Fooling around with Compellor
Steve Holt
automatic. I'm really trying to use it mostly as a leveler, and get hotter mixes to tape. Anyone have experience with this unit? What have you used it for, and what kind of results did you get? Any
"back-door" settings?
Steve Holt
http://www.inner-music.com/
Ty Ford
Hi Steve,
I've been using a Compellor for years and like it a lot. One of the reasons
I like it is that it's usually invisible.
I got it to review many years ago. I put a piece of classical music through
it, first noting where all of the dynamic peaks and valleys were. With the
Compellor ON, those peaks and valley ceased to exist. If you have players
who lack consistency, the Compellor may help, but I'm not sure it would be
my first choice for MAKING EVERYTHING REALLY, REALLY LOUD!
I also used it on a madrigals project year before last and it was as if the
performers all took a step and a half closer to the mic. (Which was judged
to be a good thing in that particular case.)
I normally use it to track commercial or narration V/O voices, just to beef
them up a bit.
I normally don't use much more than 4-8dB of gain reduction, but you can get
differnt density sounds if you make it sweat.
Hope this helps.
Regards,
Ty Ford
Ty Ford's audio equipment reviews and V/O sound files can be accessed at
http://www.jagunet.com/~tford
WillStG
>I bought an Aphex Compellor 320A to use primarily as a 2-buss, and finally
started to fool around with it tonight. What an unusual machine, with so much
complexity, and yet so much of its operation is
automatic. I'm really trying to use it mostly as a leveler, and get hotter
mixes to tape<
Be very careful - you can crush the hell out of your program feed and
not have a clue that you're doing it. This is actually a good thing for TV and
film; for smoothing out badly edited video tapes or leveling out a track of
dialogue it works very well. But a/b a lot if you decide to use it on your 2
bus for music, and be sure not to use it post fader on a group or you will
wonder why no matter how high you push your fader nothing happens. ( Heh heh...
)
That being said if you're recording instrumments with a classical or
similar acoustic music approach you might find the Compeller very useful for
bringing up the soft passages smoothly. As for settings I like to leave the
"limiter" button off, and you might start with all the knobs at 12 o'clock (
except for the expansion gate. )
Cheers,
Will Miho
NY Music & TV Audio Guy
Fox And Friends/Fox News
"The large print giveth and the small print taketh away..." Tom Waits
David Morgan (MAMS)
"Steve Holt" <simpli...@yahoo.com> wrote in message ...
> I bought an Aphex Compellor 320A to use primarily as a 2-buss, and finally
> started to fool around with it tonight. What an unusual machine, with so much
> complexity, and yet so much of its operation is
> automatic.
Considering its most common use, it has to be idiot proof - like, "set and forget".
It's basically designed for broadcast levelling / compression. I use the heck out
of it in the studio - usually leaning toward way more compression than levelling.
> I'm really trying to use it mostly as a leveler, and get hotter mixes to tape.
Levelling can be a bit unnatural and overly mess with, or even create it's
own *artificial* dynamics to the program.
Its 'sister' piece, the Aphex "Dominator" (multiband peak limiter) is what will
truly save your arse if you're just looking for hotter levels to tape. Transients
tend to get by even the better compressors, which have to be somewhat
mangled from normal compression settings to become limiters. Most of
them just aren't transparent or fast acting enough to brick-wall limit.
The limiter built into the 320A doesn't hold a candle to the Dominator, and
I don't recommend relying on it at all. My favorite position for the 'limiter'
function on the Compellor is "off" unless that's all you've got. In which case,
keep your ears open when it's working - you might not like it.
> Anyone have experience with this unit?
Around 15 or so years for me.
