Advanced Mastering - How can I make the most of the mixdown?

Have an urge to learn, or a calling to teach? Want to share some useful Youtube videos? Do it here!
Post Reply
RobC
Posts: 1833
Joined: 10 Mar 2018

27 Jun 2022

When mastering a mixdown (no stems, no going back possible!), I would do the following:

- Preparing the file, checking the volume, whether everything is there, and doing some trimming at the start and at the end if needed.
- Listening for problematic sounds and imbalances, and what needs work regarding dynamics, stereo width and depth, and mono compatibility.

Getting to work:

- Isolating the problematic sounds, if possible, and then enhancing them (for example bringing out a weak kick; or stereo-positioning a hat ~ and again I say: if possible)
The above is a very complex task, and uses mostly crossover filters, EQ, compressors (including several types and methods of compression/expansion, sidechaining, etc.), gate, saturation (with emphasis EQ, then distortion, then de-emphasis EQ), etc. including any combination of these (such as dynamic EQ for example)
- Equalizing the whole mix by gently cutting any frequency buildups, or boosting notches
- Compressing (balancing and enhancing dynamics ~ again, using any type of compression, be it upwards/downwards -compression/expansion; multi-band; parallel)
- Maybe Mid-Side processing (although I'm not a huge fan of that) ~ again, a complex task
- Centering frequencies below 80 Hz
- 20 Hz low cut
- Limiting peaks, setting -1 dB headroom
- Rendering
- Exporting the final format, usually dithered, 16 bit 44.100 Hz wav

Most importantly, always comparing with the original/previous processing, if there's any improvement (keeping in mind that in some cases, only parallel processing will sound better).
Before I forget, maybe checking reference material now and then to ensure it's heading in the right direction.

I know it's very much in a nutshell, but I'd like to ask regarding the above to-do list:

Is there anything I left out?
Is there anything I should rather not do?
Any optional tasks?

User avatar
moofi
Posts: 1024
Joined: 19 Jan 2015
Location: hear

10 Jul 2022

I wouldn´t necessarily mono everything below 80H even though it migth be adequat in many cases.

User avatar
QVprod
Moderator
Posts: 3488
Joined: 15 Jan 2015
Contact:

10 Jul 2022

I won't call myself a "mastering engineer" per se but with the little experience I do have I'd say this. What you have is a nice list. But literally all of it is conditional based on what the mix sounds like. Hard to say whether you should do more or less.

However, if you're mastering your own mixes and you find yourself doing most of the list on just one song. It's probably best to just go back to the mix and fix it. If it's a client mix, then sure do what you can to make it sound "better".

User avatar
crimsonwarlock
Posts: 2255
Joined: 06 Nov 2021
Location: Close to the Edge

11 Jul 2022

I would say that the bulk of what you are mentioning is 'mixing' and not 'mastering'. The mastering stage is not about fixing problems, it is more about giving that final touch-up to make your track sound 'commercial', 'radio-ready' or whatever other denomination suits your taste.

My own mastering stage is mainly about gluing the track. This consists basically of a compressor, an EQ (preferably a Baxandall-type), and a limiting stage.

The choice of compressor changes for each track, whatever gives the best result. My favorites so far are McDSP's C670 and Moo Tube, and Reason's own Master Bus Comp.

My mastering EQ of choice is currently Sugar-Q, but EVE AT1 and McDSP's E670 are very nice as well. However, any 'simple' EQ with just a few bands would work pretty well here.

My limiting stage consists of a clipper, followed by a limiter. I use the Omega Mastering clipper and Kratos 2 Maximizer for this.

I do experiment with additional FX in my mastering chain, like mid-side stuff. But in every case it has to do broad strokes, no surgical stuff allowed, as that should be done in mixing.
-------
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

11 Jul 2022

I agree, you describe mixing.
Mastering should NOT change the mix IMO, unless requested. I have a friend just starting out with mastering, so for some practice I gave him the tracks I just mixed and also send to Bob Olhsson for mastering. My friends master changed balances, kick came forward, other things went back into the mix. It was nice, it didn't suck, but it changed the perceived balances for the worse IMO. He was trying to fix one thing and ends up 'breaking' another - it IS a delicate balance, mastering.

