What is the best "Q" Setting for audio mastering?

[Trythanks Lightfoot]

Perhaps my question is kinda broad but in most given situations what would y'all recommend as an ideal quality factor (or bandwidth) for audio mastering?

 

[Mpumelelo von Mumhanzi]

There are absolutely no de-facto best settings for audio mastering or for mixing music. It's all about taste — as Dave Pensado likes to call it. There are however certain guidelines that come in handy when one wants their audio material to fit a specific style, trend, or whatever.

For example, sharp/gentle boosts for the bottom and wider boosts for the top end, because at low frequencies, the human ear differentiates changes in frequency better than it does at high frequencies.

In other words, it's much easier to tell the difference between 80 Hz and 100 Hz compared to 8000 Hz and 8100 Hz.

So why would someone want to use a narrow quality factor for high frequencies unless they are trying to fix something requiring a surgical EQ? It would sound piercing. Nevertheless, that might actually be one's desired sound they want after all.

And as for the bottom end, wide boosts will make the audio material sound boomy obviously depending on the bass that's currently present at the moment. If a mix is thin then there shouldn't be issues.

 

[Trythanks Lightfoot]

@Mpumelelo von Mumhanzi I recently watched this Hot Trax Production Magazine video on YouTube where Chris Athens was talking about using a narrow bandwidth at around 100 Hz and a medium bandwidth at 8 kHz.

What would you say corresponds to a narrow or medium Q i.e. when using a VST plugin. By the way what equalizer was Chris Athens using?

 

[Mpumelelo von Mumhanzi]

@Mpumelelo von Mumhanzi I recently watched this Hot Trax Production Magazine video on YouTube where Chris Athens was talking about using a narrow bandwidth at around 100 Hz and a medium bandwidth at 8 kHz.

What would you say corresponds to a narrow or medium Q i.e. when using a VST plugin. By the way what equalizer was Chris Athens using?

Well, that's a Sontec MES-432C mastering equalizer Chris Athens is using in that video.

And to be honest I don't think you can get the sound of a Sontec with a generic VST EQ plugin because some hardware gear may or may not be modified. Unless you just want the EQ curves. It's easy to clone any hardware EQ you can think of. It's not that hard to do at all.

P.S. by the way, you can't clone some's subjective hearing though. Maybe in the soon-to-come Brave New World (or the Great Reset), if you will. I suppose then it will be possible with relative ease assuming there's approval from powers-that-be to make the biotechnology available to all the masses.

 

[Tanonoka]

The Sontec MES-432C EQ has 5 different pre-defined Q settings in regards to the width of the bell curves.

Therefore it's a bit vague to understand what a narrow Q means to Chris Athens here.

A narrow bandwidth to him could be either 11 dB/octave or 15 dB/octave whilst a medium Q could be either 6 dB/octave or 9 dB/octave.

Who knows really? It's only Chris himself who can answer that one, unfortunately.

 

[Mpumelelo von Mumhanzi]

@Tanonoka Semantics, LMAO.

 

[Tanonoka]

@Tanonoka Semantics, LMAO.

Yeah, the issue of semantics makes things so confusing fo' sho'—especially in the world of professional audio. Some people use words like blunt, broad, or wide to describe EQ shapes. But then again the question becomes what's the consensus of all those terms in numerical values though?

Furthermore, you also hear other terms like beefy, punchy, meaty, bite, smack, and on and on. It's all subjective from one person to another I guess.

 

[Trythanks Lightfoot]

@Tanonoka Okay suppose I want to replicate EQ shapes of a Sontec MES-432C?

 

[Tanonoka]

@Tanonoka Okay suppose I want to replicate EQ shapes of a Sontec MES-432C?

That question of cloning Sontec EQ curves has been answered already by @Scorpio when he was talking to @Tafadzwa Twabam

 

[Trythanks Lightfoot]

That question of cloning Sontec EQ curves has been answered already by @Scorpio when he was talking to @Tafadzwa Twabam

It's kinda written in Patwah and I can't comprehend it.

 

[Tanonoka]

It's kinda written in Patwah and I can't comprehend it.

I am feeling lazy right now to translate the whole thing for you.

I'd suggest maybe @Nolwazi Kwayedza or even @Scorpio himself to respond to your query.

 

[Nolwazi Kwayedza]

I am feeling lazy right now to translate the whole thing for you.

I'd suggest maybe @Nolwazi Kwayedza or even @Scorpio himself to respond to your query.

Why me?

@Scorpio should do it.

 

[Trythanks Lightfoot]

@Nolwazi Kwayedza @Scorpio @Mpumelelo von Mumhanzi come on guys someone... I mean, please.

 

[Scorpio]

Every song is different and for this reason, there can never be a Q setting that is going to work on all mixes.

Some mixes are dull; some mixes are bright and some mixes don't need a whole lotta fixing because they sound good already except maybe in terms of commercial loudness which of course can easily be achieved these days by anyone who knows what they are doing.

After all brickwall limiting is not rocket science—it doesn't take too long to learn compared to equalization which is extremely challenging because:

• It costs money to set up your own good listening environment (i.e. room treatment)
• It costs money to purchase a decent pair of studio monitors.
• It takes so much time and dedication to come up with your own unique taste for music because that's what you will be selling to clients and that's all that matters eventually—forget the silly objective rhetoric that charlatans peddle around. You can't be objective with music otherwise there would be one supreme scientific style to rule them all if that was the case.

