New 20 band stereo width method

[wlangfor@uoguelph.ca]

Hi all, it's hard to always come up with new methods of affecting stereo width like the pros do it, but here is a 4 band EQ I have been working on that only uses 2% cpu.

It has the option of linking the above and below sliders so that if you add a peak at the top, it takes the same away from the bottom. And the reason is to keep it at unity to a degree while still allowing the gain and the width effect.

It's a little hard to get your head around what it is doing, but it is a good showcase of a width filter that does not work the same way that most do. And the other strength of it is, is that if you are to set the top to mono, bottom to stereo etc; you can basically be setting the amount of mix, so because one is split stereo and the other is mono, they make eachother a lesser width together. It's an interesting idea, and like oversampling.

So, you get the oversampling and width effect all in one.

vstplug excalibur v0010.fsm

(91.86 KiB) Downloaded 868 times

WARNING: LOUD EFFECT, USE SLIDERS IN A MINIMAL, CAREFUL WAY!!!

here's some images to describe the effect better:

detailing the effect with arrows:


screenshot:


so, I was thinking about it, and you could do almost anything with this same effect and perhaps with different filters. I guess My only concern with that idea is the damage to the audio; it's rare that certain filters do none.
But, you could always try MV's ZDF filters (link below), they're in a schematic called transferfunction demo.
http://dsprobotics.com/support/viewtopic.php?f=2&t=13964&p=46469&hilit=transfer+function#p46469

thanks to Azurstudio for the knobman knob designs btw.

[wlangfor@uoguelph.ca]

So, this was a prototype that has mixed results but what is interesting is that it reduces phase cancellation.

Because of mono occuring first, and then being mixed in; it allows a different style of sound, where technicaly both realms of both mono and stereo exist simultaneously. And this, if you had inspected this schematic as you might expect that the stereo in for the left and right was halfed so as to stay closer to unity.
Furthermore, in the newer version which I am happy with it now uses 5 db range.

A pretty unique solution I thought and probably with the ZDF it will use less CPU. So, assuming that the colour could be made shinier maybe, this could be one of the better commercial methods.

[MichaelBenjamin]

.

[wlangfor@uoguelph.ca]

i see now, this is a prototype that led to silverhawk.

yeah, originally the main purpose was to show that EQ peaking filters use less CPU than crossovers and the sound is basically more intact. If you were to think about it; in the case of being only a width filter; the knobs could be replaced by an algorithm that merely were set to a curve of your choosing; that or may not for instance even be dynamic; reacting by means of a correlator.

I guess where My curiosity had been piqued was in the case of the Soundspot MSW2, which uses dynamics and crossover to affect width. Yet this has forty bands and uses the same amount of CPU. You see My meaning, we see peaking filters; but are they instead a way of replacing crossover to achieve the same ends. And because of the ability to gain or reduce it's possible to combine them in the algorithms to both and remove width in any given specific order or fashion.

And its that functionality that could possibly change the face of width enhancers, it could also potentially make them use less CPU.

[MichaelBenjamin]

.

[wlangfor@uoguelph.ca]

i see, but RBJ style peaking filters will introduce lots of phase difference, unlike phase-agnostic approaches like in some bandsplitting linkwitz-riley approaches (correct me if i'm wrong assuming that linkwitz-riley style desplit/band_amp/and re_merge doesn't affect phase).

Yes, you're right; definitely some phase cancellation. I thought it was tolerable; but the algorithm would need to be very precise.

[MichaelBenjamin]

.

[wlangfor@uoguelph.ca]

you know where phase coherence is important?
in the bass range
if you get phase cancellations under about 99 hz it will fuckup the bass response in unpredictable ways
that's why it is so important to keep the bass range in mono, bcs human hearing cannot percieve stereo bass at all.
it is also the reason why subwoofers are mono only.

you can however make it an artistic effect, it will lead to much "wobbling"

heh, you're preaching about bass what I often preach, yet choose to forget all too often lol.