gyraf

gml opamps
« Reply #80 on: June 30, 2004, 09:25:17 AM »
There is something in the predictability of feed-forward designs that I really like in compressors. I use feedforward in both my G3 and G10 compressors for this reason.

Granted, for exact-threshold limiting nothing beats the feedback design - I think this is why most of the "old" comp/lim types uses this topology - but very often I find the feed-forward more intuitive in use.

What about an adjustable FF/FBCK control, ganged with ratio: Low ratios feed forward, and higher ratios gradually more feedback, ending in feedback-only at limiting. Two rectifier circuits would be needed before the mixing, but timing, threshold, and soforth could still be common.

Or am I missing something here..?

Jakob E.
..note to self: don't let Harman run your company..


DrFrankencopter

gml opamps
« Reply #81 on: June 30, 2004, 09:50:58 AM »
Quote
What about an adjustable FF/FBCK control, ganged with ratio: Low ratios feed forward, and higher ratios gradually more feedback, ending in feedback-only at limiting. Two rectifier circuits would be needed before the mixing, but timing, threshold, and soforth could still be common.


I think this is a great idea, except it would best be implemented in software since the timing in a feedback circuit is a function of the compression ratio.

Cheers,

Kris

gyraf

gml opamps
« Reply #82 on: June 30, 2004, 10:09:11 AM »
And then again - we're not really that interested in "absolute" values for attack/release timing - yes, these would drift a bit with FF/FBCK setting, but not so much as to be unusable. These controls are nearly always set by ear anyway..

Jakob E.
..note to self: don't let Harman run your company..

DrFrankencopter

gml opamps
« Reply #83 on: June 30, 2004, 10:17:09 AM »
Yes. but say for example that at a ratio of 4:1 you might have 50/50 Feedforward/feedback split. If you're using one knob for global attack and one knob for global release, then they should have about the same time constant, but if feedforward compressor has a time constant of T, then a feedback compressor's time constant is T/ratio, so the feedback time constant is shortened when compared to the feedforward. In order to make a blended FF/FB compressor easy to set, I think you'd need to compensate for the change in attack/release time as a function of ratio.

BTW, there's a great paper in the journal of AES by Fred Floru of THAT explaining some of the differences in these topologies.

Cheers,

Kris

SSLtech

gml opamps
« Reply #84 on: June 30, 2004, 12:25:45 PM »
The points about feedback and feed-forward are well taken, but the funny thing is that my favourite and most reached-for compressors all happen to be feed-back... From an engineering point of view, feed-forward is predictable and "does what it's told to do"... but from a recordist point of view, feed-back just ends up sounding better to me...

I wish I had an explanation.

Keith
"A waist is a terrible thing to mind"
Quote from: PRR
Ah, but that was 1999; we don't party like that any more.

gm

gml opamps
« Reply #85 on: June 30, 2004, 05:48:14 PM »
Quote from: "SSLtech"
The points about feedback and feed-forward are well taken, but the funny thing is that my favourite and most reached-for compressors all happen to be feed-back... From an engineering point of view, feed-forward is predictable and "does what it's told to do"... but from a recordist point of view, feed-back just ends up sounding better to me...

I wish I had an explanation.

Keith


Well, quite obviously I feel strongly that feed-forward designs - not only mine but also the dbx designs and some digital designs (which are somewhat more difficult to build in a feed-back topology without going to very, very high sample rates) - sound better, particularly on vocals.  It might be that the feedback designs crunch more, and probably crunch with more obvious idiosynchratic characteristics between respective devices.

But that's something else you'd be going for.  

If effective, transparent dynamic gain control is taken by itself - forget the squishing and crunching (that's not so hard to do anyway) of all of those old boxes we all know so well - then you are obliged to do feed-forward because feed-back is so unpredictable (uneven artifacts & widely variable attack and release timings across operating points and depths of compression).

George

SSLtech

gml opamps
« Reply #86 on: June 30, 2004, 06:09:50 PM »
Quote from: "gm"
Well, quite obviously I feel strongly that feedback designs - not only mine but also the dbx designs and some digital designs...<snip happens>

...okay, I'm puzzled now... I only remember seeing feed-forward dbx offerings... did they also make feed-back offerings?

Your own topology is a mystery to me, George; I've never worked at a place that permanently had one, and the only time that we ever rented one in for a client, it was quickly pressed into use and labelled "do not touch"... so I've never played with it. -I looked over the front panel once though and I recall that the assistants frowned and asked what the hell "crest factor" was... even after a few explanations they still seemed to struggle with the theory, but I digress...

John Klett's forum was down too, Matt, so I suspect that there are some server-ey type things afoot!

Keith
"A waist is a terrible thing to mind"
Quote from: PRR
Ah, but that was 1999; we don't party like that any more.

gml opamps
« Reply #87 on: July 01, 2004, 07:58:05 AM »
I remember when we were on the old server, a member recalled having a supposedly faulty 8900 compressor. It turned out that the 8900 was compressing fine, but he thought it was faulty as he couldn't hear any compression artefacts with the metering in deep GR :green: George would've been proud!


Justin
Prepare yourself. You are about to become the voice of Interplanetary Parliament.

gm

gml opamps
« Reply #88 on: July 01, 2004, 09:17:54 AM »
Quote from: "SSLtech"
Quote from: "gm"
Well, quite obviously I feel strongly that feedback designs - not only mine but also the dbx designs and some digital designs...<snip happens>

...okay, I'm puzzled now... I only remember seeing feed-forward dbx offerings... did they also make feed-back offerings?

