Arny,

Sometimes. There still seem to be soundcards out there with analog input
gain controls.

Thanks for clarifying. I know you have experience with a lot of consumer
grade cards.

--Ethan

Arny,

Sometimes. There still seem to be soundcards out there with analog input
gain controls.

Thanks for clarifying. I know you have experience with a lot of consumer
grade cards.

--Ethan

On Thu, 2 Dec 2004 18:07:45 -0500, "Arny Krueger"
wrote:

"Preben Friis" wrote in message

"Joseph Raymond" wrote in message

If the level control actually controlled the gain of the analog
preamp, then some control words would have to be sent to the sound
card's ADC to either change the reference voltage or to change the
sound card's preamp gain. Based on economic considerations, I doubt
that that is the case.

Actually, it is the case for most audio chipsets. Take a look block
diagram in the datasheet below. Between the input selector and the
A/D there is a small block called PGA (Programable Gain Amplifier)
that can be set from 0 to 22.5 dB gain in 1.5 dB steps. And this is
for a part that costs $3...!
http://www.analog.com/UploadedFiles/...7AD1881A_0.pdf

Unfortunately, I've got quite a bit of experience with this device. Not
exactly the device of choice for high quality audio production:

Anything made for the AC-97 spec is of course made for the cheap
and high-volume (as in number of manufactured units, not loudness)
consumer PC soundcard/motherboard market, and I'd be surprised to find
any such chip with really good specs. Furthermore, the "phat stereo"
blocks don't endear me to think this is high quality.
There's this quote from that datasheet: "SoundMAX is a registered
trademark and PHAT is a trademark of Analog Device, [SIC] Inc." I can
only wonder if Analog Devices has ever sued anyone in the Hip Hop
community for use of AD's trademaked word PHAT.

Test results for a typical device from the SoundMax series:

Frequency response (40 Hz - 15 kHz), +9.68, -12.97 dB

I can only wonder how that can be so bad in ANY piece of
electronics (and I can't even think of how an improper test condition
could do that either), but let's not go there for now...

Noise level, -84.6 dB (A)
Dnamic range, 82.0 dB (A)
HD, 0.0044 %
Intermodulation distortion, 11.441 %

The HD and IM figures appear wildly incongruous. How is that? I
can't imagine how the IM figure could legitimately be so large with
the HD figure so small. Were the peak values the same in each test? I
could see where clipping in the IM test could cause a big number like
that.

-----
http://mindspring.com/~benbradley

On Thu, 2 Dec 2004 18:07:45 -0500, "Arny Krueger"
wrote:

"Preben Friis" wrote in message

"Joseph Raymond" wrote in message

If the level control actually controlled the gain of the analog
preamp, then some control words would have to be sent to the sound
card's ADC to either change the reference voltage or to change the
sound card's preamp gain. Based on economic considerations, I doubt
that that is the case.

Actually, it is the case for most audio chipsets. Take a look block
diagram in the datasheet below. Between the input selector and the
A/D there is a small block called PGA (Programable Gain Amplifier)
that can be set from 0 to 22.5 dB gain in 1.5 dB steps. And this is
for a part that costs $3...!
http://www.analog.com/UploadedFiles/...7AD1881A_0.pdf

Unfortunately, I've got quite a bit of experience with this device. Not
exactly the device of choice for high quality audio production:

Anything made for the AC-97 spec is of course made for the cheap
and high-volume (as in number of manufactured units, not loudness)
consumer PC soundcard/motherboard market, and I'd be surprised to find
any such chip with really good specs. Furthermore, the "phat stereo"
blocks don't endear me to think this is high quality.
There's this quote from that datasheet: "SoundMAX is a registered
trademark and PHAT is a trademark of Analog Device, [SIC] Inc." I can
only wonder if Analog Devices has ever sued anyone in the Hip Hop
community for use of AD's trademaked word PHAT.

Test results for a typical device from the SoundMax series:

Frequency response (40 Hz - 15 kHz), +9.68, -12.97 dB

I can only wonder how that can be so bad in ANY piece of
electronics (and I can't even think of how an improper test condition
could do that either), but let's not go there for now...

Noise level, -84.6 dB (A)
Dnamic range, 82.0 dB (A)
HD, 0.0044 %
Intermodulation distortion, 11.441 %

The HD and IM figures appear wildly incongruous. How is that? I
can't imagine how the IM figure could legitimately be so large with
the HD figure so small. Were the peak values the same in each test? I
could see where clipping in the IM test could cause a big number like
that.

