I have no hardware compressor, and I'm told I could use some compression on my vocals. I've never used one before.
How do I use software compression in Ardour?
How's the latency?
Does latency even matter when recording in a studio??
I am assuming you are talking about adjusting the levels of something in post (using compressor automation). The solution is to avoid this during / pre-recording -- but that said.
Many external boxes have several aspects, which folks refer to as: Compressor, Limiter, Expander and Gate.
A Compressor only does one thing, it provides controlled attenuation based on a few controls. The first control is a threshold, anything under the threshold, the signal passes the core compression stage without any attenation. Once the signal is past the threshold, attenuation is applied to the signal.
There are more "knobs" to tweek, but that is the core. There is a ratio (aka 1:4) of how much attenuation to provide, the larger the ratio (all the way up to infinity), the more attenuation, which infinity reducing all of the signal over a threshold.
There is also an control "envelope" with parameters that are known as attack, release and a "correction" curve (also known as Knee) that defines an averaging & smooth method for the signal's level over time. A longer envelop will not respond to short spikes above the threshold or does so very mildly. A continuous signal (longer than the envelope) will provide the full attenuation as defined by the ratio.
(wordy, but short).
Also note that many purist dislike the use of compressors, and I understand (in part) why. The overly used and poorly configured compressor produces noticeable distortion. If done well, the distortion will not be distracting -- however, compressors (any dynamics modification) are a necessary evil, especially when you have a limited bandwidth such as broadcast FM radio (Signal to Noise Ratio/SNR of usually 40 or [at best] 60dB) where lower/quieter passages would be lower (in signal strength) that the noise present in the system. Those signals would be hard to understand (if at all).
Compressors help here. And can be use in other live performance scenarios. I would never run a live stage performance without a limiter on the output. Otherwise, the moment a performer drops a mic, you run the risk of delivering that sharp (and very loud) signal thru the amplifiers to the drivers and blowing them out... Ending in a highly dis-liked silence / ending the performance.
Many (most, those that are smart) broadcasters use a limiter (which is like a mini, compressor with only a threshold setting and an assuming ratio of 1:infinity with a fast acting envelope / for the purpose of protecting equipment and in things like radio broadcast, to stay under the allowed maximum signal (protect against over-modulation, the FCC doesn't like that).
TV commercials make use of heavy compression to "make their ads louder" and get your attention.
Expander is like a compressor, only somewhat backwards, in that it reduces sound (proportional as well) below a threshold... usually right above the level of noise in the system. When a person talks (or music) is present OVER the threshold, all the attenation is bypassed.
A gate is like a Expander but with a ratio of 1 to infinity again. Aka, below this level, completely remove all source noise. Hint: there will still be noise present, but only from the Gate onward in the system (up to the headphone or speakers). Professionals sometime use gates (or expanders) on each mic when miking a drummer / to help reduce pickup of sounds from adjacent instruments
Hope this provides a basic understanding of the "Compressor" family (also known as dynamics) of sound processing.
If you have JAMin installed on your computer, use that. Even with the default settings, it does an OK job of compressing vocals to within OK levels, unless your vocalist does horrible things like move around a lot.
Here, have a look at this: http://www.prorec.com/prorec/articles.nsf/8380cd588060ef03862565e2000db872/b425578c027460898625661000745390?OpenDocument
That's a decent series of articles that will help you use a compressor. After reading that, using JAMin will make lots of sense, and you'll probably be able to get your vocalist to sound consistent unless s/he does really amazingly horrible things like address the mic from the other side of the room.
Great stuff! I'll have to get some practise in before I really overstand these new( (for me) concepts
I'm running 10.4 on a MacBook, and I haven't yet seen a jaMIN build for osx86. Great article, though!
A good counterpoint
A balance... In moderation... With good taste...
Like I said, overly used and poorly configured hardware (or software) compressors have led the "purist"'s arguments (like discussed in the link) and to the purists' dislike for anything that has been "tampered" with.