[...]
Keith


Sorry, Kieth, I meant feed-forward.  dbx only made feed-forward designs that I know of.  

Typing too fast...

George

SSLtech

gml opamps
« Reply #89 on: July 01, 2004, 09:18:06 AM »
I had the opposite 'problem'... a "boss" that used to fiddle.

He liked to to-tweak the room EQ a lot. (we had a pair of KT 27-bands on the main monitors.) -After he once made an ever-so-slight adjustment and satisfied himself that the change was going to make all the difference in teh workd to the clients, I onserved that the graphics had been in bypass the entire time...! (a client who -wisely- didn't want the dammned things in circuit had been in the day before.)

After that, I had a brainwave... I took the male & female XLRs out of the back and bypassed the whole thing "invisibly".

It tiik moer than a year for him to notice... :green:

Keith
"A waist is a terrible thing to mind"
Quote from: PRR
Ah, but that was 1999; we don't party like that any more.


gm

Graphic EQ's
« Reply #90 on: July 01, 2004, 09:45:10 AM »
...IMNSHO the only correct choice with graphic EQ's.

George

Kev

gml opamps
« Reply #91 on: July 01, 2004, 05:08:23 PM »
Quote from: "SSLtech"
 After that, I had a brainwave... I took the male & female XLRs out of the back and bypassed the whole thing "invisibly".


Keith,
you are quite correct
There is often an un-patched Graphic in Monitor Land of larger PA system and the artist is sometimes invited to have a trim.

You can also find a fake set of compressors for the support band.

GM,
... sometimes we just have to use graphics BUT it does encourage a bit of slackness on our part as it is probably never the correct choice, if we had the budget and time to fix the problem as best we could.
Kev
DIY Factory

Bear

gml opamps
« Reply #92 on: July 01, 2004, 10:45:01 PM »
Remember the 80's when every coked up dipsh*t had a graphic eq built into his amp?  That's really why indie rock was better then - no money for coke or graphic eq.

Bear

mnats

gml opamps
« Reply #93 on: July 02, 2004, 07:27:50 AM »
DBX did make feedback compressors. They were used in the encoder side of their noise reduction units. The reason is interesting: when paired with a feedforward expander and placed back to back the two sides detect and respond to the same signal. Logically, it makes sense but the problem is that they were never hooked up back to back but were passed through the analog recording medium. Interesting to think of the two schemes as essentially cancelling each other out though.

SSLtech

gml opamps
« Reply #94 on: July 02, 2004, 09:12:32 AM »
Quote from: "mnats"
DBX did make feedback compressors. They were used in the encoder side of their noise reduction units. The reason is interesting: when paired with a feedforward expander and placed back to back the two sides detect and respond to the same signal.


Ah... Interesting!

-I'm going to give that some thought... only yesterday I'd said that the only easy way to do NR the wide-band dbx way, would be with feed-forward...

I'm going to have to give this some thought... "This might be a three-pipe problem, Watson!"

Keith

[edited for carelessness]
"A waist is a terrible thing to mind"
Quote from: PRR
Ah, but that was 1999; we don't party like that any more.

Learner

gml opamps
« Reply #95 on: July 02, 2004, 09:31:27 AM »
meow...........
Still into audio gears!

SSLtech

gml opamps
« Reply #96 on: July 02, 2004, 10:47:26 AM »
...

...

single cricket chirp...

...

...

:?:

Keef
"A waist is a terrible thing to mind"
Quote from: PRR
Ah, but that was 1999; we don't party like that any more.

SSLtech

gml opamps
« Reply #97 on: July 02, 2004, 01:03:07 PM »
Whoah... I just realised my mistake...

No offence was intended whatsoever. I apologise for my rather careless and prorbably insensitive remark. I was trying to convey the dawning of enlightenment and not cause any offence.

Ince again, my humble and unreserved apologies.

 :oops:

Keith
"A waist is a terrible thing to mind"
Quote from: PRR
Ah, but that was 1999; we don't party like that any more.

mnats

gml opamps
« Reply #98 on: July 02, 2004, 05:26:25 PM »
Just thinking about it conceptually I'm wondering if the instability of feedback at high ratios can be cancelled out with feedforward in a single unit that does compression only. But when I try to picture it, my brain seems to go into feedback! Maybe it needs a feedforward sidechain fitted.

If Keith wants to be Master Po or Sherlock Holmes I really don't see what the problem is. "This might be a three-pipe problem, Watson!" could be construed as racist too if you wanted to look at it that way. No apology required IMO.

Kev

gml opamps
« Reply #99 on: July 02, 2004, 06:52:03 PM »
I must be very sleepy this morning.  It has been a long week for me.

Even so I have a strong suspicion that I have missed something obvious here as I'm sure I just got left behind.

 :oops:
Kev
DIY Factory


 

Related Topics

  Subject / Started by Replies Last post
5 Replies
2247 Views
Last post February 16, 2007, 12:22:04 PM
by chrissugar
4 Replies
5788 Views
Last post December 31, 2008, 08:08:38 PM
by joe-electro
0 Replies
1160 Views
Last post June 16, 2009, 06:54:07 PM
by mcfate
7 Replies
1833 Views
Last post December 28, 2012, 12:44:07 PM
by r2d2