Then I get the mastered files from Bob and they just sound fantastic - like 'better' versions of MY mixes. More cohesive from track to track too. But still 100% my mixes.

For me, THAT is mastering (from a certified 'master'!). And whenever I ask Bob what he did, it's different every time (we've been working together over 20 years now). There is no 'formula'.
If there WAS a formula, it would be this: You listen, you decide what needs to be addressed, and you sometimes try a few different approaches to achieve the desired results.

The point being, there is almost always more than one way to address any issue you hear. Maybe you need to address the issue with EQ, or maybe a filter, or maybe compression, or even saturation.

It appears counter intuitive to me to have a list of solutions before you know the problems, let alone having only ONE solution to any particular 'problem'.
Selig Audio, LLC

RobC
Posts: 1833
Joined: 10 Mar 2018

14 Jul 2022

moofi wrote:
10 Jul 2022
I wouldn´t necessarily mono everything below 80H even though it migth be adequat in many cases.
We don't really localize sounds below that.
With speakers, as sub bass frequencies travel in the air, when they meet, they completely cancel out, when out of phase.

Most importantly, I tried binaurally making out of phase sub, to see how it sounds like. It seemed to make it weak, even if it was kind of separated to the back. I think it could be partly how our brain works. As it adapts, it may or may not center sounds. - This is just my thought though!

That said, I'd probably only high pass filter a side channel with a linear phase filter - personally, to avoid funny things an analog style filter may do. It kind of is a parallel type of processing, after all, if I'm not mistaken.
QVprod wrote:
10 Jul 2022
I won't call myself a "mastering engineer" per se but with the little experience I do have I'd say this. What you have is a nice list. But literally all of it is conditional based on what the mix sounds like. Hard to say whether you should do more or less.

However, if you're mastering your own mixes and you find yourself doing most of the list on just one song. It's probably best to just go back to the mix and fix it. If it's a client mix, then sure do what you can to make it sound "better".
Whomever can deliver a professional master, whether done on a potato, or in a million dollar studio, is a mastering engineer. (- This is funny, coming from me, when I set myself the goal: 'Once I hold a vinyl in my hands where I did everything, up to the point that the cutting engineer cuts it into vinyl, THEN I will call myself a professional' x D It will be my personal "certificate".)

I didn't mean to put the list in a way, that we have to add all of these effects in a huge processing chain, but rather 'when/if any of them is needed'. For example, if there's not much going on below 80 Hz in the side channel, then no filtering needed.

Sure, I have my mixes, but I wanted to test the waters, to simulate what happens in case of someone else's song, where I have no access to the mix, and that's why I said, there's no going back. I better experiment on my own music, first than doing something awkward with someone else's. x D
crimsonwarlock wrote:
11 Jul 2022
I would say that the bulk of what you are mentioning is 'mixing' and not 'mastering'. The mastering stage is not about fixing problems, it is more about giving that final touch-up to make your track sound 'commercial', 'radio-ready' or whatever other denomination suits your taste.

My own mastering stage is mainly about gluing the track. This consists basically of a compressor, an EQ (preferably a Baxandall-type), and a limiting stage.

The choice of compressor changes for each track, whatever gives the best result. My favorites so far are McDSP's C670 and Moo Tube, and Reason's own Master Bus Comp.

My mastering EQ of choice is currently Sugar-Q, but EVE AT1 and McDSP's E670 are very nice as well. However, any 'simple' EQ with just a few bands would work pretty well here.

My limiting stage consists of a clipper, followed by a limiter. I use the Omega Mastering clipper and Kratos 2 Maximizer for this.

