 |
 |
 |
 |
 |
 |
 |
Forums - Music techology
Subject: Compression
Viewing all 9 messages - View by pages of 10: 1
Original Message 1/9 28-Jul-98 @ 12:20 PM - Compression
Something about this is getting to me - everyone seems to advise to "compress it heavily" or "compress the hell out of it" - whether "it" be vocals, drums, guitars, 909 kicks - whatever. Now in theory I kinda know what a compressor is supposed to do - at least I think its something to do with bringing the lower-volume-signals proprtionally towards the higher-volume-signals so that the overall differences in volume are less and then, hey, it sounds better.
Trouble is, the controls confuse the hell out of me. Unlike a reverb (which is immediately obvious as an effect) the actual application or compression to sounds still baffles me. It's not only the controls - threshold/attack/release/ratio/whatever - but also the use in conjunction with gates.
If someone could clear up what each does for me in plain-speaking terms I'd be grateful. Above all, ratio disturbs me. If you say "compress the hell out of something" does that generally imply that you up the ratio? Or that you increase/descrease the threshold? Finally, would it help if I learnt how compression works on a hardware compressor - I have a sneaking suspicion that software versions of effects are generally best used after you've seen the "real thing" in action.
I also think that the whole idea may be pretty unexplainable - a case of try-it-and-see. Maybe I just can't tell at all when I hear it.
Will
Trouble is, the controls confuse the hell out of me. Unlike a reverb (which is immediately obvious as an effect) the actual application or compression to sounds still baffles me. It's not only the controls - threshold/attack/release/ratio/whatever - but also the use in conjunction with gates.
If someone could clear up what each does for me in plain-speaking terms I'd be grateful. Above all, ratio disturbs me. If you say "compress the hell out of something" does that generally imply that you up the ratio? Or that you increase/descrease the threshold? Finally, would it help if I learnt how compression works on a hardware compressor - I have a sneaking suspicion that software versions of effects are generally best used after you've seen the "real thing" in action.
I also think that the whole idea may be pretty unexplainable - a case of try-it-and-see. Maybe I just can't tell at all when I hear it.
Will
Message 2/9 28-Jul-98 @ 03:19 PM - ok, i'll give this a whirl...
first of all, you might want to check out the Studio FX Rack section - kilo has done a nice job of explaining different processors there and gives some suggestions on usage.
if you're still confused, maybe this will help...
threshold determines when the compression process kicks in. so if you set threshold real high, only loud portions of the signal will get squashed. the lower you set it, the more the signal will get affected, cus it's kicking in at lower volumes. keep in mind you don't want to set it too low, cus it will boost some quieter sounds that you don't want to hear (like tape hiss, for instance). use your ears and see what sounds best.
ratio determines, as you guessed, how much the signal gets compressed. if you want to "squash the hell" out of a signal, set ratio high and threshold low. ratio is expressed as volume in to volume out (per kilo's example - a 4:1 ratio means a 4dB signal becomes a 1dB signal). since the effect of compressing a signal will make it quieter, most compressors will have a post compression gain control so you can boost the processed signal back up. just be careful that you don't apply too much gain and clip the signal. notice that if you match the input signal and output signal for peak levels, the output signal sounds much louder (cus you've boosted the average level).
attack and release just determine how fast the process kicks in and goes off, respectively, once the signal has gone above threshold. these settings really depend upon what you're compressing - so just use your ear, or follow kilo's suggestions.
so what are you using for a compressor anyways? i compress all my samples in Sound Forge.
if you're still confused, maybe this will help...
threshold determines when the compression process kicks in. so if you set threshold real high, only loud portions of the signal will get squashed. the lower you set it, the more the signal will get affected, cus it's kicking in at lower volumes. keep in mind you don't want to set it too low, cus it will boost some quieter sounds that you don't want to hear (like tape hiss, for instance). use your ears and see what sounds best.
ratio determines, as you guessed, how much the signal gets compressed. if you want to "squash the hell" out of a signal, set ratio high and threshold low. ratio is expressed as volume in to volume out (per kilo's example - a 4:1 ratio means a 4dB signal becomes a 1dB signal). since the effect of compressing a signal will make it quieter, most compressors will have a post compression gain control so you can boost the processed signal back up. just be careful that you don't apply too much gain and clip the signal. notice that if you match the input signal and output signal for peak levels, the output signal sounds much louder (cus you've boosted the average level).
attack and release just determine how fast the process kicks in and goes off, respectively, once the signal has gone above threshold. these settings really depend upon what you're compressing - so just use your ear, or follow kilo's suggestions.
so what are you using for a compressor anyways? i compress all my samples in Sound Forge.
Message 3/9 28-Jul-98 @ 04:17 PM - RE: Compression
Yeah, I'm using Soundforge too - but with the Waves Atrack (which has a Compressor and EQ and Gate) direct-x plugin. If compression effectively averages the signal out so the whole can become louder, can you also use the alternative "normalise" in Soudforge - average RMS power or something - that seems to make the track louder than maximising the volume to the peak point.