> What have you used it for,
Everything. These days it sometimes sees a little vocal work and goes
on my two-mix quite often also - but *very* sparingly (as in2 to 4 dB of
reduction with the DVG lamp almost constantly active). Folks doing 'live'
to 2-track recording on the fly usually get a heavier dose. Although this
box is *extremely* transparent, the harder you drive it, the change in the
natural dynamics becomes audible quickly and does have a significant
impact on the clarity of the result. As has been mentioned... A/B the
before and after often.
> and what kind of results did you get?
To me, better than I ever expected. I don't care for compression in general,
but when you can see the metering indicate how well it's working, yet not
hear it.... well, I can live with that.
> Any
> "back-door" settings?
Totally dependant on your signal path. As you said earlier, the blessed thing
is practically automatic. From a +4 output of a D&R on something like a rock
mix, I'll be looking at 'drive' around 10 o'clock, 'process balance' at around 2 or
3 o'clock depending on the program material, with 'output' as necessary for the
make-up gain, and the built-in limiter bypassed. Remember, it also has 3
selectable input operating levels, -10, +4, +8. If you're overdriving the inputs,
try a higher setting.
Basically I set and forget when it comes to the 2-mix. If the program material
being mixed falls below it's normal "protective" operating level, I'll turn it off
and forget it. If it drives the unit too hard, I'll reduce the drive or lean back
toward the levelling side, and increase the make-up output. Too much of
the levelling function can leave the dreaded "pumping' anomaly - even with
the high quality of the DVG circuitry.
It's a cool tool in my book. Keep up the experimentation. Don't expect
to "hear" the device working unless you're just totally overdoing it for the
sake of using it as an effect. It's *very* transparent.
--
David Morgan (MAMS)
Morgan Audio Media Service
Dallas, Texas (972) 622-1972
_______________________________________
Main Street Studio, Garland, TX (972) 487-4045
Steve Holt
you recommend?
dr.sound
> Very helpful post. Thank you. BTW, the Dominator comes in two flavors, the
720, and the 722, the latter having switchable pre and de-emphasis EQ
curves, as well as a higher price tag. Which model do
> you recommend?
>
> "David Morgan (MAMS)" wrote:
>
> > "Steve Holt" <simpli...@yahoo.com> wrote in message ...
> >
>
The 722 with the switchable Pre and de-emphasis is for FM stations.
You should get the 720 and follow it with the Aphex Dominator.
Marti D. Humphrey C.A.S.
aka dr.sound
David Morgan (MAMS)
"Steve Holt" <simpli...@yahoo.com> wrote in message ...
> Very helpful post. Thank you. BTW, the Dominator comes in two flavors, the 720,
> and the 722, the latter having switchable pre and de-emphasis EQ curves, as
> well as a higher price tag. Which model do you recommend?
I've only used the 720. I see no need for 'emphasis' outside of broadcast use.
PS: I mentioned earlier that I might change setting on the Compellor if the
2-mix program material is driving it too hard - actually, my first option would
be to simply lower the master fader a bit and leave the unit alone.
Jay Kahrs
idiot I sold it, wish I still had it. Anyway, I used to use it all the time on
bass on vocals when the player was inconsistent because it did a great job of
removing dynamics and staying rather invisible. Also, I used to use it when
people wanted to throw a sample onto the multitrack from VHS or an audio
cassette. I borrowed a second one and tried it on the mix bus but didn't like
it. A Dominator would probably be cooler if your looking for more "squash".
---
-Jay Kahrs
Mad Moose Recording Inc.
(formerly BrownSound Studios)
Livingston, NJ
http://www.madmooserecording.com
Robert Orban
I can hear the Compellor's compressor the moment the "process balance" control
is moved away from "level only." In case after case, I've listened to radio
stations off air, observed "there's something wideband whacking away in there"
and sure enough, it turns out to be a Compellor with "compression" mode engaged,
placed ahead of the main audio processor (Optimod or whatever).
OTOH, I find them quite innocuous and useful when the "process balance" control
is set for "level only."
