Cleaning up the low-end and low-mid sections in a mix - with single track EQ, master track EQ, EQed aux effect sends and other methods

[Master Mi]

Today, I had a listen to some older and newer mixes of mine on the rather ordinary consumer hi-fi system at my mom's home when I was setting up a shelf at her place.

Of course the newest mixes I'm currently working on sounded way better than the older mixes - but even there I could hear a very small amount of unwanted low-end clutter which I didn't recognize when mixing all the stuff with my Yamaha MSP 3 studio monitors (frequency range from 65 Hz to 22000 Hz) and my Fostex subwoofer (adds lower frequency representation to around 40 Hz) and my Beyerdynamic DT 880 Pro headphones (with its comprehensively represented frequency range from around 5 to 35000 Hz).

I could also hear a bit low-end mud at soundtracks from professional bands known all over the world.
So I thought, it could rather have to do with the hi-fi system itself and its own (raised but less defined) bass representation.

But I'm still not sure since the soundtracks from famous bands - especially those from the 80s - mostly sounded like they were still on a higher level of perfection at music production.
And I think it's mostly because of the cleaner low-end.

Up until this point, I've always separated the other instruments from the bass and drums in the mix with various individual low-cut filter settings, while often allowing bass and/or drums to pass relatively freely down frequency.
But with the time I think more and more about cutting even the deeper low-end frequencies of instruments like bass and drums, which usually unchallenged cover the lowest frequency ranges in the whole mix, with a very steep low-cut filter below the frequency ranges of around 20 Hz to remove such almost inaudible low-end jumble (which one usually still perceives on rather bass-emphasized hi-fi systems or even some kitchen radios) especially for the playback on less good playback devices in maximum way.

I'm already thinking about doing this directly with a master EQ plugin, just because it might save a lot of time and you don't have to care about phase issues (in marked contrast to using steep-edged low-cut filters at standard tracks which share a similar frequency range - if my information about phase problems at this point is correct so far).

...

In this case, I could leave out the steep-edged EQ low-cut filter for the low-end frequencies at the bass and the drum elements completely (the softer, less aggressive low-cut filters of the other instruments for a better separation of frequencies between different instruments should still do their common job, of course) - and the steep-edged EQ low-cut filter in the master EQ plugin would do the whole job instead, without causing any phase issues.

So, in the EQ plugin interface it might look like this:



With the option Modus (mode), I could also set the phase line (green) from Normal (normal) to Linearphasig (linear-phase) - so, it would shift the phase of the whole frequency range and not just the phase of cutted lower frequency range.
But it should not make a really big difference at the master track at all, except of a few (not even perceptible) milliseconds in the latency of (parts of) the playback maybe.

I rather ask myself what the value of 18 Hz in the frequency field is supposed to mean in this case (the drop in the frequency range caused by the steep-edged low-cut filter with 36 dB per octave already seems to begin slightly at around 70 Hz - at 18 Hz the drop is around 10 dB and at 10 Hz the drop is already around 27 dB) - but maybe it's rather useful for EQ peak adjustments than for low-cut filter settings.

...

But what do you think...

1) Is it useful to filter out the bottom of the low-end frequency spectrum completely?
2) If yes, at which frequency you would start to radically filter out the bottom end (with at least a drop of 10 dB) - rather at 40 Hz, 30 Hz or 20 Hz?
3) Would you do it directly with a master EQ plugin like in my example?
4) Or would you be afraid of loosing some crucial audio information (maybe some playing noises or reverb/delay effects of a contrabass, an electric bass, a kick drum or a lower key from a concert grand)?
5) What is your favourite method of cleaning up the bottom end of your frequency spectrum in the mix?

...

PS: Since the topic became more extensive and more useful than expected, I changed the title from "Steep-edged low-cut filter on the master track as a general solution for solving low-end clutter issues in the mix?" to "Cleaning up the low-end und low-mid sections in a mix - with single track EQ, master track EQ, EQed aux effect sends and other methods".

Edited January 24 by Master Mi

[timaeus222]

It isn't a general solution, but I still do it on the Master Track because I will never hear or feel below 25 Hz anyway. 20 Hz is too little impact to pick. It will cut out room noise on vocal recordings, and excess low end on poorly-recorded samples.

But low-end clutter to me describes low-mid-range as well, which this does not solve. You simply have to solo each instrument that is occupying that range and decide what you want to be prominent, then scoop the others at that range (140 - 300 Hz or so, above kick drums and below regular midrange).

Edited November 2, 2023 by timaeus222

[Sengin]

If you haven't watched all of Dan Worrall's stuff, I highly recommend it: https://www.youtube.com/@DanWorrall.  He also does tutorials for FabFilter's stuff, but in a general way such that it can apply to everything ("Introduction to..." series).  Here's a few tips I've picked up on:

-Reduce the stereo width for your kick channel to mono or near-mono, and consider it for the bass (maybe don't go all the way to mono for bass)

-As far as I know, it's generally a good strategy to HPF each track up until it starts affecting the sound.  The low frequencies stack and build, especially after reverb + effects, after combining multiple tracks, etc.

-On the master, cut all the sub-bass with a HPF @ 20 Hz - you won't hear it anyway.  Consider setting it a bit higher depending on your mix - for music you don't always want the 'rumble' of low bass frequencies.

-Definitely try a HPF @ ~90Hz for just the side channel since our ears generally aren't sensitive to lower frequencies on the sides.  Depending on the specific content of your mix, this may change your stereo image for 90Hz and below (due to phase differences in the filter's slopes and target Hz) - if it sounds worse, play with the slopes of the 20Hz HPF and the 90Hz HPF or use a linear phase EQ for this specific HPF.

