VERY BASIC HELP REQUIRED ABOUT USING A COMPRESSOR!

I am not sure if the Samson can handle a mic signal or if its only for line signals.?!
lets assume it can handle a microphone:

you need to set the threshold lower than 0dB, as 0dB will actually bypass the compressor.
-10dB or -20dB will give you enough headroom for the compressor to work.. (you have to keep an eye on the gainreduction though.. I'd go for a gainreduction of 4dB as a starting point)
So primarily focus on the amount of gainreduction!
set the Attacktime to 0ms to hear the compression, and back off to 30-40ms.. whatever you like..
release time is very dependant on the vocals/feel of the song.. I'd go a bit higher than your 50ms..

If you have a good audiointerface you might as well use a software compressor to get the job done.. either during recording with input-echo enabled, or afterwards..
it might be the better solution, before trying to get the samson to work..
just dont hit your a/d-converters too hard..

I have to set the threshold to almost -40 dB for the gain reduction meter to register 4dB.  If I have it between -10dB and -20dB it doesn't register anything!

Beggars Bridge:
"Is it f*cked?"

If it is, you better find the person who did it and discipline him. Those liquids are almost sure to cause a malfunction. Some people have the weirdest needs.

OK, thanks.

Something just occurred to me.  I have CoolEdit Pro software that I already use to manipulate WAV files (cleaning up, equalisation etc), and I know that it has compression facilities.  So, instead of recording through the Samson, couldn't I just record straight onto my PC then do all the compression via CoolEdit?  Would it be adequate?

I'm using Sonar 4 Producer Edition.

OK Mike, thanks again ;)

Here is a thread you might find useful:

http://forum.cakewalk.com/tm.aspx?m=2116921

Beggars,
 Let me assure you that actually learning how to listen to compressor is probably the hardest task to learn in audio production.

 One reason it is so difficult is that often times you are thinking about one set of concerns while working with material that isn't very applicable.

 A compressor can destroy the transients on drums and it's easy to hear... it's a whole lot harder to hear the transients on the vocals of a slow ballad so if you are thinking about details like that while working on material that isn't suitable it will take a long time to rack up enough experience.


 Here's a vocal tip.

 Put 2 instances of the Sonnitus compressor on your vocal track. Use two presets to get started.

 Try the 1176 on the first one and the LA2A on the second.

 Set the 1176 up for really fast response but a very light threshold so it juts grabs the extra wild peaks... use just a little bit of makeup gain.

 Now set the LA2A up with an aggressively deep threshold and use lots of makeup gain and you should clearly hear how it squashes the vocals and makes them sound more familiar to a radio listener etc.


 Experimenet with that a bit... then swap the order.

 Then take a good look at the preset setting for those 2 emulations and figure out their characteristics... and try to get to the point where you have a feel for what happens with the settings when they change. Basically try to get to the point where you don't need the presets any more because the settings have begun to make sense.

 Then spend a few years second guessing all your compression choices... and then you will be good to go.

 Those two emulations work good on acoustic guitar as well.

 Don't use that recipe on drums!!! :-)




You can do it!!!


best regards,
mike

Hi Mark  thanks for checking out the thread I mentioned. Most of the confusion is in many of the posts that come after my first post. My first post gives an explaination of how to set up a compressors parameters correctly. (And in the correct order) It is a much more true account of how to do it. Many people on the forums don't listen or actually read what it is that is being said. They have their I don't do it that way so therefore it must be wrong and I am not going to try any other approach. Which is sad because many I am sure have not got a clue to how to set up a compressor correctly. (eg the fiddle around with all the knobs approach and hope it does something good)

But your comments Mark do bring up some issues and that is maybe that my initial approach is a bit compex in that it assumes you already have some understanding of what the parameters do in setting up a compressor. What you are saying is you are still trying to hear some of things I mentioned which I just assume to be quite clear but maybe they are not for some.

I will have a think about coming up with an approach that makes it easier and approaches the more simpler aspects of a compressor which is about levelling out volume and dynamics variations to create that more even sound.