(BTW, I seem to have an unusual sensitivity to the "spectral gain modulation"
sound of wideband compression, and I that find many people don't notice that
artifact even when it's pointed out to them. So, as always, YMMV.)
In article <F7074D21B966AA74.BD321778...@lp.airnews.net>,
ma...@airmail.net says...
Casey
<< It's a cool tool in my book. Keep up the experimentation. Don't expect
>to "hear" the device working unless you're just totally overdoing it for the
>sake of using it as an effect. It's *very* transparent. >>
Man, is that an understatement! I owned a 320a for a couple of years and
constantly went back and forth between thinking it was the greatest and
wondering what the heck it was doing and why I couldn't hear it...
I finally got rid of it for a ridiculously low price. And have kicked myself
fairly often ever since...
I still don't quite trust anything I can't hear!
Casey
David Morgan (MAMS)
"Robert Orban" <ror...@earthlinkxyxy.net> wrote in message ...
> If I might respectifully disagree:
>
> I can hear the Compellor's compressor the moment the "process balance" control
> is moved away from "level only." In case after case, I've listened to radio
> stations off air, observed "there's something wideband whacking away in there"
> and sure enough, it turns out to be a Compellor with "compression" mode engaged,
> placed ahead of the main audio processor (Optimod or whatever).
Mr. Orban,
Thank you to an extreme for taking the time to reply. I am truly appreciative
that you and your experience remain available on this forum.
If your dissagreement was regarding the Compellor's true "transparency" when in
compression mode, then my mileage doesn't vary so much. ;-)
In my application of using the Compellor as a compressor *very lightly* across
the two-mix, I reach a dilemma regarding the anomalies of 'pumping' versus the
loss of ability (for lack of a better wording) to maintain 'perfect clarity'. Though this
device remains on the top of my list in it's price range for transparency, I do indeed
hear the losses when gain reduction begins to show on the meters.
> OTOH, I find them quite innocuous and useful when the "process balance" control
> is set for "level only."
>
> (BTW, I seem to have an unusual sensitivity to the "spectral gain modulation"
> sound of wideband compression, and I that find many people don't notice that
> artifact even when it's pointed out to them. So, as always, YMMV.)
When operating the unit with the process balance heavy toward the "level" side
however, pumping becomes more apparent to me. As a matter of fact, it is
difficult to see, via metering, any compression at all taking place - given that the
input "drive" is not set to exceed +12 on the metering. (In my scenario, there
would be no other device strapped across the 2-mix. The Dominator might be
used on mixes of a hard-rock nature, applied post the Compellor).
Perhaps I am not driving the unit hard enough in "level" mode, though that seems
unlikely if pumping is already discernable before gain reduction is apparent on the
meters. At any rate, it seems I could be making a grave error - not just on my work,
but having quoted this style of application to others. To me, the rise a fall of the
average program level is more of a negative than the modulation distortion in
light doses. If there is an alternative method to achieving light compression
using only this device in "level only" mode, I could use the advice. Is there
something that I am overlooking in the setup of the Compellor ?
If your dissagreement is with the manner in which the Compellor was used (or
abused) across a stereo mix buss rather than in broadcast applications, could I
beg of you please, to give some examples of the manner in which you might
apply the device in such a scenario. Also, please correct any errors I may have
made in my response to Steven. Any operating tips would be appreciated by both
of us, I'm certain.
Thank you,
Robert Orban
>>When operating the unit with the process balance heavy toward the "level"
side
>however, pumping becomes more apparent to me. As a matter of fact, it is
>difficult to see, via metering, any compression at all taking place - given
that the
>input "drive" is not set to exceed +12 on the metering. (In my scenario,
there
>would be no other device strapped across the 2-mix. The Dominator might be
>used on mixes of a hard-rock nature, applied post the Compellor).