-Try HPF'ing your input to your reverb so that lower frequencies aren't 'verbed (probably want a shallower slope for the filter here).  Or at least ducking the wet low frequencies on a new note.

On 10/30/2023 at 6:20 PM, Master Mi said:

I'm already thinking about doing this directly with a master EQ plugin, just because it might save a lot of time and you don't have to care about phase issues (in marked contrast to using steep-edged low-cut filters at standard tracks which share a similar frequency range - if my information about phase problems at this point is correct so far).

Why care about the phase so much - are your filters not clean (e.g. there's a resonant bump at the target frequency)?  Your ears only pick up on the relative phase differences e.g. if you have different phases at a frequency on the left channel vs right channel or mid channel vs side channel (you can give some width to a mono track by using e.g. a 12dB/oct HPF on L channel but a 24 db/oct HPF on R channel at the same frequency).  But you should be able to use steep filters on individual tracks without issue (as far as phase is concerned). Phase can be a problem if the stars align and a bunch of individual waveform peaks line up, but that would very likely only be an issue for that individual note and not overall.

Edited November 15, 2023 by Sengin

[Master Mi]

On 11/15/2023 at 12:12 PM, Sengin said:

-Reduce the stereo width for your kick channel to mono or near-mono, and consider it for the bass (maybe don't go all the way to mono for bass)

I do it in a similar way.

I keep the bass in mono or very close to mono, opening it up around 1 to 5% towards full stereo - that way the bass sounds less stiff, static and monotone, and it has at least some room to move, especially if I'm using a really dry bass.

And for acoustic drums, I often use a greater stereo width of between around 20 and 50%.

 

On 11/15/2023 at 12:12 PM, Sengin said:

-On the master, cut all the sub-bass with a HPF @ 20 Hz - you won't hear it anyway.  Consider setting it a bit higher depending on your mix - for music you don't always want the 'rumble' of low bass frequencies.

That's the big question.

Guess I'm still afraid of losing some low-end audio information instead of embracing high-end sound quality in my mixes.
But when listening to some soundtracks from the 50s, it calms me down a little bit, because in the soundtracks back then wasn't too much low-end stuff at all:
 

It might have to do much more with the recording technology and the consumer audio playback devices of those days.
But even if you listen to the soundtracks of the 50s with today's audio equipment, they still sound really fresh, clean, well-mixed, highly dynamic, very controlled in the low and mid ranges and still pretty complete in the frequency spectrum.

...

In my remix of the Baywatch opening, for example, I really wanted some mighty low-end rumble in the industrial percussion section at the beginning and towards the end of the soundtrack.
Unfortunately, the latest remix version of this track that I uploaded many years ago, was still mixed on my old studio monitors (where I couldn't really hear or evaluate what was going on in the low-end and low-mids sections) without the application of my new mixing concept I've especially developed over the last two years:
 

I've already started working on an improved mix in a way that meets my current standards and mixing skills, where I even use a low-cut filter on the mighty industrial percussion and other instruments in order to clean up the low-end and mids sections in the track.
But it might take a while, because I also want to enhance the piano composition in this track and I'm mainly working on 2 or 3 other music projects at the moment.

At the moment, I might be lucky and be able to continue working on my mixing and composing projects during the week between winter service assignments and some work on the construction sites.

 

On 11/15/2023 at 12:12 PM, Sengin said:

-Definitely try a HPF @ ~90Hz for just the side channel since our ears generally aren't sensitive to lower frequencies on the sides.

Yeah, lower frequencies like from bass and kick drum often sound most effective and powerful in the center of the stereo panorama.

But there are also interesting exceptions to this rule concept.
Just have a look on the famous soundtrack "Stand by Me" by Ben E. King:
 

In this track, you have the bass far on the left side, together with a shaker and triangle as a nice contrast in the higher frequency section.
On the right side of the stereo panorama in this track, you have a lower strings section, a humming choir and a higher strings section with a violin (if I hear it correctly) as a contrast in the higher frequency section again.
But the center of the stereo panorma seems to be mainly reserved for the singer voice in this case.

Kinda unsual mixing concept with the bass panned to the side - but pretty effective in this track.

 

On 11/15/2023 at 12:12 PM, Sengin said:

-Try HPF'ing your input to your reverb so that lower frequencies aren't 'verbed (probably want a shallower slope for the filter here).  Or at least ducking the wet low frequencies on a new note.

I guess you're talking about working with aux sends for VST plugin effects such as reverb, which allow you to process the plugin effect separately from the main signal (e.g. EQing only the reverb without affecting the main signal of the instrument, synthesizer, voice, etc.) - in contrast to working with direct VST-based plugin inserts, where entire signal chains including the source signal get processed.

The thing is, I'm really used to work with direct plugin inserts or integrated effects of the VST instrument itself (where I also like the much better, faster and more accurate regulation of several settings based on much more conceivable parameters and values) 'cause I never really got aux sends to work properly in my DAW.
Whenever I tried to create an aux send and turned it on in a track where I wanted to use that effect, the DSP (the internal digital signal processor of my DAW) would suddenly go over 100% and cause huge instabilities, nasty sound artefacts or even crash-like dropouts.
This was very unusual, because with direct plugin inserts the DSP barely reaches the 50% performance mark of my DSP even in my biggest, most complex music projects - and I usually work with raw, unbounced/unfrozen MIDI tracks (needs much more DSP/CPU performance, but it's totally uncomplicated to change something in the composition while listening, mixing and editing the track).

