Compressor Meters And Maximizer Release

I was wondering what is that thing that looks like a meter between the input and output meters for? (see the pic)

I thought that would be the amount of signal that is let through over the threshold, however it does not seem to always be triggered when the signal goes over the threshold according to the meter on the left. Especially that seems to be the case when only one channel goes over the threshold, even by a large amount, the middle meter remains dormant.

Second question, I am not able to figure out how exactly maximizer works. For instance, what are the two releases that it employs?

1: gain reduction, the threshold slider is where you adjust this, for example a gain reduction of -3db means that you can use 3db gain makeup to get equal volume again (but with less peaks, so it will appear louder)
2: peak release is for very fast transient peaks, long release is for every else

You mean this meter shows how much of a reduction applied to the signal? That makes sense, however I am not seeing this reduction being applied for some reason where I think it should be. In my example I have a fairly high threshold of -4dB with a modest ratio of 2:1 and granted one might not be seeing too much of reduction applied, however I have a very short attack of .001 ms so I would expect to see some reduction especially that the signal is very tight to the 0dB.

In fact I just turned the compressor off, turned the prefade volume up a bit to let the signal clip a bit on the master at about +2dB and turned the compressor back on. The result is no difference whatsoever, it clips at exactly the same level, to thousandths of a dB so the fact that there is no gain reduction is shown on the meter appears to be right. However it shouldn’t be the case.

Yeah, that’s right, so what does it mean?

1: perhaps use more threshold? -4db doesnt mean much without knowing volume/rms levels and what kinda sound it is. or perhaps a shorter release might work?
personally i set renoise’s headroom to -0db (instead of its default -6db) to avoid confusing myself, i dont use the pre faders either and prefer gainers.

2: i am not sure, default seems to work for me ;] it might be in the manual though

I did more experimenting with the compressor and I am completely lost now.

I put both release and attack to the minimum, set the threshold to -8dB and put the ratio 4:1 leaving the makeup at 0dB. It still completely failed to bring the peaks down. In fact even lowering the threshold to -12dB made no difference according to master track peak readings.

If anyone would be willing to experiment on their own and see what results they come up with, I would be quite eager to know. The source I am using is an acoustic piano instrument with the quite fast transients on high pitched notes but not really shorter than.001 ms.

Hi… You’ve got the attack at .oo1, and the release at .001? Is that right?

I think the compressor, has become over used, and I think that it gives us all terrible production habits. Threads like this, are very good, cause they allow us to check ourselves, before we wreck ourselves… So to speak.

If we are making, “a major label pop album.” that is going to be a totally different story… But to make great music… an acoustic piano, with a 4:1 ratio is probably doing yourself more harm than good. For one thing, you are making definitely making your ears become accustomed to the compressor.

So what’s the point of a compressor in a mix? Tames peaks. Why? So the volume is not popping out, and things are balanced. Balanced is the key. Well, what about pumping and slamming compression on my drums? Yep that exists. And what about squashing my master too? Also exists.

Using a compressor to tame a transient, would take at least 10 milliseconds… ( That’s just a guess. but that’s my guess.) And having an attack of at least a millisecond or two on a piano couldn’t hurt. ( but keep in mind, I haven’t heard the piano, and I’m just going by what you are saying, "its got fast transients on the high freqs.)

Really, for accoustic piano, I’d treat it more like a vocal. With a ratio less than 2:1… ever so slight gain reduction… make up gain is optional… Usually, what I do… Is never use the make up gain. I make things in my mix, quieter instead… unless my mix is already whisper quiet… Here’s the funny thing… When your mix is pretty quiety, and things are not so gained out… You can than beat that master track… Ruthlessly! You can Duke Nukem that waveform to hell, and you will get a really loud master… You are pretty much saving all your clipping for that one final squash!

Getting a great result with a compressor is really hard… You have to think, “i’m using this to tame peaks, to balance my volume,” if you are worried about transients… There are some other things… Even lowering the velocity, and using a filter… ( I’m not big on transient shapers. )

Ok… whatever, I’m just mumbling… Accoustic piano? try an attack of like 10 or more milliseconds, and release of say 20 or 30, or even more… Seems like modern music is all about these short attacks and releases… You can probably put a 50, or 65 millisecond release on that…

http://www.soundonsound.com/

^ I bet they’ve got crazy amounts of tips on compression

If you don’t really understand how a compressor works, or what the parameters are designed to do, then I think you’re only going to frustrate yourself by experimenting with random values.

