any smaller bit-depth simply by attenuation. Did you miss that info?

[Carey Carlan]

"Peter Larsen" > wrote in news:017a3afd$0$30254
:

> Ty Roberts wrote:
>
>> I want to experiment with different bit depths and sample rates,
> [snip]
>> I took a look at Audacity, but it seems to only support 32 bit, 24
>> bit, and 16 bit.
>
> Audition supports a lot of output formats .... get the demo!

Audition supports "all of them".

You can specify the sampling rate from single digits to megahertz and bit
depth by any number of bits from 2 to 32.

What you'll find is that as you decrease bit depth, noise levels increase,
and as you decrease sampling rate, high frequencies disappear.