Until some weeks before, I have never really found out what was causing this issue - and I own a really good computer with an Intel i7 6700 processor system, 32 GB of DDR-4 RAM, a decent UR44 audio interface, top-notch DAW named Samplitude Pro X4 Suite, and more than enough free disk space.

But then I have found out that I messed up with one single setting in my DAW - the number of processor cores that should be used to process my DAW tasks.
I had set 8 cores in my DAW because I thought that my i7 6700 processor system really had 8 cores - but it only has 4 cores (which I may have confused with the 8 threads).

After changing the setting to 4 cores, I was finally able to use my first aux sends for separately processed effects plugins with smooth DSP performance and no futher issues in my DAW.

I could have also used my Origami convolution reverb from my Independence Pro FX plugin library in Samplitude.
This Origami reverb plugin also includes a 4-band parametric EQ that just affects the reverb - unfortunately, it only comes with a low shelf filter, two band-pass filters and a high shelf filter without a clear graphic display instead of providing a nice low-cut filter, several peak filters and a high-cut filter with a clear graphical interface).

It looks like this:

But with the finally functioning possibility of working with plugin-based aux effects sends, I may be able to enhance the sound quality of my mixing concept even further.
I probably won't use EQed aux sends on the main instruments in the upper frequency range (if the frequency of an instrument including reverb there might clash with the frequency of another instrument including reverb, it could make sense to EQ the whole signal chain directly in order to get a cleaner mix, or - if just the reverb is the problem - drastically reduce the reverb or replace the reverb with some nice ping-pong delay effects).
But for the instruments with the lowest frequencies in the track - like bass and bass-heavy drum elements with stronger reverb - that don't have to compete with other instruments from even lower frequency ranges, it could really be useful to filter out just the long-reverberating low-end reverb clouds (which often sounds like dull, undefined sound mud on ordinary consumer speaker systems) from the mix, while maintaining the power and assertiveness from the main signals of the bass and lower drum elements.

…

Since I currently work on a new mix (based on my new mixing concept) for my Crisis Core: Final Fantasy 7 remix called "Wings Of Freedom", I could try out a few things and provide you with sound clips from different mixing approaches - especially the old version, the new version (based on my new mixing concept), the new version with an additional master low-cut filter, and the new version with an additional master low-cut filter plus some aux reverb sends with low-cut filter for crucial instruments.

As long as winter doesn't give me its legendary white-out ultra finisher with unexpected masses of snow these days (as I have already mentioned, I also work in winter maintenance during the cold season), I'll upload a few audio samples for you soon. ))

[Sengin]

On 12/5/2023 at 9:23 AM, Master Mi said:

I guess you're talking about working with aux sends for VST plugin effects such as reverb

I'm talking about inserts.  Instead of sending your main signal through the reverb, split the signal (e.g. in Reason you'd use the spider audio merger/splitter, but in every DAW it's different) and send only one through to the reverb.  But before it hits the reverb, send it through an EQ to e.g. roll off the lows.  Then merge it (the EQ'd + reverb'd signal) with your main signal.

[Master Mi]

On 12/18/2023 at 11:17 AM, Sengin said:

I'm talking about inserts.  Instead of sending your main signal through the reverb, split the signal (e.g. in Reason you'd use the spider audio merger/splitter, but in every DAW it's different) and send only one through to the reverb.  But before it hits the reverb, send it through an EQ to e.g. roll off the lows.  Then merge it (the EQ'd + reverb'd signal) with your main signal.

If I understand it correctly, it's more or less a change of the order of the plugin insert signal chain.
Usually you use the reverb first on the signal and then the EQ on the signal + reverb - but in your case it's first the EQ that hits the signal, and after this, the EQed signal gets the reverb.

Depending on the order of the plugin inserts, the sound result will be a different one.

...

But it could be really useful if the DAW developers create a system with primary plugin slots (maybe 7 per track) and secondary support plugin inserts (maybe 3 per primary plugin slot).
So, you could put a reverb in the primary plugin slot and an EQ in the connected secondary plugin slot that will only affect the plugin effect in the primary slot (and not the source signal itself).

Since I usually treat each instrument/track individually, I'd prefer a system like this over creating several aux send tracks for each instrument track.
But I guess I will use EQed reverb aux sends only for critical instrument tracks like drums, bass and instruments with lots of low-end and low mids.

For the instruments in the frequency ranges above, I will certainly continue to EQ the complete sum signal (source signal + reverb).

...

Got the 4 audio samples almost done in between work and weekends full of sprawling Christmas preparations.

Just gimme a few more days (already working on the 4th one, where I still try to find out which further instruments besides drums, viola, acoustic guitar, and maybe also the rather dry bass in the mix would benefit by using EQed aux reverb sends on them - and especially how much EQ/low-cut filter on the reverb effect is optimal to clean up the mix without destroying its ambience).

Edited December 19, 2023 by Master Mi

[Sengin]

On 12/19/2023 at 1:49 PM, Master Mi said:

it's more or less a change of the order of the plugin insert signal chain

Not quite - there's no "in place of" here.  You can do this approach (only reverb a HPF'd signal) and then EQ afterwards too (and you'll probably want to, or at least EQ only the wet signal).  The intent of the approach I mention is to reduce the part of the signal that gets reverbed because of how reverb tends to muddy out the lows.  Of course, you are free to EQ afterwards instead, or only EQ only the wet signal, or some combination, and the result won't be the same.  There's no one-size-fits-all approach - each will sound better in different situations (different genre, different reverb plugins) - it's up to you to try multiple approaches and decide on which is best for the song in that spot.  The more approaches you have, the more likely you will find the perfect fit.

