TranceAddict Forums (www.tranceaddict.com/forums)
- Production Studio
-- another sidechain question
Pages (2): « 1 [2]

quote:

Originally posted by PutBoy I love sidechaining. I have an idea for you, if your using FL, you can create your own multiband side chain. Take a short kick sample, assign it to say 4 different mixer channels. In each mixer channel you're going to put one BP filter, one compressor and one peak controller. Then name a fifth mixer channel (depending on how many mixer channels you'd want to assign the kick to) "sidechain" or something like that, and put a parametric EQ in that. Now, you're going to want to take whatever kick you're using and using the spectroman you're going to find 4 key frequencies (or however many channels you assigned the kick to), that the kick is using. Usually around 60 hz, 200 hz, 800 hz and 3000 hz. Now, set the BP filters on the 4 mixer channels (again, if that's how many you're using) to those frequencies, with one freq for each filter ofc. And now you will want to compress that signal, or it will be too low to be used practically. Set the ratio low, and turn the gain up until you peak at just before 0 dB _on the lowest frequency mixer channel_. After you've done that you want to set the exact same setting on all the other 3 channels (see before ;P). You don't want all of them to peak at 0 dB, you just want them to be relative to each other before and after the compression. And, the more you compress, the more significant the sidechaining will be. After that. One your 'sidechain'-channels, you want to set the parametric EQ on the same frequencies. Now, assign each peak controller to their corresponding frequency, set the base to 50 %, and the vol to anything minus (but the same for all of them, the more negative the more sidechain). And voila. Multiband sidechain. It sounds softer, it's more subtle, and it just makes the kick that much more powerful. You can ofc do this with anything else, have a vocal instead of the kick and then assign a lead to your sidechain-channel, should bring the vocal way up above the lead without the lead being any less powerful.

Cool! Could you elaborate a little more on why this would be preferable as opposed to normal sidechain? You could always use all the parameters on the compressor to get the sidechain as you want, and base, volume, tension and decay on the fpc, so in what way does this give more control over the result/process?

and PS:
(yes, sidechain compression with the fpc attached to a compressor is not conventional sidechaining like in a compressor with external trigger input, but it does sound really good and a lot like it's normally done. The "best sounding" technique is not necesserily the conventional method, so fpc -> compressor is a very valid process in and of its own (and extremely handy)).

quote:

Originally posted by lenieNt Force Cool! Could you elaborate a little more on why this would be preferable as opposed to normal sidechain? You could always use all the parameters on the compressor to get the sidechain as you want, and base, volume, tension and decay on the fpc, so in what way does this give more control over the result/process?

Well, to be frank normal sidechaining would prefered because it's easier. However, I like to do this sometimes as it just sounds a little bit better and is a lot more subtle.

On you second question, I don't know really, maybe you can find out. But it's kind of obvious that it does give you more control no? Try it and see, let me know what results you have.

I have made this technique up myself btw, I don't know if it is anywhere else, or if anyone else made it up as well.

quote:

Originally posted by PutBoy Well, to be frank normal sidechaining would prefered because it's easier. However, I like to do this sometimes as it just sounds a little bit better and is a lot more subtle. On you second question, I don't know really, maybe you can find out. But it's kind of obvious that it does give you more control no? Try it and see, let me know what results you have. I have made this technique up myself btw, I don't know if it is anywhere else, or if anyone else made it up as well.

Kk gotcha.. no I just wondered cause the reason why it might be useful is maybe because of how the kick goes from high to low frequency when it hits, and therefore hits the different bands at different times. But as you can always make the compressor be triggered at different times yourself with the attack setting.. I couldn't really grasp what was the advantage of doing it multiband wise.

