Print

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

[thephatboi]

I was wondering if you guys know of a plug-in that will simply break up the stereo audio spectrum into low, mid and high bands that could be bussed to separate destinations for processing, then recombined.  I use Protools HD and am looking for something that does this, basically a "crossover plugin" with ability to route bands to separate places,  I know some people acomplish this by using duplicate tracks with hi, low and band pass eqs but I wonder about weirdness at the crossover points. Any info appreciated. thanks, Sean

[Ged Leitch]

Hi Sean, i dont know of any plugins that will buss out different bands but someone else might.
Can I ask why you want to do this???

just curious...

[Dave Davis]

I don't know of any plug-ins that do this, however it's not rocket science which is probably why there aren't any common ones out there (little to no market!).

To get 'er done:

1) Mult your source to the same number of aux tracks as bands you need via sends or directly
2) Apply hi/lo pass filters at the appropriate and desired crossover points as the first plug of the recieving aux.
3) Apply any processing you desire in slots after that
4) return these aux's to the main bus, or, recombine in an additional aux track for further processing.

Done.

While some plugs might create some wierdness at those points, it's something you can sort out by ear or instrument.  Further benefit is that you can "flavor" or not with your choice of filters/plugs, and use the crossover points as a means to filter out specific troublesome frequencies in a different way.  Its not quite the same thing as what you're asking for, but depending on the eq you select it can be a very good alternative.

-d-

[Jerry Tubb]

Dave Davis wrote on Sun, 14 May 2006 17:33

1) Mult your source to the same number of aux tracks as bands you need via sends or directly 2) Apply hi/lo pass filters at the appropriate and desired crossover points as the first plug of the recieving aux. 3) Apply any processing you desire in slots after that 4) return these aux's to the main bus, or, recombine in an additional aux track for further processing.

Hi Dave... a relevant tale from days of yore:

Back in the mid '90s I was working with a picky producer on a "roots rock" project.

One song had multiple mixes, none of which he was happy with.

Remarked that he liked one mix, except for the low bass and drums.

Another mix he liked the low bass and drums but not the rest.

So we decide to combine the two!

IIRC we used a crossover of 300-350Hz, lo-passed the one with good lows, hi-passed the other one... and it actually worked!

Both pleasantly surprised, we listened carefully for any phase problems, I think we EQ'd the low mid a little around the x-over point, and he gave it the thumbs up.

I'm not advocating that lurkers go and do this, the odds were against us, but this one time we got lucky! Tried it a few other times over the years with less than stellar results.

BTW they were digital recordings so no chance for analog "drift".

Anyone else ever try this?

Cheers   JT

[Ronny]

Dave Davis wrote on Sun, 14 May 2006 18:33

I don't know of any plug-ins that do this, however it's not rocket science which is probably why there aren't any common ones out there (little to no market!). To get 'er done: 1) Mult your source to the same number of aux tracks as bands you need via sends or directly 2) Apply hi/lo pass filters at the appropriate and desired crossover points as the first plug of the recieving aux. 3) Apply any processing you desire in slots after that 4) return these aux's to the main bus, or, recombine in an additional aux track for further processing. Done. While some plugs might create some wierdness at those points, it's something you can sort out by ear or instrument.  Further benefit is that you can "flavor" or not with your choice of filters/plugs, and use the crossover points as a means to filter out specific troublesome frequencies in a different way.  Its not quite the same thing as what you're asking for, but depending on the eq you select it can be a very good alternative. -d-

And it's the way to do it, but beware that any boosted bands that lap crossover points between auxes will increase gain and can result in clipping, this can be taken care of by attenuating the auxes before the mix down back to two, so that gain remains at unity between the original and the final eq'd version.

I'd add a number 5:

1) Mult your source to the same number of aux tracks as bands you need via sends or directly
2) Apply hi/lo pass filters at the appropriate and desired crossover points as the first plug of the recieving aux.
3) Apply any processing you desire in slots after that
4) return these aux's to the main bus, or, recombine in an additional aux track for further processing.
5. Note any increased peak gain and attenuate the aux channels so that unity gain is kept on the final.

The final mix will than be the same gain as the original, but with the eq's applied. If say he high passes one aux at 120Hz and on another aux he bandpasses from 100Hz, the frequencies between 100Hz and 120Hz will be boosted more than the freq's below 100Hz and above 120Hz. It will be the same thing as boosting between 100 and 120Hz on another aux, due to multiple buss summing increase. Of course it depends on the slope of the filter and the high pass frequency may not be the start of the rolloff, for example it could be -6, -12 or -18dB where the user altered frequency resides, but you can expect a gain increase on any bands that are duplicated on each aux.

[Ronny]

Jerry Tubb wrote on Mon, 15 May 2006 02:30

Dave Davis wrote on Sun, 14 May 2006 17:33 1) Mult your source to the same number of aux tracks as bands you need via sends or directly 2) Apply hi/lo pass filters at the appropriate and desired crossover points as the first plug of the recieving aux. 3) Apply any processing you desire in slots after that 4) return these aux's to the main bus, or, recombine in an additional aux track for further processing. Hi Dave... a relevant tale from days of yore: Back in the mid '90s I was working with a picky producer on a "roots rock" project. One song had multiple mixes, none of which he was happy with. Remarked that he liked one mix, except for the low bass and drums. Another mix he liked the low bass and drums but not the rest. So we decide to combine the two! IIRC we used a crossover of 300-350Hz, lo-passed the one with good lows, hi-passed the other one... and it actually worked! Both pleasantly surprised, we listened carefully for any phase problems, I think we EQ'd the low mid a little around the x-over point, and he gave it the thumbs up... was a happy camper! I'm not advocating that lurkers go and do this, the odds were against us, but this one time we got lucky! Tried it a few other times over the years with less than stellar results. BTW they were digital recordings so no chance for analog "drift". Anyone else ever try this? Cheers   JT