I do experiment with additional FX in my mastering chain, like mid-side stuff. But in every case it has to do broad strokes, no surgical stuff allowed, as that should be done in mixing.
Yes, I understand that now. But it depends on the situation. Somebody still might say: "Is there anything we can do to pan the hats? I have no access to the mix." So I guess it's sort of trying to mix/restore, when a single song file is all we have.
I'll keep in mind that mastering is the final polishing, while the more extreme things I mentioned are mix(engineering).

Speaking of compressors, I'm thinking about developing a dynamics processor RE, too. When it comes to coding, it's pretty fascinating to see what's going on under the hood.

Ah, Equalization on a master is still an enemy of mine. I'm getting closer to the solution, but some things are still a bit blurry. I mean, I do have dead-accurate hearing, so once I understand it all, it will be easy, but until then, I will struggle a bit.

I'm not a fan of clipping, unless it's done using emphasis and de-emphasis EQ for example. Some mild limiting can do interesting things, as I've noticed, but then the music may get accidentally louder than those -16 and -12 LU values. But no problem, that might mean extra headroom. In which case we might play around with transients, or use an expander for example.
selig wrote:
11 Jul 2022
I agree, you describe mixing.
Mastering should NOT change the mix IMO, unless requested. I have a friend just starting out with mastering, so for some practice I gave him the tracks I just mixed and also send to Bob Olhsson for mastering. My friends master changed balances, kick came forward, other things went back into the mix. It was nice, it didn't suck, but it changed the perceived balances for the worse IMO. He was trying to fix one thing and ends up 'breaking' another - it IS a delicate balance, mastering.

Then I get the mastered files from Bob and they just sound fantastic - like 'better' versions of MY mixes. More cohesive from track to track too. But still 100% my mixes.

For me, THAT is mastering (from a certified 'master'!). And whenever I ask Bob what he did, it's different every time (we've been working together over 20 years now). There is no 'formula'.
If there WAS a formula, it would be this: You listen, you decide what needs to be addressed, and you sometimes try a few different approaches to achieve the desired results.

The point being, there is almost always more than one way to address any issue you hear. Maybe you need to address the issue with EQ, or maybe a filter, or maybe compression, or even saturation.

It appears counter intuitive to me to have a list of solutions before you know the problems, let alone having only ONE solution to any particular 'problem'.
I will definitely ask how my friend feels about his mix. If he only wants that mastering clean up, or some mix corrections.

Honestly, originally when I heard a mastering result, my first reaction was: "That's it? D :" ~ and recently I started watching an idiotic youtube video (the usual clickbait, with a dumb face - but the topic seemed interesting), where the raw mix was pretty dead sounding, but then the master sounded very full. Of course, the video was full of BS, like "Once your music sounds like this, you can get on any major record label", multiple minutes of rambling during the intro, which said basically nothing, just tried hyping, and when I was sick of it all, came the "subscribe, hit the cowbell, etc" - I closed it there, lol, and I bet, statistics showed that channel, that's where other people stop watching, too. Unbelievable what trash they put out these days, which reeks from SEO. - But the reason I mentioned this was, how these "scammers" can fool people (including me, up until recently) into believing that mastering should be about all kinds of added voodoo. I suspect, that video probably deliberately added mistakes to a mix, then undid it. I'll never know. I won't watch it.

I completely agree on only changing something, when it's needed.

As for how I'd approach solving any issue ~ I'd go for whatever is the least destructive, and most transparent.

