From Newsgroup: rec.audio.pro

I'm intrigued by the new wave of recorders that
utilize one ADC set to a low gain, and one set
to a higher gain. Backed by 32bit float, it would
seem on the surface that there is no need to adjust
gain, and indeed these new models often don't have
a way to do that.

It seems to me though that there might be some
compromise there. Take very quiet sounds - sounds
that in this case might only use a half, or quarter
or even a tenth of the range of the ADC that is
more sensitive. With a normal recorder, I'd increase
the gain to approach the capacity of the ADC. Now,
I get that I don't have to worry about digital noise
because of the 32bit float, but on the analog side
I'm really just going in with gain set too low for the premp.
I mean isn't that why we generally set gain near the top?
I generally assume that the best noise performance
is achieved at gain levels that bring the signal
close to max level for the preamp. It seems like using the two
ADC's is like giving me a low/high gain switch
instead of a knob.

I'm also curious about mid range and high level
signals. Say the signal is strong enough to
fill the lower sensitivity ADC half way. Does
the signal use only the output of that one ADC
in that case or is the waveform actually patched
together using output from both?

And again, say the signal makes it just 10% of the
way into the lower sensitivity ADC. It seems like
this is the same as recording from a regular preamp
at low gain levels. The signal will have to be raised
along with the noise floor, which I guess is why we
don't usually record with too low a gain level.


Toby
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: rec.audio.pro

On 30/07/2025 1:30 am, Tobiah wrote:

I'm intrigued by the new wave of recorders that
utilize one ADC set to a low gain, and one set
to a higher gain.-a Backed by 32bit float, it would
seem on the surface that there is no need to adjust
gain, and indeed these new models often don't have
a way to do that.

It seems to me though that there might be some
compromise there.-a Take very quiet sounds - sounds
that in this case might only use a half, or quarter
or even a tenth of the range of the ADC that is
more sensitive.-a With a normal recorder, I'd increase
the gain to approach the capacity of the ADC.-a Now,
I get that I don't have to worry about digital noise
because of the 32bit float, but on the analog side
I'm really just going in with gain set too low for the premp.
I mean isn't that why we generally set gain near the top?
I generally assume that the best noise performance
is achieved at gain levels that bring the signal
close to max level for the preamp.-a It seems like using the two
ADC's is like giving me a low/high gain switch
instead of a knob.

I'm also curious about mid range and high level
signals.-a Say the signal is strong enough to
fill the lower sensitivity ADC half way.-a Does
the signal use only the output of that one ADC
in that case or is the waveform actually patched
together using output from both?

And again, say the signal makes it just 10% of the
way into the lower sensitivity ADC.-a It seems like
this is the same as recording from a regular preamp
at low gain levels.-a The signal will have to be raised
along with the noise floor, which I guess is why we
don't usually record with too low a gain level.

Toby

What are they trying to record - nuclear explosions next to ant-whispers
?!!! Surely a 'good 24' would cover all music with headroom, and most
other real-world things.

On a technical front, without deeper explanation, I worry about the 'crossover' region between the ADCs.
--
geoff
--- Synchronet 3.21a-Linux NewsLink 1.2