Driving the IK Multimedia Bus compressor

[4tune]

A question about driving the gain into this compressor, which I've recently picked up and am currently experimenting with on the master bus.

I read in a gearslutz thread that at different threshold levels, ssl bus compressors react differently for the same gain reduction. i.e. if I put a gainer before the compressor and drive different volumes into it, then adjust the threshold accordingly to get say 4dB GR each time, I should get a different sound.

I just wanted to check if this is true for the IK comp? as I am liking higher thresholds while driving higher volume into the compressor. It seems to sound less 'blocky' to me, or am I imaginig it?

[Funkybot's Evil Twin]

If the plugin has some saturation modelling, that might account for the change in sound. I guess this wouldn't be too hard to test out...

1. Take a track and feed into the IK Buss Compressor
2. Render the output
3. Take the original track and add a gain plugin set to +4db before hitting the IK Buss Compressor, then use another plugin to trim the gain -4db
4. Render the output

Now, take both renders and flip the phase. If you get silence, then it's not reacting to input signal at all. Again, my guess here is that you will NOT get silence, and there's some saturation modelling going on. I'd be surprised if there was an audible change to the actual compression behavior, but who knows.

[Burillo]

no need to render anything, you can do that on two tracks with flipped phase

[Funkybot's Evil Twin]

no need to render anything, you can do that on two tracks with flipped phase

Rendering is key!

Ok...fine...perhaps it's a completely unnecessary step.

[4tune]

Thanks both of you!

I tried this inverting polarity using the panipulator plugin http://www.recordingreview.com/blog/too ... nipulator/ and two tracks playing the same thing.


-With the compressor at exactly the same settings on either track and no gainers = silence, so all working well.

-With the +4dB before and -4dB after you get silence if the signal low and not hitting the threshold

-If you run the source high enough to trigger the compression, you hear something.
Note in this scenario the threshold of the compressor being driven by 4dB is adjusted 4dB higher so it 'should' be compressing the same GR.

So I guess the question is what is causing the sound difference?

Maybe I'll try this with some another compressor for comparison.

[4tune]

Ok I tried this on some other comps:

Standard renoise bus compressor = silence

Cytomic Glue demo = silence

Reacomp = silence

Klanghelm DC1A = louder than ik bus comp

So something seems to be happening with the IK bus comp and the DC1A at least when driven.

[Peter - IK Multimedia]

...So something seems to be happening with the IK bus comp...

Yes, the evidence of the extremely accurate, complete, and detailed modeling of the hardware on a circuit level in order to capture all of its nuances to put quite a fine point on it.

Some technical notes to consider pertinent to the tests seen above, brought to you via some of the talented folks involved in creating for you this Bus Compressor plugin:

If boosting the input level by 4dB gives a 4dB increase in output level (which is then compensated) this means there is no compression going on (and I can see that this has been noticed with other plugins and with ours too when it's not compressing so far so good there for the most part...)

When the signal hits the set threshold, instead the gain reduction achieved does depend on a few things, though:

A) By how many dB the signal is over the threshold
B) By the set ratio: if, for instance, the set threshold is 1dB above the signal's level and the level is increased by 4dB that will make it overcome the threshold by 3dBs; then if ratio is set to 2:1 the resulting GR will be 1.5dBs which would require an output compensation of -2.5 dBs, not 4...

Note, though, that the fact that the Bus Compressor is a "soft knee" type essentially makes these algebric calculations mainly theoretical. Bottom line is, whether or not this kind of test is run in a way that takes the above points about gain reduction fully into consideration, it would server further to show that the way our plugin has been modeled takes the real analog hardware behavior and its "anomalies" into account, compression curve being the first one.

[bradleyfilms]

Note, though, that the fact that the Bus Compressor is a "soft knee" type essentially makes these algebric calculations mainly theoretical. Bottom line is, whether or not this kind of test is run in a way that takes the above points about gain reduction fully into consideration, it would server further to show that the way our plugin has been modeled takes the real analog hardware behavior and its "anomalies" into account, compression curve being the first one.

Yep, this is why I love it.
Steve

[4tune]

Yes, the evidence of the extremely accurate, complete, and detailed modeling of the hardware on a circuit level in order to capture all of its nuances to put quite a fine point on it.

Some technical notes to consider pertinent to the tests seen above, brought to you via some of the talented folks involved in creating for you this Bus Compressor plugin:

If boosting the input level by 4dB gives a 4dB increase in output level (which is then compensated) this means there is no compression going on (and I can see that this has been noticed with other plugins and with ours too when it's not compressing so far so good there for the most part...)

When the signal hits the set threshold, instead the gain reduction achieved does depend on a few things, though:

A) By how many dB the signal is over the threshold
B) By the set ratio: if, for instance, the set threshold is 1dB above the signal's level and the level is increased by 4dB that will make it overcome the threshold by 3dBs; then if ratio is set to 2:1 the resulting GR will be 1.5dBs which would require an output compensation of -2.5 dBs, not 4...