User avatar
moalla
Posts: 541
Joined: 20 Oct 2017
Location: DDR WEST

14 Jul 2022

So i‘m totally unsettled how to make the mix, bass/kick 12db lower than the rest, other drums 6-8db lower, deep voices mono maybe its needed some instruments more separeted to left or right channel, also some filtering but how could it be, that mixes from the past i made with mixer 14:2 sounds better, especial i don‘t know really whats the right factor of ssl master compression or how master compression is working, maybe also cause my eve sc205 and dt880pro hp sound is to much up normal stereo systems, in things of equing i think I know and hear a liitle bit more. Selig you wrote about this „socail media shit, first flat than tune up, now i‘m totally unsettled.....
https://soundcloud.com/user-594407128
Reason12.5, Yamaha EG112, Ibanez PF10, RhythmWolf, Miniak, Ipad+SparkLE
SE2200t :arrow: VAS micpre MOTO:better repair-mod well made stuff than buy the next crap

User avatar
moalla
Posts: 541
Joined: 20 Oct 2017
Location: DDR WEST

14 Jul 2022

So i‘m totally unsettled how to make the mix, bass/kick 12db lower than the rest, other drums 6-8db lower, deep voices mono maybe its needed some instruments more separeted to left or right channel, check phase problems and fix with a dynamic eq like hornet dyn eq, also some filtering saturation but how could it be, that mixes from the past i made with mixer 14:2 sounds better, especial i don‘t know really whats the right factor of ssl master compression or how master compression is working,

Maybe also cause my eve sc205 and dt880pro hp sound is to much up normal stereo systems, in things of equing i think I know and hear a little bit more and i found in our house two old telefunken l61 speakers do you think they are usable for the downmix of electroacustic music?

Selig you wrote about this social media shit, first flat than tune up, now i‘m totally unsettled.....
https://soundcloud.com/user-594407128
Reason12.5, Yamaha EG112, Ibanez PF10, RhythmWolf, Miniak, Ipad+SparkLE
SE2200t :arrow: VAS micpre MOTO:better repair-mod well made stuff than buy the next crap

User avatar
moofi
Posts: 1024
Joined: 19 Jan 2015
Location: hear

04 Aug 2022

They only cancel each other out completely if they are out of sync by 180°. Just saying, not saying already lesser phasecancellation cannot be an issue.

And while, like said, in many cases this might be the way to go I can hear width below 80Hz, at least via headphones.

RobC wrote:
14 Jul 2022
moofi wrote:
10 Jul 2022
I wouldn´t necessarily mono everything below 80H even though it migth be adequat in many cases.
We don't really localize sounds below that.
With speakers, as sub bass frequencies travel in the air, when they meet, they completely cancel out, when out of phase.

[...]


User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

04 Aug 2022

RobC wrote:
14 Jul 2022
We don't really localize sounds below that.
With speakers, as sub bass frequencies travel in the air, when they meet, they completely cancel out, when out of phase.

Most importantly, I tried binaurally making out of phase sub, to see how it sounds like. It seemed to make it weak, even if it was kind of separated to the back. I think it could be partly how our brain works. As it adapts, it may or may not center sounds. - This is just my thought though!

That said, I'd probably only high pass filter a side channel with a linear phase filter - personally, to avoid funny things an analog style filter may do. It kind of is a parallel type of processing, after all, if I'm not mistaken.
1st paragraph - ANY two frequencies when out of phase, will cancel, not just sub frequencies and even if the source is a single sub woofer playing by itself (because reflections).

2nd Paragraph - linear phase filter ALSO do funny things, even more so in some cases of pre-ringing. Check out this video by the amazing Paul Frindle from SSL/Sony Oxford fame (and also the guy who first said to suggest leaving at least 10 dB headroom on all channels in digital mixers, thanks Paul!).