>
>Perhaps I am not driving the unit hard enough in "level" mode, though that
seems
>unlikely if pumping is already discernable before gain reduction is
apparent on the
>meters. At any rate, it seems I could be making a grave error - not just
on my work,
>but having quoted this style of application to others. To me, the rise a
fall of the
>average program level is more of a negative than the modulation distortion
in
>light doses. If there is an alternative method to achieving light
compression
>using only this device in "level only" mode, I could use the advice. Is
there
>something that I am overlooking in the setup of the Compellor ?
I think it's more an issue of context. In broadcast, the Compellor is
usually employed as an AGC, and it's expected to level out VERY wide
variations in level (perhaps up to 15 or 20dB), usually caused by jocks who
are too busy (or careless) to watch their level carefully.
In LEVEL mode, the Compellor sounds to me as innocuous as a "hand on the
pot." That is, one might also hear equivalent manual gain riding (and one
might object to it), but the Level mode does not cause the constant change
in _texture_ that the Compress mode does.
>
>If your dissagreement is with the manner in which the Compellor was used
(or
>abused) across a stereo mix buss rather than in broadcast applications,
could I
>beg of you please, to give some examples of the manner in which you might
>apply the device in such a scenario. Also, please correct any errors I may
have
>made in my response to Steven. Any operating tips would be appreciated by
both
>of us, I'm certain.
I think the Compellor's compressor is audible because, among other things,
it is not only wideband by also has a very high compression ratio and this
is not adjustable. I'm sure that Donn Werrbach (the Compellor's designer) is
aware of the artifacts produced by its compression, because he desiged
another device (the Dominator) that will do fast compression and limiting
much more smoothly, being a multiband device.
If you _want_ a wideband compression sound, then certainly the Compellor is
one available "flavor." But if you want to stay with Aphex products and want
the most "invisible" compression available from that company, the Dominator
is definitely to be preferred, IHMO.
Because the Compellor's compressor has very few "color" controls, there is
little that the operator can do to affect the sound other than using it in a
sidechain, ala Dolby.
Bob Orban
Richard Kuschel
>From: ror...@earthlinkxyxy.net (Robert Orban)
>Date: 03/06/2001 8:08 PM Mountain Standard Time
>Message-id: <3aa5a646$0$51...@wodc7nh6.news.uu.net>
Not really wanting to take issue, but it is my experience that the Compellor is
a compressor with a very low ( not high) compression ratio.
As the Process knob is turned from level to compress the ratio increases
slightly and the attack speed increases. There is also a limiter function on
the Compellor, but it is fixed and doesn't really work all that well.
The Dominator is not a compressor. It is a multiband peak limiter that allows
different density settings. Use the least density that does not show
distortion it the Low frequencies.
As a limiter, the ratio is very high and pretty much a brick wall.
I have found that the Dominator by itself does indeeed limit and works very
well for protecting speakers ( and ears on in ear monitors) though the
resulting sound is not very natural.
The Combination of a Compellor and a Dominator works very well to control
dynamics transparently.
The controls are not intuitive and a lot of listening is required to get the
best from this combination.
I use this combination for getting classical work to translate well on
television.
Richard H. Kuschel
"I canna change the law of physics."-----Scotty
Bob Ross
> a compressor with a very low ( not high) compression ratio.
FWIW, the Compellor Owner's Manual states:
"It is a soft knee compression characteristic, with the ratio varying from 1.1:1
at threshold (30 dB below nominal 0VU), increasing to 20:1 at 0VU."
David Morgan (MAMS)
"Robert Orban" <ror...@earthlinkxyxy.net> wrote in message ...
Thanks for the reply, Bob. It just shoed up on my server today.
> I think the Compellor's compressor is audible because, among other things,
> it is not only wideband by also has a very high compression ratio and this
> is not adjustable. I'm sure that Donn Werrbach (the Compellor's designer) is
> aware of the artifacts produced by its compression, because he desiged
> another device (the Dominator) that will do fast compression and limiting
> much more smoothly, being a multiband device.
I definitely 'hear' the comp. Not as bad as many, IMVHO.
