Order of applying Effects etc

[karlp]

Is there a correct order in which plugins or external effects like compression, reverb should be applied to a mix.

Any advice appreciated.

Karl

[PeeTee]

ReniuR...you should be going comp-eq-verb. Eq before comp is a no-no.

[ReniuR]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by PeeTee:
ReniuR...you should be going comp-eq-verb. Eq before comp is a no-no.<HR></BLOCKQUOTE>

Thanx for the tip. I'll remember it images/icons/smile.gif

[QuikDraw]

First of all, there's no such thing as a no-no! Secondly, it depends on what you want to accomplish. It is often a good idea to put a hi-pass filter in front of compressors/limiters. In many cases if you don't filter first you'll get poor results due to the fact that the bass freqs contain a lot more energy than higher freqs. It'll compress things that really shouldn't be compressed due to the extra bass energy that you're going to filter out anyway.

Mike

[snoopy]

I agree with quik. I often EQ before compress on stuff like guitar tracks. One of the nice features about 5.1 is that you can drag plugins, so I suggest swapping the order around and see what you like best (this will vary from track to track),

[marcusb]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by QuikDraw:
It is often a good idea to put a hi-pass filter in front of compressors/limiters.<HR></BLOCKQUOTE>

In all correctness you'd be wanting to put a hi-pass filter in the SIDECHAIN unless you wanna pull the bottom end out of all your tracks.

Specific orders really only need to be considered with 'processors' as opposed to effects. A processor is designed to take an input, process it and output the processed signal. Examples of processors are eq's, gates, compressors, de-essers, auto-tune images/icons/smile.gif

Processors are designed to modify the signal and and to not be mixed with the original signal. An effect on the other hand creates a new signal from an input and are designed to be used mixed with the original signal (ie. reverbs, delays etc ...)

These are general rules and of course CAN be broken. However more specific to your question you need to consider what happens to your signal when you run processors in serial (ie. at your insert points) ... for example an EQ before a compressor is significantly different to a compressor followed by an EQ. Both setups have their purpose, but you need to train yourself to think about what is happening to your signal when you run processors in serial.

Back to the EQ before the compressor example if for instance you setup your eq, and then your compressor all is working nicely, then you go back and change your EQ .. say add a boost in the upper mids .. all of a sudden your compressor is operating differently because it is tracking differently with the EQ boost.

When it comes to effects however there are defintinely no rules, this is where you can experiment and come up with crazy sounds.

Hope that makes a bit more sense.

Cheers,
Marcus

[ReniuR]

It all depends on what you want it to sound like. I like to go, for example, eq-comp-verb etc.

Experiment, and see what sounds best to you.

[karlp]

Thanks for your help.

I record live brass bands onto ADAT (at present 8 track but looking to 16). Then mix on PTLE adding compression on percussion and reverb on main instruments. I do no other type of recordings.

I have 2* Rode NT2 and have just got a Neumann U87ai (getting another soon). Pre amps are TLA 5001 (maybe get a better one when I get the other Neumann).

My weak point in the ADAT 4.03 but when the Alesis HD24 is out I am thinking of one of those.

I have just bought Golden Ears (music training course).

AS I said before I do not make any money out of this (infact lose a fortune) but it is fun and I want to give the people who ask me to record them the best I can.

Thanks guys.

Karl

[dsteele]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR> PeeTee: ReniuR...you should be going comp-eq-verb. Eq before comp is a no-no. <HR></BLOCKQUOTE>

I believe a lot people assume this for a simple, incorrect reason.

When you are tracking it is very nice to sometimes compress with outboard gear, in order to get the signal hotter before presenting it to the A/D converters.

However once the signal is already digital, this reasoning doesn't apply and you are pretty much free to use a different order
according to how it sounds.

D.

[lwilliam]

FWIW, my first plugin slot often has a HPF on it; prior to the comp or other eq.

It would be pretty unusual (except for a bass instrument) to want the comp to respond to frequencies below about 80-90hz.

[ August 30, 2001: Message edited by: lwilliam ]

[Mr T]

Do agree with the Eq low frequencies before comp. thing. Unless you want your comp. to pump like hell and "eat " the other frequencies, do this!