But ok I think I got the point now.. so not only do you assign one fpc to each frequency, but also, on the channel you want to have sidechained, you compress by using the gain controllers on three of the bands on the parametric EQ as sidechain inputs, and not a compressor? Then I can see why it is multiband compression (although it's more like multiband automated gain reduction then, and not compression)

Cool method! (if I got it right) Another way to do it is to use the fruity multiband compressor I guess, and not the parametric eq. That would bring more correct multiband sidechaining with compression and not gain reduction as a result. Luckily you have exactly 3 bands in that compressor. If you do it that way you won't really need three channels with the fpc on each. You could use just one fpc to rule all the thresholds on each band of the multiband compressor, and use attack, knee and release on each to get the timing right. Or more prefarably, have 3 fpc's on that one channel to also govern each threshold differently (with the settings on the fpc).

quote:

Originally posted by lenieNt Force Kk gotcha.. no I just wondered cause the reason why it might be useful is maybe because of how the kick goes from high to low frequency when it hits, and therefore hits the different bands at different times.

That's it excactly. It's useful if what you are sidechaining is broad. For example if you're sidechaining everything in the song, then this way of doing it would be very helpful

quote:

Originally posted by lenieNt Force you compress by using the gain controllers on three of the bands on the parametric EQ as sidechain inputs, and not a compressor?

That's the way I do it, but you could ofc use the multiband compressor instead.

quote:

Originally posted by PutBoy I love sidechaining. I have an idea for you, if your using FL, you can create your own multiband side chain. Take a short kick sample, assign it to say 4 different mixer channels. In each mixer channel you're going to put one BP filter, one compressor and one peak controller. Then name a fifth mixer channel (depending on how many mixer channels you'd want to assign the kick to) "sidechain" or something like that, and put a parametric EQ in that. Now, you're going to want to take whatever kick you're using and using the spectroman you're going to find 4 key frequencies (or however many channels you assigned the kick to), that the kick is using. Usually around 60 hz, 200 hz, 800 hz and 3000 hz. Now, set the BP filters on the 4 mixer channels (again, if that's how many you're using) to those frequencies, with one freq for each filter ofc. And now you will want to compress that signal, or it will be too low to be used practically. Set the ratio low, and turn the gain up until you peak at just before 0 dB _on the lowest frequency mixer channel_. After you've done that you want to set the exact same setting on all the other 3 channels (see before ;P). You don't want all of them to peak at 0 dB, you just want them to be relative to each other before and after the compression. And, the more you compress, the more significant the sidechaining will be. After that. One your 'sidechain'-channels, you want to set the parametric EQ on the same frequencies. Now, assign each peak controller to their corresponding frequency, set the base to 50 %, and the vol to anything minus (but the same for all of them, the more negative the more sidechain). And voila. Multiband sidechain. It sounds softer, it's more subtle, and it just makes the kick that much more powerful. You can ofc do this with anything else, have a vocal instead of the kick and then assign a lead to your sidechain-channel, should bring the vocal way up above the lead without the lead being any less powerful.

Interesting, but what a fuck about!! I get the subtlety by changing the sidechain trigger and the wet / dry balance on the compressor. Going to add this technique of yours to the list though, along with vocoding percussion.

quote:

Originally posted by PutBoy That's it excactly. It's useful if what you are sidechaining is broad. For example if you're sidechaining everything in the song, then this way of doing it would be very helpful That's the way I do it, but you could ofc use the multiband compressor instead.

yap yap

Just got done reading this thread, and a few thoughts:

The side chain is paramount to all EDM production. It comes in various forms including volume automation. Its main usage and functionality for basslines has to do with clearing up the bottom end of a mix, so that when your bassline is playing, it is compressed when the kick comes in, thus making more room for those 2 low end items to "fit" together. The compressor knows when to signal because the producer assigns that signal with a sample of a hihat or kick. The best results for a good pumping (2 or 3:1 ratio, 72 attack, 93 release) is using the shortest sample possible. Of course your bassline must be EQ'd down with a low cut of about 250, otherwise your kick will have no punch and its affect in the mix will be diminished substantially (the kick and the bassline cannot compete for frequency room in the mix). Ideally the side chain should not be a "noticable" characteristic, but rather used just enough to achieve clarity and room in the mix so all the elements can fit and interact properly. Not over done.

