Lines: 19
X-Admin:
From: (Richard Kuschel)
Newsgroups: rec.audio.tech
Date: 29 Dec 2003 14:49:56 GMT
References:
Organization: AOL http://www.aol.com
Subject: Bit-resolution and Clipping?
Message-ID:



Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?




I think what you are asking is if it can handle a higher input level and the
answer to that is no.

The dynamic range may be greater and the resolution at low levels will be a lot
better.

The main advantage to the higher resolution and dynamic range of a 24 bit file
over 16 bit is that you can record at a lower level with better resolution and
not run into the danger of clipping in the first place.

Richard H. Kuschel
"I canna change the law of physics."-----Scotty

Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

The dynamic range may be greater and the resolution at low levels
will be a lot better.

Right, high resolution and a low noise floor go together.

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Some people have been confused about this, thinking that both the noise
floor and resolution can both be higher. High resolution and low noise are
inseparable. High noise and low resolution are inseparable.

Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

The dynamic range may be greater and the resolution at low levels
will be a lot better.

Right, high resolution and a low noise floor go together.

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Some people have been confused about this, thinking that both the noise
floor and resolution can both be higher. High resolution and low noise are
inseparable. High noise and low resolution are inseparable.

Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

The dynamic range may be greater and the resolution at low levels
will be a lot better.

Right, high resolution and a low noise floor go together.

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Some people have been confused about this, thinking that both the noise
floor and resolution can both be higher. High resolution and low noise are
inseparable. High noise and low resolution are inseparable.

Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

The dynamic range may be greater and the resolution at low levels
will be a lot better.

Right, high resolution and a low noise floor go together.

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Some people have been confused about this, thinking that both the noise
floor and resolution can both be higher. High resolution and low noise are
inseparable. High noise and low resolution are inseparable.

"Arny Krueger" wrote in message
...
Some people have been confused about this, thinking that both the noise
floor and resolution can both be higher. High resolution and low noise are
inseparable.

Not necessarily true. A soundcard like the Audigy 1 SE has much less than 16
bits resolution even when recording at 24 bits. A better soundcard will have
greater S/N ratio at 16 bits than the Audigy 1 SE does at 24 bits. Your
headroom (or footroom) will be governed by the analog performance of the
card as much as the number of (noise) bits being recorded.

TonyP.

"Arny Krueger" wrote in message
...
Some people have been confused about this, thinking that both the noise
floor and resolution can both be higher. High resolution and low noise are
inseparable.

Not necessarily true. A soundcard like the Audigy 1 SE has much less than 16
bits resolution even when recording at 24 bits. A better soundcard will have
greater S/N ratio at 16 bits than the Audigy 1 SE does at 24 bits. Your
headroom (or footroom) will be governed by the analog performance of the
card as much as the number of (noise) bits being recorded.

TonyP.

"Arny Krueger" wrote in message
...
Some people have been confused about this, thinking that both the noise
floor and resolution can both be higher. High resolution and low noise are
inseparable.

Not necessarily true. A soundcard like the Audigy 1 SE has much less than 16
bits resolution even when recording at 24 bits. A better soundcard will have
greater S/N ratio at 16 bits than the Audigy 1 SE does at 24 bits. Your
headroom (or footroom) will be governed by the analog performance of the
card as much as the number of (noise) bits being recorded.

TonyP.

"Arny Krueger" wrote in message
...
Some people have been confused about this, thinking that both the noise
floor and resolution can both be higher. High resolution and low noise are
inseparable.

Not necessarily true. A soundcard like the Audigy 1 SE has much less than 16
bits resolution even when recording at 24 bits. A better soundcard will have
greater S/N ratio at 16 bits than the Audigy 1 SE does at 24 bits. Your
headroom (or footroom) will be governed by the analog performance of the
card as much as the number of (noise) bits being recorded.

TonyP.

TonyP wrote:
"Arny Krueger" wrote in message
...
Some people have been confused about this, thinking that both the
noise floor and resolution can both be higher. High resolution and
low noise are inseparable.