Another similar approach is to use "de-emphasis EQ" - EQ is (I forget the mathematical term [edit: the term is "linear"]) 'non destructible' and reversible - a 6dB cut at 150Hz then a 6dB boost at 150Hz leaves you exactly where you were.  This means you can e.g. cut, reverb, then boost, and "de-emphasize" the lows that get reverbed.  Same works for e.g. distortion - you can emphasize fun frequencies by boosting, adding distortion, then cutting. 

 

On 12/19/2023 at 1:49 PM, Master Mi said:

I will certainly continue to EQ the complete sum signal (source signal + reverb)

Why restrict yourself?  I don't think this is such a black-and-white situation where EQ'ing the input+wet signals identically is always preferred.  I perhaps even find it beneficial to usually assume I will need to EQ the wet signal individually - this lets me keep a clean initial hit, but then ducking out the verb can make space for other tracks (especially in the all-important mids, or to reduce hissing/esses).  Then of course I am free to EQ the input+wet signal together if needed.

Personal preferences and all that, but I find myself usually staying away from aux sends as each track is its own thing and usually needs custom tailoring.  With sends, the only thing you can change on a per-track basis is the volume of the send.

Edited December 26, 2023 by Sengin

[Xaleph]

On 12/18/2023 at 5:17 AM, Sengin said:

I'm talking about inserts.  Instead of sending your main signal through the reverb, split the signal (e.g. in Reason you'd use the spider audio merger/splitter, but in every DAW it's different) and send only one through to the reverb.  But before it hits the reverb, send it through an EQ to e.g. roll off the lows.  Then merge it (the EQ'd + reverb'd signal) with your main signal.

In Reason you have the side chain filtering as well in the mixer.  I really don't use this as often as I should, I usually do a parallel track for reverb and EQ the reverb (and then sidechain the main onto the parallel track if necessary), or I just use a reverb vst that already has ducking and EQ already built in.

[Master Mi]

On 12/20/2023 at 4:57 AM, Sengin said:

Why restrict yourself?  I don't think this is such a black-and-white situation where EQ'ing the input+wet signals identically is always preferred.  I perhaps even find it beneficial to usually assume I will need to EQ the wet signal individually - this lets me keep a clean initial hit, but then ducking out the verb can make space for other tracks (especially in the all-important mids, or to reduce hissing/esses).  Then of course I am free to EQ the input+wet signal together if needed.

Personal preferences and all that, but I find myself usually staying away from aux sends as each track is its own thing and usually needs custom tailoring.  With sends, the only thing you can change on a per-track basis is the volume of the send.

Nah, I'll definitely use separately EQed aux reverb sends - but not for every instrument...

I think I will handle it just as I wrote before:

On 12/19/2023 at 10:49 PM, Master Mi said:

But I guess I will use EQed reverb aux sends only for critical instrument tracks like drums, bass and instruments with lots of low-end and low mids.

For the instruments in the frequency ranges above, I will certainly continue to EQ the complete sum signal (source signal + reverb).

The main reason for EQing the entire sum signal for the higher frequency instruments is firstly the fact that the instruments and sounds in the higher frequency range are often much more competitive (... so you may need to cut more frequencies in generel to clean up the track - then why not cutting straightly the entire sum signal?), and secondly the fact that reverb in the higher frequency ranges doesn't cause much of a problem for human ears (higher frequency reverb doesn't blur the soundtrack like low frequency reverb does).

...

Besides...

After the last week of work with the crappiest weather conditions (lots of rain, mud and almost a storm) on the building site, many hours of Christmas preparations, a merciless workout, the final cleaning of my cozy palace, a somehow relaxed and interesting Christmas Eve (as a big surprise my uncle visited my mother and talked about his trip to Japan and his experiences with the Japanese culture and the people there) and also a lot of boring small talk (I even took a big break from further family events and could finally enjoy working on my music projects), I managed to finish the 4 audio samples.

Maybe I'll already upload them in the next few hours or tomorrow morning. ))

[Sengin]

8 hours ago, Master Mi said:

so you may need to cut more frequencies in generel to clean up the track - then why not cutting straightly the entire sum signal?

Because of intermodulation (how the amplitude of a frequency determines how much of an effect is applied, and how that frequency's amplitude affects other frequencies).  It is not the same to cut and then reverb as it is to reverb and then cut - that is, reverb is not a linear operation.  If you are being wary of a specific frequency region because it can be crowded, if you cut first the reverb may sound more natural than if you cut afterwards (where it may sound like something is missing or "off" because the reverb was applied to a different signal at this point).

That said, I'm just giving you options.  Doing it one way over another is not always better - it depends on the mix and the sound you are going for.  I'm just letting you know there is a difference in cutting before a reverb and cutting after and why it is different.

 

On 12/24/2023 at 9:58 AM, Xaleph said:

parallel track

Ah yep, you are right - forgot about that way.

Edited December 26, 2023 by Sengin

[Master Mi]

10 hours ago, Sengin said:

Because of intermodulation (how the amplitude of a frequency determines how much of an effect is applied, and how that frequency's amplitude affects other frequencies).  It is not the same to cut and then reverb as it is to reverb and then cut - that is, reverb is not a linear operation.  If you are being wary of a specific frequency region because it can be crowded, if you cut first the reverb may sound more natural than if you cut afterwards (where it may sound like something is missing or "off" because the reverb was applied to a different signal at this point).

That said, I'm just giving you options.  Doing it one way over another is not always better - it depends on the mix and the sound you are going for.  I'm just letting you know there is a difference in cutting before a reverb and cutting after and why it is different.

Ah, I guess we were talking about two different things in this case.