A quote from Paul in the comments for the video:
[i"]But the crucial thing to remember with linear phase filters is that when set for such low frequencies they will have a significant time preamble due to their symmetrical impulse responses. This will mean that some parts of the sound spectrum will happen relatively in advance of the rest - and therefore smear and muddy the sound of percussive sounds like drums significantly. I.e. they will not sound so tight and well defined.
So the bottom line is; if you MUST roll off the bottom end you are still sonically better off using a conventional filter or EQ (and correcting for possible clipping), because at least this will avoid smearing and softening the sound. I hope this helps?"[/i]

And absolutely NO, linear phase filters are not in any way "parallel" (parallel EQ is another thing entirely,, happy to talk about that and provide examples if you're interested).
Linear phase and minimal phase filters both use the same basic filter design (shifting phase to alter the spectrum) as analog but linear filters are compensated for phase by delaying ALL frequencies to align (super simplification), which is why the overall latency is so high compared to minimum phase filters.
Selig Audio, LLC

RobC
Posts: 1833
Joined: 10 Mar 2018

07 Aug 2022

moofi wrote:
04 Aug 2022
They only cancel each other out completely if they are out of sync by 180°. Just saying, not saying already lesser phasecancellation cannot be an issue.

And while, like said, in many cases this might be the way to go I can hear width below 80Hz, at least via headphones.
I rather meant, if for example you pan a kick slightly, then there will be content in the side channel (Mid/Side processing). That's how I think - that the Side Channel content is always out of phase. The difference between the two L & R channels.

Try filtering the mid channel below 80 Hz, and move it to the side channel. With binaural listening, you will get the illusion that it's coming from behind you, but it sounds worse.

They say, deeper frequencies pull the stereo image to the center, when panned for example. That's kind of why I dropped my idea of hard panning the bass instrument to the left, and the same with the kick to the right ~ while centering both below 80 Hz. There were other rules to it, and it sounded interesting ~ but mostly only when the track got busy, and every instrument and drum was playing.
selig wrote:
04 Aug 2022
1st paragraph - ANY two frequencies when out of phase, will cancel, not just sub frequencies and even if the source is a single sub woofer playing by itself (because reflections).

2nd Paragraph - linear phase filter ALSO do funny things, even more so in some cases of pre-ringing. Check out this video by the amazing Paul Frindle from SSL/Sony Oxford fame (and also the guy who first said to suggest leaving at least 10 dB headroom on all channels in digital mixers, thanks Paul!).


A quote from Paul in the comments for the video:
[i"]But the crucial thing to remember with linear phase filters is that when set for such low frequencies they will have a significant time preamble due to their symmetrical impulse responses. This will mean that some parts of the sound spectrum will happen relatively in advance of the rest - and therefore smear and muddy the sound of percussive sounds like drums significantly. I.e. they will not sound so tight and well defined.
So the bottom line is; if you MUST roll off the bottom end you are still sonically better off using a conventional filter or EQ (and correcting for possible clipping), because at least this will avoid smearing and softening the sound. I hope this helps?"[/i]

And absolutely NO, linear phase filters are not in any way "parallel" (parallel EQ is another thing entirely,, happy to talk about that and provide examples if you're interested).
Linear phase and minimal phase filters both use the same basic filter design (shifting phase to alter the spectrum) as analog but linear filters are compensated for phase by delaying ALL frequencies to align (super simplification), which is why the overall latency is so high compared to minimum phase filters.
I'll be the first to say that I can still be wrong, but maybe I wasn't clear with a lot of things, so let's take a look at how I understood and meant things:

The first part, I meant in case of out of phase stereo. In the past, I experimented with creating music using simplified stereo effects. Inverting the polarity of one channel was one such effect. I used it on bass. The result with "Hi-Fi" speakers was a thin sound, where the sub woofers were working hard, yet the sub bass disappeared in the air. If I turned my head 90 degrees, the sound completely disappeared.
I do understand how M/S processing and (mono) out of phase works ~ but maybe I wasn't clear then that I meant out of phase stereo / side channel content.
On stereo speakers, below 80 Hz it's pretty much impossible to reproduce such content in an "audible" way, so I don't see the point of leaving it in.
With binaural listening, it might be audible, but it actually makes bass sound weaker than when centered.

