Effect like a scratched CD?

[Hambo]

I used the search and I know this has got to be a processed effect added in after recording, but just on the off chance, does anyone have an idea how to replicate it?

Its like the hold function on a digital delay I suppose, but really really fast, faster even than a real scratched CD I think.. thinking of some of the tracks on Melt Banana - Cellscape 37 in particular.

[Meanderthal]

 Uhh stuttering pedal?

[Hambo]

I have something like that, not quite the ticket.

More like it takes the same fraction of a second in a chord and plays it back over at incredibly fast rate, whilst completely cutting the real time guitar signal - sounds totally unnatural. I heard it also on Peeping Tom - We're not alone, Used to massive effect just before a change.

[Meanderthal]

 Hmmm... not too long ago there was a circuit posted by choklitlove called the Hysterysis Oscillator that was more or less a 555 based extremely choppy tremolo that... got really really fast. Very interesting, could do ring modulator sounding stuff, etc.

But that probably ain't it either...

I'm afraid I'm not familiar with the musical references, although I'm sure it's interesting stuff...

[Sam]

Uhh stuttering pedal?

I get the impression that Hambo is looking for a way to construct an effect that somewhat resembles the hold function on the DD-3 but with an extremely short loop. Like a suttering CD - chopping time, not signal strength...

BTW, Melt Banana rocks!

[R.G.]

Not too hard.

All's you gots to do is have a digital delay with some control logic that normally doesn't get mixed with the dry sound at all, and when you press a button it recycles the last N milliseconds of audio forever and puts that out on the output instead of the dry sound.

Piece of cake. Conceptually anyway. A bit of hacking on a digital delay pedal should do it.

Analog delays will lose quality with each recirculation and fade into noise.

[Sam]

Not too hard.

All's you gots to do is have a digital delay with some control logic that normally doesn't get mixed with the dry sound at all, and when you press a button it recycles the last N milliseconds of audio forever and puts that out on the output instead of the dry sound.

Piece of cake. Conceptually anyway. A bit of hacking on a digital delay pedal should do it.

Analog delays will lose quality with each recirculation and fade into noise.

That's it.

[Hambo]

Hehe Meanderthal, When I said I had something like that, it was exactly that hysteresis oscillator, another fun toy. And yes I have been reading and potching for a while. 

Yes RG, that seems like the idea, I'm going to have to learn a LOT more about how these things work first I suspose, and actually get a cheapo digidelay at some point. I circuit bent my analog delay and got pretty much what you describe there, its like an angry robot noise, which is fairly wicked in itself. Unfortunately my knowledge doesnt extend very far into the actual workings of these things.. yet.

[Processaurus]

As the meanderthal said, the Boss DD2 and DD3 do it, unfortunately it holds and loops what you were playing just before you hold the button down, which seems counterintuitive.  With the aim to use it rhythmically, rather than just a weird sound, its difficult, because to get it on rhythm you need a sense of how fast the delay will be and then play that increment of time before hitting the switch, then hit switch hopefully in time, if its a fast delay its hard to get it as solid and naturally as you could get it if, say you were able to play and hit it in tempo, simultaneously. 

Maybe an electronic switch that switches the feedback and mix like RG suggested, but if it counted the length of the current delay buffer after the mechanical switch was hit before doing its thing, and switched back right when you let up on the switch.

[R.G.]

Unfortunately my knowledge doesnt extend very far into the actual workings of these things.. yet.

That's the right attitude! Go get it!!

[Processaurus]

Here's a hack i'm going to try on a dd2 someone loaned me (what, is that rude?):

Found this schematic for the dd-2/dd-3 (same circuit by the way, name was a marketing thing because they found a way to make the dd-2 much cheaper but didn't want dd-2 buyers to lose face).

You could get the hold with the shorter times by hacking the 2p4t mode switch (fast, med, slow, hold).  If you could find a tiny 2p6t and kluge it in there that would be amazing and pro, but you could easily add a sub-miniture spdt center off toggle switch to make the hold use not only the long setting (200ms-800ms) but the other two.  The only quirk is the fastest setting would be in the middle.

look at the bottom half of the switch on the scheme, the s/m/f positions go to address pins on the controller chip (the BIG 64 pin one) to determine which of the three settings will take effect.  Theres a little table on the scheme that shows whats going on with the address pins and the modes (thanks boss!).  The hold position goes somewhere in addition, but also via diode D9 goes to the long setting's address.  So disconnect D9 anode, and hook that to a throw on the SPDT center off toggle, hook the toggle's common up to where you just disconnected D9 from.  Then make a new diode  go from the free throw (anode end) on the toggle to the cathode of D11 (cathode end).  That should do 'er.  Fix 'er right up.

All this does is make it so the other addresses combinations can happen when the hold mode is activated.  Both address pins high, its long, one of them high, its medium, and when none are high (when the toggle is in the center, and voltage goes through neither diode) its the short setting.

Just heard the Melt Banana guy's solo/noise CD by total coincidence yesterday, he uses the skip devastatingly I must say.

[Processaurus]

now for a question, I see the VCO around IC 8 in the upper right corner, that makes the master clock of 1.28MHz to 5.18MHz from the CV from the delay time pot.  Would it be possible with 40xx series logic like shift registers and/or 555 timers to produce a semi precise delay on the footswitch like i was rambling about before?  use a set/reset flip flop whos output is inserted to where the footswitch went before, the logic sets it, when you let up on the footswitch it resets it.  A perfect job for a uC, I know...

Maybe it would be easier to read the CV from the delay time pot and calibrate a less precise switching delay from that?  Or learn to live with it? Or buy something else?

