[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Date Index][
Thread Index][
Author Index]
Foot controller for a dream looper
The Looperlative stares at me saying "use me". The Walker brothers keep
encouraging me. I look at foot controllers trying to decide what would
match my use pattern. At long last, I think I can map out what I would
like to be able to do. Now the question becomes, how close will the
Looperlative actually let me get and is there a foot controller that will
help me get closer (or even be capable of doing it). As a starting point,
I know I've got issues because my ideal control configuration needs to
know more status information about the Looperlative than is available to a
controller or than is available to commands on the Looperlative, so I
would need to model the internal logic from the Looperlative and hope that
the model was correct. I also need to distinguish between short and long
presses. So, it's a long way to go. But I figured I would post this and
see what reaction it gets.
My current thinking is centered around a 5 x 3 button matrix similar to
the Line6 M13. It could reasonably be successfully revised to a 4 x 4
matrix.
The basic configuration is to run the eight loops as four pairs though
much of this work work with other groupings like 8 independent loops or
two groups of four or various mixtures. Basically assume the loops are in
groups. A group shares settings for reverse, half-speed, and scramble. At
most one loop at a time within a group plays. All of the loops in a group
share a cycle length but they may consist of different numbers of cycles.
This is supported via eight buttons -- probably arranged as two rows of
four -- corresponding to the eight loops. Tap a button and if that loop
isn't empty and isn't playing, it starts playing and stops playing any
other loop from the group (probably quantized to the cycle boundary for
the loop pair or even to the loop boundary for the currently playing
loop). Tap a button for a loop that is playing and it stops the loop
playing.
Tap and hold a loop select button and it does the same as above but it
copies the other loop in the pair into this loop. (Alternatively, it could
copy this loop into other one in the pair. Either way, the idea is that
the pair now contains two matching loops so we can go manipulate one and
then come back to the original.)
Whichever loop select button is tapped last becomes the target loop and
its group becomes the target group.
This set up alone now puts the player (me) in a position to mix up to four
different parts with two different but potentially related versions of
each part.
Now we turn to the action buttons:
* Sync Rec/Dub +++: This button performs a number of duties depending on
the status of the loop.
If the loop isn't playing, then tapping the button goes into
record mode. If the rest of the group isn't empty, it syncs to the cycle
length for the group.
If the loop isn't playing and the group is empty, it syncs to the
master cycle length (set with the first recorded loop).
If the loop isn't playing and the group is empty, then a long
press triggers record on the down but disables syncing.
If the loop is playing, then tapping the button goes into overdub
mode.
If the loop is playing, then a long press of the button undoes or
redoes the last overdub on this loop. As a result overdub doesn't engage
for a tap until the button is released.
If the loop is recording, then tapping the button switches into
overdub -- ideally at 100% feedback (see EDP safe mode).
If the loop is recording, then a long press deletes the recording.
This is not undoable.
If the loop is overdubbing, then a tap exits back to play mode.
If the loop is overdubbing, then a long press exits back to the
play mode undoing the overdub (but leaving it available for redo).
* Multiply
A tap doubles the length of the loop. A second tap triples. A
third tap quadruples. (Each off of the state before we started
multiplying.) Any actions other than tapping the button break the sequence.
A long press erases all of the loops in the group. A second long
press while the group is empty resets everything.
* QReplace
A short tap does a QReplace.
A long press accesses a page in which the buttons now all pick
replacement time divisions. These divisions are also the divisions for
scramble.
* Half-speed
A short tap turns half-speed on or off for the group.
A long press accesses a page in which one can determine what the
actual speed shift is.
* Reverse
A short tap turns reverse on or off for the group. If the current
loop is empty, it will record forward and then playback backwards when
closed.
A long press toggles the alt send status for the group.
* Scramble
A short tap turns scramble on or off for the group.
A long press does a SUS scramble.
* Fade
A short press turns global fade on or off. When fade is on,
transitions for starting and stopping loops and possibly when switching
between loops in a group does a fade.
A long press invokes a Fade All which will fade everything down or
if everything is already faded down will fade everything up.
There are other features that would probably be interesting. I'm not
thrilled with doubling the alt send control onto the reverse button, but I
didn't have a good alternative given the button budget I had allowed
myself. It might be nice to change sync modes as well (loops v cycles v
immediate). It probably all would need some real experience playing it to
tune the behaviors, but there's the general model.
Given another button I might also look at supporting replace mode directly
with a button so that one can hit a button to start overdubbing with
feedback at 0% and then hit another buttons (rec/dub) to shift to
overdubbing at some other feedback level. This can be done with a feedback
control pedal, but then one has to pay attention to where the feedback is
set when jumping between loops.
For loop level, I'd like to have essentially a foot controlled mod wheel
that could push the volume level up or down. Feedback can probably be a
standard footpedal or other continuous controller though as noted above, I
would like this to be the feedback for secondary overdubs as opposed to
the feedback on first loop closure (or at least I'd like an option to have
it behave this way).
Clearly the underlying loop recording and playback engine on the
Looperlative is capable of this. Most of the operations are possible on
the Looperlative. (Exceptions: Changing QReplace division with a command;
having scramble tie to QReplace; copying into a specific non-blank-loop --
but those seem to have more to do with what the commands support rather
than the loop engine.) But it takes a control system that is aware of the
current status of the Looperlative and is short-press/long-press aware and
those both seem like steeper hurdles.
Any advice on how close I can get?
Mark
P.S. I targeted the M13 form factor because it seems to be a good
compromise between having lots of buttons and not consuming horrendous
amounts of floor space. My real ideal in all of this would put the looper
in the box itself.