MStereoprocessor as 'per-instrument' mixing tool

[astramistil]

Is there risk of artifacts if I use MStereoProcessor on all individual track/bus? I tried with guitar buss/drums buss and on an incomplete mix but its hard to tell. I want to minimize the use of the 'pan' as main stereofield control (which I feel bring often un-cohesive, thin results). So I want your advise

Problematic:

1-MStereoProcessor on Master:
a) When I expand high-mid, often the cymbals sound way too far, but Guitar stand exactly where I want. Also, distortion often appears because it modifies the whole range instead of certain instruments on certain frequency pocket
b) I like to have guitar low-mid a little widen, but bass almost mono centered in that area
c) etc.

2-MStereoSpread:
a) **Only have one filter (for lows), but I want to try something like this on Guitar bus (panned center): Low band mostly centered, low-mid 10-20% widen, high-mid 40-70% (and more on chorus), high (as mush as my ear likes) -- Is that a good idea? Or can bring alot of problems?
b) The 'Spread' option is not what I'm looking for *unless I'm missing something* (for now, i want to use MStereoProcessor mainly on instrument bus)
c) Is it me or to achieve a wide sound it colors sound alot on punchy instruments? (even with Band at 100, spread not set to high); I hear some weird ''phaser stuff'' (like guitar palm mute). So I guess i should keep it for less 'punchy' instruments?

3-MStereoGenerator:
a) The EQ as a kind of per-hz Dry/Wet control is cool
b) Too subtle for what Im looking for (but great on ind. instr)

=======================================================

So MStereoProcessor could let me gain better control on stereofield. But:

4-MStereoProcessor on individual bus(/tracks not mono):
a) Im afraid that if I 'widen' high-mid from a guitar at a 'crossover' of 2k, and a HH at 3k with different widen amount (considering stereo content already present in these sounds) that weird things could happen (am I right?)
b) If I use upsampling and linear phase, will artifacts be minimal?
c) Otherwise, why is it consider only a mastering tool?

** 5- Should I consider in that case using exclusively M/S Eq/MSpectral as a way to do what I want?
a) I mean, it would be more precise than MStereoProcessor, but can I achieve similar results without artifacts?
b)*Does raising a high-mid band in 'Side' mode = raising ''widening'' for high-mid band in MStereoProcessor?

Thank you, I hope what Im trying to achieve is clear enough so you guys can help!

[jmg8]

My personal choice is MStereoSpread (MSS) as it has 3 stereo widening types.
1. Increasing the gain of the side signal. This is the method used in MStereoProcessor (MSP)
2. A delay based winding aka Haas effect. This one sounds phasey and is the least transparent. This is also in MSP.
3. Spectral based widening. This is the most transparent and only found in MSS.

I would recommend using the spectral option on individual elements in your mix. Multiband isn't essential here as there is a focus parameter that controls how much widening is done to the higher and lower frequencies. This method should be pretty transparent but if you do get any transient smearing, the min frequency parameter should help.

[astramistil]

Thanks for your answer, but I tried MSS and I have the impression that it color sounds too much (even with focus and freq well set), I mean, I already have stereo content (except for voices) in my material

I would like to 'place' it making different 'pocket' of hz. As I said: Low range guitar mostly centered, mid little wide, high-mid +wide, high +++wide (for each instruments)
---------------------------------
With MSP, I can accomplish that easily on different instrument busses, but Im afraid there could be trade-offs I havent heard, or think of (in later stage of mixing or mastering)

2-So would a M/S Eq the answer + a widener (as the one in MSS) post-EQ? Or is it totally different processing/results from what MSP do?

[MeldaProduction]

MStereoProcessor does pretty much the same thing as MAutoDynamicEq with M/S, but with a crossover and a few more features. It does NOT create artificial stereo (except for the delay thing). MStereoSpread and MStereoGenerator DO create artificial stereo, MSS is more powerful, MSG is more natural. MSS can be natural too, but you must not use extreme settings. Here's more info:

https://www.meldaproduction.com/guides/stereo

[astramistil]

''MStereoProcessor does pretty much the same thing as MAutoDynamicEq with M/S'' thanks exactly what I wanted to know

So if I understand well, I can use MSP on every instrument bus and get more control over stereo *without* artifacts (as I am only using 'widen' for each band, and sometimes exciter which I guess I could add after with MMSaturator),

but I could get even more control with MAutoDynaEq (because of curves)?
1-so should I use ''M+S'' mode and ''boost'' part of the spectrum I want using band M or S depending on my needs?
2-but I guess if I do that, i will have to use AGC? Or should I avoid it in that case?
3-Or would ''S'' mode would be easier for instance to get a tight Guitar (and I know it depends on the case but I have to give a concrete example):

Lows: HP cut
Mid-Lows: Shelf-cut
Mid-Highs: Boost a little
Hihgh: Boost more

4-Would this get me similar results to MSP (using only widen per-band), or is the crossover the magic thing happening here and *why*?

