DIY synth bass?

[David]

Way back when while I was still doing bar gigs as a bassist, I was really jealous of keyboard players and the ridiculous bass they could get from their synths.  My poor little bass couldn't come close to that fatness.  With the advent of pedals like the Deep Impact, perhaps the tide is turning a little bit.  I'm wondering if there might be a DIY solution to this also.  I'd like to find out if it would be possible to build a circuit to:

Take my bass input note
Generate the same note on another "channel"
Take the generated note and flat it (or sharp it) about 3% or so (to do pitch-shift doubling)
Mix it all back together
Send to the output

I've done a little research on this forum and in other places, and it looks like a phase-locked loop could help with the "generate the same note" part.  Does anyone have any ideas on how to do the "pitch shift"?  Is this idea even worth pursuing or is there maybe an easier way to do it?

[Mark Hammer]

What you suggest is probably more workable with bass than guitar from a tonal standpoint, since a well-filtered oscillator output could seem more like a second bass than it could seem like a second guitar.

From another standpoint, though, bass is problematic since it takes a little while to detect a bass note than a guitar note, based on the time between successive peaks.  That delay may be disruptive to your playing or it may not.  What may be more disruptive is the fact that it would be an exclusively mono system.  Although bass players may generally *pick* one note at a time, that doesn't mean that note N is automatically killed the moment note N+1 is picked.  So there is a stylistic and usability issue with any sort of simple pitch detection system, despite the fact that the tone produced might be entirely acceptable.

There was an interesting article by Bernie Hutchins in Journal of the Audio Engineering Society in 1981 I think, on achieving a "fat" sound with synths.  You will remember that some of the earliest polyphonic synths felt sort of bland, and it was a common complaint amongst players that they just seemed lifeless in comparison to acoustic instruments.  Hutchins explored how much modulation was needed to achieve a "fat" sound, and concluded that something like 3-5 concurrent sources (It's over 20 years since I read the paper so I may have the numbers wrong, but it certainly was not as many as 6 and was clearly more than 1) of modulation would achieve a pleasing tone that felt like it had some liveliness to it.  Three to five sources does not mean that they are all going full crank.  It may mean something like subtle slow changes in pulse-width modulation at the same time as note-wise filter and amplitude modulation with a wee bit of chorus.  In the case of bass, a bit of chorus over top of a fretless drone achieves that in spades.

Maybe you need to spell out what you mean by "fat" a little more so that some suggestions can be prompted.

[David]

OK, 2 examples:

1.  The synth bass on Gary Wright's "Love is Alive"

2.  The synth bass on Donna Summer's "On the Radio"

[Mark Hammer]

OK, clearer picture now.

One possible strategy is this.  Run the bass to a splitter.  The splitter goes to a mixer and to an envelope-controlled filter.  The envelope-controlled filter goes to a chorus and finally to the mixer.

Basses can often sound meatier with a bit of animation, rather than with more in-your-face effects.  That's why the gist of this suggestion is to combine a straight and effect signal together.  Things like flangers, Univibes and chorusses can also sound great with bass if you chop off a big portion of the low end from the wet/effect channel of the pedal.  This way, the effect is essentially acting on the mids and highs but leaving the fundamentals intact.

I don't know if you've used it before, but the MXR Envelope Filter has a variable attack time.  It can also be modified (see the A-Gua project at Tonepad) to have a variable Q or filter emphasis.  With slow attack and minimal filter emphasis it seems to do a startling fretless bass "drone" simulation of the way a note blooms shorting after picking.  Add a wee bit of fuzz in front to give it more harmonic content to bloom to, tack a chorus on the output to thicken it up and you have a very rich fat sound.

I suspect a big part of the great synth bass sounds you refer to is the capacity to hard-sync oscillators together such that the onset of one square wave an octave above or below resets the other oscillator (or 2 if it's a bigger machine).  That provides a sort of pulse-width modulation that is relatively aperiodic.  Tim Escobedo may have some useful circuits for you.  Check out his folkurban circuitsnippets site.

[David]

One possible strategy is this.  Run the bass to a splitter.  The splitter goes to a mixer and to an envelope-controlled filter.  The envelope-controlled filter goes to a chorus and finally to the mixer..