If you have a good percussive groove, it's worth a shot to see if you can side chain that because it does free up room in the mix and offer support to the 4 on the floor beat (make sure and signal your compressor to side chain on the off-beat of the kick drum, although use that advice with discretion). It can even make it a little more groovy, try it out. Not all loops should be side-chain, though. Most percussive grooves that drive a track are made up of numerous loops and samples. You can try side chaining the master percussion, or go into individual loops and try varying side chain methods to each loop accordingly. Try varying ratios, attacks and releases on all your compressors that do side chaining. There is no limit to the possibilities you can try and you can get a really robust professional sounding mix the more control you have over dynamics. Although a side chain can never replace a good, swingin', well-grooved percussive loop so don't expect to get fantastic drums just by knowing how to side chain.

One of the examples I should mention where you do want an obvious sounding side chain would be in some huge trance pads. Or in any track where you need a good pumping effect. Always used a well-rested ear to establish the threshold because if your ears are exhausted it's hard to get this right.

Take a listen to some Daniel Kandi tracks, he uses Reason 3 exclusively. He does not use the M-Suite compressor with sidechain function in Reason, use does volume automation and (obviously) it sounds fucking fantastic.

quote:

Originally posted by PutBoy That's it excactly. It's useful if what you are sidechaining is broad. For example if you're sidechaining everything in the song, then this way of doing it would be very helpful That's the way I do it, but you could ofc use the multiband compressor instead.

I can definately see how this would make the kick really prominent now, when using the eq instead of a compressor. The ducking will completely follow the kicks characteristic. Like Evolve140 says sidechain volume automation may very well have an even better effect than sidechain compression at times. And I can see how this will make it really subtle if thats what you want, by sidechaining only the freq where the kicks energy currently is at. So when the kick rolls down in freq, the sidechain rolls with it. Excellent technique. If you were to use attack times on a multiband compressor to get this perfect instead, you'd have to struggle alot more. With an eq steered by the kick divided in three parts the attack times will be instant according to where the kick is at, and that way make it subtle and really work to only make the kick prominent as opposed to making an audible ducking effect.. Also controlling the tension and decay with the fpc.

quote:

Originally posted by Evolve140 Of course your bassline must be EQ'd down with a low cut of about 250, otherwise your kick will have no punch and its affect in the mix will be diminished substantially (the kick and the bassline cannot compete for frequency room in the mix).

Hm, bass highpassed at 250 might not be such a good idea. Most of the kick's energy in EDM comes really low, 55-75 hz with timbres much higher, that don't interfere that much. If you keep your low-end bass at around 80hz, and have a healthy sidechain, it should work pretty fine.

Sometimes I layer 3 basslines, low end comes from 80-125 range, to add some low-colour there's another one playing 125-180 range and then the "main" bass going from 180 up, and the kick remains strong, aswell with the bass.

Having your bass going from 250 up only will make a track really weak in its low end.

What you guys refer to isn't sidechaining, it's called ducking. It's been existing for at least 20 years. All people including myself call it now SC, but it's not. The Sidechain means that the source feeding the effect on channel A is coming from channel B.

Behringer were making those Intelligate units 15 years ago. This is what everyone was using to slice pads and long sounds.

You know the rule with compression. Some of you noticed the effects when applying ducking on pads with vocals, which brings clarity.

Also the pumping pad effect, but you don't need a compressor for that with 4/4 kicks, you can easily use an autopan or tremolo effect set to center position so both sides are in sync, you put a reverse sawtooth as the waveform synced to 1/4, and smooth it a little.


You can try the basic uses:

1�)Send a channel A to an aux 1 with a reverb on it, put a compressor after the reverb, and set the sidechain to channel A. As heard in Benny B's and Rank 1 tracks

2�) Channel A is a vocal, copy it to channel B, save it as a separate file, reverse it, reverb it with 10 sec decay and 100% wet, bounce it, re-reverse it, and put it back on Ch B and remove all plugins, now compress channel B with Sidechain feed from CH A, and don't forget to replace the reversed region correctly. As heard in a zillion vocal trance records.

3�) Filter and Gate do sidechain too. Use a hihat pattern to feed those placed on a pad channel and enjoy the results. As heard in LSG, for example.

