Adjusting Surround in a DAW for better ballance and EQ

[J. PUPSTER]

A lot of discussions and posts here about some titles; (many for personal taste) having low rear channels or being too bright sounding etc.
I'm curious to know how folks here deal with this by a digital audio workstation (DAW); as one example:
Do you adjust rears for both tracks as a whole, or just certain sections/instruments?

This can be a wide ranging discussion on this whole aspect of enjoying our hobby.

And just to add some extra flavor into the discussion, here's an interesting chart I found here that's giving me ideas about fiddling with various tunes.

 

[Sonik Wiz]

Sure, I'll take the bait....
Not only would it be interesting to learn how people use a DAW for this but if they even bother or need to do this. After all, if the rear chs are a bit loud, just grab the remote & dial them down a bit.

For myself this kind of manipulation is a middle step & depends what I start out with & what I want to end up doing. If I'm doling stereo to surround on the PC I am not a purist I will do whatever it takes to make it sound best on my system. Maybe not yours, just mine. AA 3 has 4 different types of EQ programs & they all fit different needs but I usually make curves on the 20 band. When start a project like this I begin with making the front chs sound good on their own & sometimes that includes the AA3 Pan/Expand effect to increase width. Many of the upmixes I make are from J Pop & that can sound strongly bright with thin bass so that's is easily adjusted. In a David Bowie Greatest Hits it seemed each song was either too bright or boomy so that took a bit of doing. After the front chs are tuned in Ilisten to the rear chs solo & see what needs to be changed there. Generally I will normalize front & back chs independantly to -1dB & then evaluate together & if they don't balance (usually the rears are louder) I will adjust till it all seems to be balanced. If matching to video I will do dts to go with the mpg but if strictly audio it would be multi-ch FLAC.

As you know I am fond of pre-synth & before my SMv2/Chase RLC-1 combo I would do that in 2 ch mode on the PC. Usually stereo decoded QS or DPL II will have a stronger front end & I adjust the pre-synth to the right amount to bring it into balance & when you do that it has a nice wrap around effect.

If I am using a digital source I can just use my Anthem to change bass/treble/balance but an analog 5.1 signal say from Oppo or Surround Master is not processed digitally in the Anthem so I have only balance & of course volume to adjust. Going through the Chase RLC-1 to the SMv2 I can adjust bass/treble their & wrap around effect. Still it's not as precise as what you can do on the PC. A million times easier.

Interesting that we independently posted a similar music/frequency chart:

https://www.quadraphonicquad.com/forums/threads/smell-of-smoke.27687/#post-470149

 

[J. PUPSTER]

Sure, I'll take the bait....
Not only would it be interesting to learn how people use a DAW for this but if they even bother or need to do this. After all, if the rear chs are a bit loud, just grab the remote & dial them down a bit.

For myself this kind of manipulation is a middle step & depends what I start out with & what I want to end up doing. If I'm doling stereo to surround on the PC I am not a purist I will do whatever it takes to make it sound best on my system. Maybe not yours, just mine. AA 3 has 4 different types of EQ programs & they all fit different needs but I usually make curves on the 20 band. When start a project like this I begin with making the front chs sound good on their own & sometimes that includes the AA3 Pan/Expand effect to increase width. Many of the upmixes I make are from J Pop & that can sound strongly bright with thin bass so that's is easily adjusted. In a David Bowie Greatest Hits it seemed each song was either too bright or boomy so that took a bit of doing. After the front chs are tuned in Ilisten to the rear chs solo & see what needs to be changed there. Generally I will normalize front & back chs independantly to -1dB & then evaluate together & if they don't balance (usually the rears are louder) I will adjust till it all seems to be balanced. If matching to video I will do dts to go with the mpg but if strictly audio it would be multi-ch FLAC.

As you know I am fond of pre-synth & before my SMv2/Chase RLC-1 combo I would do that in 2 ch mode on the PC. Usually stereo decoded QS or DPL II will have a stronger front end & I adjust the pre-synth to the right amount to bring it into balance & when you do that it has a nice wrap around effect.

If I am using a digital source I can just use my Anthem to change bass/treble/balance but an analog 5.1 signal say from Oppo or Surround Master is not processed digitally in the Anthem so I have only balance & of course volume to adjust. Going through the Chase RLC-1 to the SMv2 I can adjust bass/treble their & wrap around effect. Still it's not as precise as what you can do on the PC. A million times easier.

Interesting that we independently posted a similar music/frequency chart:

https://www.quadraphonicquad.com/forums/threads/smell-of-smoke.27687/#post-470149

Yeah, hadn't even seen that post of yours yet, funny!

