Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Jerome Malsack]

I am going to add the note from Rane on Compression. 

Bass

It is common for the sound mixer to reduce the bass signal because it overwhelms the total system. Use a compressor to smooth a bass sound by lessening the variations between the strings and increasing the sustain.

Typical settings for a bass guitar are a ratio of 4:1, with a fast attack of 25 ms and a slow release of around 500 ms. These settings produce a strong, smooth bass line to start with, and then adjust further as necessary. A hard knee setting is often preferred since all you want is to tame the excessive peaks and leave everything else alone.

Placement of the compressor in the signal chain depends on how you want it to function. For just the input signal, it goes after the bass guitar (if it has a line-level output) and before the preamp. If it is to function as a limiter to protect the speakers in the bass rig, it goes between the preamp and the power amp. Another method is to insert the unit into the effects loop of the preamp. This affects the bass signal by the preamp first, then the compressor limiter, and finally out to the power amp.

Guitar

Here are a few suggestions on how to achieve a lower volume without sounding as if you are playing out of a transistor radio. Set a slow (~500 ms) attack time with a medium to slow (100 ms – 500 ms) release and a relatively low threshold. Experiment from these initial settings. Do not over compress the high frequencies or the pick attack will sound slurred.

One of the favorite uses of compression by guitarist is to increase the sustain, or duration of a note after it is played. Jeff Beck, Carlos Santana and Gabor Szabo used sustain to great acclaim, although they did it the old-fashioned way of creating feedback by aiming the guitar pick-ups at the loudspeaker and then jamming over it (using a hard limiter to prevent damage). A compressor creates a similar effect. Set a high ratio and low threshold for long sustain, along with fast attack and slow release. Again, experimentation produces the best results.

Drums

When using compressors on drums and percussion, it is polar opposite time. Two popular approaches exist for changing the sound character. Both yield pleasant results and allow you to dramatically or subtly change the nature of the sound. Regardless of which approach you try or end up liking, the vast majority of compressor use on drums is limited to snare and bass drum.

The first approach is to use the attack time of the compressor to modify the drum sound. For example, by using a slower attack time along with a 2:1 ratio on a snare drum and by varying the threshold you can make very dramatic changes in the tonal character. By using a slower attack time the initial "crack" of the drum gets through -- great if the drum heads are old. This was a very popular technique for getting a beefy consistent snare sound for many years and is heard on many rock ballads. By only adjusting the attack time and ratio you will be amazed at how many variations of tone are possible, even without a stitch of EQ.

Reducing the leading edge of drum hits is another popular use of compressors. Try ratios between 2:1 and 5:1 accompanied by fast attack and release times. Listen carefully while changing the attack time to find the final setting.

Cymbals need a fast attack but a slow release to allow the sustain through. Begin with a low ratio of 2:1.

Bass drums are difficult to capture consistently due to drumming technique and other issues. It is common to use a compressor to keep the bass drum at a consistent level and tonality. Since many mixes are created with the rhythm section as the foundation, it is important to keep those elements even and consistent. This involves higher ratios (between 4:1 and 10:1) with a fast attack and release time.

To prevent turning the drum sound into pure mush, it is important to use a monitoring system (or the main sound system) having good low frequency performance. Use high ratios to keep the level consistent, and use the fastest attack possible without destroying the drum punch.

 

[Jason Lucas]

It depends what you call the transient. Is it the instantaneous moment of highest signal? The span of the highest slope of amplitude? Consider that for every hit, a compressor is goin to be hitting that kick  at a different point in the hit (1,5,10,30 milliseconds in). As long as the kick is longer than the attack time, every one will be landing at a different point in the hit. It will sound different, whether you believe you've let most of the "transient" through.

You'll hear a transient even with a 0 attack. What does it sound like? Turn it up to 10 ms - now it sounds different. Turn it up to 30 - I bet you'll find it sounds different. Nothing wrong with using a 10 ms attack on kick, but it will sound different at 30 in my experience. You should be able to easily replicate this. I'm not saying "omg ur destroying your transients" I'm only saying that kicks often sound different at 30 then at 10. Try it out, I'm certain you'll recognize a different sound. Change the threshold to match the GR as you do it.

