Update: version 0.0.2 fixes the fast-click problem reported by NaviRetlav.
I have developed a workaround to the bug described here that may be suitable for some situations.
It takes the form of a Combinator containing two Thor instances. When the bug is triggered the output from the first Thor is interrupted and this triggers the second Thor to generate a short pulse. This pulse is then used to gate the triggering of the first Thor's Step Sequencer, which has the effect of correctly resetting the Voice and reopening it.
This Combinator can be used to provide a reset pulse to any Thor in your workspace that relies on an open Voice, such as one that is using the Shaper to affect an external sound source (routed into the Thor). This pulse is generated whenever the bug occurs, thereby providing a way to recover from it.
A test-bench for this workaround can be found here (0.0.2). Simply open the file and you'll see the CV Debug combi on the left, and the Reset combi on the right. The top-right Delay device in the CV Debug (labeled "Reset Out") combi shows the state of the pulse. Hit Stop a few times to see the value jump from 1127 to down to 1000 briefly. You can adjust the length of the pulse with the first knob on the combi.
To use this pulse to reset another Thor, simply take this "Reset Out" signal into a CV Input on your Thor and use it as the Scale control (@ 100%) for your "LFO2 -> S.Trig" rule that you are using to retrigger the Step Sequencer. Look at the first Thor in the CV Reset combi for an example of this.
If you find this workaround useful or broken I'd appreciate some feedback please.
There is a small bug in this setup.
ReplyDeleteIf you double or more times hit stop button with mouse it will not reset properly.
It can reset only once.
Can you fix that ?
Nice catch, thanks. How quickly are you clicking stop to get that problem? I have to click *really* fast to cause it, but you're right, it's a bug. I'll have to think about how to fix it.
ReplyDeleteI've got a new version that I can't seem to break with fast-clicking stop - please give it a try:
ReplyDeletehttp://dl.dropbox.com/u/360506/blogs/offwhitenoise/bugs/Dual%20Thor%20Voice%20Reset%200.0.2.reason
The trade-off is that the pulse width has been reduced significantly.
It's not a problem with fast clicking ( that I already solved alone ) but a problem with multiple stops.
ReplyDelete1st stop makes full reset
2nd stop makes half reset
3rd stop makes again full reset
4th or more doesn't work at all !
Just try hitting stop and look at red lights in CV spider.
I just tried new version and it works a bit better.
ReplyDeleteNow only second stop doesn't work.
Hmmm, I wonder if we're seeing the same thing. The original bug is such a weird one that I wonder if different speed computers show different results.
ReplyDeleteI'm not sure I understand what you mean by multiple stops. I can hit 'stop' any number of times and it seems to work every time, whether I do it slowly or quickly. Are you saying that the second time you click stop it simply doesn't work properly?
This will be difficult to debug if I can't see the same behaviour. I just wish Propellerhead would fix this bug, they've known about it for a while (I reported it before Reason5 came out).
To recreate all what I said you need to set time marker in the middle of time line in sequencer, then hit play m wait a bit and hit stop ( with mouse ) , time marker will stop and reset will work, then hit stop again and time marker will back to previous position and reset will not work. One more stop, and marker will back on beginning, and everything will be fine again.
ReplyDeleteJust test it one more time.
I use Reason 6.0.1
You found what I mean ?
Ok, I see what you mean now - I have *no idea* why it does that! I'm not even sure that can be fixed with this sort of workaround. Ideally someone from Propellerhead will read this and look into it. I really hope so anyway because Thor's shaper is almost useless with this bug present :(
ReplyDelete