> If you _want_ a wideband compression sound, then certainly the Compellor is
> one available "flavor." But if you want to stay with Aphex products and want
> the most "invisible" compression available from that company, the Dominator
> is definitely to be preferred, IHMO.
I use them as a pair. Never considered the Dominator as "compression", which
is why I use the Compellor very lightly followed by the Dominator.
My questionable habit came from hearing more pumping on the level side.
> Because the Compellor's compressor has very few "color" controls, there is
> little that the operator can do to affect the sound other than using it in a
> sidechain, ala Dolby.
Thanks for the words. I'll be experimenting with this more. It's a shame that
one (me) can fall into a pattern and overlook experimentation after having
used a device in a certain fashion for several years. I'll try leaning toward
levelling and experimenting with both devices.
Richard Kuschel
That isn't what my Owner's Manual says. From Page 2-2 Operational
Charecteristics.
"The level detector has a "soft knee" compression characteristic, with the
ratio varying from 1.1to 1 at threshold (30 dBbelow nominal VU), increasing to
3 to 1 at 0VU. The attack and release times are automatically varied from 5 to
50 ms for attack and 200 to 1000 ms for release
The leveling process provides a slow and smooth control over the program level
platform. It has a fairly steep ratio (20to 1) with variable attack and release
times to simulate the way the ear perceives loudnessover long periods. The
attack and release times vary from preset values of 2.5s and 5s, respectively.
The circuit's most useful purpose is to maintain a consistant compression
depthand therefore quality regardless of the long term program input level.
Certain sources such as public addrewss systems , can have extreme, quick level
changes. Those sources require faster release times on the leveler and the
speed is switchable from the front panel.
The peak limiter has a fixed threshold 12dB abovenominal 0VU with an attack
time of 1 microsecond and release time of 10 miceoseconds."
I recognize Bob Orban's expertise in this field and if I am confused about
implementation, I invite further clarification.
Bob Ross
I wrote:
> >FWIW, the Compellor Owner's Manual states:
> >
> >"It is a soft knee compression characteristic, with the ratio varying from
> >1.1:1
> >at threshold (30 dB below nominal 0VU), increasing to 20:1 at 0VU."
> >
> >
>
then
Richard Kuschel wrote:
>
> That isn't what my Owner's Manual says. From Page 2-2 Operational
> Charecteristics.
>
> "The level detector has a "soft knee" compression characteristic, with the
> ratio varying from 1.1to 1 at threshold (30 dBbelow nominal VU), increasing to
> 3 to 1 at 0VU. The attack and release times are automatically varied from 5 to
> 50 ms for attack and 200 to 1000 ms for release
>
> The leveling process provides a slow and smooth control over the program level
> platform. It has a fairly steep ratio (20to 1) with variable attack and release
> times to simulate the way the ear perceives loudnessover long periods. The
> attack and release times vary from preset values of 2.5s and 5s, respectively.
> The circuit's most useful purpose is to maintain a consistant compression
> depthand therefore quality regardless of the long term program input level.
> Certain sources such as public addrewss systems , can have extreme, quick level
> changes. Those sources require faster release times on the leveler and the
> speed is switchable from the front panel.
>
> The peak limiter has a fixed threshold 12dB abovenominal 0VU with an attack
> time of 1 microsecond and release time of 10 miceoseconds."
I assume your Compellor is a more current 320 or 320A model?. My Compellor is a
much earlier version, circa 1986, hence the discrepancy. I noticed the ad copy
for the 320A states the same specs as your manual. With the exception of that one
sentence that I quoted (above) regarding the 20:1 at 0VU, my manual says word for
word what yours does. It's probably not a missprint, since it reiterates that
spec in section 7.1 Operating Instructions as well as 4.0 Functional Description.
Seems safe to assume that the older Compellors would be even *more* audible than
the newer ones due to this higher compression ratio...yet even my ancient unit has
been blessed as "the invisible compressor" by every engineer who's used it.