Yep, done this, same technique that Dave is relating. I've split mono recordings into 8 different tracks and eq'd for each instrument on each track to get a pseudo stereo final. Same thing though, any bands that overlap increase in gain, so this may be the reason why you are lucky sometimes and not others and you mentioned that you eq'd around the x-over points, likely you attenuated rather than boosted. You have to be aware when using this technique that summed same frequencies will increase gain by +6dB. The gain increase continues with added busses, but will decrease in gain relative to the added bus, for example if you double a 1k tone, you'll get a 6dB increase, triple it and your gain will increase by about 4dB to +10dB not +12dB as you might expect(figuratively speaking) add another buss, gain increases by 3dB. It works that way, so that the +6dB increase will be on the first doubling and any subsequent added buss will increase gain, but decrease in ratio of gain from the original to the first doubling.  

[ATOR]

I've experimented with this. Three multed tracks with hp and lp filters on it. When you're setting up your crossover-filters it's good to have a normal stereo copy of the track for reference. You could phase reverse it and play it with the crossovered tracks, if they cancel out you're good.

I thought I would be nice to have different compressors on each band but I didn't like it at all.

[HansP]

I would propose a subtractive method:

create a highpassed and a lowpassed track, then SUBTRACT both from the original and you have a very exact midrange that will melt with the other 2 without phase problems.

[Dave Davis]

Ronny, I believe your number 5 is in the paragraph that followed. "While some plugs might create some wierdness at those points, it's something you can sort out by ear or instrument."

Something like Spectrafoo or PAZ will catch it.  I didn't add it as a step because it's not really processing, just checking and part of every job we do.

Hans, GREAT idea!  It would be easy to do in real time in PTHD by adding busses for the inverted-summing/subtraction.

-d-

[cerberus]

HansP wrote on Mon, 15 May 2006 06:34

I would propose a subtractive method:

good thinking.

i would use the gain controls on a multiband compressor. the crossovers are actually designed for this task.

jeff dinces

[Ronny]

HansP wrote on Mon, 15 May 2006 06:34

I would propose a subtractive method: create a highpassed and a lowpassed track, then SUBTRACT both from the original and you have a very exact midrange that will melt with the other 2 without phase problems.

Works ok for some apps and in his case of 3 eq channels, but when you want to get as much isolation between instruments as in the create a psuedo stereo image scenario from mono, that I was relating, boosting the instrument fundamentals and cutting all other instrument ranges out of that channel gives the best isolation, because you have greater dynamic level adjustment with a combination of boosts and cuts. It really depends on what you want to achieve, but especially in the case of  digital eqs compared to analog eqs, there is much less reason to do subtractive mixing and more reasons "for me" to ustilize the full dB range to cut freq's that I don't want and boost the freq's that are wanted on the same eq curve.

Understood, Dave on your following sentence.

[thephatboi]

Thanks for the great ideas guys. I assume to make sure the crossover points are as smooth as possible you would want to use the same type of eqs and same slopes on lowpass, bandpass and high pass tracks? I was just worried that the CO points would either lose or gain a bit of the frequencies that overlap or don't overlap. Would you recommend setting the CO points of the various eqs exactly where the others left off, for ex. if my lowpass is set to 120 then should the lower end of my bandpass be at 120 or would that give me a bit of a boost in that spot? Sorry if that is obvious, just want make sure I get it right. Thanks again, S

[MattGray]

thephatboi wrote on Thu, 18 May 2006 14:27

Thanks for the great ideas guys. I assume to make sure the crossover points are as smooth as possible you would want to use the same type of eqs and same slopes on lowpass, bandpass and high pass tracks? I was just worried that the CO points would either lose or gain a bit of the frequencies that overlap or don't overlap. Would you recommend setting the CO points of the various eqs exactly where the others left off, for ex. if my lowpass is set to 120 then should the lower end of my bandpass be at 120 or would that give me a bit of a boost in that spot? Sorry if that is obvious, just want make sure I get it right. Thanks again, S

I'm coming into this topic late, however I have tried this out last year after watching the demo of the Neve Masterpiece, I figured there had to be a way of setting up a similar architecture in Pro Tools HD. I tried it using the Waves Lin Phase EQ & I set up another aux where I could check that the original mix (phase inverted) nulled perfectly with the new mix I was splitting up. It worked remarkably well & cancelled perfectly. Then I tried to take it one step further & use my Studer A807 hardware inserted through my Cranesong HEDD's converters to run tape saturation just over the low band I'd created. Well this caused all sorts of problems with phase & even with ADC on & hardware latency manually dialled in it wasn't usable so I abondoned the idea & decided to start saving for the real 'Masterpiece' lol.

But that doesn't mean that you can't use plug-ins in ADC mode to achieve a similar result. Using say McDSP's AC2 plugin over the bottom or Cranesong's Phoenix. It also opens other possibilities like compressing & pseudo stereo etc. I guess when you are messing with this sort of stuff you have to be very careful that you don't screw with the integrity of the audio too much & that you are making it sound better & not worse. I think these type of approach could be useful if you needed a lot of control & you needed to virtually re-mix a bad demo that someone has brought in for mastering.

Matt