Note, though, that the fact that the Bus Compressor is a "soft knee" type essentially makes these algebric calculations mainly theoretical. Bottom line is, whether or not this kind of test is run in a way that takes the above points about gain reduction fully into consideration, it would server further to show that the way our plugin has been modeled takes the real analog hardware behavior and its "anomalies" into account, compression curve being the first one.

Thanks Peter!

nice to have an official response on this, and also the detail with it!

I'll have to digest the points a bit more but always good to confirm that my ears aren't tricking me.


Just a small point with all this in mind it would be nice to have an input drive parameter on the GUI to speed trying to home in on the sweet spot.

Cheers!

Yep, this is why I love it.
Steve

Certainly seems to make it a more interesting comp!

For anyone interested this is the post at gearslutz I was refering to earlier:
http://www.gearslutz.com/board/1832080-post20.html

[4tune]

When the signal hits the set threshold, instead the gain reduction achieved does depend on a few things, though:

A) By how many dB the signal is over the threshold
B) By the set ratio: if, for instance, the set threshold is 1dB above the signal's level and the level is increased by 4dB that will make it overcome the threshold by 3dBs; then if ratio is set to 2:1 the resulting GR will be 1.5dBs which would require an output compensation of -2.5 dBs, not 4...

Just looking at this again in the test I did the only difference between the two plugin settings was the threshold.

The gain was set to compensate the threshold difference (+/-4) only, not the gain reduction. Both compressors were compressing to the same ratio and theoretically the same GR.

So for A)

The dB signal over the threshold readout was exactly the same

Code: Select all

Undriven plug:   No Gainer    ->    threshold -10dB
Driven plug:     Gainer +4dB  ->    threshold -6dB  -> Gainer -4dB 

for B)

All other parameters were the same.


Interestingly I just fired the test up again and if you switch the grit button on, on one of the compressors, the difference becomes more audible still, which I guess should be expected too!

[Peter - IK Multimedia]

Interestingly I just fired the test up again and if you switch the grit button on, on one of the compressors, the difference becomes more audible still, which I guess should be expected too!

Yes indeed.... this makes me smile

Also, thank you for the kind words and appreciation of the detail from our response (it wasn't mine overall, though I did flesh it out a bit.. I haven't worked directly on these plugins' creation but I have a somewhat unique and hopefully worthwhile perspective on hardware and plugins based on the meat of my education and experience being formed during a very transitional/pivotal period for recording technology with a great hybrid of analog and digital and I'm happy for that timing and now being able to relish in a lot of the more cumbersome parts the process like razor blades etc being unnecessary in my work flow thanks to these advances in technology! In other words I'm old but not TOO old.)

Thanks also for the suggestion and your additional detail to explain your testing and methodology further. I'm glad the response from me/IK wasn't seen as challenging your methods as I was merely pointing out a few things that our folks noticed when looking at the thread and test information. Mainly, though, it resulted in some pride too since your tests did basically affirm their hard work and that maniacal attention to detail resulting in capturing some of the more elusive qualities of the gear that is modeled.

[4tune]

Interestingly I just fired the test up again and if you switch the grit button on, on one of the compressors, the difference becomes more audible still, which I guess should be expected too!

Yes indeed.... this makes me smile

Also, thank you for the kind words and appreciation of the detail from our response (it wasn't mine overall, though I did flesh it out a bit.. I haven't worked directly on these plugins' creation but I have a somewhat unique and hopefully worthwhile perspective on hardware and plugins based on the meat of my education and experience being formed during a very transitional/pivotal period for recording technology with a great hybrid of analog and digital and I'm happy for that timing and now being able to relish in a lot of the more cumbersome parts the process like razor blades etc being unnecessary in my work flow thanks to these advances in technology! In other words I'm old but not TOO old.)

Thanks also for the suggestion and your additional detail to explain your testing and methodology further. I'm glad the response from me/IK wasn't seen as challenging your methods as I was merely pointing out a few things that our folks noticed when looking at the thread and test information. Mainly, though, it resulted in some pride too since your tests did basically affirm their hard work and that maniacal attention to detail resulting in capturing some of the more elusive qualities of the gear that is modeled.

Hey it's all good, I learnt something doing the tests and when is there ever not misunderstanding or further explanation needed when discussing technicalities?.. particularly on a forum

Agree it's a great time to be a musician these days even if a bit overwhelming at times!

I Look forward to see what you guys at IK and similar companies come out with next!

[4tune]

@ Peter

I've just checked out the CS manual for the bus comp and this is what it says about the threshold

• THRESHOLD:this control sets the minimum level above which the signal is compressed. Lower settings of this control result in a more compressed signal. This control has no fixed settings; it offers continuously variable values from +20 to -20.

Is this: "it offers continuously variable values from +20 to -20" referring to what we are talking about in this thread?

If so could I suggest that you add something like "different threshold values will give varying compression characteristics even when hitting the same gain reduction"