Not necessarily true. A soundcard like the Audigy 1 SE has much less
than 16 bits resolution even when recording at 24 bits.

Right, but my point is that they don't have REAL 24 bit resoltuion. They may
have 24 bit digital data paths, but the least significant bits are noise,
which is a symptom of lower resoltuion.

A better
soundcard will have greater S/N ratio at 16 bits than the Audigy 1 SE
does at 24 bits.

Agreed. For example, a LynxONE or a CardDeluxe.

Your headroom (or footroom) will be governed by the
analog performance of the card as much as the number of (noise) bits
being recorded.

I have no argument with that at all.

However, I don't call an Audigy2 a 24 bit resolution sound card. I call it a
card that meddles with 24 bit data, but as more like 15 bits of actual
resolution.

TonyP wrote:
"Arny Krueger" wrote in message
...
Some people have been confused about this, thinking that both the
noise floor and resolution can both be higher. High resolution and
low noise are inseparable.

Not necessarily true. A soundcard like the Audigy 1 SE has much less
than 16 bits resolution even when recording at 24 bits.

Right, but my point is that they don't have REAL 24 bit resoltuion. They may
have 24 bit digital data paths, but the least significant bits are noise,
which is a symptom of lower resoltuion.

A better
soundcard will have greater S/N ratio at 16 bits than the Audigy 1 SE
does at 24 bits.

Agreed. For example, a LynxONE or a CardDeluxe.

Your headroom (or footroom) will be governed by the
analog performance of the card as much as the number of (noise) bits
being recorded.

I have no argument with that at all.

However, I don't call an Audigy2 a 24 bit resolution sound card. I call it a
card that meddles with 24 bit data, but as more like 15 bits of actual
resolution.

TonyP wrote:
"Arny Krueger" wrote in message
...
Some people have been confused about this, thinking that both the
noise floor and resolution can both be higher. High resolution and
low noise are inseparable.

Not necessarily true. A soundcard like the Audigy 1 SE has much less
than 16 bits resolution even when recording at 24 bits.

Right, but my point is that they don't have REAL 24 bit resoltuion. They may
have 24 bit digital data paths, but the least significant bits are noise,
which is a symptom of lower resoltuion.

A better
soundcard will have greater S/N ratio at 16 bits than the Audigy 1 SE
does at 24 bits.

Agreed. For example, a LynxONE or a CardDeluxe.

Your headroom (or footroom) will be governed by the
analog performance of the card as much as the number of (noise) bits
being recorded.

I have no argument with that at all.

However, I don't call an Audigy2 a 24 bit resolution sound card. I call it a
card that meddles with 24 bit data, but as more like 15 bits of actual
resolution.

TonyP wrote:
"Arny Krueger" wrote in message
...
Some people have been confused about this, thinking that both the
noise floor and resolution can both be higher. High resolution and
low noise are inseparable.

Not necessarily true. A soundcard like the Audigy 1 SE has much less
than 16 bits resolution even when recording at 24 bits.

Right, but my point is that they don't have REAL 24 bit resoltuion. They may
have 24 bit digital data paths, but the least significant bits are noise,
which is a symptom of lower resoltuion.

A better
soundcard will have greater S/N ratio at 16 bits than the Audigy 1 SE
does at 24 bits.

Agreed. For example, a LynxONE or a CardDeluxe.

Your headroom (or footroom) will be governed by the
analog performance of the card as much as the number of (noise) bits
being recorded.

I have no argument with that at all.

However, I don't call an Audigy2 a 24 bit resolution sound card. I call it a
card that meddles with 24 bit data, but as more like 15 bits of actual
resolution.

"Arny Krueger" wrote in message ...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out clipping?

"Arny Krueger" wrote in message ...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out clipping?

"Arny Krueger" wrote in message ...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out clipping?

"Arny Krueger" wrote in message ...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out clipping?

(Curious) writes:

"Arny Krueger" wrote in message ...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out clipping?

