Author Options:

Do I need to invert phase a second time? Answered

I'm building a guitar pedal and I am inverting phase with the initial buffer (TL072).

I am then splitting the signal into two parts, dry (no effect) and wet (effect). Each of these will be output as separate channels (left and right) and not mixed back to mono.

I plan to send each channel through its own 386 amplifier chip to boost it to a level suitable for headphones.

So, what I am wondering, is, should I use the 386 to invert the phase a second time back to normal? Or should I use an op amp at unity gain to do that before I send it to the 386? Or does it not matter if I am not mixing the two signals back to mono?



Best Answer 8 years ago

Perhaps I'm more confused at why you're inverting the phase in the first place if phase reversal is the only thing this pedal does.  If you were to perform or record with an effect like this, you will end up with all sorts of unpredictable effects - none of which you'd find useful or desirable.  If you invert the right channel a second time, then no net change has taken place and your guitar will sound exactly the same.

Imagine you're in your living room with a stereo.  It's a typical setup, with two speakers, they're aimed properly, and you're sitting in the "sweet spot".  Your guitar effect is playing on the stereo, and you notice that it's much quieter in the mix as before, and the tone is very different from how you originally recorded it.  The reason for this is due to the fact that you are experiencing partial cancellation.  The direct sound from the speakers is exactly 180 degrees out-of-phase, and therefore cancels completely.  The remaining sound (the only thing you're able to hear) is reflections off of surfaces in the room, and depending on the arrival times of each wave, the direction it comes from, and the material it bounces off of, you'll hear absorption and cancellation at certain frequencies which will differ in every position in the room.  It will inevitably sound as if your guitar is being sucked out of the mix.

Now imagine you're in your living room with a surround sound.  Again, a typical setup, with five speakers and a sub, and the receiver is set to Dolby Pro Logic.  If the guitar is mixed equally left and right, your guitar can either end up dead center in the rear speakers, muddied up with the reverb from the rest of the mix (reverb has scrambled phase elements, which are interpreted by Pro Logic as rear channels) or canceled out entirely.  If there is any difference in your guitar's volume from left to right, this can be interpreted by the receiver in several ways depending on how strict it's implementation of Pro Logic; it could end up panning to some channel you didn't intend, or it could end up canceling out altogether.  Keep in mind that, unlike the previous example, the cancellation is performed electronically in the receiver before it hits the speakers - this means that if there is any cancellation, you will not hear your guitar at all.

In a live setup, the first example will apply.  If you're using headphones, this can be perceived differently depending on the listener; some will hear it as spaciousness, others will hear it as a sound emanating from inside their head (which sounds novel at first, but becomes fatiguing in a hurry).

Don't get me wrong - I'm all for creativity, incidental music, and "noise art".  However, since there aren't many of the above scenarios that are even remotely pleasant, I wouldn't recommend designing a pedal that achieves an effect like this.


8 years ago

I tend to think it might matter.

I accidentally wired two speakers out-of-phase (one was unmarked) in a 2x10 cab, and there was a definitely auditory difference. Actually a pleasant change at lower volumes, but the interaction definitely effected the sound.

The effect was NOT pleasant at higher volume--There was loud resonance/reinforcement in some note ranges, etc.

While headphones are different--the actuators aren't in the same cabinet--your head does have open cavities, etc., that would mix the two signals together. Whether or not it sounds good or bad...who knows?

Might be a fun experiment.

And clearly even separate speaker cabs would interact, depending on where you were standing.

BTW, feeding an inverted signal into a speaker should sound identical to your ears, being that it's only a half-cycle shift. But in practice most people find that there's a right way and a wrong way to hook up their speakers--I suspect that the movement of a voice coil isn't symetrical in each direction...


8 years ago

Why do you need to invert it a second time? Our ears can't tell the difference if it's inverted, it'll only matter if you're adding together an inverted and non-inverted signal at some point as they'll partially/completely cancel out.....


Answer 8 years ago

.  Inverting the signal to a one speaker system is not noticeable. Inverting one side of a stereo system (or mono with multiple speakers) is usually noticeable, especially during an A-B comparison.
Speaker phasing at Google.


8 years ago

There is no need at all to worry.
If only one of the signals was inverted, you'd get a funky effect much like what Re-design and Jayefuu mentioned.
Having wet and dry in left and right channels will be an odd effect, and I'd imagine a rather cool one if they are balanced well!


8 years ago

I've never run across this question before.

FWIK and from what I could research, you should get the phases back into sync.  With each ear processing the sound but in 180 deg. phase difference your mind will notice something different.  You may hear the true quality of the music but it may seem that the music is coming from another direction rather than in front of you.  We detect direction of sound by comparing the volume of the sound but more importantly by comparing it's arrival at each ear.

Id' be interested in hearing your test results by trying it both ways using scientific method and on several subjects.

It would make a nice Science Fair project.