Real World Interfaces logo

In June 2012, the subject of TB-303 Slide and Gate times has arisen on the Synth DIY mailing list:

Perhaps some of what I wrote below is wrong.  I plan to review the discussion and update this page accordingly.

- Robin  2012-06-10


Investigating the Slide function of the TB-303

A few years ago I investigated exactly how the TB-303's sequencer worked
when sliding notes.  I wrote this to a friend and Devil Fish fiend Rob Joyner.
Rob passed it on to the Analogue Heaven mailing list.

This was in response to persistent rumours that the sequencer anticipates the
next note and starts sliding before that note starts.

Robin Whittle 31 May 1999

Back to the main Devil Fish page.

Lets say you have 16 1/16 th notes programmed.  Each note occupies 
six clock pulses - lets call them pulses 0 to 5.

The gate goes high at the start of clock pulse 0 - the positive edge.

The gate goes low half-way through clock pulse 3 - on its negative 

That is for a normal, 1/16 note.  So there are 3.5 clock pulses on 
and 2.5 off.

If there is a slide programmed on that note, then the slide circuit 
is actually activated on the *next* note. The slide starts at 
the start of clock-pulse 0 and ends at the end of clock-pulse 5.

The other function of slide is that the note it is programmed on does 
not turn off, but the gate stays high and runs into the next.

Two diagrams.  They are wide - so don't let the lines wrap in your 
email reader.  Also, of course, you need courier text, not some 
propotional spaced font.

Firstly, three 1/16 notes. no slide:

Clock edges:   / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \

Clock number:  0   1   2   3   4   5   0   1   2   3   4   5   0   1   2   3   4   5   0     

Note boundary: |                       |                       |                       |   


Gate:        __==============__________==============__________===============_________====  

Now with slide programmed on note B, which ties it to note C.
Note B plays at it's normal pitch, but it keeps playing. 
Then at note C's starting time, there is no new note - as
far as the gate is concerned - because the gate is just 
staying on.  However it is a new note as far as the CV is
concerned - because it changes to whatever CV is programmed
for note C.

Clock edges:   / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \ / \

Clock number:  0   1   2   3   4   5   0   1   2   3   4   5   0   1   2   3   4   5   0     

Note boundary: |                       |                       |                       |   


Gate:        __==============__________=======================================_________====  

Slide        __________________________________________________========================____

So the CV seen by the VCO - which is the output of the slide circuit - 
is sliding from pitch B towards pitch C, during note C. We don't
hear the last part of this - since the gate is off the volume drops
to nothing very quickly - unless it is an accented note, in which 
case the Main EG drives the VCA as well.  In this case, in the TB-303
the Main EG has a fixed short time.  In the Devil Fish, this time is
variable and can be quite long.

Now, back to the slide system. What happens if the note B is more 
than a 1/16 note?  Then we look at what happens of the note following
the slid note is longer than 1/16 note.

Investigation reveals the algorithm:

1 - The slide signal will go high 1/16 note after the start of 
    slid note. This includes when the slid note is 1/16 note long, 
    as in the above example.

So if the slid note is 2/16 long, then the slide actually goes
active half-way through it - though I don't think it would make
any difference, since the slide circuit is already putting out 
the same voltage as the CV it gets from the DAC.

2 - The slide signal will remain active until the end of the note
    which follows the slid note.  This includes if it is 1, 2 or 
    more 1/16 ths notes long.

As far as I can tell, thats it.

Now, while we are at it, lets kill another rumour that wastes time:

   "The TR-808's internal sequencer has its own special groove."

This is not true.  The TR-808 plays in perfect time with the clock 
pulses.  Thats all there is to it.

You wrote:

> Someone was telling me that they've been 
> trying to convince you to do some live performace work but I can't 
> remember who it was. 

Me either!

> I've heared it said on this list b4 that magic elves inhabit the 
> workings of these little boxes? 

The *synthesiser* of the TB-303 certainly has its magic little elves - 
and so do lots of people, elves in their ears, bats in their belfry . . 
Best to direct that energy into generating music, not into analysis.

Karl Kraus wrote:

    Art is light synthesis. 
    Science is spectral analysis.

I am happy to do the science to support the art - as long as I get
to do the art too.  People with ghosts in their gizzards, fairies in
their fingers and all the rest are to be cherished and gently 
encouraged to keep playing with art - and not to get too distracted
by the science.

However, who am I to say what is going on with their TB-303?  I have
only looked at mine.  For all I know, a well cared for TB-303 does
develop its own groove and does manifest precognition of notes yet
to be slid.  

Please pass this on.

Return to the main Devil Fish page.