Possibly. More bits provide more dynamic range, i.e., more range between
the loudest possible signal and the noise floor. That extra range can be
provided by increasing the loudest possible signal while keeping the
same noise floor, decreasing the noise floor while keeping the same
loudest possible signal, or something in between.
--
Randy Yates
Sony Ericsson Mobile Communications
Research Triangle Park, NC, USA
, 919-472-1124

(Curious) writes:

"Arny Krueger" wrote in message ...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out clipping?

Possibly. More bits provide more dynamic range, i.e., more range between
the loudest possible signal and the noise floor. That extra range can be
provided by increasing the loudest possible signal while keeping the
same noise floor, decreasing the noise floor while keeping the same
loudest possible signal, or something in between.
--
Randy Yates
Sony Ericsson Mobile Communications
Research Triangle Park, NC, USA
, 919-472-1124

(Curious) writes:

"Arny Krueger" wrote in message ...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out clipping?

Possibly. More bits provide more dynamic range, i.e., more range between
the loudest possible signal and the noise floor. That extra range can be
provided by increasing the loudest possible signal while keeping the
same noise floor, decreasing the noise floor while keeping the same
loudest possible signal, or something in between.
--
Randy Yates
Sony Ericsson Mobile Communications
Research Triangle Park, NC, USA
, 919-472-1124

(Curious) writes:

"Arny Krueger" wrote in message ...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise floor that
is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out clipping?

Possibly. More bits provide more dynamic range, i.e., more range between
the loudest possible signal and the noise floor. That extra range can be
provided by increasing the loudest possible signal while keeping the
same noise floor, decreasing the noise floor while keeping the same
loudest possible signal, or something in between.
--
Randy Yates
Sony Ericsson Mobile Communications
Research Triangle Park, NC, USA
, 919-472-1124

Curious wrote:
"Arny Krueger" wrote in message
...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

Because FS is the end of the line.

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise
floor that is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out
clipping?

The additional headroom is obtained at a cost - lower average recorded
levels.

Curious wrote:
"Arny Krueger" wrote in message
...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

Because FS is the end of the line.

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise
floor that is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out
clipping?

The additional headroom is obtained at a cost - lower average recorded
levels.

Curious wrote:
"Arny Krueger" wrote in message
...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

Because FS is the end of the line.

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise
floor that is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out
clipping?

The additional headroom is obtained at a cost - lower average recorded
levels.

Curious wrote:
"Arny Krueger" wrote in message
...
Curious wrote:

Can a wave file with stronger bit-resolution handle more decibels
w/out clipping than a wave file with weaker bit-resolution?

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

Because FS is the end of the line.

The main advantage to the higher resolution and dynamic range of a
24 bit file over 16 bit is that you can record at a lower level with
better resolution and not run into the danger of clipping in the
first place.

Right, you can record with more headroom, and still have a noise
floor that is well below the average signal levels.

Doesn't "more headroom" mean it can handle louder input w/out
clipping?

The additional headroom is obtained at a cost - lower average recorded
levels.

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

#snip#

Doesn't "more headroom" mean it can handle louder input w/out clipping?

Possibly. More bits provide more dynamic range, i.e., more range between
the loudest possible signal and the noise floor. That extra range can be
provided by increasing the loudest possible signal while keeping the
same noise floor, decreasing the noise floor while keeping the same
loudest possible signal, or something in between.

True enough.

You have to look at the whole signal chain, including the rest of the
electronics. In consumer-electronics space, most line-level signal
paths these days use a 2-volt peak-to-peak analog signal as their
"maximum output, 0 dB" reference. This is the voltage that most
home-model CD players (for example) will generate for a
maximum-amplitude signal, and you can't exceed this. Even if you can
generate a higher voltage, it's quite possible that the next device
"downstream" will be overloaded by it. If you try to shove 10 volts
peak-to-peak into a typical consumer ADC, it'll overload and clip no
matter how many bits of resolution is may have internally.

In the pre-CD era, 1 volt P-P was considered full-scale for a
line-level input. The first CD players sometimes turned out to be
less than fully compatible with older-generation preamps and
receivers, whose line-input stages would be forced into nonlinearity
or clipping by voltages higher than they were designed to handle.