As for Linear Phase EQ/Filter ~ I had topics on that on here, too. There was a time that I swore by them, but I quickly realized that the smearing it does, is pretty bad.
One thing I don't get with analog style filters ~ especially the ones that get close to brick wall style filtering, that the ones I used, seemed to do an extreme boost at the set frequency. For a long time, I believed, it's an artifact of IIR filters. But it could be that those, which I used, were designed to add resonance. It was useful for designing a kick, or a snare, but not always desired by me. I prefer a clean analog style filter.

I didn't mean that a Linear Phase filter would do any type of parallel processing by default. I meant that in Mid/Side processing, I high pass filter the Side channel, then merge with the Mid channel again. As far as I knew, Mid/Side processing is kind of parallel.
However, if we high pass filter the side channel with an analog style EQ, it can break especially the panning of the mix (if I remember correctly from a Dan Worrall video). If so, then a Linear Phase filter might be a better solution, and the smearing could be even beneficial in this case. Because if the transients are smeared, they kind of get pushed to the center, and they can sound more punchy. I noticed this when I hard compressed the side channel, when experimenting with preparing a mix for vinyl.

I'm always curious ~ I only heard of creating a shelf EQ, with the help of parallel filtering. And multi-band processing of course - in that case though, I think even the Dan Worrall video said that Linear Phase crossovers/splits could be superior.

I wouldn't use a Linear Phase filter live, though. I'd pre-render a multi-band split.

That said, the smearing can somewhat be tamed when designing one shot sounds. Simply by rendering (a multi-band split for example) and cutting the pre-ringing.

At least this is how I thought things work. And again, I could be wrong.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

07 Aug 2022

RobC wrote:
07 Aug 2022
I rather meant, if for example you pan a kick slightly, then there will be content in the side channel (Mid/Side processing). That's how I think - that the Side Channel content is always out of phase. The difference between the two L & R channels.

Try filtering the mid channel below 80 Hz, and move it to the side channel. With binaural listening, you will get the illusion that it's coming from behind you, but it sounds worse.

They say, deeper frequencies pull the stereo image to the center, when panned for example. That's kind of why I dropped my idea of hard panning the bass instrument to the left, and the same with the kick to the right ~ while centering both below 80 Hz. There were other rules to it, and it sounded interesting ~ but mostly only when the track got busy, and every instrument and drum was playing.
The first part, I meant in case of out of phase stereo. In the past, I experimented with creating music using simplified stereo effects. Inverting the polarity of one channel was one such effect. I used it on bass. The result with "Hi-Fi" speakers was a thin sound, where the sub woofers were working hard, yet the sub bass disappeared in the air. If I turned my head 90 degrees, the sound completely disappeared.
I do understand how M/S processing and (mono) out of phase works ~ but maybe I wasn't clear then that I meant out of phase stereo / side channel content.
On stereo speakers, below 80 Hz it's pretty much impossible to reproduce such content in an "audible" way, so I don't see the point of leaving it in.
With binaural listening, it might be audible, but it actually makes bass sound weaker than when centered.

I didn't mean that a Linear Phase filter would do any type of parallel processing by default. I meant that in Mid/Side processing, I high pass filter the Side channel, then merge with the Mid channel again. As far as I knew, Mid/Side processing is kind of parallel.

However, if we high pass filter the side channel with an analog style EQ, it can break especially the panning of the mix (if I remember correctly from a Dan Worrall video). If so, then a Linear Phase filter might be a better solution, and the smearing could be even beneficial in this case. Because if the transients are smeared, they kind of get pushed to the center, and they can sound more punchy. I noticed this when I hard compressed the side channel, when experimenting with preparing a mix for vinyl.

