|
|
Post by young Protoboard on Oct 10, 2019 19:03:26 GMT
After an arduous wait, I finally received my Starter Rack 2 yesterday. I'm still in the process of learning the ropes and familiarizing myself with the ins and outs of each module. I noticed some strange from the Beat Divider, and I'm not sure if it's working as designed.
From what I can tell, like many beat dividers, AE's Beat Divider works by counting incoming clock pulses and holding the output HIGH or LOW for a certain number of pulses. The number of pulses counted to output HIGH and LOW is selected by the DIV potentiometer.
Here's the alleged aberrant behavior: my Beat Divider outputs HIGH for more clock pulses than it outputs LOW. For example, the 1/32 setting gives HIGH for 2 pulses, then LOW for 1 pulse duration. The same behavior is found in other settings, eg. 1/16 = (4 pulses HIGH, 2 pulses LOW); 1/8 = (7 HIGH, 5 LOW), and so forth. The asymmetrical HIGH/LOW behavior is present in all DIV settings, and from all clock sources (pulse and square wave, of all pulse widths that give regular behavior).
Is it normal for the Beat Divider to count asymmetrically? It seems really limiting in usefulness to have a Beat Divider count off-beat of a power of 2; it doesn't synchronize with any 4-count measures, doesn't repeat with an 8- or 16- step sequencer of any kind, etc. Further, it might be giving a "1/32nd note," but the duration of a 1/16th note and its accompanying rest is definitely NOT twice the 1/32nd note. It's not musically or mathematically "correct" by its labeling or its function.
I expect (and I'm not certain of my validity here because I'm new to modular synthesis) the Clock Divider's behavior to be as follows:
1/32 = (2 HIGH, 2 LOW);
1/16 - (4 HIGH, 4 LOW);
1/8 = (8 HIGH, 4 LOW);
and so forth.
So, what's going on with my Clock Divider? is it designed to be asymmetric and is working fine? Or is mine malfunctioning?
|
|
|
|
Post by admin on Oct 11, 2019 13:01:28 GMT
Hi young Protoboard, I believe that the Beatdivider module is really only made to work with MIDI clock coming from the BUS CLK. That's how I'm using it all the time with my Keystep. Keystep goes into MIDI on the Master module, then patch BUS CLK into CLK on the Beatdivider and then everything is going well together. If you wanted to divide other clocks, say from an LFO for example you would be better off using the Divider or new MM-Divider module. |
|
|
|
Post by sycophante on Oct 11, 2019 13:23:07 GMT
After an arduous wait, I finally received my Starter Rack 2 yesterday. I'm still in the process of learning the ropes and familiarizing myself with the ins and outs of each module. I noticed some strange from the Beat Divider, and I'm not sure if it's working as designed. From what I can tell, like many beat dividers, AE's Beat Divider works by counting incoming clock pulses and holding the output HIGH or LOW for a certain number of pulses. The number of pulses counted to output HIGH and LOW is selected by the DIV potentiometer. Here's the alleged aberrant behavior: my Beat Divider outputs HIGH for more clock pulses than it outputs LOW. For example, the 1/32 setting gives HIGH for 2 pulses, then LOW for 1 pulse duration. The same behavior is found in other settings, eg. 1/16 = (4 pulses HIGH, 2 pulses LOW); 1/8 = (7 HIGH, 5 LOW), and so forth. The asymmetrical HIGH/LOW behavior is present in all DIV settings, and from all clock sources (pulse and square wave, of all pulse widths that give regular behavior). Is it normal for the Beat Divider to count asymmetrically? It seems really limiting in usefulness to have a Beat Divider count off-beat of a power of 2; it doesn't synchronize with any 4-count measures, doesn't repeat with an 8- or 16- step sequencer of any kind, etc. Further, it might be giving a "1/32nd note," but the duration of a 1/16th note and its accompanying rest is definitely NOT twice the 1/32nd note. It's not musically or mathematically "correct" by its labeling or its function. I expect (and I'm not certain of my validity here because I'm new to modular synthesis) the Clock Divider's behavior to be as follows: 1/32 = (2 HIGH, 2 LOW); 1/16 - (4 HIGH, 4 LOW); 1/8 = (8 HIGH, 4 LOW); and so forth. So, what's going on with my Clock Divider? is it designed to be asymmetric and is working fine? Or is mine malfunctioning? I checked yesterday and noticed my Beat Divider has the same behaviour. I've noticed though that it gives me 4 counts dividing when I take the signal out of output B and put the switch on "triplets", so I think maybe my outputs have been inverted in the making ? |