I'm glad you said that, because I was about to say "Well of course it will sound different, the GR will be different." Obviously 30ms sounds different from 10ms, you're allowing more of the signal to come through.

BTW Jerome I've actually read Rane's note on compression. I don't think I agree with the 500ms attack setting on the guitar though (at least not with the compressor I'm using) That's way slower than what I would find useful for almost any purpose.

[Tim McCulloch]

In radio a slow attack time tracks a song as it slowly gets louder at the begining then a slow release time as the song slowly fades away. This keeps the "overall" level as loud as possible. Of course this is in recorded music but the settings still need to be there to track both fade in and out of the music. Sometimes longer TC are used.
Add to that an old optimod on the end for 3 band compression and songs don't seem to have fade in/out.
The longer times are needed in cases like that. If someone is selling a device it should be able to do as many jobs as possible for the greatest return on investment.

Douglas R. Allen

I remember when Bob Orban's first OptiMod® came out.  Great, now FM could sound like bad AM.... or it could sound pretty discreet.  Some station engineers were more attuned (NPI) for 'loudest on the dial' while others wanted their stations to be more listenable.

[Tim McCulloch]

With compressors I've never really had a need to use an attack setting slower than, say, 10ms. 10ms seems to normally be slow enough to let an entire transient through. For drums I seem to stick to around 7ms.

But most of the time I just completely minimize the attack setting because when I use compression my goal is to reduce the peaks. With a slower attack setting the peaks still seem to get through.

So I'm kind of puzzled as to why the attack setting on compressors can go all the way up to like 200ms. When would you ever want an attack setting that slow?

Just curious. I've been told a slower attack can sound more "natural", but to me it seems like the opposite. Using a slow attack seems to emphasize the sudden, loud peaks that are much more irritating to my ear than gradual increases in volume.

EXPERIMENT AND LISTEN.

Use different kinds of recorded instrument sounds, vocals and unmastered program mixes.  You can do this over headphones in the privacy of your den, living room or garage.

There is your homework assignment.  Get 'er done.

[Jerome Malsack]

I'm glad you said that, because I was about to say "Well of course it will sound different, the GR will be different." Obviously 30ms sounds different from 10ms, you're allowing more of the signal to come through.

BTW Jerome I've actually read Rane's note on compression. I don't think I agree with the 500ms attack setting on the guitar though (at least not with the compressor I'm using) That's way slower than what I would find useful for almost any purpose.

That may be true for Live sound issues, The document from rane may refer to a more studio and recording environment. 
With live sound you would be trying to work around the room sound, reverb, and problems areas in the acoustics. 

The Rane document provides a guideline for use and with drums you will be compressing to provide a consistent sound for the percussive instruments providing the rhythm.  Crushing this to death removing the dynamics is not always a good thing.  This would be dependent on the music style being worked.     

[Samuel Rees]

I'm glad you said that, because I was about to say "Well of course it will sound different, the GR will be different." Obviously 30ms sounds different from 10ms, you're allowing more of the signal to come through.

BTW Jerome I've actually read Rane's note on compression. I don't think I agree with the 500ms attack setting on the guitar though (at least not with the compressor I'm using) That's way slower than what I would find useful for almost any purpose.

500 ms will leave the guitar transients fully untouched and really alter the average volume more. Maybe some day on some guitar, somewhere, you'll want to do that. I know your trying to think about what is useful for your service, but I really believe you're boxing yourself out of a wide world of attack time experimentation.

Try the kick attack. Report back. I'm confident you'll hear it. Also listen for the low frequency content, fast attacks can in my experience, clamp down on very LF content even if the transient is seemingly coming through plenty.

[John Roberts {JR}]

There is a reason they put adjustable knobs on compressors instead of preset drum-guitar-vocal switches (while I did design one with a 3 position fast-med-slow switch).

The result of applying compression is subjective so YMMV.

JR

[ProSoundWeb Community]