I'm just throwing out the question, does anyone know how to detect rising or falling CV? I'd prefer it if I could get a positive result for rising and a negative for falling, but I could just use a CVA-7 to adjust it.
I'm not sure what CV devices might let me temporarily hold a value for comparison, other than perhaps Lolth and Miranda. The min/max detection on a CVA-7 might work, except that it has a 1 second minimum auto reset on those, and I think I'd need to update quicker than that.
Has anyone run into this before?
Need Help: How to detect rising or falling CV?
If you ain't hip to the rare Housequake, shut up already.
Damn.
Damn.
- Raveshaper
- Posts: 1089
- Joined: 16 Jan 2015
Hmm. I guess I am lacking a key piece of information here, but it seems to me this would be covered by:
cv_last = cv_this
cv_this = rgiv (n)
cv_diff = cv_this - cv_last
No? Use EMI to analyze it this way and spit out -1, 0, or 1 based on cv_diff to be mapped elsewhere?
cv_last = cv_this
cv_this = rgiv (n)
cv_diff = cv_this - cv_last
No? Use EMI to analyze it this way and spit out -1, 0, or 1 based on cv_diff to be mapped elsewhere?
Enhanced by DataBridge v5
I mean, in the rack. I am putting my Remote development on hold for a while currently.Raveshaper wrote: ↑24 Apr 2018Hmm. I guess I am lacking a key piece of information here, but it seems to me this would be covered by:
cv_last = cv_this
cv_this = rgiv (n)
cv_diff = cv_this - cv_last
No? Use EMI to analyze it this way and spit out -1, 0, or 1 based on cv_diff to be mapped elsewhere?
I wasn't sure if there was a way to do what I'm wanting to mostly in Thor's programmer or not. I think I'll have to use the DVC-1 Digital Voltage Comparator, but that's not ideal because It does (more or equal) or (less or equal). There's nothing that does straight up comparison -- the DVC-1 should really have 5 states to compare. (gt, gte, e, lte, lt.)
The other logic REs I have, Truth and Elementary Logic Gates just work on pulses, not on values.
I'll figure this out eventually, just wish I could avoid REs.
If you ain't hip to the rare Housequake, shut up already.
Damn.
Damn.
For this to work you need more than one sample in time, so there will always be some delay in detecting rising vs falling signals.
It’s simple math otherwise, meaning what you do is add two consecutive samples/batches and compare them. If the first value is less than the second, the signal is rising. If the first is greater than the second, it’s falling. Obviously if they are equal the signal is neither.
To do this you need to ‘hold’ the first sample so you can compare it to the second, then output the result which should take three batches (192 samples):
1: take first sample
2: compare to second
3: output result
Doing this with stock devices is going to be tricky, especially getting the timing tight (and correct), and probably IF possible, would be extremely bloated and inefficient.
Sorry I don’t have more help, but I’m guessing the very best solution would be an RE specifically designed to do this. The second best would be combining a sample/hold with a comparator, but there’s still getting the timing right since I’m not sure there are REs that help with that.
Sent from some crappy device using Tapatalk
It’s simple math otherwise, meaning what you do is add two consecutive samples/batches and compare them. If the first value is less than the second, the signal is rising. If the first is greater than the second, it’s falling. Obviously if they are equal the signal is neither.
To do this you need to ‘hold’ the first sample so you can compare it to the second, then output the result which should take three batches (192 samples):
1: take first sample
2: compare to second
3: output result
Doing this with stock devices is going to be tricky, especially getting the timing tight (and correct), and probably IF possible, would be extremely bloated and inefficient.
Sorry I don’t have more help, but I’m guessing the very best solution would be an RE specifically designed to do this. The second best would be combining a sample/hold with a comparator, but there’s still getting the timing right since I’m not sure there are REs that help with that.
Sent from some crappy device using Tapatalk
Selig Audio, LLC
not sure if this might be helpful. I stumbled across it last week looking for an RE for my giant 'AI drummer' combinator, but decided it probably wouldn't suit my needs.
https://shop.propellerheads.se/rack-extension/now/
per the description:
https://shop.propellerheads.se/rack-extension/now/
per the description:
Now detects a rising edge in the CV signal above the Threshold setting. Each detection triggers a count.
-
- Moderator
- Posts: 1851
- Joined: 14 Sep 2015
- Location: Paris, France
It's super easy with DVC-1 (19 coins, by Ravenherz) and Lolth (free, by Jiggery Pokery). Simply create a 1ms delay with Lolth, output both the original signal and the delayed one in DVC-1, compare them, output 1 if CV2 > CV1 or 0 if not.
There's obviously a delay of about 1ms, depending on your application it may or may not be noticeable. You'll have to decide what to do if the signals are equal (either you consider them rising, falling, or add a third "0" value). You could also use it to output a third CV if the signal is raising and a fourth if it's falling, for instance. Here's the combinator.
There's obviously a delay of about 1ms, depending on your application it may or may not be noticeable. You'll have to decide what to do if the signals are equal (either you consider them rising, falling, or add a third "0" value). You could also use it to output a third CV if the signal is raising and a fourth if it's falling, for instance. Here's the combinator.
- Raveshaper
- Posts: 1089
- Joined: 16 Jan 2015
This was a good challenge and I finally came up with a completely stock solution for you. Written in 9.5 format.
Simply take Gate 1 and Gate 2 of "This - Last" BV512 and use them as your positive and negative triggers.
This patch works best if the CV source is wrapped inside. I'm narrowing the delay time down to 1 ms as the rate of the modulation increases.
Max delay time when Rate / Res is at zero needs fine tuning for some waveforms but the rest works the way you needed.
Simply take Gate 1 and Gate 2 of "This - Last" BV512 and use them as your positive and negative triggers.
This patch works best if the CV source is wrapped inside. I'm narrowing the delay time down to 1 ms as the rate of the modulation increases.
Max delay time when Rate / Res is at zero needs fine tuning for some waveforms but the rest works the way you needed.
- Attachments
-
- Raveshaper - Slope Detection.cmb.zip
- (2.46 KiB) Downloaded 39 times
Enhanced by DataBridge v5
-
- Information
-
Who is online
Users browsing this forum: No registered users and 28 guests