U-he Hardware - CVilization Eurorack Module
-
News From The Sky News From The Sky https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=241844
- KVRist
- 183 posts since 20 Oct, 2010
I read your description of how it worked and literally laughed out loud. It's brilliant, which is exactly what I would expect from you guys. And I'm not even the least bit interested in Eurocrack, just curious to see what you've been up to.
-
- KVRian
- 549 posts since 9 Aug, 2005 from USA
Well said.News From The Sky wrote:I read your description of how it worked and literally laughed out loud. It's brilliant, which is exactly what I would expect from you guys. And I'm not even the least bit interested in Eurocrack, just curious to see what you've been up to.
-
- KVRAF
- 7357 posts since 9 Jan, 2003 from Saint Louis MO
I'm still very much looking forward to this one. Both my Maze and Mimetic Digitalis are in some danger of replacement, for being brilliant modules that I don't actually find myself using much in practice. CVilization may be just the thing to do it. 
-
- KVRAF
- 12354 posts since 7 May, 2006 from Southern California
I've got a 10HP spot reserved in my sequencing case for this. Such an incredibly comprehensive module but it seems very easy to understand.
-
- u-he
- 28063 posts since 8 Aug, 2002 from Berlin
After assembly of our first 10 pre-series models (we're very careful with our first design...) was delayed due to a part which wasn't available anymore, this week the soldering of the through hole parts has begun. We should have them next week, which should be incredibly helpful for finalizing and polishing the firmware.
It really is quite a journey.
It really is quite a journey.
-
- addled muppet weed
- 105853 posts since 26 Jan, 2003 from through the looking glass
Urs wrote: ↑Fri Dec 07, 2018 7:11 am After assembly of our first 10 pre-series models (we're very careful with our first design...) was delayed due to a part which wasn't available anymore, this week the soldering of the through hole parts has begun. We should have them next week, which should be incredibly helpful for finalizing and polishing the firmware.
It really is quite a journey.
-
- KVRAF
- 12354 posts since 7 May, 2006 from Southern California
I know the feeling. I recently saw the first hardware product, which I designed, come to life. It was an exhausting/exhilarating experience. I wish you all the luck and much success!
-
- KVRAF
- 7357 posts since 9 Jan, 2003 from Saint Louis MO
It seems like various supply/availability issues have been a real problem the last year or so
Do you have any more hints about what's in the other three modes? If I were to throw together a quick wishlist, it'd include a step sequencer 
-
- u-he
- 28063 posts since 8 Aug, 2002 from Berlin
All with a grain of salt, as we have literally zero feedback due to lack of available test subjects:
The second mode is a sequential switch, or merely, 4 sequencers with up to 8 steps each. Each step can either be a set value (as in a normal sequencer) or an individually attenuated choice of any 4 inputs. Programming works with the 8 encoders, either live or step by step (live mode is tricky, might be dropped). Clocking with just 2 CV/Clock inputs is the hard part - so on top of just clock division and clock iteration, we're experimenting with negative clocks (square LFO...) and different clock thresholds to extract multiple patterns out of one input. If it works out, it'll be available on all modes.
The third mode is what we call a "quad sequence multiplier". It works just like a sequencer, but you can record the input of each track, and more than 8 steps. It's more performance orientated than the 2nd mode. Instead of the manual editing you get randomization/mutation options similar to Turing Machine or Noise Ring. Resetting recalls the originally recorded sequence. It works as "S&H delay" when permanently recording, so your sampled input arrives n clock divisions later at the output. It works as plain S&H for each sequence set to 0 steps. It can also just be a random source, and (IIRC) outputs are normalled to next lane's inputs. Stuff like that, still a bit experimentation required.
The fourth mode is going to be a scanning mixer. But instead of one output which scans through several inputs, it works from the input side. Each input can be "panned" through the 4 outputs, even circular. Panning can either be manual, automatic, randomized or CV controlled. With 96kHz/24bit resolution and +/-10V, this should be suitable as audio mixer for quadrophonic setups.
Each of these modes has all the same quantization (scales), S&H and glide options of the first mode. Well, S&H is inherent to the "Sequence Multiplier".
More modes are possible, of course, but we first want to see how we go with these.