Bizarre.
/Bob
Robert Orban
<FE6C66DFF83B236B.C1A451D7...@lp.airnews.net>,
ma...@airmail.net says...
>
>
>
>"Robert Orban" <ror...@earthlinkxyxy.net> wrote in message ...
>
>Thanks for the reply, Bob. It just shoed up on my server today.
>
>> I think the Compellor's compressor is audible because, among other things,
>> it is not only wideband by also has a very high compression ratio and this
>> is not adjustable. I'm sure that Donn Werrbach (the Compellor's designer)
is
>> aware of the artifacts produced by its compression, because he desiged
>> another device (the Dominator) that will do fast compression and limiting
>> much more smoothly, being a multiband device.
>
>I definitely 'hear' the comp. Not as bad as many, IMVHO.
>
>> If you _want_ a wideband compression sound, then certainly the Compellor is
>> one available "flavor." But if you want to stay with Aphex products and
want
>> the most "invisible" compression available from that company, the Dominator
>> is definitely to be preferred, IHMO.
>
>I use them as a pair. Never considered the Dominator as "compression", which
>is why I use the Compellor very lightly followed by the Dominator.
When folks start talking about using 2-3dB "compression" on an overall mix (as
I have seen many times in this newgroup), then the difference almost becomes a
matter of semantics. This is more typical of the depth of gan reduction
usually used when one does peak limiting.
>
>My questionable habit came from hearing more pumping on the level side.
>
>> Because the Compellor's compressor has very few "color" controls, there is
>> little that the operator can do to affect the sound other than using it in
a
>> sidechain, ala Dolby.
>
>Thanks for the words. I'll be experimenting with this more. It's a shame
that
>one (me) can fall into a pattern and overlook experimentation after having
>used a device in a certain fashion for several years. I'll try leaning
toward
>levelling and experimenting with both devices.
I should also comment that I am mainly familiar with the older version of the
Compellor, and I was not aware that the final ratio on the compressor had
evidently been changed from 20:1 to 3:1 when the model number changed.
Bob Orban
Richard Kuschel
>is a
>much earlier version, circa 1986, hence the discrepancy. I noticed the
>ad copy
>for the 320A states the same specs as your manual. With the exception of
>that one
>sentence that I quoted (above) regarding the 20:1 at 0VU, my manual says
>word for
Yes, mine is a 320.
I fell in love with the Compellor long before the 320 was built though.
I first heard one in 1984, or maybe I should say I didn't hear it.
I agree with " invisible compressor " remark.
Ty Ford
The string follows his response:
Regards,
Ty Ford
Dear Ty,
It is inexplicable to me that there would be audible pumping without any
indication of gain reduction, whether the process control were in pure leveling
or compression.
The entire idea of the Compellor is to combine high ratio, slow moving leveling
with low ratio, faster moving compression. Bob was completely wrong about the
compression in the Compellor being high ratio. The onset of compression starts
at 1.1:1 and increases slowly to 8:1 at 20dB of pure compression. Using a
combination of leveling and compression prevents the compressor being at a high
ratio for any length of time.
I would recommend that the process balance be set at about 10:30 to 11 o'clock.
Use fast leveling. Decide how much you want to bring up the lowest level signal
(2, 4, 10, 20dB?). With a 0Vu input, generate that amount of gain reduction.
Adjust the output for 0Vu out and away you go.
Bob was also wrong about the Dominator. It is not a compressor, it is a peak
limiter only. The combination of the Compellor and Dominator is the most
transparent way to control dynamics- the Compellor for average level and the
Dominator for peaks.
If the Dominator is use after the Compellor I recommend turning off the limiter
in the Compellor.
I hope that this was of some help.
Best regards,
Marvin
Ty Ford wrote:
> Marvin, Donn,
>
> I thought you might like to see and/or respond to Bob's comment.