You are talking about the direct insert plugin effect order:
Y1) EQ before reverb... will make a different sound result (maybe even cleaner as well - also might save some processing power of the CPU or internal DSP of the DAW) than...
Y2) reverb before EQ....

So I finally get what YOU are talking about - and thanks for the reminder at this point, because (if I remember correctly) I usually took the Y2 route.
This might have to do with my work habits when composing, arranging and mixing, where I often take a suitable instrument, then try to fit it into the ambience of my imagination with reverb, delay, chorus and other stereo/pan/room effects, and often do the fine mixing with the EQ stuff last.
That's probably the main reason for my plug-in insert order with the EQ at the end.

So if the source signal is "A", the EQ is "B" and the reverb is "C", the two ways of signal processing would result in different equations...

I'm neither a math geek nor a signal chain processing expert, but in this case the equations of processing the stuff for the two ways could be kinda close to my following creations of equations (guess it's still not the best and most accurate way to transribe the signal processing chains into an abstract terms - but perhaps it is enough for a rough imagination of the different results in the two different versions of processing the signal):

Y1 = sound result 1 (EQ before reverb)
Y2 = sound result 2 (reverb before EQ)

A = source signal
B = EQ
C = reverb

Y1 = AB + C*(AB) = AB + ABC = A (B + BC)
Y2 = AC + B*(AC) = AC + ABC = A (C + BC)

Let's take numeric values instead of the variables, something like: A = 2, B = 3, C = 5

Y1 = 2*3 + 5*(2*3)
Y2 = 2*5 + 3*(2*5)

Y1 = 36
Y2 = 40

Different numbers, different sound results on both ways.

"quod erat demonstrandum" :D

(Dude, I really hope I won't radically fool and disgrace myself with the math stuff here - if a math wizard 'n' tech sage reads this, feel free to correct, improve and transcend my light-footed pigeon-level equations.)


So, this was about the stuff you were talking about.

...

But I was talking about a different thing when I wrote:
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
"But I guess I will use EQed reverb aux sends only for critical instrument tracks like drums, bass and instruments with lots of low-end and low mids.
>>> only aux effect reverb sends

For the instruments in the frequency ranges above, I will certainly continue to EQ the complete sum signal (source signal + reverb).
>>> only direct plugin inserts for the instrument track (I didn't have the plugin order in mind here when writing about processing the sum signal - I could have also written "I will certainly continue to put a reverb plugin insert after the sum signal (source signal + EQ)" instead - my main focus here was just about using the plugins as direct plugin inserts in the instrument tracks.)

The main reason for EQing the entire sum signal for the higher frequency instruments is firstly the fact that the instruments and sounds in the higher frequency range are often much more competitive (... so you may need to cut more frequencies in generel to clean up the track - then why not cutting straightly the entire sum signal?), and secondly the fact that reverb in the higher frequency ranges doesn't cause much of a problem for human ears (higher frequency reverb doesn't blur the soundtrack like low frequency reverb does)."
>>> In this case, direct plugin inserts for these instrument tracks would save a lot of time compared to creating additional aux effect send tracks for each individual instrument track (unless you want to work with entire instrument groups where you use one aux effect send for the entire group).
-------------------------------------------------------------------------------------------------------------------

...

I don't know how much knowledge and experience you have with aux effect sends.

But when you work with aux effect sends, you really have 2 different tracks there - the instrument track (let's say track number 1) and another aux effect send track (could be track number 40, for example)...
Both tracks (and this is the great feature of working with aux effect sends) can be processed completely differently - different pannings, different inserts etc.
You just need to activate the aux effect send in the aux slots of your instrument track to route the effect to the instrument (otherwise the aux effect send track doesn't "know", to which instrument track it should put the aux send reverb effect or other effects) - and of course you need to set the relative ratio or intensity/level strength (in dB) of the aux track in relation to the level strength of the instrument track directly in the aux slot of the instrument track (if the level strength ratio between both tracks is set, the routed aux effect sends will get louder or quieter as the instrument track gets louder or quieter).

And here comes the big one - for the case you want to radically clean up your mix (especially the blurring reverb) without losing the original character and frequency range of your instrument.

Just as an example...
You're composing a complex ambient soundtrack with an acoustic guitar that has a really cozy, warm tone (you definitely want to keep the full frequency spectrum of this particular instrument) - but as soon as you apply reverb to this instrument, it completely messes up the other instruments (bass, drums and a few other instruments that play in the lower frequency range).

So, dry acoustic guitar sounds great in the mix - acoustic guitar with reverb makes the mix messy and muddy.
A problem which could be solved with the magic of aux effect sends.

Remember?
Two completely different tracks - the instrument track (which we will leave as it is - without any plug-in inserts in this case) and the other aux effect send track (into which we will insert the reverb and EQ only this reverb).

We want to keep the full frequency range, warm tone and clean sound of the guitar in the mix - so, we don't use any EQ plugin insert oder reverb insert on this instrument track.
Yep - sounds warm and clean, but still dry as hell.

So we still need some reverb (but a radically cleaned up reverb without the problematic low-frequency reverberation) for the ambience.
And of course we'll only put the reverb plugin in the plugin slots of the separate aux effect send track, because if we EQ the reverb in the separate aux effect send track, it won't affect the source signal of the instrument (two separate signal chains - one for the instrument track, one for the aux effect send track).

It is not of primary importance here whether the reverb is switched before the EQ in the signal chain of the aux effect send track or whether the EQ is processed before the reverb.
The important and really helpful feature here is that you can EQ only the reverb for the instrument without EQing/touching or changing the instrument itself.

So...
If you do it wisely, you can get a instrument with its full frequency range and original sound character together with a decently low-cut-filtered reverb by using aux effect sends.