But a good start might be first practice using the compressor over a single individual track. Music is a much more complex signal compared to just a bass for example. Try an attack of say 10 ms, Release of 200 or 300 ms or so and start with a ratio of say 2 or 3: 1 Now start with your threshold control set high so the compressor is not doing anything. Lower it slowly and observe the Gain Reduction indicator. Keep lowering the threshold until the GR is showing a certain amount of GR eg 3 db to start with. The key here is listen to what is going on at the same time. Try lowering the threshold a lot so the compressor is really working on the track. Once again listen and see how it effecting the sound. Try higher ratios with this approach too and see how they also effect the sound.

Don't forget that compressors by their very nature will always be attenuating the signal so use of the makeup gain control is essential to get the compressed signal level back up to the uncompressed signal level.

This might be getting complex but here is another thread I started about actually removing the music and only hearing sound when the compressor actually does anything. (for setting up purposes only of course!) I am surprised that no one has really caught on (or unwilling) with this approach. It is quite amazing and very revealing actually.

http://forum.cakewalk.com/tm.aspx?m=2282782

Hmmm...

Just tried to play about with the Sonitus compressor.  I haven't used it before, as I usually use CoolEdit Pro to work on individual wavefiles, but I had one or two problems:

The audition only lasts a couple of seconds!  How can I get it to play longer?

Also - during the short time that it was playing, there was no activity in the level meters.

Mark I am definitely a fan of opening every track of a multi track session in a separate editor such as Adobe Audition etc. Especially on tracks that do have a few silly peaks that stand out too much. It is better in my opinion to bring them down to match the overall level of the track and perhaps add a db or two to the whole track then bringing the whole thing up a bit.

Just parking a compressor over a track to control a few silly out of control peaks is adding the effect of that and sound etc to the overall track. The compressor might catch the peaks but it might effect the rest too.

I like limiting the peak to the average level of the track rather than subtract gain from the peaky area as well. It sounds more natural. Then you can park a compressor doing a much lighter overall compression on the track and it will all sound better. Lower ratios tend to make the sound bigger but not quite as well contained. But if you have levelled out the track already then you can get the bigger sound happening and it will be nicely contained as well.

Yes, I believe so.

OK fellers.  After a few days of having my head immersed in the intensive world of learning about compression by trial-and-error, I've pretty much got to a level of knowledge that should be good enough to meet my needs.

I've decided not to compress individual tracks, as the evening out of volume is generally outweighed by the artificial sound which comes into play (apart from maybe the bass, as it seems to deal with uneven levels without making much difference to the sound), and I will to continue to get rid of level peaks manually.

However, after lots of experimenting and hours and hours of comparisons, I will be adding some fairly soft compression to the overall mix; I did actually settle on a CoolEdit preset in the end (RealAudio compander - with 3.8:1 above -9dB, 1.1:1 below -9dB and 2:1 expansion below -40dB), but I didn't chose it just because it was a preset, but because these settings just sounded the best out of the many I tried.  I will be using it as my default compression for my ensemble of largely acoustic instruments from now on, although - of course - I will still have the option to deviate and tweak away from it if required.

Thanks for the help!

Hi Mark. I agree with Bristol here too. I was not tryng to convey the idea the individual compression on tracks is not a good thing. I was only saying that using compressors set for limiters to catch out of control peaks is better handled with editing.

As Bristol  says very transparent compression is possible on individual tracks. Many instruments will benefit from it eg bass for example. (just use lower ratios and simply bring the threshold down for small amounts of gain reduction eg 3 to 4 db.)  Keep an eye on the Attack settings so you are not killing transients. It is too fast attack settings that get compressors to jump too fast over the music (or tracks) and ruin it. (if you read my explaination of setting up compressors in that thread it the Attack that is set first)

You say you find it hard to hear sometimes what compressors are doing to the sound. Experiment with the Attack setting. It is one of the most powerful of all the compressor parameters. It is the one that destroys the music first. Listen to the leading edge of the sound or the attack transient while you play with the Attack. When it is too fast the leading edge of the sound will quickly be destroyed. But slow it down and the attack transients come back fast.