One thing that's bugging me (and I don't have anything in particular in mind -just yet) is say you need some things raised in the rears, but along with it (if you do it in whole) you're potentially screwing with the mix even more, as say there's a nice guitar lick going on in the rear left but it's also mixed along with some accompanying synthesizer or drum parts. Then you're also raising those along with the guitar (potentially) ruining your main goal of just bringing out that guitar.

Also as far as EQ and volume adjustments on the fly; I use an analog direct hookup to my AVR which doesn't let me do those kinds of changes, so I figure, just work it out on the DAW and re-save to whatever i.e. 24/96 flac and play that back when wanted.

 

[Sonik Wiz]

Yeah, hadn't even seen that post of yours yet, funny!

One thing that's bugging me (and I don't have anything in particular in mind -just yet) is say you need some things raised in the rears, but along with it (if you do it in whole) you're potentially screwing with the mix even more, as say there's a nice guitar lick going on in the rear left but it's also mixed along with some accompanying synthesizer or drum parts. Then you're also raising those along with the guitar (potentially) ruining your main goal of just bringing out that guitar.

Also as far as EQ and volume adjustments on the fly; I use an analog direct hookup to my AVR which doesn't let me do those kinds of changes, so I figure, just work it out on the DAW and re-save to whatever i.e. 24/96 flac and play that back when wanted.

Oh you mean precision surgery....
In the example you gave I can think of two different ways to do that. In AA3 edit view zoom in and select a good sample of that guitar part. Go to Window>Frequency Analysis & it will show the frequency spectrum of that guitar selection. Then go back to one of the graphic EQ plug in's & boost by desired amount. Yes it affect the other sounds in that freq range so you can't go crazy & whether you apply it just individual guitar riffs or over all is up to you.

The other way is to really zoom in on the guitar selection & use tools>amplitude & compression > envelope.This lets you select a start/stop & gain in the middle of that selection. If you simply selct on the wafeform that guitar bit & boost the level it will be abrupt & noticeable. Using the envelope pug in lets your start at unity gain, boost the guitar, and return to unity gain which makes it much smoother. Here you can see how I boosted the level of the left ch only using the tool:

I had a similar challenge & by similar I mean opposite with a Chisato Moritaka up mix. It was with a music video, a song called The Blue Blues.The set upwas a typical smoky bar lounge with Moritaka singing backed up by a typical jazz quartet band.When finished up mixing the high hat cymbal was searingly bright & loud out of the right front. So on each & every cymbal hit I used the envelope tool to redice the level by 10dB. It worked perfect.

 

[PmikeK]

I adjust loads of albums. I recently reworked Oasis whats the story. MANY albums that some perceive as being poor mixes can be made into pretty decent surround albums.

 

[jimfisheye]

I treat it like any mastering job.
I listen for what I think the intention was in the mix.
You can usually tell what someone was going for. The tinny, shrill sound of a treble brightened master is usually obviously not what a mix was going for. You can tell it was done after the fact to make the mix louder. Bright sounds loud and it goes hand in hand with the compressor slamming that lets you boost the volume.
Just one example.

Maybe a more obvious example might be recording a gun shot on your phone. When you hear that played back, you aren't wondering if the distortion artifacts from the little phone microphone diaphragm being slammed against its mount are part of the intended sound. You know what a gun shot sounds like more or less in real life and you get what's going on. If you were to do restoration or mastering on such a recording, you'd approach it with that perspective.

So, you hear stuff like that and you find familiar landmarks when you master.

If I hear what I believe is an imbalance between channels in a recording (surround or even stereo) that I believe was an after the fact handling mistake, I'll address that. Often there are clues. A stereo recording skewed to one side is usually more obvious but you find similar clues in surround.

The treble eq brightening thing...
I'll strictly use subtractive eq. (Which should go without saying... We don't want added distortion from boosting an eq!) I'll normally use linear phase eq and parallel tracks ("New York" style). Low pass a 2nd parallel copy of the program with a linear phase eq to remove the entire region hit by the errant eq boost. Mix this track in with the original to bring up the rest of the program to where they boosted the top end to. Mind your gain staging as you go. Mind your phase between instances and make sure you vet your linear phase eq's! This gives me no added damage over their errant eq boost while putting it back into what would have been the original mix balance. (As well as you can guess after the fact based on your experience.)

 

[jimfisheye]

The above would be trying to restore a mix to its original fidelity and presentation.