Thanks for the tip, Mark!  Much easier for me to understand than a PLL!

Basses can often sound meatier with a bit of animation, rather than with more in-your-face effects.  That's why the gist of this suggestion is to combine a straight and effect signal together.  Things like flangers, Univibes and chorusses can also sound great with bass if you chop off a big portion of the low end from the wet/effect channel of the pedal.  This way, the effect is essentially acting on the mids and highs but leaving the fundamentals intact...

I don't know if you've used it before, but the MXR Envelope Filter has a variable attack time.  It can also be modified (see the A-Gua project at Tonepad) to have a variable Q or filter emphasis.  With slow attack and minimal filter emphasis it seems to do a startling fretless bass "drone" simulation of the way a note blooms shorting after picking.  Add a wee bit of fuzz in front to give it more harmonic content to bloom to, tack a chorus on the output to thicken it up and you have a very rich fat sound.

So how would a Zombie do with bass frequencies?

I suspect a big part of the great synth bass sounds you refer to is the capacity to hard-sync oscillators together such that the onset of one square wave an octave above or below resets the other oscillator (or 2 if it's a bigger machine).  That provides a sort of pulse-width modulation that is relatively aperiodic.  Tim Escobedo may have some useful circuits for you.  Check out his folkurban circuitsnippets site.

I'll take a look.  Thanks!

[Nasse]

:? Pure sine wave at bass freq is sometimes what several people think as "deep". But is it "fat" sounding :wink: ? Maybe using strings tuned higher than normal bass and then tracking the pitch with PLL would be easier... but really dunno, I have not done the mathematics, maybe it is impossible. One thing I know that some of my oldish cheap synth modules can not do the deepest bass, no no no...

I have been thinkin some schems made with cmos logig shift registers or counters that converts freq fed in to sinewave (output freq is input freq divided 50 times or something)...

My eldest daughter got together with a bass player, and the guy sort of gave me his bass guitar (well I gave him money... but the bass is "in the family" now) it is a very nice Tokai Precision copy. Maybe some time I´ll learn some bass lines...

Envelope controlled resonant lowpass filter might get you into synth territory, but you can not play notes over wide range without some filter tracking...

:oops: I have this wife and kids and mortgage and day job and I´m too old but maybe those 18" reggae bass style subwoofers might be cool with a Precision copy... :mrgreen: and some local surf bands use Ampeg SVT (lotts of bottles) with 6x10 cab

[Mark Hammer]

How would a Zombie do?  Well, once you get the ticking problem beat, you need to "thin out" the delay signal.

There is a 1uf cap on the output to get rid of DC in the wet+dry signal before the next pedal in line, but nothing else that would limit low end between the input op-amp (where the bias voltage for the BBD is introduced) and the point where wet and dry are mixed.

You certainly can't introduce the bias voltage the way that John Hollis does here and limit low-frequency content going into the BBD (which would have been better for avoiding distortion issues, but it's too late now), so to restrict the wet signal so that you have primarily dry signal for the low end, you need to add in some simple highpass filtering after the BBD but before the mixing point.  I'm going to suggest that you try to insert a .022uf cap between the output of the second op-amp and R10 (a 10k resistor).  This should create a low end rolloff around 720hz, which will leave the fundamental of most notes undisturbed, but leave some "swirl" in the mids and highs.  The simplest way to do it is to lift the op-amp side of R10 and install one leg of the cap there, soldering together the two free leads of the cap and resistor.

[David]

Mark:

Thanks so much for the ideas.  I was almost sure there was something I had missed that would sink this boat before it even got away from the dock.

Now, what AM I gonna call this thing?  The synthetic synth?  Naahhh...

[Mark Hammer]

Well if it's REALLY fat, you could say its collagenic, or cell-u-light.  Stick it in a box and call it "The Luv Handle".  Or, in honour of David St. Hubbins' lyric to "Big Bottom": The Mud Flap. :lol:

Enough with the plumpness jokes.  I hope it does what you want.

[moosapotamus]

What would a bass synth be without attack/decay/sustain controls?

http://omega.tellus.vallentuna.se/anders/pdf/envelope.pdf

~ Charlie