Sing Wah - any success?

[Dave_B]

I'm fairly new in the community, so the Sing Wah (on Geofex) is new for me also.  After searching through the archives, I couldn't find any posts mentioning success after building it.  

Out of curiosity, have any current posters tried to build at least part of this and tune it to the frequencies R.G. listed?  My guess is that the problems stem from component tolerences throwing off the filter tuning, but that's just a guess.  

FWIW, I'm just striking up conversation.  As cool as the Sing Wah appears, I don't think it will be on my breadboard until long after I've built some other stuff.

[Ry]

I have built it (you probably noticed my posts in the archives about it).  I still have it on a breadboard.  It doesn't quite do what I expected it to.  I've tried tuning everything as closely as possible to the correct components and it still doesn't sound much like a human voice.

I know that some people were putting envelope filters in parallel to try to get a more vocal sounding effect.  I don't recall how those attempts turned out.

Ry

[spudulike]

I know that some people were putting envelope filters in parallel to try to get a more vocal sounding effect.  I don't recall how those attempts turned out.

Ry

Search for FrankenQuack (IIRC)

[edit] Found the thread here http://www.diystompboxes.com/sboxforum/viewtopic.php?t=23334&highlight=frankenquack

[Ry]

That's the thread I was thinking of.  Very cool!  Thanks for the link.

Ry

[Dave_B]

I have built it (you probably noticed my posts in the archives about it).  I still have it on a breadboard.  It doesn't quite do what I expected it to.  I've tried tuning everything as closely as possible to the correct components and it still doesn't sound much like a human voice.

Yes, those were your posts alright.    From what I'm reading, it looks like R.G. designs a lot of things that may not always get prototyped.  I'm not sure if this is one of them or not, but I can't find any mp3's that demonstrate the thing.  

Regarding the circuit itself, I was looking through Lancaster's filter cookbook yesterday.  I don't pretend to know a lot about bandpass circuit design, but I'm trying to understand why the EQ filters couldn't use different resistors to help keep the capacitor sizes managable.

[Ry]

This is one of those things that has been rattling around in my skull for a couple weeks now.  It's interesting that you posted this at the same time that I've been thinking about it.

Here's my totally untested and not-very-well-thought-through idea for why this might not sound much like a voice:

The seek wah has discrete steps from formant to formant, rather than slurring one to the next.  Human speech tends to slide from one sound to another, rather than moving in discrete steps...so maybe if we can smooth out the sequencer steps with some sort of envelope driving both formants, we might get what we're after.

I forget why RG chose different value caps over resistors, but it was probably just a design choice.  The Seek Wah has variable resistances for the filters, maybe he didn't want to appear to be too similar in his design.

RG has loads of great ideas and, like the rest of us, limited time to perfect them.  I don't know if he's ever tried to get this one to work himself.  My understanding at the time this design was posted was that he hadn't breadboarded it.