Then there's altering and upmixing...
The lids are off and the worms are on the march now!

A surround mix of course gives you more opportunity to alter things with the multi-channel source.

The other way around...
Separating already mixed content to separate channels can be done to different degrees. Success ranges from fully unsuccessful to magical "How the heck did you pull that off?!" and everything in between.
Sound elements in different frequency ranges can be separated by eq. Often easily.
In more recent times, spectral editing tools can let you separate sound elements that overlap in frequency domain. A spectral editor puts time on the x axis and frequency on the y axis. iZotopeRX is the tool for this.
Acts of what seems like genuine magic are actually possible with this tech now but still a labor of love to pull off seamlessly.

There are still the good old fashioned phase tricks. You can separate mono content from a stereo pair with a stereo L/R to stereo MS flip.

A multiband compressor can still pretty much let you remix an already mixed mix. Not like going to the actual multitrack. Of course there are limitations! But you can alter a mix pretty significantly by dividing it into separate frequency bands and having a compressor on each band to play with. Take a 5 channel surround mix with already more discrete separation between channels and now you can really alter things!

 

[J. PUPSTER]

I'm looking at the iZoptope RX-7 software, but at the pricing, I feel I have way too much more "basic" knowledge to obtain before tackling that package. But then sometimes you just have to make a leap and dig in and learn from trial, error and lots of mistakes.

Here's a project I want to take on sometime. It's the recent offering from Analogue Productions SACD (yes only stereo- but still) of the title Bluesey Burrell.
https://www.discogs.com/Kenny-Burrell-With-Coleman-Hawkins-Bluesy-Burrell/release/13999200
I've mentioned this before here a couple times, that I feel Coleman Hawkins' sax is too out front (prominent) to the rest of the band. I figure since it's headliner is Kenny Burrell that his guitar should be out front. But, perhaps that was considered the correct sound for that period (as it was Kevin Gray that mastered it.) So at some point when I feel I'm ready, I just want that sax to come down in the mix so it doesn't drown out the rest of the band. I just wish someone else here that has that title would chime in and give me their take on that, in case I'm just all wrong or sensitive to that range of horn or something. I'm sure you more experience guys will be happy to guide me once I take that on.

 

[jimfisheye]

I'm looking at the iZoptope RX-7 software, but at the pricing...

The kind of surgery you do with spectral editing is on the extreme end of things!

More meat 'n potatoes kind of stuff is handled by a DAW app. (That's Digital Audio Workstation FYI.)
Reaper is a really really good choice for that! Yes it's the flagship app and it's complex... but you only need to learn and use what you need. Their hobbyist license option (with the exact same unrestricted app) and their lenient 'still evaluating' button if you haven't paid yet are a gift you don't often see.

Any and all multitrack manipulation, editing, mixing, and mastering is happiness and light in Reaper. You can even run live sound with it (set for extreme low latency for real time audio) while simultaneously recording multitrack. Reaper has a crude early stage spectral editing feature actually but it's a toy compared to iZotopeRX. Spectral editing is going off the deep end of things though. I'm told their Windows version is the most high functioning Windows audio app anyone has ever seen too. (In case that applies.)

 

[Sonik Wiz]

I'm looking at the iZoptope RX-7 software, but at the pricing, I feel I have way too much more "basic" knowledge to obtain before tackling that package. But then sometimes you just have to make a leap and dig in and learn from trial, error and lots of mistakes.

Here's a project I want to take on sometime. It's the recent offering from Analogue Productions SACD (yes only stereo- but still) of the title Bluesey Burrell.
https://www.discogs.com/Kenny-Burrell-With-Coleman-Hawkins-Bluesy-Burrell/release/13999200
I've mentioned this before here a couple times, that I feel Coleman Hawkins' sax is too out front (prominent) to the rest of the band. I figure since it's headliner is Kenny Burrell that his guitar should be out front. But, perhaps that was considered the correct sound for that period (as it was Kevin Gray that mastered it.) So at some point when I feel I'm ready, I just want that sax to come down in the mix so it doesn't drown out the rest of the band. I just wish someone else here that has that title would chime in and give me their take on that, in case I'm just all wrong or sensitive to that range of horn or something. I'm sure you more experience guys will be happy to guide me once I take that on.

Are you considering working on this in 2 ch stero & let the SM do the rest? Or doing a PC upmix to multi?

 

[J. PUPSTER]

Are you considering working on this in 2 ch stero & let the SM do the rest? Or doing a PC upmix to multi?

Just a title I think about often (stereo only) to do some adjusting to the sound; at this point no plans on up-mixing this one.