[jeffinmarin]

Great thread. What do you guys think about using a limiter AND compression on the same track??? For example, I have a track with comp. but need to limit a couple of spikes. Using more comp. colors the sound,so why not use limiting for the spikes? Also,what do you people think of the stock plugins.Please, no more harping about D-verb. How bout the other plugins? images/icons/grin.gif images/icons/grin.gif

[ReniuR]

Personally, I think the stock plugs are fine to get you started (and then some, if you use them right). If you need something better, get outboard gear. The comp + limiter thing is a good question. I have done that before, using the limiter first, to cut off the high peaks, then a compressor to squeeze it a bit more and add that compressor tone. It worked pretty well for what I was going for. The guitar on some of the Anonymous Threats tracks on our site were done that way.

(we've been having problems with our page today, so plz try later if it doesn't work)

[ August 31, 2001: Message edited by: ReniuR ]

[jeffinmarin]

RENIUR, Thanx for your help. i`m wondering if you (and others) are saying that outboard processors are "better" than most REASONABLE optional plug-ins. I`m fairly familar w/ outboard stuf, but I don`t know much about history.I mean other co.`s s/w plug-ins. By reasonable,I mean price wise. Again,THANX... images/icons/grin.gif

[marcusb]

I would avoid sticking a limiter on the track if there are only a few spikes. Why not do one of the following (in my own order of preference):

1) reduce the gain of the spikes using the audiosuite gain plugin

2) apply limiting to just the spikes

3) dip the spikes with fader automation (this will be post compression of course)

Cheers,
Marcus

[odysseys]

I would like to comment to all these that there are no rules in using any of the tools we have in any order.If,for instance put an eq with +6 db at 3-6 khz and a compressor after,you'll have a de-esser,and if you do the opposite you'll hear something else.The point in using all these stuff is getting the SOUND you want some time in a particular song.Most of us cannot imagine a sound and make it real (as we imagined it) in a modular synth consisted of 100+ modules (most sound physists can).So we experiment most of the time and if we find something interesting then we apply it to music.Few times we know what we are looking for and we do it blindly (How can i make my voice solid as a rock? Oh,i'll put a limiter! images/icons/rolleyes.gif ).Finally,i'm trying to say that we should experiment with tools without taking very seriously the "theory" but at the same time try to learn as much as we can about the tools we have so the experimentation is no longer a blind trip,but an exciting course into the world of sound images/icons/blush.gif hrr,psiii,hrrr,psiii

[ReniuR]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by jeffinmarin:
i`m wondering if you (and others) are saying that outboard processors are "better" than most REASONABLE optional plug-ins.<HR></BLOCKQUOTE>

All I'm really gonna say on that matter is that I prefer outboard gear. IMO, it gives a certain quality that plugins and software can't re-create. But, whatever you are looking for (sound-wise) will vary your tastes of sound. I hear EVERYBODY under the sun bitch about D-verb, but I kinda like it for some things. Everything has a use to somebody.

Keep having fun with it.

[karlp]

I like the idea of outboard gear and have got a Lexicon PCM81 and a DBX Quantum.

As yet no EQ or compressor - can you guys recommend a suitable one up to 1000GBP each.

I find the plugins not very good but I am looking at the Waves Native Bundle as it is reported to be good.

Thanks you all for your advice - very much appreciated.

Karl

[editor]

I do mainly mastering, the general rule of thumb with the guys I know is eq, comp, limiter with dither, Although, whatever sounds good to you is what you might want to use.

Regards

e

[billythekid]

Yes that’s a hard question; I like using the EQ after the compression but never before. I think when ever I use the EQ before the compression I get a amp- type of vocal not as real as doing it the other way around but that’s only my thought. But that’s a good question that I will be back to read every ones answer to thank’s images/icons/grin.gif

[bassmac]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by lwilliam:
FWIW, my first plugin slot often has a HPF on it; prior to the comp or other eq.<HR></BLOCKQUOTE>

What plug do you use for your HPF?

[lwilliam]

1-Band Digirack EQ set as HPF if it's realtime only. If I'm sure I'll need it, then I'll use the RenEQ 2-band or the Q4 and just activate the low band as a HPF - they're better sounding than the digirack EQ.

I often use a/s HPFs on non-bass instruments - it really helps keep the low end of the mix tight and clean.

[marcusb]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by odysseys:
I would like to comment to all these that there are no rules in using any of the tools we have in any order.If,for instance put an eq with +6 db at 3-6 khz and a compressor after,you'll have a de-esser<HR></BLOCKQUOTE>

This is TOTALLY incorrect ... say your sibilant problems are around 8.5k, so you add a +6db boost at 8.5k, you then run into a full range compressor, you set it so that it's reading around 6db gain reduction on your now INCREDIBLY peaky 'esses what happens to your audio you are reducing the level of your entire signal by 6db, so what you now have is a 6db quieter signal with a massive 6db boost at 8.5k and your compressor is pumping like mad on the esses.

the correct way to do de-essing is to use an EQ in the SIDECHAIN! ie. you don't wanna add a gain boost to the actual SIGNAL you are just making the problem about 300% worse ...

a sidechain point allows you to insert an EQ before the compressor's detector circuit, here is where you add the EQ boost, what the compressor hears (and subsequently reacts on) is the EQ'd signal, but it is processing the un-eq'd signal

I suggest you read up on sidechain keying in your plugin documentation.

Marcus

[odysseys]

You don't make sense! Firstly,as for the sibilant frequencies i set a range that meets a broad band of voices that i have heard (not excluding yours).Secondly,if you set the thresold of the compressor to trigger it just above the program (so it's been triggered ONLY by the sibilant peaks) you get an apropriate decrease of sibilance by the ratio you have put.I think you're talking about expanders here! The side chain method is a way too,but i prefer multiband for the purpose

[lwilliam]

If you just boost the EQ into the compressor, you have a high frequency bump in the whole signal. You will then have an eq'ed track with compression. The 6db boost is there whether or not it's compressing.

De-essing is done only by using the sidechain input. By using the sidechain, you are telling the compressor to compress ONLY on the 6db peaks. At levels below the threshold, the track is not affected by that EQ boost. In fact, even at levels above the threshold, the eq boost doesn't affect the EQ of the track, just whether the compressor kicks in or not.

Compressing an EQ'd track might help with some sibilance because the compressor is working harder on the frequencies that are boosted, but it's not a de-esser. It's still an EQ'd track being compressed. De-essing doesn't put EQ on the track.

Hope that makes sense... images/icons/smile.gif

[kalle74]

Marcusb and lwilliams are right on the sidechain thing.

[marcusb]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR>Originally posted by odysseys:
Firstly,as for the sibilant frequencies i set a range that meets a broad band of voices that i have heard (not excluding yours)<HR></BLOCKQUOTE>

In a multitrack situation or even in a mastering situation you should be tuning your eq in as tightly as you can to the specific problem frequency NOT a broad range of frequencies that may apply to a bunch of different voices. You'll end up taking all the top end out of the vocal leading to a lispy lifeless vocal track. Over a mix you'll totally kill your cymbals.

Marcus

[ September 05, 2001: Message edited by: marcusb ]

[odysseys]

Iwilliam you are right except one thing.When you set the RIGHT thresold,the compressor reacts only when the signal passes the thresold you set,so only the part of the vocal with sibilance is affected.The +6 dB thing is to help the compressor identify more easily the problem frequencies so you can do with them whatever you want.Markus,i think that sibilance is more than a few frequencies but i agree that for mastering it's better to work with small Qs on eq

[KaiXXX]

I use filters before compression and EQing afterwards. The reasoin I do this is the following: I filter the signal to get rid of unwanted lo-fre. rumble or (hi-freq) hiss. For the lo-cut I always hit it on the mic pre. But for OH I might use a plugin, because I need it at a higher frequency.
Then I leave the filters as they are. This way compressors won't react on rumble I filter out anyway, but are not influenced by the EQing that might change during the mix.

For gating I'd rather edit the track than using the plugin (safes resources, too images/icons/wink.gif ). But if I'm gating, I'll duplicate the track, drag it a little forward and use it for sidechaining. The gate opens a little before the signal, this works killer (might also use EQing for the dupl. track. Hey, this should work for the de-essing stuff, too!)

I'd rather apply limiters to the stereo bus. It makes not too much sense to cut the peaks of a single track which you might mix at -10dB. Just limit the peaks of the whole mix not of the single tracks (an exception might be the lead vocals. Exception what from, actually?).

Since comps/limiters change the sound of the track, there is another method called 'gain riding'. Just volume-automate the track, then send the output of that track to an aux track and apply some limiting or compression there. This is especially useful when there are volume changes within the track. This way the comp hasn't to make these up, going -9dB all the time at the louder parts.

[PeeTee]

I don't consider a filter to be an equalizer. I don't know anybody who does.

[Park Seward]

You can consider an equalizer a filter. You might say a filter is an equalizer that only cuts.

Some like to limit the sound going to the A/D so as not to get the "brick wall distortion". In fact, some mic-pres have built-in limiters for this reason. I would rather have a limit than digital distortion. Also, some think you can get a better sound the harder you hit the A/D, as long as you stay below "0".

In audio processing in radio, we would always use a compressor first followed by a peak limiter. The limiter was set so we could turn up the modulation without peak distortion. Some AM peak limiters actually "turned around" negative peaks to positive peaks. You were limited to 100% negative modulation but could go 125% positive mod. Again, this made the station louder.

Most of the signal processing was tuned to the engineer's taste. They made it sound like they wanted. It was usually pretty heavy. The mod meter barley moved! Some station consultants even had their own secret and magic compresson settings!

If you want to control bass and treble frequencies separately, either get a multi-band compressor or make your own. Duplicate your track three times. Put a filter on each track and have the first one a low pass (just bass), the middle track just the middle frequencies and the last track just the high frequencies. Then put a comp/limiter on each track. By adjusting the relative gains of the three tracks, you can get an EQ balance. With this way, the bass won't effect other frequencies. Play with the overlap of the crossover on the filters and adjust the settings on the comps differently. You may want a tighter bass, for instance (faster attack and release.

I usually put a compressor on every strip and then on the master for rock music. It's because I like the sound of compression, not just to control levels. For vocals, I like the Waves Renaissance compressor and would not add a limiter to that since it has one built-in. With the standard plug-ins, I do add the comp and limiter together for vocals.

[marcusb]

Compressors on every channel ?? Geeezus, over-compression is one of the biggest problems with a large majority of recorded music these days.

The biggest problem being that with all the new DAW based engineers about 1 in 10 of them don't even know what a compressor does (aside from "makes things louder") let alone actually hearing the improvement, or most of the time the damage, they are making to their signal.

Believe it or not, maintaining reasonable dynamic range will actually make your music punchy ... think about it logically for something to sound LOUD you need to hear it in the context of something being SOFT preceding it. Rock music can have crescendos and decrescendos, let your music breathe .. PLLLLLLLEEEEEEASE images/icons/smile.gif

Marcus

[Park Seward]

I've been recording with compressors since the 70s. Like I said, I like the sound.

I didn't know that was a problem. I guess I missed those rules somewhere...

[ReniuR]

yeah, me too. I guess that is what the faders and automation are for then?

[where02190]

OK I gotta put my 2 cents worth in here.
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:<HR> Iwilliam you are right except one thing.When you set the RIGHT thresold,the compressor reacts only when the signal passes the thresold you set,so only the part of the vocal with sibilance is affected.The +6 dB thing is to help the compressor identify more easily the problem frequencies so you can do with them whatever you want.<HR></BLOCKQUOTE>

Time to go back to recording 101 odysseys. The comressor has no idea what it is and is not supposed to react to in full bandwith comression, EXCEPT ANY SIGNALS THAT ARE OVER THE THRESHOLD. In your scenerio, your resultant output is identical to the input except the entire signal bandwidth is reduced by the parameters set on the compressor. The resultant track will still have the +6 high end, that is reduced in equal proportion to the rest of the spectrum, and I don't even want to think about how that is going to sound.

A de-esser is governed by an external or internal eq/filter, to only compress a certain frequency range of the incoming signal, by placing the eq/filter into the sidechain address of the detector circuitry of the compressor.
Broadband compression affects the entire mix, regardless of it's frequency content. Once the comp reached threshold, it compresses everything. Given that lower frequencies have more energy in comparison to higher frequencies of the same spl, before your compressor reacted to the high end boosting over the mid/low levels, your vocal woudl be so painfully sibilent, I doubt it would even be intelligible.

[Park Seward]

BTW, I also like too much reverb, too much blonde hair on my women and too much sugar in my tea.

I drive too fast on the freeways and enjoy too much JD in my Coke.

[odysseys]

Sorry,I have other things in my head and i s....d up.I delete what i said except my first reply to the topic (not including the deessing thing)
I'll be more responsible when i discuss with all of you.
Sorry again.

[Jason from MaggieJack]

If you use the Waves renaisance eq it has a function that adjusts the trim (input signal) to the highest setting without peaking, which is actually normalizing the signal first thing in the chain. It really works great. After that compression or multi-band comp, then effects, last limiter. I would probably put a gate before the compression.

Peace,
Jason images/icons/tongue.gif

[APAULOS]

Did someone say WAVES. There support of ptle windows is the best thing that has happened to digi/pc since... well... digi/pc. Have you guys heard the reny col. 2? try it you'll like it www.waves.com (http://www.waves.com)