So, yes, when you increase the number of bits in your encoding, you
_can_ leave the amplitude of the least-significant-bit the same, and
(in effect) add all of the headroom at the high-amplitude end of the
scale. This will allow you to _represent_ higher signal levels
without going into digital clipping - but in a real-world
recording/playback system with analog components, you may not be able
to generate and carry the higher voltages required without overloading
various analog devices in the recording/playback chain. It's
embarrassing when your audio interconnect cables start arcing-over at
the connectors :-)

Or, you can leave the maximum full-scale signal level the same, and
add headroom at the low end of the scale. This reduces the noise
floor of the digital system. You can then record lower-level signals
(without running into quantization noise as quickly). Alternatively,
you can attenuate your analog input signal somewhat and record
everything at a lower average level (compared to full-scale output).
This _will_ reduce the tendency to clip during the analog-to-digital
conversion (although it won't do anything to address analog clipping
or overload prior to this point). By pushing the analog signal down
to a lower average level, it may make the presence of analog noise in
your electronics or converter somewhat more of a problem.

There's no free lunch, alas.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

#snip#

Doesn't "more headroom" mean it can handle louder input w/out clipping?

Possibly. More bits provide more dynamic range, i.e., more range between
the loudest possible signal and the noise floor. That extra range can be
provided by increasing the loudest possible signal while keeping the
same noise floor, decreasing the noise floor while keeping the same
loudest possible signal, or something in between.

True enough.

You have to look at the whole signal chain, including the rest of the
electronics. In consumer-electronics space, most line-level signal
paths these days use a 2-volt peak-to-peak analog signal as their
"maximum output, 0 dB" reference. This is the voltage that most
home-model CD players (for example) will generate for a
maximum-amplitude signal, and you can't exceed this. Even if you can
generate a higher voltage, it's quite possible that the next device
"downstream" will be overloaded by it. If you try to shove 10 volts
peak-to-peak into a typical consumer ADC, it'll overload and clip no
matter how many bits of resolution is may have internally.

In the pre-CD era, 1 volt P-P was considered full-scale for a
line-level input. The first CD players sometimes turned out to be
less than fully compatible with older-generation preamps and
receivers, whose line-input stages would be forced into nonlinearity
or clipping by voltages higher than they were designed to handle.

So, yes, when you increase the number of bits in your encoding, you
_can_ leave the amplitude of the least-significant-bit the same, and
(in effect) add all of the headroom at the high-amplitude end of the
scale. This will allow you to _represent_ higher signal levels
without going into digital clipping - but in a real-world
recording/playback system with analog components, you may not be able
to generate and carry the higher voltages required without overloading
various analog devices in the recording/playback chain. It's
embarrassing when your audio interconnect cables start arcing-over at
the connectors :-)

Or, you can leave the maximum full-scale signal level the same, and
add headroom at the low end of the scale. This reduces the noise
floor of the digital system. You can then record lower-level signals
(without running into quantization noise as quickly). Alternatively,
you can attenuate your analog input signal somewhat and record
everything at a lower average level (compared to full-scale output).
This _will_ reduce the tendency to clip during the analog-to-digital
conversion (although it won't do anything to address analog clipping
or overload prior to this point). By pushing the analog signal down
to a lower average level, it may make the presence of analog noise in
your electronics or converter somewhat more of a problem.

There's no free lunch, alas.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

#snip#

Doesn't "more headroom" mean it can handle louder input w/out clipping?

Possibly. More bits provide more dynamic range, i.e., more range between
the loudest possible signal and the noise floor. That extra range can be
provided by increasing the loudest possible signal while keeping the
same noise floor, decreasing the noise floor while keeping the same
loudest possible signal, or something in between.

True enough.