5-If you suggest me to try ''3-'', do you think MSpectralDynamics in ''S'' (or 'M+S'?) mode would deliver tighter results?

(dont worry Ive everything at least 10 times , theres no shortcut taken here; but for most of us experimenting, sometimes it can get overwhelming ''in practice'' when we thought a vst would act in a certain way while reading the doc but doesnt, so we need more insight in what the vst is doing)

Thanks for your patience and time!

[jmg8]

MAutoDynamicEQ (MADEQ) will be much better for this. Multiband is quite limited as all the bands have to share the same crossover type and you can only have 4 bands in MSP.
In MADEQ you can use the auto feature to measure one instruments frequency content and then remove it from anothers.
If you apply this to just the mid/side independently you will be able to quite accurately make room for each instrument in the stereo field as you described.

[astramistil]

Thanks

Ill try (I have never used AutoMode for defining which band to potentially sidechain, maybe because earlier version didnt convinced me (but last time I tried it was like version 5 four years ago!)

But I guess the ''dynamics'' feature would make all kind weird movement in the stereo (unless a very slow attack&release ), Ill check that too

-Do you think I should use ''S'' mode or ''M+S'' mode (with bands M or S accordingly)?
-AGC?
-Boost (acting as a D/W) or cut (to avoid phasing; I actually dont know if boosting bands in MADEQ add phasing problem like other EQ) ?

[steve2KVR]

...jng8...brilliant use of MADEQ!...nice, very very nice.../s~

[astramistil]

(I know its not polite to bump but:

jmg8 and Vojtech ,

(yes I know hours of trial-and-errors; but dont have all time I would want so if I can get advises/tips it can cut alot of wasted time)

1-Do you think ''dynamics'' can be used without de-balancing stereofield? (I tried with 2 guitars only sidechaining dynamically the ''Side'' part (with slow attack and release), but havent tried in full mix yet, and not sure if I should do this to certain instr. only etc).

So is ''instant *subtle* dynamic widening'' can cause later (or actual) problems in the overall stereofield? (my aim with M/S EQ is always to make 'space'/uncluttering, bring aliveness and get mostly a natural sound while adding depth)

2-Do you think I should use ''S'' mode or ''M+S'' mode (with bands set to M or S accordingly)?

3-The use for ''Mid zero Side'' or ''S zero M'' would be as an aux D/W I guess?

4-AGC or not? (cause I guess the aim of M/S EQ is to act as a per-band widening ''D/W'')?

5-Boosting or cut (getting equivalent widening results)? (to avoid phasing; I actually dont know if boosting bands in MADEQ add phasing problem like other EQ)

Thanks!

[jmg8]

1. I would definitely not use this technique dynamically. It will make instruments move around in the stereo field which will not sound natural. Just keep it as normal static EQ for M/S processing.

2. I would use M + S and then set your band accordingly. As my technique would be to boost and cut M/S at the same time (I will explain in more detail bellow)

3. Don't know what this is to be honest. I've never needed it. I'm sure it's pretty simple though. Just hover your mouse over it an click CTRL + H for info.

4. Yes I would definitely use AGC as your goal is not to change the level but to change it's stereo position. I almost always have AGC on, it helps to not trick you into thinking louder is better. Keep it on AGC while you are adjusting parameters and then turn it off and press SET when your done.

5. Boost or Cut. Both.
MADEQ is not linear phase so will cause phasing. This should be very minimal though so long as you don't go too much with it. I wouldn't boost more than about 4db personally.

Ok, so my technique would be the following:
Fist use multi analyser on all your tracks in the mix, try to identify which tracks belong where in the stereo field. First find the dominant ones that have a definite space and leave them there. There is not point in trying to force and instrument out of it's space too far as you will not get natural results. So for example a kick and vocal would be almost all mono where a pad might be almost all stereo. Once these main tracks have been identified look for other tracks that are fighting for the same space. Ask yourself if that competing sound would be better wider or narrower? Every mix is different, but when you have a good idea where everything will go, do the following.
Let's say that your bass and kick want the same space in the stereo field. But it feels like the bass would sound better wider and the kick narrower. Open MADEQ and set to M + S mode and turn AGC on. Put on kick and analyse source, move to bass and analyse target. Click the not sound like button. Now reduce the dry/wet until the max amount of boost/cut is around 4db. Now right click all the bands and change them to MID. Now do the exact opposite for the kick. Which is to reduce the SIDE content of frequencies only present in the bass.
That's it. Don't worry to much wether you cut or boost mid or side it will depend on the mix and where there is space for things to fit. Just don't over do it or you may get phasing. It's a subtle difference that will have a cumulative effect on the mix and make a noticeable difference. The stereo meter will be your best friend here, but in the end trust your ears.

[astramistil]

Wow mega thanks!

There should definitively be a complete tutorial from Melda covering that technique (for mixing a full band song). From what Ive heard (tutorial from Brainworx) and tried, I think M/S Eqing makes all the difference between a bedroom mix and a professional mix in terms of ''aliveness'', breathing and uncluttering of the lows.