I use a compressor on the bus to "glue" elements of the mix together. I use SSL but there are a few out there that can achieve this gluing effect. I want to use exclusively Melda plugins, but can't find the right way to achieve this. Simply copying the same settings into MDynamics like attack, release, ratio etc doesn't get the same sound.
So my question is what is the glue in a glue compressor what makes them better suited to buss compression?
Is it saturation?
Is it a certain transfer curve of the attack?
I'm not trying to emulate completely the SSL, but I would like to get a gluing effect from any of the Melda plugins.
Thanks.
Glue Compressor
-
- KVRist
- 140 posts since 27 Jul, 2015
Here's an excellent article about bus compression:
http://www.soundonsound.com/sos/may08/a ... ession.htm
I tried to achieve "glue" compression with some Melda plugins, but since I'm still learning all these mixing techniques, I couldn't achieve anything useful. In my attempts the effect was either too subtle, or too perceptible, I couldn't find the correct balance. I decided I would get back to it when more familiar with the technique.
I think it also depends on the kind of sound you're after: many of the renowned glue compressors seem to somewhat color the sound, whereas Melda plugins are usually very transparent (e.g. I don't think any of the compressors have saturation settings; only the limiters). Maybe this can be achieved by adding MAmp in the chain?
http://www.soundonsound.com/sos/may08/a ... ession.htm
I tried to achieve "glue" compression with some Melda plugins, but since I'm still learning all these mixing techniques, I couldn't achieve anything useful. In my attempts the effect was either too subtle, or too perceptible, I couldn't find the correct balance. I decided I would get back to it when more familiar with the technique.
I think it also depends on the kind of sound you're after: many of the renowned glue compressors seem to somewhat color the sound, whereas Melda plugins are usually very transparent (e.g. I don't think any of the compressors have saturation settings; only the limiters). Maybe this can be achieved by adding MAmp in the chain?
-
- KVRAF
- Topic Starter
- 2696 posts since 9 Jul, 2015 from UK
Thanks, good article. Was an interesting read. I usually use an SSL or vari mu. The later uses a tube. I experimented a little with using MDynamics and MAmp for the make up gain. I've had another idea too!:
I think a lot of the sound is down to the attack and release shapes, MDynamics allows us to draw in our own custom shape. There is a plugin by DMG audio called 'Compassion' that emulates vintage compressers. It has a visual attack and release curve window, that changes depending if you load different compressor models. I might have a go at trying to draw these curves into MDynamics and see how it sounds.
I think a lot of the sound is down to the attack and release shapes, MDynamics allows us to draw in our own custom shape. There is a plugin by DMG audio called 'Compassion' that emulates vintage compressers. It has a visual attack and release curve window, that changes depending if you load different compressor models. I might have a go at trying to draw these curves into MDynamics and see how it sounds.
Jason @ Melda Production
-
- KVRist
- 140 posts since 27 Jul, 2015
It's true that the plugin I currently use add a bit of non-linearity to the compression curve when in bus mode. But I think another aspect is the "dynamic" distortion, i.e. a (subtle) distortion that is applied only to the compressed part (and proportionally to the compression factor).
I'm not sure how to achieve that with Melda plugins. Maybe with MXXX?
I'm not sure how to achieve that with Melda plugins. Maybe with MXXX?
-
- KVRist
- 140 posts since 27 Jul, 2015
When the input signal is under the compression threshold, there is no distortion. As soon as it's over the threshold, distortion kicks in, and the distortion amount is proportional to how much above the threshold the signal is. (I guess with a soft knee and non-linear compression curve, things get more complicated.)
At least that's my understanding of what it does.
At least that's my understanding of what it does.
-
- KVRAF
- Topic Starter
- 2696 posts since 9 Jul, 2015 from UK
Ok, I understand.
I have been doing some more research and found some more interesting things to try.
Firstly the difference between feedforward and feedback. Most 'musical' bus compressors are feedback (SSL bus compressor for example)
A compressor in simplest form has a level detector which controls a VCA (level) if the detector is after the VCA it is a feedback compressor.
In MXXX we can use the follower (not the global mod, rather the in module fx version) placed after the amp fx to control the output of the amp. Leaving the input static would equal more distortion relative to the input level, right?
I have been doing some more research and found some more interesting things to try.
Firstly the difference between feedforward and feedback. Most 'musical' bus compressors are feedback (SSL bus compressor for example)
A compressor in simplest form has a level detector which controls a VCA (level) if the detector is after the VCA it is a feedback compressor.
In MXXX we can use the follower (not the global mod, rather the in module fx version) placed after the amp fx to control the output of the amp. Leaving the input static would equal more distortion relative to the input level, right?
Jason @ Melda Production
-
- KVRist
- 140 posts since 27 Jul, 2015
I don't have MXXX so I can't test this.
I think the result might be very different though, because feedback in hardware is almost instantaneous, whereas in MXXX I guess there is a delay of at least 1 sample. This is why many modern synth feature "zero delay feedback" (zdf) filters that use predictive algorithms. Though in filters the difference is mostly heard at extreme settings, so maybe the 1 sample delay doesn't matter here...
I think the result might be very different though, because feedback in hardware is almost instantaneous, whereas in MXXX I guess there is a delay of at least 1 sample. This is why many modern synth feature "zero delay feedback" (zdf) filters that use predictive algorithms. Though in filters the difference is mostly heard at extreme settings, so maybe the 1 sample delay doesn't matter here...
-
- KVRAF
- Topic Starter
- 2696 posts since 9 Jul, 2015 from UK
Yes, I read that the 1 sample delay doesn't matter as its too quick. Also we can make this even smaller with up sampling.
I'm going to do some testing today. I'll share what I find.
I'm going to do some testing today. I'll share what I find.
Jason @ Melda Production
-
- KVRist
- 140 posts since 27 Jul, 2015
Yes, thinking a bit more about this, the 1 sample delay only matters in filters because the output signal itself is fed back into the input. Here it's only for modulation purpose, so it's irrelevant. Especially since in a bus compression scenario you probably don't want a fast attack.
-
MeldaProduction MeldaProduction https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=176122 - KVRAF
- 14339 posts since 15 Mar, 2008 from Czech republic
How about a screenshot of what you have and what you want?
-
- KVRAF
- Topic Starter
- 2696 posts since 9 Jul, 2015 from UK
A screenshot will not really show anything, as I only have an amp loaded into MXXX, that's it.
What I am trying to do is modulate the amp with a follower. But the follower needs to listen to the output of the amp.
Like a feedback compressor.
At the moment I can only get the global follower to listen to the input of MXXX.
Is there a way of doing it with the module follower?
What I am trying to do is modulate the amp with a follower. But the follower needs to listen to the output of the amp.
Like a feedback compressor.
At the moment I can only get the global follower to listen to the input of MXXX.
Is there a way of doing it with the module follower?
Jason @ Melda Production
-
MeldaProduction MeldaProduction https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=176122
- KVRAF
- 14339 posts since 15 Mar, 2008 from Czech republic
Actually you can do that but note that there will be delay. You can do that by generating feedback by any of the processors and then using the follower mode in a modulator you can set the input to the feedback channel.