Why are there no 'MIDI' LFOs in Live ?!?
Posted: Tue Sep 30, 2008 6:36 pm
...this is a long story. I try to make it short.
First of all, we (Ableton) would want to have modulation devices like LFOs in Live too. So, why the are we not doing it ????
The reason is a technical reason that leads to a conceptual problem, and we need to
solve the conceptual problem, which is very tricky, first. Then we can address the
technical problem, which is a huge effort. It's basically rewriting an important part of the engine.
So, you want to ask what is the problem? The problem is the ambiguity between
automation data, which is e.g. the parameter changes you can record in the arranger, and modulation data, which is the clip envelopes.
In the current internal data structure, each parameter can be controlled by exactly
one automation and one modulation. The automation is always directly the
movement of a 'knob' in the pannel, and this movement can be created by a
moving a knob with the mouse, or by an automation curve in the arranger, or by a
midi mapping, or by a macro control. The modulation which currently only can be
provided by a clip envelope belongs to that parameter and is independent from the
automation. The point where these things come together is the parameter itself.
When we invented the clip envelopes for Live 3 this structure seemed to be a good
concept, since the user does not need to do anything but drawing a clip envelope,
and he/she still can change the parameter that is modulated and there is no further understanding needed.
A better system from a todays perspective would allow any parameter to be
controlled in a relative or absolute fashion by many sources, and therefore allow for things like:
-automation of parameters within a clip, and not just modulation.
( -> automation recording not just of MIDI CCs but all parameters from within a clip )
-more then one modulation source per parameter ( -> LFOs !!! )
-control of a single parameter from various sources with variable curves. ( e.g. macro controls )
This sounds all very exciting, and would improve Live dramatically. However, it
would be the one single most complex re-design of the software since version 1.0.
It would change not only the way the engine handles automation data, but would
also require new interface strategies. It raises questions like: Is there automation
in the clip or in the arrangement or both? If so, which one wins if you move/delete
things? How to visualize multiple modulations? Can an LFO be modulated by itself?
What happens at overdub, ... and many more little things that can drive you nuts
if you sit in front of a whiteboard and try to solve them...
We planned to address the automation story for Live 7 but we had to realize that it
would be too risky at that time, and there were too many open questions. We are
still continue thinking about it, because it is obviously very very desirable.
However, it might turn out that the changes would be so drastic that it would be
impossible to open older documents in that new version and things like that.
Things we simply cannot do anymore, too many users rely on us.*
We could find a it-somehow-works solution for the LFO idea. But it would be a
dirty hack, and would make it even harder to finally come up with a better solution
for all the above mentioned wishes.
So, at the current state of affairs, it is not too likely that the LFOs will come soon.
Now you can hang the messenger,
Cheers, Robert.
*as a side note, this happens to a lot of companies: Mac OS 9 had to die because
there was no way to maintain and improve it anymore. Protools has this huge
legacy of the TDM hardware, that makes it hard to re-write the audio engine
without making customers who payed a gozillion for their old systems totally
mad...
First of all, we (Ableton) would want to have modulation devices like LFOs in Live too. So, why the are we not doing it ????
The reason is a technical reason that leads to a conceptual problem, and we need to
solve the conceptual problem, which is very tricky, first. Then we can address the
technical problem, which is a huge effort. It's basically rewriting an important part of the engine.
So, you want to ask what is the problem? The problem is the ambiguity between
automation data, which is e.g. the parameter changes you can record in the arranger, and modulation data, which is the clip envelopes.
In the current internal data structure, each parameter can be controlled by exactly
one automation and one modulation. The automation is always directly the
movement of a 'knob' in the pannel, and this movement can be created by a
moving a knob with the mouse, or by an automation curve in the arranger, or by a
midi mapping, or by a macro control. The modulation which currently only can be
provided by a clip envelope belongs to that parameter and is independent from the
automation. The point where these things come together is the parameter itself.
When we invented the clip envelopes for Live 3 this structure seemed to be a good
concept, since the user does not need to do anything but drawing a clip envelope,
and he/she still can change the parameter that is modulated and there is no further understanding needed.
A better system from a todays perspective would allow any parameter to be
controlled in a relative or absolute fashion by many sources, and therefore allow for things like:
-automation of parameters within a clip, and not just modulation.
( -> automation recording not just of MIDI CCs but all parameters from within a clip )
-more then one modulation source per parameter ( -> LFOs !!! )
-control of a single parameter from various sources with variable curves. ( e.g. macro controls )
This sounds all very exciting, and would improve Live dramatically. However, it
would be the one single most complex re-design of the software since version 1.0.
It would change not only the way the engine handles automation data, but would
also require new interface strategies. It raises questions like: Is there automation
in the clip or in the arrangement or both? If so, which one wins if you move/delete
things? How to visualize multiple modulations? Can an LFO be modulated by itself?
What happens at overdub, ... and many more little things that can drive you nuts
if you sit in front of a whiteboard and try to solve them...
We planned to address the automation story for Live 7 but we had to realize that it
would be too risky at that time, and there were too many open questions. We are
still continue thinking about it, because it is obviously very very desirable.
However, it might turn out that the changes would be so drastic that it would be
impossible to open older documents in that new version and things like that.
Things we simply cannot do anymore, too many users rely on us.*
We could find a it-somehow-works solution for the LFO idea. But it would be a
dirty hack, and would make it even harder to finally come up with a better solution
for all the above mentioned wishes.
So, at the current state of affairs, it is not too likely that the LFOs will come soon.
Now you can hang the messenger,
Cheers, Robert.
*as a side note, this happens to a lot of companies: Mac OS 9 had to die because
there was no way to maintain and improve it anymore. Protools has this huge
legacy of the TDM hardware, that makes it hard to re-write the audio engine
without making customers who payed a gozillion for their old systems totally
mad...