Keep in mind that for the best results, the feeding material MUST be panned to CENTER and most preferably MONO, unless that's what you want.

Some VI's also feature SC. Logic users, you can use Ultrabeat to slice and dice entire regions feeding Ultrabeat in time. ES2 makes it possible to route the sidechain signal to an osc parameter such as Pitch/Filter, whatever.

These are some basic uses of SC. There are plenty more. See SC as an external modulation protocol, as E140 said. Sky is the limit.

Example, copy your reverb aux to another one. decay to death, put a filter or a gate that gets fed dynamically by a 909 closed-open hat sequence with variable amplitude. Bounce that, and loop that.

quote:

Originally posted by Tarpex Hm, bass highpassed at 250 might not be such a good idea. Most of the kick's energy in EDM comes really low, 55-75 hz with timbres much higher, that don't interfere that much. If you keep your low-end bass at around 80hz, and have a healthy sidechain, it should work pretty fine. Sometimes I layer 3 basslines, low end comes from 80-125 range, to add some low-colour there's another one playing 125-180 range and then the "main" bass going from 180 up, and the kick remains strong, aswell with the bass. Having your bass going from 250 up only will make a track really weak in its low end.

My bet is when he says low cut he actually means low pass.. Lol... low cut bass at 250 and you wont have any bass left :P Probably the stupidest recommendation I've ever heard.. but like I said he probably means low pass.. if not he should do some more research on what bass is, where it's at in the freq spectrum :P

Some very interesting stuff here guys. Thanks for not treating your methods like cold war secrets! I don't use much SC so I would consider it a weakness. Would love to learn more about it so I can use it when needed. Keep the informative posts coming and thanks in advance

quote:

Originally posted by Lolo What you guys refer to isn't sidechaining, it's called ducking. It's been existing for at least 20 years. All people including myself call it now SC, but it's not. The Sidechain means that the source feeding the effect on channel A is coming from channel B. Behringer were making those Intelligate units 15 years ago. This is what everyone was using to slice pads and long sounds. You know the rule with compression. Some of you noticed the effects when applying ducking on pads with vocals, which brings clarity. Also the pumping pad effect, but you don't need a compressor for that with 4/4 kicks, you can easily use an autopan or tremolo effect set to center position so both sides are in sync, you put a reverse sawtooth as the waveform synced to 1/4, and smooth it a little. You can try the basic uses: 1�)Send a channel A to an aux 1 with a reverb on it, put a compressor after the reverb, and set the sidechain to channel A. As heard in Benny B's and Rank 1 tracks 2�) Channel A is a vocal, copy it to channel B, save it as a separate file, reverse it, reverb it with 10 sec decay and 100% wet, bounce it, re-reverse it, and put it back on Ch B and remove all plugins, now compress channel B with Sidechain feed from CH A, and don't forget to replace the reversed region correctly. As heard in a zillion vocal trance records. 3�) Filter and Gate do sidechain too. Use a hihat pattern to feed those placed on a pad channel and enjoy the results. As heard in LSG, for example. Keep in mind that for the best results, the feeding material MUST be panned to CENTER and most preferably MONO, unless that's what you want. Some VI's also feature SC. Logic users, you can use Ultrabeat to slice and dice entire regions feeding Ultrabeat in time. ES2 makes it possible to route the sidechain signal to an osc parameter such as Pitch/Filter, whatever. These are some basic uses of SC. There are plenty more. See SC as an external modulation protocol, as E140 said. Sky is the limit. Example, copy your reverb aux to another one. decay to death, put a filter or a gate that gets fed dynamically by a 909 closed-open hat sequence with variable amplitude. Bounce that, and loop that.

Yea.. sidechaining once wasn't what it is commonly known as now, a musical effect. It was first invented for radio speech as an automated ducking of the music when the host was speaking (if I don't remember completely wrong...) And as we all know its still in use for that today..

quote:

Also the pumping pad effect, but you don't need a compressor for that with 4/4 kicks, you can easily use an autopan or tremolo effect set to center position so both sides are in sync, you put a reverse sawtooth as the waveform synced to 1/4, and smooth it a little.