In our case, we can have a nice, warm and cozy acoustic guitar with an untouched frequency range in combination with an ambient but clean guitar reverb, where just the low frequencies of the separately processed guitar reverb have been heavily low-cut-filtered.
And as a result, the guitar reverb shines much brighter and won't mess up the mix anymore.

...

I hope that I was able to make it a little clearer what I was referring to in my previous comments.

Edited December 26, 2023 by Master Mi

[Master Mi]

PS: Since the topic became more extensive and more useful than expected, I changed the title from "Steep-edged low-cut filter on the master track as a general solution for solving low-end clutter issues in the mix?" to "Cleaning up the low-end und low-mid sections in a mix - with single track EQ, master track EQ, EQed aux effect sends and other methods".

...

And since the upload feature below the comment field is working again (big thanks to DarkeSword at this point for restoring the upload fuction), I can provide some audio samples to show different mixing approaches or possible stages of mixing in my next comment.

[Master Mi]

Some first steps of improving the mixing quality - using EQ filters and using aux effect/reverb sends instead of direct effect/reverb plugin inserts
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

So, finally I can start with the promised audio samples of my Crisis Core: Final Fantasy 7 remix at 4 different mixing approaches or mixing stages (only excerpts, but in which you can hear some crucial parts regarding the mixing quality - the final remix might still get some changes in the composition)...

Here they go:


1) The old version
-----------------------

 

 

This version is about 6 to 7 years old and was mixed on my former studio monitors (Presonus Eris E3.5 studio reference speakers in combination with my still existing Fostex PM-SUBmini 2 subwoofer) and on my Beyerdynamic DT880 Pro studio headphones connected to my Steinberg UR44 audio interface.

I didn't have the biggest mixing experience back then and didn't have as fine an ear as I do now (which may have something to do with the fact that my studio monitors in particular at the time couldn't show me the necessary details and the critical things in the mix - they had a very pleasant and powerful sound, but they rather varnished my tracks and made them sound kind of finished at a still unfinished stage).

I didn't even use dedicated EQ VST plugins back then (just a couple of 3-way vintage EQs and timbre controls in the VSTi and synth interfaces) because I thought that EQing (especially if you overdo it) turns the natural sound of an acoustic instrument into something very artificial and that the more EQ you use, the more your hearing constantly adapts to the new timbre until you can't tell when the bass is getting too heavy and trebles are getting too shrill - and so on.

I was thinking more about how it is still possible for so many different instruments to play together with the whole room reverb in large concert halls without the overall listening impression sounding muddy (I just wanted to understand this concept on a fundamental level and try to implement it in my DAW).

I rather suspected that it might be due to the well-placed positioning of the instruments or the size and spaciousness of the concert hall for a more relaxed expansion of the audio waves or acoustic signals, or that it might also be related to certain differences between real acoustic, analog sound signals and digital sound signals.

But back then I didn't even get to the kinda obvious thought that also great concert halls often get an enormous acoustic treatment on a highly professional level.
And if you talk about large bass traps, acoustic panels for the walls, floor and ceiling, or even about special materials for the seating, you could also say that these acoustic tools work in a similar way like a low-cut filter of EQ VST plugin, for example (only with the big difference that with a good EQ plugin you can react much better, much faster and without huge expenditure to changes in the soundscape).

...

But despite my hesitant moves at mixing, I still had the base of a mixing concept back than - a concept which should bring at least some great possibilities for dynamics, untouched/undestroyed signal peaks for better sound quality and a natural sound of the instruments, and a concept for sophicated loudness regulation of soundtracks.

I'm talking about mixing at EBU R 128 standards (a recommendation of the European Broadcasting Union) developed by really farsighted audio engineers, who wanted to bring back lots of dynamic, sound quality and a very good loudness regulation to audio and audiovisual content (no more annoying loudness jumps between different soundtracks and other audio programs) at broadcasting as a reaction to the ongoing phenomenon called "loudness war", mainly caused by a growing use of compressors in the ad and music industry to sound louder than the competitors.

With the help of this, every single soundtrack and audio program will be mastered to a target level of - 23 dB +/- 1 dB in context of the full scale (for a better imagination: 0 dB is the point no sound signal can exceed and where a sound signal turns into clipping).
Loudness is something like the perceived sound pressure level measured over a certain amount of time.
So in order get a mix towards a loudness target level of -23 dB, you need a loudness meter and you have to measure your soundtrack from the very beginning to the very end with your loudness meter (cause the target level of - 23 dB is always an average value - so, your soundtrack might start at - 33 dB or even will be at - 20 dB in the middle of the track - if it is at -23 dB at the end, it's fine so far).

Of course there are also limits for the maximum dynamics at EBU R 128 mixing (so, an untamed sound of a gunshot after a soft piano melody won't blast your ears after mixing it at EBU standards) defined in terms like "Maximum Short-term Loudness Level" (should not exceed -18 dB), "Maximum Momentary Loudness Level" (turns red in my EBU-adjusted loudness meter as soon as it reaches - 15 dB) or "Maximum True Peak Level" (should not exceed - 1 dB), but I usually just keep a fleeting eye on these parameters (because I mainly create music and no heavily dynamic cinematic special effects or stuff like that).

Many modern soundtracks are already mixed at target levels of -15 dB to -12 dB or something like that, leaving no big headroom for the signal peaks above (guess that was the time where sound surgery like peak compressing or brickwall limiting started, and where the dynamic of soundtracks has decline more and more).
So, soundtracks and audio programs mastered at EBU standards are around 50% to 60 % as loud as lots modern music - that's a similar loudness like the loudness of original sound mixes from the 80s.

