#370

Yes I do access crow through Druid.

I have tried pulling the resistors up and down, resetting crow, running First, all with no change in behavior.

Do you by chance have a crow and JF’s and an i2c connector? Maybe you could verify my results! :grinning:

#371

I do have all of those things together and I’m more proficient today with them then I ever have been (getting my druid practice in!). But I’m not keen on fiddling with wires behind the panels tonight. I was going to suggest sending crow a ^^kill command from druid and seeing if it fixes your issue. Or if not, there are several commands that can leave JF in a different tuning, so after the ^^kill you could try these (one at a time, I’d suggest):
ii.jf.mode(0)
ii.jf.run_mode(0)
ii.jf.transpose(0)
ii.jf.god_mode(0)

Then there’s the intone ratios set by ii.jf.retune... but I don’t know the defaults!

1 Like
#372

Fiddling with i2c cables is a real pain so I understand. Thanks for the command ideas. I’ll try those out. I just ordered one of these: Maybe it will fix the issue!

#373

Good news! The ii.jf.mode(0) command gets v/8 jack back to working without having to unplug anything. I wonder if this could also help with the Ansible 3.0 firmware leader mode issues? I’ll have to do some more experiments!

Thanks!

Edit: Also now I can recreate the issue but when I do a normal power cycle the v/8 jack works as it should again. Strange.

2 Likes
#374

EDIT: Corrected by trent below. JF.MODE is a setting that sticks through power cycles.

It puts JF into a radically different state: 6 voice “Synth” in sound mode, rhythmic modulator “Geode” in shape mode.

#375

Ahhh. I did run the command to put it in mode 1 previously so it was staying in mode 1 through powercycles then I guess. How many modes are there, just 0 and 1?

Thanks

#376

affirmative
and twenty chars

1 Like
#377

ii.jf.mode will not persist between power cycles. It is required every time you restart.

It is far more likely that you had uploaded a script which set the mode in the init script.

If you post a script that is causing the issue I can try and debug it!

1 Like
#378

Hi @Galapagoose,

It seems that for this particular issue which was basically:

Sequencing JF with awake on Norns will cause the v/8 jack to not work correctly until running ii.jf.mode(0) command. Even power cycling would not reset JF’s back to working correctly (which now seems to be working though).

Now there is another issue I am having with Crow, JF’s, and Ansible regarding the new 3.0 ansible firmware. If you wouldn’t mind checking that new firmware thread I have posted the issue there.

Thank you!

#379

Oh this is my mistake. In Synthesis mode, the v8 jack (and the Time knob) set the timing of the internal envelope per voice. This is expected behaviour, instead expecting pitch values to be sent via i2c.

The Just Type documentation gives a rundown of the changes in synthesis mode, albeit in the Teletype syntax. https://www.whimsicalraps.com/pages/just-type

1 Like
#380

Have we figured out how many I2C messages per second is typically stable? Is there a built-in buffer/delay on crow? I have been getting stable results at 500-1000 msg/sec but it seems that even at 1 msg/sec I am knocking other devices off the bus. TXi seems especially sensitive. My comms are basically TT<->TXi and F8R<->crow<->ER-301 on the same bus.

Crow often tells me that the I2C lines are low and I should activate pull-ups. I have a powered backpack though so I shouldn’t need the extra pull-up from crow. The warning happens regardless of the crow pull-up setting though.

I will probably divide my bus but it makes me a bit sad to isolate TT and ER-301 if I don’t have to.

#381

If you have a test case that fails that involves only: crow, w/, jf, txi, or txo, I’m willing to investigate. The reality is this requires some firmware level debugging and I haven’t prioritized it, mostly because there’s a very tiny audience that use such a dense bus.

It is an interesting technical challenge to me though, so if you can provide a reliable crash case with those modules, let me know and I’ll investigate.

#382

Not sure what I would be reporting but I have a crow, jf, w/, ansible chain that I can use to help out if it helps.

#383

Ok, I think I can retool my recent script to use high speed crow<->TXi and then see if TT<->JF stays stable. I had it fail in synth mode earlier after about 5 minutes. It started sending little sad blips instead of full notes.

#384

Hi, i want to ask regarding the First script. my understanding is, i only need to put clock on input 1 for it to work (spitting out melody and envelopes). but it seems i can’t get pitch out from output 1 and 3 (no melody). and i only get 4 envelope out of 8 trigger everytime from output 2 n 4. my clock source is from marbles.

i have updated my crow, and it’s in recent version 1.0.3.
its weird because, when i test uploading all script from bowery using druid everything seems works. i tried the scripting with druid sample & hold example without problem, i even tried duplicating @Galapagoose 's workshop material from last week Maps with Trent session, and i can get same result without problem. i also have used my crow with norns and max for live without problem. only with the First script that i’m having a problem.

just curious what could be the problem (i know i must be missing something obvious, but couldn’t figure it out what it is) i tried ^^R from druid and the ^^F several times, but still get the same result with First. any suggestion or idea?

thanks

1 Like
#385

First uses psuedo randomness to generate the melodies & rhythmic patterns, based on the serial number of the microcontroller in each device. The nice thing is that means each module plays different melodies out of the jacks, but has the downside that it’s possible for the generated melodies to be uninteresting. I think that might be what’s happened here – your crow is very calm in its nature :slight_smile:

You could likely change this by uploading a modified First.lua with a slight change to the initialization code. You can download First here, then change line 61 from lcg(unique_id()) to lcg(unique_id()+1). I believe that will cause a complete recalculation of the patterns!

If this doesn’t work let me know & I’ll think about what else could be at fault.

7 Likes
#386

Woah!!! I know I read that First is seeded with the individual crowd’s serial number, did not realize that it could be further modified by adding to that seed!

#387

oh wow that’s a very unique way to personalize the module :smiley:
thank you so much for your response.

also want to add, the things is, it’s not uninteresting melody, but rather no melody, the pitch doesn’t move anywhere, it just stays all the time :sweat_smile: .
but anyway i’ll try your suggestions and will let you know what happen.

btw, i really enjoy your session on twitch, and learn a lot. looking forward to the next one. thanks so much for doing it :pray:

2 Likes
#388

On my module I usually need to apply some nonzero CV to input 2 for First’s Melody to vary more than one or two notes. Generally I do get an envelope per trigger though I believe.

2 Likes
#389

I also need to do this to hear more notes, try sending a constant voltage or an LFO to input 2. Positive and negative voltages seem to add notes of a different scale too.

1 Like