I'm always curious ~ I only heard of creating a shelf EQ, with the help of parallel filtering. And multi-band processing of course - in that case though, I think even the Dan Worrall video said that Linear Phase crossovers/splits could be superior.
You say “side channel is always out of phase”, but fail to account for the conversion from M/S to stereo. Do you not always listen in stereo? If you understand M/S you understand the phase inversion is totally necessary for the process to 100% recreate the original stereo image. There is no “out of phase” in the side channels that is not 100% accounted for once converted back to stereo. Sure, if you listen to the side on it’s own, it’s 100% polarity inverted. But no one hears this signal on it’s own…

You mention bass is non directional, but you still seem to believe it must be mono below 80 Hz - why does it even matter if it’s non-directional!?! Adding another unnecessary filter in the audio path that is not going to change what we cannot hear doesn’t make sense to me. Hard panning bass to one side is indeed extreme, but stereo bass frequencies are common except when mastering for vinyl. The “rule” of making everything under 80 Hz ‘mono’ was strictly for vinyl mastering, at least originally.

Mid side is NOT parallel processing, any more than adding an EQ to only one channel of a stereo image is “parallel” processing.

Adding a HP filter to the sides breaks panning because changing ANYTYHING about the mid side signals individual channels will affect panning - not because the filter is minimal phase, but because of how mid/side works which is often misunderstood.

And yes, all EQ is built using a multiple audio paths that technically run “in parallel” - but it’s not “parallel processing” any more than parallel EQ is “parallel processing”. Just because the word “parallel” is used doesn’t mean it fits the definition of parallel processing, which is IMO always used to describe something where the user has control over the balance of the dry vs wet signal. Shelf EQ does not allow this, parallel EQ does not allow this, mid/side processing does not allow this. Hope I’m being clear (and accurate), this can be a complex subject to say the least! :)
Selig Audio, LLC

RobC
Posts: 1833
Joined: 10 Mar 2018

07 Aug 2022

selig wrote:
07 Aug 2022
You say “side channel is always out of phase”, but fail to account for the conversion from M/S to stereo. Do you not always listen in stereo? If you understand M/S you understand the phase inversion is totally necessary for the process to 100% recreate the original stereo image. There is no “out of phase” in the side channels that is not 100% accounted for once converted back to stereo. Sure, if you listen to the side on it’s own, it’s 100% polarity inverted. But no one hears this signal on it’s own…

You mention bass is non directional, but you still seem to believe it must be mono below 80 Hz - why does it even matter if it’s non-directional!?! Adding another unnecessary filter in the audio path that is not going to change what we cannot hear doesn’t make sense to me. Hard panning bass to one side is indeed extreme, but stereo bass frequencies are common except when mastering for vinyl. The “rule” of making everything under 80 Hz ‘mono’ was strictly for vinyl mastering, at least originally.

Mid side is NOT parallel processing, any more than adding an EQ to only one channel of a stereo image is “parallel” processing.

Adding a HP filter to the sides breaks panning because changing ANYTYHING about the mid side signals individual channels will affect panning - not because the filter is minimal phase, but because of how mid/side works which is often misunderstood.

And yes, all EQ is built using a multiple audio paths that technically run “in parallel” - but it’s not “parallel processing” any more than parallel EQ is “parallel processing”. Just because the word “parallel” is used doesn’t mean it fits the definition of parallel processing, which is IMO always used to describe something where the user has control over the balance of the dry vs wet signal. Shelf EQ does not allow this, parallel EQ does not allow this, mid/side processing does not allow this. Hope I’m being clear (and accurate), this can be a complex subject to say the least! :)
Okay, I think I know now where I got confused. Most likely when I studied mastering for vinyl, I misunderstood that 80 Hz limitation.
I indeed forgot that only the stereo mix is what matters in the end.

That said, it's rather with headphones and in ears, where stereo sub below 80 Hz becomes audible. Some get sick from it, while I felt that it made the bass sound a bit weaker.

In the tutorial video, a regular HPF was gently used on the side channel (I think at just 20 Hz or so) to deliberately break hard panning, and give it depth. However, while I haven't tested it, for this effect, a Linear Phase filter might be pretty much useless.

Yeah, the dry/wet signal explains best what's actually parallel processing.

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 5 guests