The really cool thing is that you don't have to care about the signal peaks when mixing at EBU R 128 loudness standards because they always more than enough headroom.
Even in the master track the signal peaks will barely scratch the - 5 dB mark (and I don't use compressors or limiters in my soundtracks).
Saves a lot of time at mixing/mastering and provides a good uncompressed signal.

...

But enough small talk about the early foundations of my mixing concept.
Let's go to the next audio sample for showing the next stage of mixing I'm currently working on.

...


2) The new version with my over the last years developed new mixing concept
--------------------------------------------------------------------------------------------------

 

 

This is the new version I'm currently working on.

It's based on my newly developed mixing concept, which I already used for my Goldfinger remix called "Safe 'N' Sane Skater Heaven Superman", and it is mixed on my new professional Yamaha MSP3 studio monitors (Yamaha MSP series is the professional product line of the Yamaha HS studio monitor series, so it comes with an even better audio definition and a flatter, more natural and relaxed sound) in combination with my Fostex PM-SUBmini 2 subwoofer as well as with my Beyerdynamic DT 880 Pro studio headphones (finally with the silver ear cups giving me a much more neutral and natural listening experience) connected to my new Lake People G109-P headphone amp (which can finally drive these high-impedance headphones with ease).

Cause of the really outstanding audio resolution and truthfulness of the Yamaha MSP3 studio monitors (they also have no annoying hissing or humming noises at the teeters or woofers, so they are perfect as a relieable near-field studio monitor solution), I can hear much more details in my tracks and I can play much more with the reverb without being afraid of messing up the mix with that.

In addition to the EBU loudness standards as the early foundations of my mixing concept for better dynamics, peak signals and sound quality, I also used EQs with low-cut filters in the single tracks of this mix the first time.

Another big core of my new mixing concept is the use of a really helpful 2-channel surround feature of my DAW, which encodes surround information into a stereo signal and where you can place instruments and other sound signals in a graphical interface as you wish.
Wheter hifting a signal more to one side or to the front or back, getting a signal only to the sides and sparing out the center completely, reducing the stereo width by dragging the two stereo objects more together and to the center - no problems within a short amount of time with this useful tool.
So, it's no only good for creating a great imagination of depth and spaciousness in your mix - it's also good for separating the competing frequencies of your instruments by different placement of these, and for radically cleaning up the mix in this way.

If you want to read more or even get a little audiovisual impression of this feature, I'd recommend the thread "Creating a realistic impression of depth in stereo mixes" (especially under my comment with the title "Visual tools for creating a realistic impression of depth in stereo mixes") in the Music Composition & Production section of this forum.

...

I guess this was a nice and really succinct summary of the main elements of my new mixing concept.

...


3) The new version + master track EQ with low-cut filter
---------------------------------------------------------------------

 

 

This version is almost the same like the second audio sample, but in addition to that I also used a steep-edged low-cut-filter on the master track this time (low-cut filter with 36 dB per octave - starts slowly around 50 Hz, at 20 Hz it lowers the frequencies about 10 dB, and at 15 Hz it lowers the frequencies already about 20 dB).

You'll maybe hear a little difference between audio sample 2) and 3) - but it's not like the far bigger jump from 1) to 2).


I'm not sure if I will use a master track low-cut filter as a general device in the future of my mixing concept - still afraid of losing some crucial audio information between 30 and 50 Hz (especially the kinda earthy power of the bass and kick drum - not the reverb).

...


4) The new version + master track EQ with low-cut filter + aux reverb sends with low-cut filters for specific instruments
---------------------------------------------------------------------------------------------------------------------------------------------------

 

 

This final audio sample is based on the version for the previous audio sample.

But it has one big difference.
Instead of using EQ and reverb plugins as inserts for all instruments, I picked 4 instruments with the most critical (lower) frequency ranges, especially due to possible reverb mud, and used EQed aux reverb send with decent low-cut filters on the reverb effects for these instruments.

The 4 instruments on which I used aux reverb sends on are:

- the drums (whole drum kit)
- the electric bass
- the viola
- and the acoustic guitar playing the chords

I think that this made a bigger perceivable difference and cleaned up the bottom of the mix with the low-end and low-mid reverberation really well.

...

So, now I'm really curious about your opinions regarding the different mixing approaches (or the possible stages of mixing) and the sound quality.
And I would like to know which mixing results you like best or which might be the most promising ones.

 

Edited April 8 by Master Mi

[Master Mi]

How do you pan aux reverb sends in the context of the panning of the instrument or main signal source?
-----------------------------------------------------------------------------------------------------------------------------------

After finally getting aux effect sends working in my DAW some time ago, I've increasingly integrated them into my mixing approach due to the improved clarity, sound quality and enhanced sound design possibilities they can bring to a mix.
But there's one thing I'm still not quite sure about - how to arrange them in the stereo panorama (especially in the context of the panorama of the instrument or source signal) in the best possible way.