[Hambo]

Cripes Proc!

I dont have the faintest idea what youre on about in that second post.
However, the Boss you're saying that pins 56-57 and 54-55 being connected at the switch generate the GENERAL delay times in the region 50-200ms or 200-800ms, the S isnt connected to anything so it defaults to the 12.5-50? the actual delay is set by a range pot working within the scope of each of these.

But the hold function is different isnt it? no pins are connected when the switch is set to H and yet the table says its 200-800ms, its set by pressing the footswitch twice then it repeats everything in between the two presses if I'm thinking it right, indeed the top half of that switch on H setting does go over to an almost direct connection to the footswitch. So what does that 200-800ms actually mean in relation to HOLD?
What I mean is, and I'm having trouble expressing it - the HOLD function is not so much a function of the delay effect in the box, its like a separate thing and you would need to be able to stomp the switch twice in the time of a heartbeat to make it work like that.

I wouldnt have any qualms about ripping a bossDD2/3 apart, and I do need a decent digital delay, its just a matter of time before I can afford to buy one on top of all the other stuff I have going at the moment.

But its worth a try, glad u heard agata btw, the guys pretty amazing in my book and that effect really makes me instantly wet my pants!

[slacker]

But the hold function is different isnt it? no pins are connected when the switch is set to H and yet the table says its 200-800ms, its set by pressing the footswitch twice then it repeats everything in between the two presses if I'm thinking it right,

No the hold works like this, you set the sampling time with the D.Time pot which gives anywhere between 200ms and 800ms. Then you press the stomp switch down and it plays back however many milli seconds you set with the D.Time pot for as long as you hold the stomp switch down. When you release the stomp switch it stops playing back. At the fastest setting it sounds a bit like a CD skipping, but a mod to make it faster would be cool.

I was bored, so here's a sound clip using hold with the fastest setting http://www.eskimo.plus.com/fxstuff/cdskip.mp3

[Hambo]

Ah I must be thinking of something else..
Starting to look very promising.

So can you still play stuff underneath while its holding?

[slacker]

Yeah you can play over the top while it's holding. You can also mess about with the delay time whilst it's playing back, to speed it up or slow it down. That what I did at the end of that clip.
If you're quick enough you can also release the stomp and quickly press it again and it will grab what you just played and layer that on top of what it was already holding.

[Processaurus]

However, the Boss you're saying that pins 56-57 and 54-55 being connected at the switch generate the GENERAL delay times in the region 50-200ms or 200-800ms, the S isnt connected to anything so it defaults to the 12.5-50? the actual delay is set by a range pot working within the scope of each of these.

But the hold function is different isnt it? no pins are connected when the switch is set to H and yet the table says its 200-800ms,

When the switch is set to hold, +9 goes through D9 (like a one way street) to the long setting throw on the switch, which then goes through two respective diodes to the two address pins to get the long setting.

I wonder if the DD3 doesn't use the analog feedback path with the hold, if it tells the chip to stop sampling and just play back what's in the buffer forever.  it seems like doesn't degrade over time (or is it just that clean?)..

The DD3 is a great one to hack, because it makes sense like an analog pedal, with an analog feedback path, and an analog  control voltage to set the delay time.  There's some thing pete cornish makes that is a DD3 lashed together with some other stuff in a giant box, its funny to see the telltale 4 tiny knobs huddled together on this plateau of an enclosure.

[boyersdad]

Hambo, I really dig your musical taste! I'm hardcore into the Patton/Kang/Spruance/MeltB/Secret Chiefs/Estradasphere thing and know what you're talking about.

Almost more so that figuring out how we the underclass can do it, I wonder how they do it! Especially live, like MB do it! WTF? They are insane, and I want to learn it all

[slacker]

I wonder if the DD3 doesn't use the analog feedback path with the hold, if it tells the chip to stop sampling and just play back what's in the buffer forever.  it seems like doesn't degrade over time (or is it just that clean?)..

I'm pretty sure it just plays back whatever is in the buffer repeatedly,  I've left mine running for minutes at a time and the signal doesn't degrade at all. The feedback pot doesn't do anything in hold mode either.

[Processaurus]

Hey gents, this mod worked!  I hacked my friends dd2 (without total authorization, exactly  ), its pretty cool, you can switch back and forth between different speeds on the hold while its working, and it has enough in the buffer for the long setting, and plays back shorter sections of that with the other two speeds, but when you go back to the long setting the sample is exactly like you left it.  So the slacker was right about there being no analog feedback in the hold mode.

It was difficult finding the correct diodes on the main board.  Its such a jungle in there.  This was a very time intensive/complex  pedal for boss to make, lots of hand wiring, the components are very close together and standing vertically, there's a funky foil shield inside, etc.  The easier way to find the diodes is to follow the wires from the mode switch, to their number on the board.    A full size DPDT (no SPDTs left) fit above the effect level knob.

The only drawback is its natural to want to fiddle with the switch and the delay time while its holding.  It needs to duct taped to a guitar, I think.

As a side project I'd been wanting to do to a delay I stuck an effects loop in the regeneration path while the dd2 was open, it was tricky fitting an extra jack in there, a switchcraft enclosed one won't work, its too thick, with it the pcb can't fit in the box.  Confidently I could maintain one of those won't fit anywhere in the pedal, not even the battery compartment.  I ended up using one of these (with a NC tip connection to route the feedback normally when there isn't a plug in it), which was slim enough to fit in front of the mono output jack, and clear the PCB.  I haven't got to try that part yet, but I'm sure it will sound plenty psychedelic.