-----
http://mindspring.com/~benbradley

Loren Amelang wrote in message . ..
On 1 Dec 2004 14:42:55 -0800, (Joseph
Raymond) wrote:
I don't know for sure the answer to your question. But I have
wondered about the very same thing and believe that it is math. In
other words, when you change the settings on the mixer on the PC
screen, you perform either a multiplication or division on the digital
word at the output of the ADC. And the level is the level at the
output of the mult/div relative to the number of bits available.

If the level control actually controlled the gain of the analog
preamp, then some control words would have to be sent to the sound
card's ADC to either change the reference voltage or to change the
sound card's preamp gain. Based on economic considerations, I doubt
that that is the case.

In any event, the question is what is the 0 dB input setting of the
software mixer? Is it with the slider in the middle or some other
position? In other words, how can you set the mulitplying factor to
1? Then, keeping that fixed, adjust the output level of your mic mixer
going into the LINE IN while monitoring the software VU meters for the
largest signal without clipping.

To add complexity to these questions, the level control you see in
your recording app is not always a second view of your Windows Volume
Control. When I start Audiotools, its sliders are at 0dB, regardless
of where my input Windows Volume Control sliders are set.

I find that cutting the gain at the Windows Volume Control provides
much better results than an equivalent reduction within Audiotools.
Even though the analog input to my sound interface is constant, if I
crank the Windows Volume Control up and make a comparable gain
reduction within Audiotools, the waveform appears compressed. Crank
the Windows Volume Control even higher and the recorded waveform is
clipped, even though the values in the digital file are nowhere near
maximum. Same behavior is observed recording with Audacity.

I've been assuming this is all happening in the digital domain, where
one would hope math is math, but it clearly it is not! Either the
Windows Volume Control must operate in the analog world before the
ADC, or it must be able to multiply as well as divide, and it must be
able to multiply so much that it can clip in the digital domain.

Another way of saying that would be to say that the (possibly digital
counterpart of the) "0dB" setting of the Windows Volume Control must
be well below its maximum setting. (At least with some audio
hardware...)

To add one more variable, I've found that Alex Mina's AC97 Mixer app,
which talks directly to AC97 audio hardware, can reach gain settings
the Windows Volume Control does not offer. At least with some
hardware...

According to its datasheet, the AKM AK4584 chip in my Transit USB has
a +18dB Input Programmable Gain Amp (0.5dB steps) ahead of the ADC.
With this IPGA set at 0 dB, it is supposed to take 3.0 Vpp of audio to
produce digital full scale. Cranking the gain up to +18 raises the
noise floor 10dB, so it sounds like this really is an analog
amplifier. And I can clearly get clipping with way less than 3.0 Vpp,
so the Transit drivers must have this IPGA set above 0 dB.

Then there are the 0/14/20/26 dB steps that M-Audio offers as "Mic
Boost" in their control panel. I don't see how that is implemented, or
how it relates to the +18 dB IPGA in the audio chip. And I haven't a
clue how the input Windows Volume Control is hooked up to all of this.

So is there a way to tell whether the Windows Volume Control for a
particular interface is operating before or after the ADC? I have a
feeling it is different for different audio devices, and possibly for
different driver versions...

Loren

That is the $64 question! Aside from doing it indirectly by
experimentation and analyzing the digitized waveform, I imagine that
there exists a Windows Audio API that defines all the possible
commands that Windows can send to the audio chip. I see in the chip
below that they say it meets MS PC2001 audio requirements. Perhaps
knowing what MS PC2001 is is a place to start.

http://www.cirrus.com/en/pubs/proDat...CS4205_PP3.pdf

Then you would have to know which chip is on your PC or sound card.
And I guess lastly, you would have to know if the chip has met Windows
Hardware Quality Labs testing. I have no idea what that entails.
And, when you loaded your drivers, it would be nice to not see any
messages appearing asking whether you want to proceed because it does
not meet Windows something or other.

Lately I just started using a Tascam US-122. It has an LED indicating
whether the front end is being overloaded and pots to adjust the input
level. It would be nice if this were the only way to adjust the input
level into the ADC. However, after this discussion, I have some
doubts!


Joe

Loren Amelang wrote in message . ..
On 1 Dec 2004 14:42:55 -0800, (Joseph
Raymond) wrote:
I don't know for sure the answer to your question. But I have
wondered about the very same thing and believe that it is math. In
other words, when you change the settings on the mixer on the PC
screen, you perform either a multiplication or division on the digital
word at the output of the ADC. And the level is the level at the
output of the mult/div relative to the number of bits available.