I've had various thoughts on this and am currently drawing on different approaches, for example:

A) Panning the aux reverb send just like instrument/source signal (so that the relationship between the instrument/source signal and its reverb send is not torn apart too much - could be useful if too much is already happening in the mix outside the panning of the instrument/source signal, but also disadvantageous if there is already too high a density of musical events in the area of the panning of the instrument/source signal)

B) Bringing the aux reverb sends away from the center of the panorama to the sides (might increase the carity of the track a lot - since I rarely pan instruments fully to one side due to the loss of spatial information of instruments panned like this, it could be useful to rather pan reverb effects more or even fully to the sides)

C) Panning the aux reverb send a little bit more to the side (on the same side as the source signal - e.g: source signal is panned + 3dB to the right, then the reverb send could be panned + 10 dB to the right - can be useful if the instrument in its panorama area needs a little more punch/less reverb and there is still some room for the reverb send in the panorama a little bit further to the right)

D) Panning the aux reverb send fully to the side (on the same side as the source signal - amplifies the effect of variant B somewhat, but the spatial relationship between the source signal and its reverb send is lost to a greater extent)

E) Panning the aux reverb send fully to the opposite side (the far extreme on the opposite side of the source signal, for example, if the instrument is panned + 3dB on the right side, the aux reverb send of the corresponding instrument is panned hard left - this does not seem to disturb the spatial relationship between source signal and reverb to such an extreme extent, while at the same time, the assertiveness of the instrument/source signal drastically increases in the mix)

D) Panning the aux reverb sends just where you've got lots of free space, how it fits your needs as a sound creator and like it sounds best (guess this sounds like some kind of a text book answer)

So, how do you handle the panning of aux reverb sends?
Do you have a general solution for this or do you rather use an individual approach depending on the instrument/signal source, music genre or the specific intention of sound design?

…

For showing some practical stuff let's go to some further audio samples of my Crisis Core remix I'm currently working on.

During the last weeks I could make some huge progress with this track - not just with the mixing, but also by composing lots of new stuff after playing, recording some further melodies for old and new instruments via MIDI keyboard and finally editing the content with the MIDI editor of my DAW.

Although the audio samples are still a few steps behind the actual mixing and compository state of my remix, the mixing of most instruments was already done at this point.
But I still had an issue with the electric guitar, where I wasn't sure how to pan the guitar reverb send the best possible way.


Just to give you an fundamental idea of the mixed instruments which appear in the following audio samples without any further changes (if I remember correctly):

- electric bass (plays almost fully in the center with a slight stereo width of around 2 % in order to make the bass sound a bit broader and with a less stiff
spatial impression)
>>> aux reverb send of the electric bass (a minimal scoring stage convolution reverb with low-cut filter) is panned like the instrument itself (still not sure if the mix sounds better when bringing the subtl bass reverb more to the sides)

- acoustic drums (play a bit more in the background between center and sides - stereo width should be around 50 % - source signal also has a reverb insert with a subtle EQed concert hall convolution reverb, which also makes the kick drum more powerful)
>>> aux reverb send of the acoustic drums (a subtl cathedral convolution reverb with heavy low-cut filter to add some airy vibes to the drum kit) is panned to the sides, leaving out the center (you will hear much more of this great effect shortly after the intro of my remix - it's not in the following audio samples)

- viola (panned around + 3 dB to the left side, source signal contains delay effect and also a smaller low-cut filter)
>>> aux reverb send of the viola (a subtl cathedral convolution reverb with a smaller low-cut-filter) is panned like the instrument itself

- acoustic guitar chords (fully panned to the sides, leaving out the center, a smaller treble boost from a vintage EQ is used on them)
>>> aux reverb send of the acoustic guitar chords (a delayed hall reverb with moderate low-cut filter) is panned in a similar way like the instrument

- trumpets (one of the new sections I've composed for this remix, panned around + 7dB to the left side, source signal contains a heavy low-cut-filter and a subtl, already low-shelved cathedral convolution reverb)
>>> no aux reverb send is used for this instrument

...

And now comes the critical part with the raw clean electric guitar, where I'm still looking for the best solution concerning the mixing.
In all audio samples the electric guitar source signal (panned around + 3 dB to the right side) goes through my guitar amp plugin Vandal (with a smaller overdrive stomp box, a special Alnico cabinet simulation and a stronger ping-pong delay effect) and a moderate low-cut-filter.
The used aux reverb sends for the electric guitar after audio sample 1 contain a subtle cathedral convolution reverb with a heavy low-cut-filter.

The differences in the electric guitar section are shown in the following audio examples with different mixing approaches for this instruments:


1) No aux reverb send - guitar reverb comes via direct plugin slot insert from guitar amp plugin
-----------------------------------------------------------------------------------------------------------------------

 

 

...


2) Aux guitar reverb send panned like the guitar
------------------------------------------------------------

 

 

…


3) Aux guitar reverb send panned about 7 dB more to the right side than the guitar
--------------------------------------------------------------------------------------------------------

 

 

....


4) Aux guitar reverb send fully panned to the right
---------------------------------------------------------------

 

 

…


5) Aux guitar reverb send fully panned to the left (the far extreme on the opposite side of the source signal)
---------------------------------------------------------------------------------------------------------------------------------------

 

 

…


Since I didn't like the sound results I got with a standard method for mixing electric guitars (like panning the source signal fully to one side and panning the reverb send fully to the opposite side - as it seems to have been done with the electric guitars at some points in this really awesome Maniac Mansion remix composition: https://www.youtube.com/watch?v=v-6Le36mlDA - but somehow I can't stand the sound from mixing appoaches where instruments - especially lead instruments - are fully put to the sides), I almost think the mixing approach from audio sample 5 works best in this case (especially at the point where the trumpets kick in).

…

But let me know your opinion about this topic and my different mixing approaches.

…

Besides, I just thought it couldn't even harm to additionally upload the...


6) Latest update of the remix section showed in the previous audio samples
-----------------------------------------------------------------------------------------------

 

In this version (which is based on audio sample 5) I slightly enhanced the trebles and brilliance of the trumpets and the electric guitar (so they can shine a bit more in the mix and so, they are also put more to the front as the lead instruments of this part), and I spiced up the drums section with a few variations.

 

 

Edited April 8 by Master Mi