Unfortunately, life functions differently. In the real world, there *IS* naturally occuring compression (of sound) in many aspects of sound reinforcement [SR] (and to sound engineering).
Take the dynamic voice coil... Used in microphones to [compression] drivers (aka horns), etc. They are not linear.
The voice coil [VC] is the 'motor' of a driver (for those outside of SR, think speaker). The VC is suspended inside the gap using the spider and attached to the diagram which is further suspended using a "surround". This setup has a physical limit that it will travel. Also the VC wire inside the gap, when it leaves the gap is less effective.
1 volt into the VC, 1mm of travel
2 volts in, 2mm travel
3 volts in, 2.8mm travel
4 volts in, 3.5mm travel
5 volts in, 4.0mm travel
6 volts in, 4.1mm travel
7 volts in, 4.1mm travel
8 volts in, 4.1mm travel
The numbers above depict a pattern, different drivers will have different amounts of travel and *acceptable* levels of input voltage. There is also something known as thermal compression, when the VC wire gets hot and thus the resistance of the wire also increases / reduces current flow. Thermal compression is not an issue for 'dynamic' microphones, but the travel and gap/wire flux field still is.
Point is, the output transducer naturally produces compression (attenuation) of a source signal.
A purist wants to minimize any compression (if at all) in the source material (and I understand why, I dislike overly used compression myself!). However there are limits, in output devices, in transport (over wires, terrestrial broadcast, in source storage), etc.
From a musical standpoint, the dynamic of one instrument might be too wide and another musical instrument (source) *masks* the first one. Compression on the first instrument, but not the second can help to level the competing devices for clarity and so both can be heard -- also proper mixing (riding the sliders correctly) is another alternative (unless the dynamics happens too fast).
Again in summary, compression in moderation.
The 'counter point' reference is a decent one to understand... but another issue (which is addressed inside some MP3 encoders/players) is replay gain.. Which attempts to match some average "level" across different recordings. While that deals with dynamics, it isn't directly related to compression.
In short, compression is very useful for the voice, because unlike so many other instruments we use in our songs (amped up guitars, keyboards, synth modules etc etc) The dynamics of the human voice is all over the place. If you have a look at the recorded waveform of a voice compared to these, you'll see what I mean. Compression is often used with a 'makeup gain', which essentially makes your louder parts a little softer and your softer parts louder, so there is a more even dynamic in your vocals. The reason you need to do this with instruments that don't change much in the background, is that if your voice level drops a little it can easily get lost 'in the mix'. So a good balanced amount of compression and make up gain can help your vocals to sit properly in the mix. Also it can be worth doing a bit of eq either on the backing track or the vocals, to try and remove/boost frequencies that are conflicting with each other (although boosting frequencies should be avoided as much as possible, as this adds it's own filter noise problems).
Andy, the "make up gain" control on almost all hardware compressors is merely an overall output volume control, that boosts the whole compressed output level. Often, when compression is applied, unity gain is lost, and the final output is lower that the nominal gain upon entry to the unit. The make up gain control does just that, it "makes up for" the lost gain.
The same thing could be accomplished by merely turning up the volume of the affected track anywhere after the fact. It's just convenient to have it all in one box.
The make up is usually associated with a separate small booster amplifier built-in.
Yes, you are right. So by increasing the overall level after compression you are in fact increasing the volume of the audio that was under the threshold, while bringing the compressed audio that was above the threshold back up to it's original level (or thereabouts depending obviously on how much gain you add). This in turn lowers dynamic range and achieves a more solid sounding vocal.
Pushing up the fader on the track after compression is okay as long as you have a strong enough signal to begin with and your compression doesn't lower the gain too substantially. Sometimes pushing your fader all the way up may still not be enough, so that is why there are makeup gain controls.
Your subscriptions & donations are critical help that make it possible for full-time development of Ardour to continue. Your support is critical and much appreciated.
August goal US$4500 (US$54k/yr)
Subscribers 1016 US$1599/month
Support Ardour, get free upgrades: pay $1, $4, $10 or $50/month for a year: