PDA

View Full Version : Multiple high-pass and compression or just one?


Jakay
09-25-2013, 12:58 AM
Just looking for advice here.

I have a bunch of vocals all bussed into one "vocals" aux track. Let's say there's six of them. Each one of them has a high-pass filter and light compression on it.

Now what I'd like to know is what the effect of summing these is in regards to these processes. My guess for compression is that it's somewhat obviously double compressing the final result. As for the high-pass though, I don't really know.

Does applying a high-pass (low-cut, whatevs) to each signal do anything if I'm applying an identical high-pass to the final summed aux track?

I've been turning this over in my head for a day or two and don't have any insight beyond "it's probably unnecessary to high-pass each track initially". But I would like to know if anyone has any insight on this.

Compression I can understand in so far as I might want different settings for different styles, but hypothetically would one final compression do the same as six initial (and assuming identical) compressions?

This might all be a bit too nitty gritty to matter too, I realize. Geekin out perhaps.

melloj
09-25-2013, 03:28 AM
Just looking for advice here.

I have a bunch of vocals all bussed into one "vocals" aux track. Let's say there's six of them. Each one of them has a high-pass filter and light compression on it.

Now what I'd like to know is what the effect of summing these is in regards to these processes. My guess for compression is that it's somewhat obviously double compressing the final result. As for the high-pass though, I don't really know.

Does applying a high-pass (low-cut, whatevs) to each signal do anything if I'm applying an identical high-pass to the final summed aux track?

I've been turning this over in my head for a day or two and don't have any insight beyond "it's probably unnecessary to high-pass each track initially". But I would like to know if anyone has any insight on this.

Compression I can understand in so far as I might want different settings for different styles, but hypothetically would one final compression do the same as six initial (and assuming identical) compressions?

This might all be a bit too nitty gritty to matter too, I realize. Geekin out perhaps.
Now what I'd like to know is what the effect of summing these is in regards to these processes. My guess for compression is that it's somewhat obviously double compressing the final result. As for the high-pass though, I don't really know.

No. You are not doubling the effects of plugins by simply sending them to an aux. To double the effects of plugins you have to stack them in the same insert eg. Some engineers will stack compressors on vocals by summing the ratios: for instance instead of setting one compressor to 5. Set it two comps to 2.5. This gives the required ratio without making the compressors work hard.


Does applying a high-pass (low-cut, whatevs) to each signal do anything if I'm applying an identical high-pass to the final summed aux track?

Technically yes. You are processing 6 individual tracks with 6 individual frequency responses and amplitude. A sum of all six channels means a single hi pass filter has to deal with the frequency response of all six channels at once. Practically though you wouldn't notice much difference.

Compression I can understand in so far as I might want different settings for different styles, but hypothetically would one final compression do the same as six initial (and assuming identical) compressions?

No. For the same reasons as above. Using individual compressors on different tracks would give different results to putting one compressor on the aux channel. simply because each individual channel has it's own frequency and amplitude. A sum of all these channels gives a different frequency and amplitude response. One compressor can't deal with a sum of all six signals in the same way as one compressor on each individual channel.

But...how does it sound????? Don't get too hung up about this sort of thing. If you are getting the results you want and it sounds good, then it is correct.

musicman691
09-25-2013, 04:10 AM
Just looking for advice here.

I have a bunch of vocals all bussed into one "vocals" aux track. Let's say there's six of them. Each one of them has a high-pass filter and light compression on it.

Now what I'd like to know is what the effect of summing these is in regards to these processes. My guess for compression is that it's somewhat obviously double compressing the final result. As for the high-pass though, I don't really know.

Does applying a high-pass (low-cut, whatevs) to each signal do anything if I'm applying an identical high-pass to the final summed aux track?

I've been turning this over in my head for a day or two and don't have any insight beyond "it's probably unnecessary to high-pass each track initially". But I would like to know if anyone has any insight on this.

Compression I can understand in so far as I might want different settings for different styles, but hypothetically would one final compression do the same as six initial (and assuming identical) compressions?

This might all be a bit too nitty gritty to matter too, I realize. Geekin out perhaps.First thing I'm wondering about: are you talking about having the filter on the 6 individual tracks being summed by the aux track AND a filter on the aux track covering the same freqs (effectively double filtering)? If so then the effect you'd see would be a steeper rolloff of freqs due to the double filtering. Not a bad thing if that's what you want.

Also if you're talking the same thing with compression with compressors on both the feeder and summed tracks you could have double compression depending on how the compressors are setup. As long as you're not in the red on the meters then you could set the compressors on the feeder tracks for a light compression and the one on the aux summer track to act as a glue for the feeder tracks. I'd use a compressor on the aux summer track either way to catch any peaks that happen due to the summing.

nst7
09-25-2013, 07:16 AM
I suggest trying things both ways, or combining them. For compression, I've found that putting a compressor on the background vocal buss (which sums all of them) can give a glue or unifying sound to them, that's not acheived by just putting compressors on each one. You could also try a combination of both - light compression on individuals, and light compression on the buss.

albee1952
09-25-2013, 10:57 AM
If you are applying the same EQ to all, then you could use a single EQ on the summed AUX track to achieve the same sound. As for compression, I would compress each individual track. If you only compress the group, then the loudest single voice will push everything else "down". If you want to compress the group, WITH compressors on each track, that might smooth things out a bit more. Its not unusual(apologies to Tom Jones:D) to use more than 1 compressor on a signal. I often use 2 on my lead vocals(BF76 working lightly to shave off the peaks, followed by SMACK! to give more compression and a bit of character).

Jakay
09-25-2013, 07:52 PM
Steeper roll off... that was the concept I just couldn't reach...

Thanks everyone, that shed some insight on it for me.

As for the compression, the idea was to tame each track, and then glue the final vocals together, which seems in agreement with everything you all said.

With the EQ, I got mixed answers with "no effect" and "steeper roll-off" both of which make sense to me, but I'm not clear on this still.

To be more specific, which I wasn't, upon reading my post, I am indeed applying EQ a second time on the aux channel, effectively double stacking as this is post insert output bussing rather than aux sends to an auxiliary channel.

I'm not going to lose sleep over this, and playing around with what sounds best to me is what I do best, but I am very much in the learning curve throws right now, so whatever bits of understanding I can integrate into my end-game is immensely useful, long-term, to me right now, as it will affect everything I do from here on out.

All your responses were helpful though, and thanks for that.

"no effect" vs "steeper roll-off" is what I'm struggling with. The rest makes sense. Steeper roll-off makes more sense to me, and fwiw, I'm inclined to roll with that.

Led
10-15-2013, 02:29 AM
If you apply a high pass filter twice, once on the source track and again on the aux, everything under the corner frequency will be reduced twice ie -
if on both you have a -6db per octave hpf set @ 100Hz, you reduce everything under 100Hz by 6db per octave on the first track, then when it goes through the aux you are reducing every under 100Hz by -6db per octave again. So after going through the second filter what was reduced by 6db is now reduced by 12db. So with filters in series the slope of db per octave becomes additive.

On the compression thing, compressors in series have the effect of multiplicative ratios. A 4:1 comp followed by a 2:1 gives you 8:1 in effect over the threshold rather than 6:1. As always use your ears, but worth understanding. If it sounds good it usually is good.

Stig Eliassen
10-15-2013, 08:45 AM
Also take into account that some of these EQ plugins sounds different depending on how hard they're driven (audio input level). So, there could very well be a sonic difference between routing 6 clean vocal tracks to an aux with an EQ instatiated, and using the same EQ settings on the individual tracks.

But, you know... gain staging is key, regardless of workflow and setup. Just throwing that in there. :D

Jakay
10-16-2013, 02:10 PM
Yeah, the additive and multiplying factor of those processes makes sense now. I think I'd intuitively reached that conclusion, but seeing the straight up math of it rings crystal clear.

As for gain staging, that's good advice too. It wasn't until last month's Sound On Sound that I even knew that some of these plug-ins produce different results based on how hard you drove them. And properly gain staging EQ and compression makes a whole more sense now, especially concerning threshold levels.

And here I used to think normalizing was the best thing ever...