However, if ever one should make a decision to master their own music (assuming they are a novice still getting their feet wet) without screwing things up, I would recommend a Q of 0.707 (and that's almost close to two octaves) but then again these VST plugins are so different and this is what I was saying to @Tafadzwa Twabam the other day.

You can't just grab a random EQ and type in 0.71 thinking the results will be the same as far as the subject of EQ curves is concerned.

 

[Mpumelelo von Mumhanzi]

@Scorpio A quality factor of 0.349672 is probably much better.

 

[Scorpio]

@Scorpio A quality factor of 0.349672 is probably much better considering how broad it is.

Q 0.349672 works tremendously well for boosts when applied to Bark 17, 21, 23 (i.e. what some folks consider to be EQ pressure points). It can also be used at 24 kHz or even 28 kHz to add top-end sheen where necessary. In fact, most analog EQs if not more than 50% of them have that type of broadness in relation to the EQ curves.

Q 0.707 is less likely to create a mess to either Bark 1 or Bark 2 even though it's slightly wider for that range—obviously, it depends on the genre.

In summary, Q 0.71 is a better recommendation than Q 0.349672 especially when someone asks a question about a general-purpose EQ curve without the need to expagorate things tew much. Otherwise, we will be here until the cows come home.

 

[Mpumelelo von Mumhanzi]

@Tanonoka Okay suppose I want to replicate EQ shapes of a Sontec MES-432C is that possible?

It depends on what EQ you will be using.

 

[Trythanks Lightfoot]

@Mpumelelo von Mumhanzi I mean on FabFilter, iZotope, DMG and Flux Epure.

@Scorpio what are EQ pressure points?

 

[Mpumelelo von Mumhanzi]

@Mpumelelo von Mumhanzi I mean on FabFilter, Izotope, DMG and Flux Epure.

@Scorpio what are EQ pressure points?

I will just translate to iZotope Ozone EQs. But that doesn't mean iZotope EQs will sound like a Sontec even though the EQ curves will snap precisely. Sometimes it slightly varies when you compare two Sontecs side by side. Anyways...

Sontec​	iZotope Ozone/Neutron​
5 (dB per octave)	0.58
6 (dB per octave)	0.92
9 (dB per octave)	1.58
11 (dB per octave)	1.95
15 (dB per octave)	2.85

N.B., use Ozone's EQ analog regular bell curves instead of the proportional bell curves for best results. And if you wanna use Ozone EQ's digital mode just make sure it's set completely to minimum phase (i.e. 100% on that slider whatever. I can't remember what it's called from the top of my head) instead of linear phase or mixed phase mode—that's it really, you are good to go.

Sontec bottom shelf​	iZotope Ozone Equalizer (Baxandall shelf)​
50 Hz	130 Hz
100 Hz	200 Hz // this is also quite similar to a Pultec EQ @ 20 Hz

• 100 Hz on Sontec is also slightly similar to a Bax EQ at 131 Hertz except the latter has a dip of -0.5 dB at 400 Hz, Q 0.7071
• What some people call a Baxandall shelf is simply a first-order shelving filter you can do that with any contemporary digital EQ.

Sontec top shelf	iZotope Ozone Equalizer (Baxandall shelf)
It's fixed	3.7 kHz

 

[Scorpio]

@Mpumelelo von Mumhanzi I mean on FabFilter, Izotope, DMG and Flux Epure.

@Scorpio what are EQ pressure points?

Just think of EQ pressure points as common go-to center frequencies for stupid loudness wars and these are:

1.2 kHz (with a 6 dB/oct shelf, or 12 dB/oct Q 0.48 or even 18 dB/oct Q 0.219 shelves on FF Pro-Q)

1.6 kHz (with a blunt bell curve)

2.0 kHz (with a 6 dB/oct shelf, or 12 dB/oct Q 0.48 or even 18 dB/oct Q 0.219 shelves on FF Pro-Q)

3.5 kHz (with a blunt bell curve)

5.0 kHz (with a 6 dB/oct shelf, or 12 dB/oct Q 0.48 or even 18 dB/oct Q 0.219 shelves on FF Pro-Q)

6.5 kHz (with a blunt bell curve)

Anything above 3.5 kHz and below 300 Hz isn't of primary priority because music is first and foremost "all about the midrange" this should be understood wholeheartedly because every speaker or headphone reproduces that frequency range. We can always disagree and that's fine if you are slow.

I am not saying the low-end or the top-end doesn't matter at all. It kinda does but... I know I am right even though I don't claim to know it all. However, y'all need not argue with me because everything else is wrong... I mean best believe I ain't trolling either at least not yet. ?

Lest I forget, I'd say in this case a blunt bell slope = Q 0.304601 assuming gain is between 1 to 2 dB, and from 2 to 4 dB use Q 0.349672 whilst medium bell slope = Q 0.7071 because doing so you will manually mimic those so-called proportional musical Q values.

If you use FabFilter then the slope Q 0.304601 becomes Q 0.430705 (or just Q 0.431) whilst the slope Q 0.349672 becomes 0.494436 (or just Q 0.5) and finally, Q 0.7071 that's 1 assuming all these bell curves are set to 12 dB per octave.

P.S. alternatively for medium bell slope you can also use Q 0.556 (id est, approx. Q 0.82 on FabFilter ProQ) which is slightly similar to the widest bell on the Sontec MES-432C equalizer. And for interest's sake, FabFilter "Qs" are similar to Weiss EQ1 "Qs" just so you know.