If the level control actually controlled the gain of the analog
preamp, then some control words would have to be sent to the sound
card's ADC to either change the reference voltage or to change the
sound card's preamp gain. Based on economic considerations, I doubt
that that is the case.

In any event, the question is what is the 0 dB input setting of the
software mixer? Is it with the slider in the middle or some other
position? In other words, how can you set the mulitplying factor to
1? Then, keeping that fixed, adjust the output level of your mic mixer
going into the LINE IN while monitoring the software VU meters for the
largest signal without clipping.

To add complexity to these questions, the level control you see in
your recording app is not always a second view of your Windows Volume
Control. When I start Audiotools, its sliders are at 0dB, regardless
of where my input Windows Volume Control sliders are set.

I find that cutting the gain at the Windows Volume Control provides
much better results than an equivalent reduction within Audiotools.
Even though the analog input to my sound interface is constant, if I
crank the Windows Volume Control up and make a comparable gain
reduction within Audiotools, the waveform appears compressed. Crank
the Windows Volume Control even higher and the recorded waveform is
clipped, even though the values in the digital file are nowhere near
maximum. Same behavior is observed recording with Audacity.

I've been assuming this is all happening in the digital domain, where
one would hope math is math, but it clearly it is not! Either the
Windows Volume Control must operate in the analog world before the
ADC, or it must be able to multiply as well as divide, and it must be
able to multiply so much that it can clip in the digital domain.

Another way of saying that would be to say that the (possibly digital
counterpart of the) "0dB" setting of the Windows Volume Control must
be well below its maximum setting. (At least with some audio
hardware...)

To add one more variable, I've found that Alex Mina's AC97 Mixer app,
which talks directly to AC97 audio hardware, can reach gain settings
the Windows Volume Control does not offer. At least with some
hardware...

According to its datasheet, the AKM AK4584 chip in my Transit USB has
a +18dB Input Programmable Gain Amp (0.5dB steps) ahead of the ADC.
With this IPGA set at 0 dB, it is supposed to take 3.0 Vpp of audio to
produce digital full scale. Cranking the gain up to +18 raises the
noise floor 10dB, so it sounds like this really is an analog
amplifier. And I can clearly get clipping with way less than 3.0 Vpp,
so the Transit drivers must have this IPGA set above 0 dB.

Then there are the 0/14/20/26 dB steps that M-Audio offers as "Mic
Boost" in their control panel. I don't see how that is implemented, or
how it relates to the +18 dB IPGA in the audio chip. And I haven't a
clue how the input Windows Volume Control is hooked up to all of this.

So is there a way to tell whether the Windows Volume Control for a
particular interface is operating before or after the ADC? I have a
feeling it is different for different audio devices, and possibly for
different driver versions...

Loren

That is the $64 question! Aside from doing it indirectly by
experimentation and analyzing the digitized waveform, I imagine that
there exists a Windows Audio API that defines all the possible
commands that Windows can send to the audio chip. I see in the chip
below that they say it meets MS PC2001 audio requirements. Perhaps
knowing what MS PC2001 is is a place to start.

http://www.cirrus.com/en/pubs/proDat...CS4205_PP3.pdf

Then you would have to know which chip is on your PC or sound card.
And I guess lastly, you would have to know if the chip has met Windows
Hardware Quality Labs testing. I have no idea what that entails.
And, when you loaded your drivers, it would be nice to not see any
messages appearing asking whether you want to proceed because it does
not meet Windows something or other.

Lately I just started using a Tascam US-122. It has an LED indicating
whether the front end is being overloaded and pots to adjust the input
level. It would be nice if this were the only way to adjust the input
level into the ADC. However, after this discussion, I have some
doubts!


Joe

"Ethan Winer" ethanw at ethanwiner dot com wrote in message

Arny,

Sometimes. There still seem to be soundcards out there with analog
input gain controls.

Thanks for clarifying. I know you have experience with a lot of
consumer grade cards.

I believe that the best analog attenuator *ever* was the CS 3310, which was
developed for high end consumer cards. It had over 93 dB dynamic range and
could handle like 20 or 30 vrms, and cut it down into something palatable
for your typical sound card ADC.

"Ethan Winer" ethanw at ethanwiner dot com wrote in message

Arny,

Sometimes. There still seem to be soundcards out there with analog
input gain controls.

Thanks for clarifying. I know you have experience with a lot of
consumer grade cards.

I believe that the best analog attenuator *ever* was the CS 3310, which was
developed for high end consumer cards. It had over 93 dB dynamic range and
could handle like 20 or 30 vrms, and cut it down into something palatable
for your typical sound card ADC.