With regard to using overall compression on your mix be careful of presets. Presets have no idea about the material your are presenting and also they do not know anything about the average level that might be present in a signal so they don't work that well. You will always have to tweak a preset anyway.

For overall compression try a much lower ratio eg 1.3:1 and set the Attack for 10 to 15 ms or so. This will allow transients to poke through. Then simply lower the threshold slowly for 2 to 3 db of gain reduction. Then it will sound very transparent on your mix and add just a little glue to start the process of bringing it all together. You may need to fine tune the Release parameter to suit the rhythm of the music. When you are compressing over the whole mix, this is a good time to visit the concepts I have mentioned here:

http://forum.cakewalk.com/tm.aspx?m=2282782

This is the null checking approach to compression. By removing the music, the action of the compressor becomes very apparent. Incorrect or poor attack and release settings (but also other parameters too of course) create jerky responses out of silence as the sounds ramps up and also as it ramps down to silence again. Fine tuning these parameters can improve the way the music comes in and goes away. You are only hearing these movements out from silence hence they are so clear. Then when you switch the null off and back to normal it will sound more musical.

Well the transients I speak of are the very first part of any sound. If you pluck an acoustic guitar string (with a pick) it is the crisper pluck part of the sound right at the start not the rest of the sound. In the case of a drum a snare has a very high and fast attack transient but not much else. A cello playing a long slow note however will have no attack transient to speak of.  A tubular bell being hit with a hammer has a fast attack transient followed by a lot of sustain.

With your attack setting I see some confusion there. You say you set the attack to 5ms and started increasing. If you are turning the attack control clockwise from there you are in fact slowing down the attack part of the compressor. You will not notice anything. But it is the lower attack settings (eg anti clockwise setting) that get faster. What happens to the sound when you set the attack to its lowest value eg fastest response eg fully anticlockwise usually. Have a listen there and you should hear the compressor destroying that attack part of the sound.  So compressor attack settings as low as 1 ms or under are going to jump over the attack transients very quickly.  (unless you have one of these dumb compressors that show attack getting faster in a clockwise direction, is that the case? It is not normal)

When a compressor is working it is firstly deciding wether a signal has reached the threshold that you have set or not. So higher thresholds mean that most of the music will slip under without being touched. A lower threshold means the music is touching and going over it more often.

OK once the threshold is met then the compressor will do its thing but it waits first for a certain time before it jumps over on the signal. This waiting is the attack setting. So if your attack is set to its fastest setting then (ie fully anti clockwise) then it does not wait at all and will jump onto the attack transient part of the signal as fast as it can. Compressors are seriously fast and can jump onto even the fastest of attack type sounds eg a snare. That lovely sharp crack right at the start will turn into a soggy woosh instead. (being a drummer for many years (40) has put me in front of one of the best sounding snare drums in the world eg Sonor.) So I really notice when compressors are jumping too fast over drum sounds because that is not how they sound in real life.

Once the attack time has been waited out then the compressor applies the ratio setting. So if we have a ratio of 4:1 then the compressor will divide the gain change of the input signal by a factor of 4. Lets say our input goes up by say 8 db after that then our output only goes up by 2 db instead. Now once the signal falls below the threshold then the compressor releases the sound (release) This setting is also important as it determines how fast or slow the compressor will ease the signal back to where it was ie under the threshold. This is why the tempo of the music comes into play a bit. If the release setting is too slow then the output from the compressor will still be a little low and on its way up at a time when another louder part might be wanting to trigger the threshold again.  The idea with release is to aim to get the compressor back to normal before the next louder part of the signal wants to trigger the threshold again. Too fast release settings can cause some strange sounds too. The idea is to aim for slower releases where you can because they are much less noticable.

So let's look at the attack time being waited out before the compressor does its thing. If you have the attack set for say 15 or 20 ms then it means the compressor won't do anything for that time even though the threshold has been well and truely crossed. The signal passes right though. So in a way we can construct attack transients to our liking. That is why when you slow down the attack settings (eg higher attack values) the compressor tends to leave the attack transients of the track or the mix alone much more so and now tends to act on the sustanined part of the sound more. Or it jumps onto the part of the sound which tends to just come after the initial attack transient.