Feedback in audio esp wrt op-amps.

[Eeyore]

Don't worry, I'm not about substiting all opamps just because I _think_
I can get better tone. It's mostly for things like programmable
compressors build on a NE570+TL082, to be substitutd for a NE572+NE5534.
I'd like to adopt a better, more natural sounding (sideband) compressor
design without breathing artefacts, though I'd have to expect facing
some problems integrating it into this existing programmable design.
Wouldn't mind references to a good sideband compressor, though..

The NE570 and 572 are fairly poor performers designed originally for telephony
IIRC. Have you looked at using dedicated high quality audio VCAs ?

Graham

 

[Scott Dorsey]

The NE570 and 572 are fairly poor performers designed originally for telephony
IIRC. Have you looked at using dedicated high quality audio VCAs ?

Or using the gain-control grid on a 6BE6! You need two of them out of
phase to deal with the linearity issues, but you need to do that with
the NE570, anyway.

THAT makes some nice VCA chips that are available from Mouser and are
good performers, with no need to parallel them and totem-pole them like
with the NE570.
--scott

 

[Mark]

Don't worry, I'm not about substiting all opamps just because I _think_
I can get better tone. It's mostly for things like programmable
compressors build on a NE570+TL082, to be substitutd for a NE572+NE5534.
I'd like to adopt a better, more natural sounding (sideband) compressor
design without breathing artefacts, though I'd have to expect facing
some problems integrating it into this existing programmable design.
Wouldn't mind references to a good sideband compressor, though..

Breathing effects are all about the attack and decay rates and overall
amount of compression ( not the particular op amp or VCA). For ham
radio applications consider two compressors in cascade, first a slow
AGC with 15 dB compression that acts to keep the overall level
correct. Follow that with a fast limiter or clipper with 6 dB
clipping or limiting (consider even a simple clipper here) that chops
off the peaks so you can increase the average.

This was the standard setup in radio stations, a slow AGC follwed by a
fast limiter. The slow AGC makes up for board (bored) operator errors
so that the limiter sees about the same input all the time and the
limiter does the dirty work to make audio louder by clipping or
limiting the peaks.

Since you are looking for communications quality and your SSB
transmitter probably has several % distortion anyway..., the choice of
VCA is not important.

You gotta get the time constants and ratios etc correct however.

Have fun.

Mark

 

[Mogens V.]

So true! The 572 part is a bit to the better side of things.
Nope, haven't yet looked at alternatives. Just started looking into
things to improve on; the crappy compressor is first on the list.

Or using the gain-control grid on a 6BE6! You need two of them out of
phase to deal with the linearity issues

Yes, I've thought about a tube-based solution. One (more) problem is
there's no space for additional tubes in the box, so I'd have to house
them externally with signals and controlwires in and out.

but you need to do that with the NE570, anyway.

Uh oh, hadn't realized the linearity problem..
That the NE570 is hooked up for a max 30:1 ratio doesn't exactly help on
liniarity and breathing

THAT makes some nice VCA chips that are available from Mouser and are
good performers, with no need to parallel them and totem-pole them like
with the NE570.
--scott

I'm looking at the That; thanks both for mentioning it.

Actually, for guitars, I have more use of a sustain function, operating
at levels lower than, and after, the initial natural attack.
The above mentioned max 30:1 compression ratio really is useless, and
only serves to completely demolish a guitar's attack and definition,
even for the heaviest metal. Further, it makes using an envelope
follower to control parameters in proceeding effects less accurate.

It's interesting to observe that many metal players, as they grow in
musical understanding and playing capabilities, tends to drift away from
the overcompressed tone, playing with much more expression and effects
liven up, even when severe distortion is used.

 

[Mogens V.]

Breathing effects are all about the attack and decay rates and overall
amount of compression ( not the particular op amp or VCA)..

Yup, and WRT this, the NE570 wasn't exactly too controlable for music
processing.

Since you are looking for communications quality and your SSB
transmitter probably has several % distortion anyway..., the choice of
VCA is not important.

Ehm, nope, not for SSB apps.. it's just that NE57x was very widely used
in music applications of ye olde times, in this case a guitar processor.

 

[Mark]

Yup, and WRT this, the NE570 wasn't exactly too controlable for music
processing.


Ehm, nope, not for SSB apps.. it's just that NE57x was very widely used
in music applications of ye olde times, in this case a guitar processor.

OK I'm sorry I mis-understood you then...

What did you mean by _sideband_ here....

"I'd like to adopt a better, more natural sounding (sideband)
compressor design without breathing artefacts....."

Mark

 

[Mr.T]

OK I'm sorry I mis-understood you then...

What did you mean by _sideband_ here....

"I'd like to adopt a better, more natural sounding (sideband)
compressor design without breathing artefacts....."

I would guess he actually meant side-chain.
I'm pretty sure he wasn't talking about amateur radio anyway

MrT.

 

[Mogens V.]

I would guess he actually meant side-chain.
I'm pretty sure he wasn't talking about amateur radio anyway

MrT.

yup, maybe I used the wrong description, so to be more verbal:
I mean a compressor not directly in the signal path, but rather where
the compressed part is summed onto the direct signal, leaving the
original dynamics untouched, so the compressed part 'blends in' only
after the initial attact/decay part, thus effectively acting as a
sustain function.

This is very much needed when creating guitar sound. Maybe mostly OT in
this discussion, but still..
AFAIU, many 'normal' in-line (guitar) solid state compressors tend to
create third harmonic distortion at the somewhat fast attack times
mostly used for guitars, plus some breathing/pumping (design problem).
At least for my own taste, I'd prefer using and preserving the natural
dynamic envelope of the instrument, coupled with the ability to prolong
the sustain part of the envelope, with a somehat natural release.
It allows for a much more expressive playing style.

Of cause I could use some external ADSR-based device for more control,
but this doesn't really fit with my otherwise programmable rack - at
least it won't fit my finances hense my attempt to rework it.
My preamp only has comp.ratio, threshold and level parameters, so this
is what I'm left with for another design.

 

[Arny Krueger]

I would guess he actually meant side-chain.

That was my interpretation.

 

[Mark]

yup, maybe I used the wrong description, so to be more verbal:
I mean a compressor not directly in the signal path, but rather where
the compressed part is summed onto the direct signal, leaving the
original dynamics untouched, so the compressed part 'blends in' only
after the initial attact/decay part, thus effectively acting as a
sustain function.
...

OK that's not sidechain either.... sidechain is when one audio signal
is applied to the control part of the compressor and is used to
controal another audio signal. i.e announcers voice used to lower the
gain of (or duck) the music is the usual example.

What you are asking for is done with an ordinary compresser and then
use an ordinary mixer to combine the wet (compressed) and dry
(original uncompressed) signal. You can use the two faders of the
mixer channels to vary the amount of wet / dry to your taste.

Mark