>
> Regards,
>
> Ty Ford
>
> In Article <3aa5a646$0$51...@wodc7nh6.news.uu.net>, ror...@earthlinkxyxy.net
> >I think the Compellor's compressor is audible because, among other things,
> >it is not only wideband by also has a very high compression ratio and this
> >is not adjustable. I'm sure that Donn Werrbach (the Compellor's designer) is
> >aware of the artifacts produced by its compression, because he desiged
> >another device (the Dominator) that will do fast compression and limiting
> >much more smoothly, being a multiband device.
> >
> >If you _want_ a wideband compression sound, then certainly the Compellor is
> >one available "flavor." But if you want to stay with Aphex products and want
> >the most "invisible" compression available from that company, the Dominator
> >is definitely to be preferred, IHMO.
> >
> >Because the Compellor's compressor has very few "color" controls, there is
> >little that the operator can do to affect the sound other than using it in a
> >sidechain, ala Dolby.
> >
> >Bob Orban
Robert Orban
says...
I just saw Marvin Caesar's reply about the Compellor where he says:
"The entire idea of the Compellor is to combine high ratio, slow moving
leveling with low ratio, faster moving compression. Bob was completely wrong
about the compression in the Compellor being high ratio. The onset of
compression starts at 1.1:1 and increases slowly to 8:1 at 20dB of pure
compression. Using a combination of leveling and compression prevents the
compressor being at a high ratio for any length of time."
I must confess to extreme confusion. We now have two numbers on the table
(3:1, 20:1) for the ultimate compression ratio from two Compellor manuals,
neither of which agrees with the number in Mr. Caesar's email. In the 320, is
it 3:1 or 8:1?
Further, I would observe that 8:1 compression _is_ a very high ratio audibly,
and doesn't sound very different from infinity:1. It's the difference between
8dB input level change's producing 1dB output level change vs. 0dB output
level change.
And finally, I must say once again that I trust my ears when it comes to
dynamics processing, and I have heard the "Compellor compressor pumping sound"
on the air over and over again, at least when the older model Compellor was
used as a pre-processor ahead of a main broadcast processor. This sonic
signature is _very_ characteristic and well-defined. It always cleared up when
the Compellor in question was set to "level only," and I think that it works
fine as a broadcast pre-processor with that setting. My intent was not to
attack the product per se, but rather the way it is often used in broadcast,
and to observe that I could hear the fast compression.
This is not surprising. I have never heard a fast, wideband, high ratio
compressor that I _couldn't_ hear working, and, in fact, I have never found a
way to design such a beast myself. The problem is always "spectral gain
modulation," which usually manifests itself as the bass' audibly pumping the
midrange up and down. Because the ear is far less sensitive to bass than to
midrange, it doesn't perceive this wideband control as a natural way to
automatically control levels.
One common band-aid for this is to roll off the bass response of the sidechain
so that the frequency response of the sidechain approximates the sensitivity
of the ear as a function of frequency. I don't consider this to be true
"wideband" compression anymore, however, because it no longer controls
electrical levels consistently--material with lots of bass will always come
out of such a compressor hotter than bass-shy material. If one needs a
combination of low audible self-modulation and consistent level control, then
multiband compression is usually a better solution, because changes in
spectral balance are usually less jarring to the ear than spectral gain IM.
Ty Ford
I'll observe that the Compellor (as well as any compressor) can be made to
pump. The 17 years I spent at various AM and FM radio stations gave me
enough experience to know that the "louder is better" attitude of many
broadcasters means the adjust gain reduction gear WAaay too aggressively.
I have found the Compellor very useful and very invisible. Invisible to the
degree that many people not familiar with what it does think it's not really
working.
As to the differences in figures Bob talks about above. I don't have an
answer other than changes with different models.
I have forwarded the message to Marvin Caesar for a reply.
Regards,
Ty Ford