I'm going to go and read back through Kilo on compression, but i've read it before. Maybe I'm just not doing it right. I'll have another go tonight but thanks anyway....
will
I'm going to go and read back through Kilo on compression, but i've read it before. Maybe I'm just not doing it right. I'll have another go tonight but thanks anyway....
will
Message 4/9 28-Jul-98 @ 04:17 PM - RE: Compression
Yeah, I'm using Soundforge too - but with the Waves Atrack (which has a Compressor and EQ and Gate) direct-x plugin. If compression effectively averages the signal out so the whole can become louder, can you also use the alternative "normalise" in Soudforge - average RMS power or something - that seems to make the track louder than maximising the volume to the peak point.
I'm going to go and read back through Kilo on compression, but i've read it before. Maybe I'm just not doing it right. I'll have another go tonight but thanks anyway....
will
I'm going to go and read back through Kilo on compression, but i've read it before. Maybe I'm just not doing it right. I'll have another go tonight but thanks anyway....
will
Message 5/9 28-Jul-98 @ 06:14 PM - RE: Compression
I DONT ADVISE COMPRESSING ALL THE MATERIAL AT ALL..... WHOEVER SAID COMPRESS HEAVILY ALL THE MATERIAL IS NOT CORRECT...... the compressor section gives some good starting points for settinmgs for instruments.... if you compress everything, your mix sounds crap, dull and without any dynamics.... normalising is totally different.... check the Glossary section..... the compressor section of the FX-Rack, couldnt be any clearer.... the point is... in most cases, you're NOT supposed to hear compression working at all... it is supposed to be transparent, unless applied as an effect.... so it's pointless compressing something expecting to hear a result... the point is.... you're not supposed to in most cases.......
Message 6/9 28-Jul-98 @ 09:38 PM - further clarifications...
if you're just working with samples which will then be used in a sampler (and subsequently filtered and processed and whatnot), then compression is a good idea in most cases. most of my samples i'll hit with a 2 or 3:1 - this gives you a nice hot signal to start with before you start subtracting stuff.
when you're compressing entire tracks, you really got to be selective and decide what needs it. i wouldn't recommend exporting an entire track to SF and compressing it there and then bringing it back into the mix. you want to be able to hear the entire mix while you fiddle with the compressor. that's where you'll hear the difference, in terms of how well it sits in the mix.
i wouldn't recommend compressing an entire mix unless you're really comfortable with mastering techniques, or if you don't mind fragging your mix...
normalizing i usually do after compressing, to make sure the peak value is at 0dB (or whatever is just below clipping). i wouldn't bother with average RMS, cus you might clip the signal. normalizing that way, if i remember right, involves a static volume change to the entire file based on the average, not the peak, volume. either way, as kilo said, it's totally different than compression.
when you're compressing entire tracks, you really got to be selective and decide what needs it. i wouldn't recommend exporting an entire track to SF and compressing it there and then bringing it back into the mix. you want to be able to hear the entire mix while you fiddle with the compressor. that's where you'll hear the difference, in terms of how well it sits in the mix.
i wouldn't recommend compressing an entire mix unless you're really comfortable with mastering techniques, or if you don't mind fragging your mix...
normalizing i usually do after compressing, to make sure the peak value is at 0dB (or whatever is just below clipping). i wouldn't bother with average RMS, cus you might clip the signal. normalizing that way, if i remember right, involves a static volume change to the entire file based on the average, not the peak, volume. either way, as kilo said, it's totally different than compression.
Message 7/9 28-Jul-98 @ 11:24 PM - RE: Compression
Normalizing a sound to a particular RMS has several advantages. I prescan all my files first, find the RMS and Peak values, and generally will normalize the RMS up to the point that the peaks top out at 0db. This way I keep more of the original dynamic of the tune while maximizing the overall db level. Some stuff (like drums) I'll normalize the RMS to the point the Peak value would go above 0db (i.e., to the point the peaks should clip), but I'll apply dynamic compression to keep the peaks to 0db or less. This makes the overall sound much hotter while keeping you from outright clipping. It dunna work for everything, but it certainly has its uses.
As a side note, I try to mix using as little compression and/or normalization as possible. My favorite mixes are the ones I pull into Sound Forge, scan the levels and find the tune is peaking at 0db or just below, and my RMS is around -16db. For overall music (complete tunes with drumms, bass, riffs, etc.) -16db is fairly comparable to 'production' CD levels. There's a preset in Sound Forge's normalize menu at -16db RMS. It makes for a good starting point. I would advise prescanning first, just to check out how much your gonna change yer dynamics. As an example, an ambient tune that has a unedited RMS of -21db with peaks at -3db, might be best to not normalize at all, but if you wanted to eek out a bit more volume, I mesef would probably not go above -18db RMS normalization. That gives your average sound a whopping 3db (roughly 150% of yer original volume level - I think) of boost and brings your peaks up, right to 0db, or thereabouts.