 

[jimfisheye]

You know, in the long and short of things...

A recording like this:
Original mix -> errant volume war mastering damage -> restoration mastering

... can very well come out sounding closer to what the actual original mix sounded like than trying to fix something with generation loss from heavy analog generation damage and/or analog encoder/decoder generation loss.

Depends on the level of destruction. There are worst case examples out there where an average vinyl pressing with a couple rounds of tape generation before it even hit the vinyl and then a quality issue with the pressing itself is closer to what the master mix sounded like than the CD edition with volume war mastering. Your time is better spent on Ebay than behind the mixing board!

 

[Sonik Wiz]

I'm looking at the iZoptope RX-7 software, but at the pricing, I feel I have way too much more "basic" knowledge to obtain before tackling that package. But then sometimes you just have to make a leap and dig in and learn from trial, error and lots of mistakes.

Here's a project I want to take on sometime. It's the recent offering from Analogue Productions SACD (yes only stereo- but still) of the title Bluesey Burrell.
https://www.discogs.com/Kenny-Burrell-With-Coleman-Hawkins-Bluesy-Burrell/release/13999200
I've mentioned this before here a couple times, that I feel Coleman Hawkins' sax is too out front (prominent) to the rest of the band. I figure since it's headliner is Kenny Burrell that his guitar should be out front. But, perhaps that was considered the correct sound for that period (as it was Kevin Gray that mastered it.) So at some point when I feel I'm ready, I just want that sax to come down in the mix so it doesn't drown out the rest of the band. I just wish someone else here that has that title would chime in and give me their take on that, in case I'm just all wrong or sensitive to that range of horn or something. I'm sure you more experience guys will be happy to guide me once I take that on.

I was totally unfamiliar with this album. I just sampled all of the songs on HD Tracks. FWIW the guitar was very much up front with all else subdued. Check it out & see how it compares to what you have:

https://www.hdtracks.com/bluesy-burrell-rudy-van-gelder-remaster

 

[J. PUPSTER]

I was totally unfamiliar with this album. I just sampled all of the songs on HD Tracks. FWIW the guitar was very much up front with all else subdued. Check it out & see how it compares to what you have:

https://www.hdtracks.com/bluesy-burrell-rudy-van-gelder-remaster

Well, two things here- One, those samples are pretty low Res. actually at 44.1/24 from 2014? and I'm listening to the new 2019 SACD with a new mastering. And Two, most of those samples have the front end of the songs, which start off fine until Coleman's Sax kicks in and takes over the show later in the songs.

 

[J. PUPSTER]

Here’s something I’ve noticed before on resulting artifacts with up- mixes; is something called Sibilance ...
https://www.quadraphonicquad.com/forums/threads/sm2-vs-specweb-vs-penteo-16-pro.27384/post-457442
And there actually is an app (plug-in) for that available in RX-7 Standard & Advanced.

https://www.izotope.com/en/products/rx/features/de-ess.html

 

[J. PUPSTER]

Here’s something I’ve noticed before on resulting artifacts with up- mixes; is something called Sibilance ...
https://www.quadraphonicquad.com/forums/threads/sm2-vs-specweb-vs-penteo-16-pro.27384/post-457442
And there actually is an app (plug-in) for that available in RX-7 Standard & Advanced.

https://www.izotope.com/en/products/rx/features/de-ess.html

Sibilance is even represented in the chart in my first post above. For vocals (where I notice it) it shows up the most at 6-10khz. Instead of messing with multiple adjustments in AA3 you can adjust with a couple sliders (including Spectral tilt) and Preview and Compare buttons in De-ess much easier.

 

[Sonik Wiz]

Well, two things here- One, those samples are pretty low Res. actually at 44.1/24 from 2014? and I'm listening to the new 2019 SACD with a new mastering. And Two, most of those samples have the front end of the songs, which start off fine until Coleman's Sax kicks in and takes over the show later in the songs.

Good points indeed. I don't think I will spend ~$20 just to hear what happens after the sample......
But I would like to hear your results when you get something your happy with.

 

[jimfisheye]

My comment from the peanut gallery would be that if you are generating sibilant artifacts (probably from comb filtering) by doing something, going after the root cause will get you further than going after the symptom. Deesser to manipulate the artifacts? Now you've got breathing artifacts! Make whatever is causing the artifact in the first place stop instead.

Ultimately though, A/B against the original (carefully level matched if you are introducing gain changes) often. If whatever you did both improves on what the complaint was and does no damage to any other element, that's really the bottom line.