Latency should be well below 1 ms. It's not quite as fast as O_c, but again, that's due to using a 24bit/96kHz converter, allowing for crazy +/- 10V range.
The second mode is a sequential switch, or merely, 4 sequencers with up to 8 steps each. Each step can either be a set value (as in a normal sequencer) or an individually attenuated choice of any 4 inputs. Programming works with the 8 encoders, either live or step by step (live mode is tricky, might be dropped). Clocking with just 2 CV/Clock inputs is the hard part - so on top of just clock division and clock iteration, we're experimenting with negative clocks (square LFO...) and different clock thresholds to extract multiple patterns out of one input. If it works out, it'll be available on all modes.
The third mode is what we call a "quad sequence multiplier". It works just like a sequencer, but you can record the input of each track, and more than 8 steps. It's more performance orientated than the 2nd mode. Instead of the manual editing you get randomization/mutation options similar to Turing Machine or Noise Ring. Resetting recalls the originally recorded sequence. It works as "S&H delay" when permanently recording, so your sampled input arrives n clock divisions later at the output. It works as plain S&H for each sequence set to 0 steps. It can also just be a random source, and (IIRC) outputs are normalled to next lane's inputs. Stuff like that, still a bit experimentation required.
The fourth mode is going to be a scanning mixer. But instead of one output which scans through several inputs, it works from the input side. Each input can be "panned" through the 4 outputs, even circular. Panning can either be manual, automatic, randomized or CV controlled. With 96kHz/24bit resolution and +/-10V, this should be suitable as audio mixer for quadrophonic setups.
Each of these modes has all the same quantization (scales), S&H and glide options of the first mode. Well, S&H is inherent to the "Sequence Multiplier".
More modes are possible, of course, but we first want to see how we go with these.
Latency should be well below 1 ms. It's not quite as fast as O_c, but again, that's due to using a 24bit/96kHz converter, allowing for crazy +/- 10V range.
-
- Banned
- 7624 posts since 13 Nov, 2015 from Norway
Awesome stuff Urs. 
EnergyXT3 - LMMS - FL Studio | Roland SH201 - Waldorf Rocket | SoundCloud - Bandcamp
-
- addled muppet weed
- 105853 posts since 26 Jan, 2003 from through the looking glass
-
- KVRAF
- 7357 posts since 9 Jan, 2003 from Saint Louis MO
-
- u-he
- 28063 posts since 8 Aug, 2002 from Berlin
I can not wait to try it myself. We still only have this single precious prototype, but next week should have a whole bunch of them, some of which should end up in my racks for in depth experimentation.
-
- addled muppet weed
- 105853 posts since 26 Jan, 2003 from through the looking glass
yeah, along with the second "seq" mode, looks like a killer module.
the fourth mode could be great fun too, specially if you have multiple outs, throwing chunks of audio to different fx chains
-
- u-he
- 28063 posts since 8 Aug, 2002 from Berlin
It actually solves two very different problems for me:
1. I want to route and mix and s&h and quantize various means of sequencing to various oscillators/voices. I don't want any more quantizers which aren't quad and which don't have s&h built in. I also don't want to wast half a row on a 4x4 matrix mixer.
2. I love sequential switches, but I dread that they are always 1->4 or 4->1 (or even 8->1). They're bottlenecks. You either give up 3 sources, or you forgo 3 targets. Having a 4<->4 sequential switch ought to solve that issue. Again, 10HP, not half a row.
The other applications are bonus for me. Originally we wanted to make them bread and butter to make it a very "agreeable" module, but then I thought it'd be more consistent and more useful if these were complements to 1 & 2 with different paradigms.
1. I want to route and mix and s&h and quantize various means of sequencing to various oscillators/voices. I don't want any more quantizers which aren't quad and which don't have s&h built in. I also don't want to wast half a row on a 4x4 matrix mixer.
2. I love sequential switches, but I dread that they are always 1->4 or 4->1 (or even 8->1). They're bottlenecks. You either give up 3 sources, or you forgo 3 targets. Having a 4<->4 sequential switch ought to solve that issue. Again, 10HP, not half a row.
The other applications are bonus for me. Originally we wanted to make them bread and butter to make it a very "agreeable" module, but then I thought it'd be more consistent and more useful if these were complements to 1 & 2 with different paradigms.