|
|
|
Post by young Protoboard on Oct 11, 2019 19:25:39 GMT
Hi young Protoboard, I believe that the Beatdivider module is really only made to work with MIDI clock coming from the BUS CLK. That's how I'm using it all the time with my Keystep. Keystep goes into MIDI on the Master module, then patch BUS CLK into CLK on the Beatdivider and then everything is going well together. If you wanted to divide other clocks, say from an LFO for example you would be better off using the Divider or new MM-Divider module. Thanks, carek! I'll purchase a MIDI cable to connect my glorious bog-standard third-hand RadioShack keyboard, so that I can verify this. I'll definitely be building a binary divider in the interim and to complement all non-MIDI patches. |
|
|
|
Post by young Protoboard on Oct 11, 2019 19:31:41 GMT
After an arduous wait, I finally received my Starter Rack 2 yesterday. I'm still in the process of learning the ropes and familiarizing myself with the ins and outs of each module. I noticed some strange from the Beat Divider, and I'm not sure if it's working as designed. From what I can tell, like many beat dividers, AE's Beat Divider works by counting incoming clock pulses and holding the output HIGH or LOW for a certain number of pulses. The number of pulses counted to output HIGH and LOW is selected by the DIV potentiometer. Here's the alleged aberrant behavior: my Beat Divider outputs HIGH for more clock pulses than it outputs LOW. For example, the 1/32 setting gives HIGH for 2 pulses, then LOW for 1 pulse duration. The same behavior is found in other settings, eg. 1/16 = (4 pulses HIGH, 2 pulses LOW); 1/8 = (7 HIGH, 5 LOW), and so forth. The asymmetrical HIGH/LOW behavior is present in all DIV settings, and from all clock sources (pulse and square wave, of all pulse widths that give regular behavior). Is it normal for the Beat Divider to count asymmetrically? It seems really limiting in usefulness to have a Beat Divider count off-beat of a power of 2; it doesn't synchronize with any 4-count measures, doesn't repeat with an 8- or 16- step sequencer of any kind, etc. Further, it might be giving a "1/32nd note," but the duration of a 1/16th note and its accompanying rest is definitely NOT twice the 1/32nd note. It's not musically or mathematically "correct" by its labeling or its function. I expect (and I'm not certain of my validity here because I'm new to modular synthesis) the Clock Divider's behavior to be as follows: 1/32 = (2 HIGH, 2 LOW); 1/16 - (4 HIGH, 4 LOW); 1/8 = (8 HIGH, 4 LOW); and so forth. So, what's going on with my Clock Divider? is it designed to be asymmetric and is working fine? Or is mine malfunctioning? I checked yesterday and noticed my Beat Divider has the same behaviour. I've noticed though that it gives me 4 counts dividing when I take the signal out of output B and put the switch on "triplets", so I think maybe my outputs have been inverted in the making ? Good to hear that it's not just my module behaving like this for non-MIDI clock input. Thanks. I wonder if your triplet output B shows this behavior for all DIV settings? It seems possible that a certain DIV setting and thus ratio (eg (7. HIGH, 5 OFF)), when divided further by the triplet, just so happens to divide down evenly, but entirely due to coincidence? This seems more likely than a production error where your outputs A and B are installed incorrectly. |
|
|
|
Post by NightMachines on Oct 12, 2019 8:53:55 GMT
Like admin said, the Beat Divider is meant to divide MIDI clock. MIDI clock runs at a resolution of 96ppq, pulses per quarter note or per “beat”. This is usually way too fast (or rather too “high res”) for modular applications, where we often want a resolution of 32, 16 or even 8 pulses. To get there from 96 we have to divide by a multiple of 3, thus all of our Beat Divider modules work as intended. Although to avoid this confusion which has been brought up a bunch of times already, it would have been good to have named it simply “MIDI Clock Divider”. |
|
|
|
Post by robertlanger on Oct 12, 2019 9:04:13 GMT
Hi Guys, the focus with the BEAT DIVIDER module is only on the rising edges; to be honest, I didn't care much about the pulse length as it didn't matter IMO. But, yes, offbeat is an argument I didn't think of. And yes, as NightMachines stated, MIDI CLOCK DIVIDER would be a more clear name now, after other dividers are in the module range. |
|