Wikipedia has a pretty nice explanation of the basics here:

Obviously Renoise’s compressor doesn’t have all of the features mentioned there, but you can pay attention to the descriptions of the Threshold, Ratio, Attack/Release, and Makeup gain.

I do understand how compressor works, I am a qualified sound engineer for ten years now.

I do not understand how compressor in Renoise works, because it does not do what it is expected to do. The bus compressor seems to do its job quite right, however the basic compressor does not seem to take anything into account.

try splitting the piano signal so you can use different compressor setting for the low/high of the piano., or only compress the highs, normaly piano’s don’t need much compression if at all, unless its dance music ;]

-edit

just saw you have experience, so never mind above ;]

Man, I know what I need a compressor for in case of this piano. Instrument compression is not what I was asking about, the question was about an operation algorithm of a specific device in a DAW, which happens to be a compressor.

My apologies then. When you mentioned that you had both attack and release set to 0.001ms, this just threw me off a little bit, because the compressor will indeed be quite useless in that configuration.

Perhaps this is just a quirk of Renoise’s native compressor design (quite likely). I can’t recall if I’ve seen many other compressors that allow similar settings for attack/release. Most plugins I’ve seen do not allow release to be set so low, for example.

Nevertheless, if you use some ‘sensible’ attack and release values, then the compressor should still perform its job quite well under normal circumstances. I can say that I’ve personally never had any huge problems with it.

Just as a comparison, if I A/B Renoise’s compressor against Kjaerhus Classic Compressor using identical settings in each, I get very similar results overall, give or take a few inevitable minor differences in how each algorithm behaves.

Generally speaking, when I’m trying to dial in Renoise’s compressor, I first tend to set the ratio to max (LIMIT). For me, this just makes it easier to set the threshold roughly where I want it to kick in. After that it’s just fine-tuning everything to achieve the overall desired character.

Your mileage may vary, obviously.

I presume you did the testing only with the compressor and no other (send)devices in the chain? (Just to exclude side-effects from other devices)

Sorry, I thought you were asking bout, “how to use,” but I will leave my OP intact…

Cheers

There was a gainer before the compressor, which I took into account, and also a send afterwards for a reverb, which was muted anyway. Nothing else.

That was purely for the sake of purity. I had a signal just touching on the +2dB mark, so decided to use some compression to soften the peaks without limiting. My math was very simple:

set the threshold to -4dB, which gives us a range of 6dB to the peak of the signal, then apply a ratio of 2:1, so that we should get around the ideal peak of -1dB with a very short attack, which then would be relaxed slowly until it gets to 0dB. The release is not really relevant in this case, as long as it is not too long.

However the basic compressor did not let me arrive at these results, it did not seem to do any compression at all. At the same time the bus compressor works exactly according to the math, taking its minimal attack time of 10ms into account.

am thinking now that it was poor choice of forum section to post this question in, but anyway.

These are the settings that I am using do what I described just above. As we can see, both compressors have pretty much identical settings. The only difference is that the bus compressor works exactly according to the plan, while basic compressor does not compress at all, according to the peak readings from the master.

3442 Clipboard Image.gif I

EDIT: Actually, looked into the settings and peaks again, the bus compressor perform darn close to theoretically predicted results.

What does the audio peak at, without the compressor? Should the threshold not start there? I never see movement on the meters, on any comp, unless the audio is it least hitting the threshold… or am I totally wrong?

Now that I am looking at Renoise… I believe the center bar is, “gain reduction.”

These compressors are kicking in with their threshold right about there…

Uploaded with ImageShack.us

Uploaded with ImageShack.us

Renoise project file…

https://www.box.com/s/d2de0bb935406f3f0969

It peaks at about +2dB for the left channel and about +0.3dB on the left without any compression. If I pass it through the basic compressor, nothing changes. If I pass it through the bus compressor, it hits around -1dB on the left channel, as per my estimations. It happens on a six column track, btw, I have a wild guess that it has something to do with it.

Well… Compressors are very funny. For example, I have a favorite compressor I love to use for my drums… However, sometimes this compressor does not compress so well, and I use a different compressor for my drums! :D

Actually, compressor are like microphones! You can not have, “just one microphone.” It is so true. And also… I like the Renoise compressors, being similar in sound to the Kjaerhus… That is a very good comparison.

But you know… If there is, “a bug,” with the compressor, I would not know. The only thing I really compress is drums… and for that I use a pspaudioware…

But I do think, if I ever get to make, “very bad major label music,” then I am going to squash everything to bits… Squash, pummel, destroy!

B)