The main thing is to try it, listen, compare, to actually get your best mix. Louder is definately not better. It's just louder. If all yer tunes are massively loud at -6db or sumpin' your ears gonna get tired faster, the tune will have less 'life' and your mix may sound a bit drab, albeit it will be loudly drab. It may sound lame, but getting the mix right in the first place, before adding compression, before normalizing, etc., etc., will almost assuredly give your tunes more life and sparkle than squashing everything down in a compressor. Use fresh ears, too. If you been mixin' a tune down for four hours, and you haven't took a break, get out of the room. Go outside, drink a brew, smoke a spliff, practice Yoga, say hi to yer friends, and so on. When you come back to it, you'll be hearing more of the tune and you may realize that all the extra reverb, delay, compression, flanging, distortion, harmonizing, etc., that you had orginally planned to do before the break is now unnecessary.
A final note and more opinion, listen to yer mixes at low volume while yer about other things. For mesef, low volume monitoring is far less fatiguing, and I am listening more to the overall impact of the tune. I notice annoying bits easier, I can tell much faster when a section is too low/too high db wise, and me ears dunna ring at the end of a day.
If any of yous haven't yet, I highly suggest checking out Kilo's advice in the Mixers & Mixing, Studio tips and tricks, and Studio techniques section of this site. I have saved myself many hours of 'reinventing the wheel' thanks to herr Kilo's articles.
Peace all
As a side note, I try to mix using as little compression and/or normalization as possible. My favorite mixes are the ones I pull into Sound Forge, scan the levels and find the tune is peaking at 0db or just below, and my RMS is around -16db. For overall music (complete tunes with drumms, bass, riffs, etc.) -16db is fairly comparable to 'production' CD levels. There's a preset in Sound Forge's normalize menu at -16db RMS. It makes for a good starting point. I would advise prescanning first, just to check out how much your gonna change yer dynamics. As an example, an ambient tune that has a unedited RMS of -21db with peaks at -3db, might be best to not normalize at all, but if you wanted to eek out a bit more volume, I mesef would probably not go above -18db RMS normalization. That gives your average sound a whopping 3db (roughly 150% of yer original volume level - I think) of boost and brings your peaks up, right to 0db, or thereabouts.
The main thing is to try it, listen, compare, to actually get your best mix. Louder is definately not better. It's just louder. If all yer tunes are massively loud at -6db or sumpin' your ears gonna get tired faster, the tune will have less 'life' and your mix may sound a bit drab, albeit it will be loudly drab. It may sound lame, but getting the mix right in the first place, before adding compression, before normalizing, etc., etc., will almost assuredly give your tunes more life and sparkle than squashing everything down in a compressor. Use fresh ears, too. If you been mixin' a tune down for four hours, and you haven't took a break, get out of the room. Go outside, drink a brew, smoke a spliff, practice Yoga, say hi to yer friends, and so on. When you come back to it, you'll be hearing more of the tune and you may realize that all the extra reverb, delay, compression, flanging, distortion, harmonizing, etc., that you had orginally planned to do before the break is now unnecessary.
A final note and more opinion, listen to yer mixes at low volume while yer about other things. For mesef, low volume monitoring is far less fatiguing, and I am listening more to the overall impact of the tune. I notice annoying bits easier, I can tell much faster when a section is too low/too high db wise, and me ears dunna ring at the end of a day.
If any of yous haven't yet, I highly suggest checking out Kilo's advice in the Mixers & Mixing, Studio tips and tricks, and Studio techniques section of this site. I have saved myself many hours of 'reinventing the wheel' thanks to herr Kilo's articles.
Peace all
Message 8/9 28-Jul-98 @ 11:26 PM - RE: Compression
Having said that, you CAN use compression to get some fucked up sounds...
Message 9/9 29-Jul-98 @ 12:37 AM - RE: Compression
Oi Noize!
I agree, and that's mainly what was trying to get at, compression on the overall mix needs to be as Kilo says, 'trasparent.' But for making mucked up noises, there is no limit really. Using old guitar stomp FX, gates, compressors, all of that gives you a broader pallette with which to experiment. Breathing basses, monster drums, phone line effect vocals, etc., can all be tweaked through good creative use of compression. Again, herr Kilo has much of this covered in the Studio Tips and Tricks section.
Peace all
I agree, and that's mainly what was trying to get at, compression on the overall mix needs to be as Kilo says, 'trasparent.' But for making mucked up noises, there is no limit really. Using old guitar stomp FX, gates, compressors, all of that gives you a broader pallette with which to experiment. Breathing basses, monster drums, phone line effect vocals, etc., can all be tweaked through good creative use of compression. Again, herr Kilo has much of this covered in the Studio Tips and Tricks section.
Peace all
Viewing all 9 messages - View by pages of 10: 1
There are 9 total messages for this topic
Reply to Thread
You need to register/login to use the forum.
Click here to Signup or Login !
[you'll be brought right back to this point after signing up]
Back to Forum