Thats a really good tip you had there actually.. Funny I've never thought of synced wah wah effects or similar in the context of sidechain substitution before.

quote:

Originally posted by lenieNt Force Yea.. sidechaining once wasn't what it is commonly known as now, a musical effect. It was first invented for radio speech as an automated ducking of the music when the host was speaking (if I don't remember completely wrong...) And as we all know its still in use for that today.. Thats a really good tip you had there actually.. Funny I've never thought of synced wah wah effects or similar in the context of sidechain substitution before.

Good stuff, though I would use ducking to describe the effect used in voice overs and narrations where the vocals duck the music behind it so you can hear what they're saying. Technically speaking though he is correct. I would also like to mention this gentleman is reiterating what I mentioned in my post, where the effect is merely a practice in achieving mix clarity. Whereas the original topic does pertain to side chaining, the implications are vast so I do appreciate everyone's comments here because this is a very broad topic! Cheers.

quote:

Originally posted by Tarpex Hm, bass highpassed at 250 might not be such a good idea. Most of the kick's energy in EDM comes really low, 55-75 hz with timbres much higher, that don't interfere that much. If you keep your low-end bass at around 80hz, and have a healthy sidechain, it should work pretty fine. Sometimes I layer 3 basslines, low end comes from 80-125 range, to add some low-colour there's another one playing 125-180 range and then the "main" bass going from 180 up, and the kick remains strong, aswell with the bass. Having your bass going from 250 up only will make a track really weak in its low end.

I do mean bass as in the bass instrument and *not* the bass from the kick drum. The kick drum does require very low frequencies of course.

quote:

Originally posted by lenieNt Force My bet is when he says low cut he actually means low pass.. Lol... low cut bass at 250 and you wont have any bass left :P Probably the stupidest recommendation I've ever heard.. but like I said he probably means low pass.. if not he should do some more research on what bass is, where it's at in the freq spectrum :P

Yes, Low pass... I have heard of different producers cutting off low ends at certain points, ie., 125, 150, 250... what ever it takes to make sure you can mix it in with the kick.

quote:

Originally posted by Evolve140 Good stuff, though I would use ducking to describe the effect used in voice overs and narrations where the vocals duck the music behind it so you can hear what they're saying. Technically speaking though he is correct. I would also like to mention this gentleman is reiterating what I mentioned in my post, where the effect is merely a practice in achieving mix clarity. Whereas the original topic does pertain to side chaining, the implications are vast so I do appreciate everyone's comments here because this is a very broad topic! Cheers.

Ye it's ducking, but thats where the concept of sidechain compression first was introduced.

This gentleman is Airwave btw

quote:

Yes, Low pass... I have heard of different producers cutting off low ends at certain points, ie., 125, 150, 250... what ever it takes to make sure you can mix it in with the kick.

Yeap, but then you most probably will have a sub-bass under that.. Often you build up your lower end of many layers of bass working together right. With no bass instrument under 250hz it will usually sound very weak, and even weaker on bigger PA. If one wants to get the kick/balance right one should make a cut in the bass where the kick is most prominent (40-70hz). Rather a notch cut than high pass, so as to not lose the depth of the bass. And then you can just low-cut the whole mix at 30 hz afterwards.

quote:

Originally posted by Evolve140 Yes, Low pass... I have heard of different producers cutting off low ends at certain points, ie., 125, 150, 250... what ever it takes to make sure you can mix it in with the kick.

Low pass doesn't cut off low end - it allows the low frequencies to pass through, meaning the high end is cut off.

A lot of trance songs have basses which have frequencies below those of the kick. You don't need to completely separate the bass and kick, frequency-wise. You can also do this with programming - for example, with an offbeat bass, it doesn't matter if the bass is as prominent at 65Hz as the kick - they don't play at the same time, so aren't competing for that space.

Powered by: vBulletin
Copyright © 2000-2021, Jelsoft Enterprises Ltd.