In article writes:

Lately I just started using a Tascam US-122. It has an LED indicating
whether the front end is being overloaded and pots to adjust the input
level. It would be nice if this were the only way to adjust the input
level into the ADC. However, after this discussion, I have some
doubts!

Those are real pots and actually attenuate the signal going from the
input jack to the A/D converter. I don't think that it uses the
Windows mixer for record level adjustments.



--
I'm really Mike Rivers )
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me he double-m-eleven-double-zero at yahoo

In article writes:

Lately I just started using a Tascam US-122. It has an LED indicating
whether the front end is being overloaded and pots to adjust the input
level. It would be nice if this were the only way to adjust the input
level into the ADC. However, after this discussion, I have some
doubts!

Those are real pots and actually attenuate the signal going from the
input jack to the A/D converter. I don't think that it uses the
Windows mixer for record level adjustments.



--
I'm really Mike Rivers )
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me he double-m-eleven-double-zero at yahoo

(Mike Rivers) wrote in message news:znr1102117998k@trad...
In article writes:

Lately I just started using a Tascam US-122. It has an LED indicating
whether the front end is being overloaded and pots to adjust the input
level. It would be nice if this were the only way to adjust the input
level into the ADC. However, after this discussion, I have some
doubts!

Those are real pots and actually attenuate the signal going from the
input jack to the A/D converter. I don't think that it uses the
Windows mixer for record level adjustments.

Thanks for that info Mike. That suits me just fine Btw, I bought
this box partly on what you posted in this NG and am quite pleased
with it especially since it has 24 bit DACs. The only bad thing is
that dog **** Cubasis software they threw in which didn't work well
with my XP machine. Fortunately, my old SoundForge Studio 6.0 worked
very well with it. Then, to get the full 24 bits I bought n-Track
Studio which is working quite well so far. Sooooooo for $275 it's a
good deal IMO.

Joe

(Mike Rivers) wrote in message news:znr1102117998k@trad...
In article writes:

Lately I just started using a Tascam US-122. It has an LED indicating
whether the front end is being overloaded and pots to adjust the input
level. It would be nice if this were the only way to adjust the input
level into the ADC. However, after this discussion, I have some
doubts!

Those are real pots and actually attenuate the signal going from the
input jack to the A/D converter. I don't think that it uses the
Windows mixer for record level adjustments.

Thanks for that info Mike. That suits me just fine Btw, I bought
this box partly on what you posted in this NG and am quite pleased
with it especially since it has 24 bit DACs. The only bad thing is
that dog **** Cubasis software they threw in which didn't work well
with my XP machine. Fortunately, my old SoundForge Studio 6.0 worked
very well with it. Then, to get the full 24 bits I bought n-Track
Studio which is working quite well so far. Sooooooo for $275 it's a
good deal IMO.

Joe

In article writes:

Lately I just started using a Tascam US-122.

The only bad thing is
that dog **** Cubasis software they threw in which didn't work well
with my XP machine. Fortunately, my old SoundForge Studio 6.0 worked
very well with it.

I didn't think too much of the Cubasis program either. I guess it
worked OK on with my computers (one XP and one Win2000), it just
didn't work well with me (that's the personal me, not Windows ME).
After proving to myself that the program was indeed functional, I went
to using other software with it when I acutally wanted to use the
interface.

Hey, they put something in the package that let a ground-zero beginner
do something with the US-122 without buying anything else, and that's
better than nothing. These days you can get a copy of Traction for
free (there's a thread or two about it here) if you want to try a
different approach to building songs, but I tried N-Track Studio
myself and that seems to make a fair amount of sense.


--
I'm really Mike Rivers )
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me he double-m-eleven-double-zero at yahoo

On 2004-12-05, Mike Rivers wrote:

Hey, they put something in the package that let a ground-zero beginner
do something with the US-122 without buying anything else, and that's
better than nothing. These days you can get a copy of Traction for
free (there's a thread or two about it here)

Free-ish. But it looks like the arrangement is still licensed to a
single hardware configuration, which means the life expectancy of this
free version of tracktion will be that of whatever machine you install
it on, and beyond that you will be out of luck.

Opinions vary on the relevance of this point, but I consider that to be
toxic. Mackie would indirectly be exerting some control of my art, by
arbitrarily limiting my access to what I create. Totally unacceptable...

I won't bestow the appellation "free" onto anything that seeks to limit
my use of it in any way, whatsoever. Others may restrict their
definition of the word "free" merely to "available at no monetary cost."

But when someone exerts control of my art, the cost is a part of my soul.

I'm totally serious about this.