You have to look at the whole signal chain, including the rest of the
electronics. In consumer-electronics space, most line-level signal
paths these days use a 2-volt peak-to-peak analog signal as their
"maximum output, 0 dB" reference. This is the voltage that most
home-model CD players (for example) will generate for a
maximum-amplitude signal, and you can't exceed this. Even if you can
generate a higher voltage, it's quite possible that the next device
"downstream" will be overloaded by it. If you try to shove 10 volts
peak-to-peak into a typical consumer ADC, it'll overload and clip no
matter how many bits of resolution is may have internally.

In the pre-CD era, 1 volt P-P was considered full-scale for a
line-level input. The first CD players sometimes turned out to be
less than fully compatible with older-generation preamps and
receivers, whose line-input stages would be forced into nonlinearity
or clipping by voltages higher than they were designed to handle.

So, yes, when you increase the number of bits in your encoding, you
_can_ leave the amplitude of the least-significant-bit the same, and
(in effect) add all of the headroom at the high-amplitude end of the
scale. This will allow you to _represent_ higher signal levels
without going into digital clipping - but in a real-world
recording/playback system with analog components, you may not be able
to generate and carry the higher voltages required without overloading
various analog devices in the recording/playback chain. It's
embarrassing when your audio interconnect cables start arcing-over at
the connectors :-)

Or, you can leave the maximum full-scale signal level the same, and
add headroom at the low end of the scale. This reduces the noise
floor of the digital system. You can then record lower-level signals
(without running into quantization noise as quickly). Alternatively,
you can attenuate your analog input signal somewhat and record
everything at a lower average level (compared to full-scale output).
This _will_ reduce the tendency to clip during the analog-to-digital
conversion (although it won't do anything to address analog clipping
or overload prior to this point). By pushing the analog signal down
to a lower average level, it may make the presence of analog noise in
your electronics or converter somewhat more of a problem.

There's no free lunch, alas.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

I think what you are asking is if it can handle a higher input level
and the answer to that is no.

Agreed.

Why?

#snip#

Doesn't "more headroom" mean it can handle louder input w/out clipping?

Possibly. More bits provide more dynamic range, i.e., more range between
the loudest possible signal and the noise floor. That extra range can be
provided by increasing the loudest possible signal while keeping the
same noise floor, decreasing the noise floor while keeping the same
loudest possible signal, or something in between.

True enough.

You have to look at the whole signal chain, including the rest of the
electronics. In consumer-electronics space, most line-level signal
paths these days use a 2-volt peak-to-peak analog signal as their
"maximum output, 0 dB" reference. This is the voltage that most
home-model CD players (for example) will generate for a
maximum-amplitude signal, and you can't exceed this. Even if you can
generate a higher voltage, it's quite possible that the next device
"downstream" will be overloaded by it. If you try to shove 10 volts
peak-to-peak into a typical consumer ADC, it'll overload and clip no
matter how many bits of resolution is may have internally.

In the pre-CD era, 1 volt P-P was considered full-scale for a
line-level input. The first CD players sometimes turned out to be
less than fully compatible with older-generation preamps and
receivers, whose line-input stages would be forced into nonlinearity
or clipping by voltages higher than they were designed to handle.

So, yes, when you increase the number of bits in your encoding, you
_can_ leave the amplitude of the least-significant-bit the same, and
(in effect) add all of the headroom at the high-amplitude end of the
scale. This will allow you to _represent_ higher signal levels
without going into digital clipping - but in a real-world
recording/playback system with analog components, you may not be able
to generate and carry the higher voltages required without overloading
various analog devices in the recording/playback chain. It's
embarrassing when your audio interconnect cables start arcing-over at
the connectors :-)

Or, you can leave the maximum full-scale signal level the same, and
add headroom at the low end of the scale. This reduces the noise
floor of the digital system. You can then record lower-level signals
(without running into quantization noise as quickly). Alternatively,
you can attenuate your analog input signal somewhat and record
everything at a lower average level (compared to full-scale output).
This _will_ reduce the tendency to clip during the analog-to-digital
conversion (although it won't do anything to address analog clipping
or overload prior to this point). By pushing the analog signal down
to a lower average level, it may make the presence of analog noise in
your electronics or converter somewhat more of a problem.

There's no free lunch, alas.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!