I had the exact same issue when biverting my first dmg using a hhl v1 bivert chip.

Problem was solved by reheating and adding a little solder to the bivert chip. I pushed the tip of the soldering iron down to touch the points on the gb pcb and added some more solder. I also reheated the solder on the lifted pins and made sure it looked fine.

I must have had a bad contact somewhere. Hope this helps.

cyberic wrote:

Wouldn'it be even more simple if lsdj itself would send a CC corresponding (somehow) to the phrase number?, or to the song row number?

Yes, that would be great. This would let LSDJ be master for a DAW or other sequencers/gear listening to row number.


On play, LSDJ could send both row number + a message containing the current tempo, and then ticks for a few beats.

The aboy can send midi clock using the tempo message the first beat, and the tick interval after that.
It would be even more awesome if new ticks were sent if tempo is changed during a track, keeping the sync.


Maybe I'm overlooking something.. But this should be possible, shouldnt it?

What about adding a CC command out per beat instead of tick? That way the GB to aboy stream wouldnt be overloaded. Also you could stop sending CC after a few beats.

I actually tried this but couldnt get the sending of midi clock out of the aboy stable enough.

The Gameboy Zero sounds pretty close to what you want:
https://www.raspberrypi.org/blog/game-boy-zero/

37

(21 replies, posted in Nintendo Handhelds)

cyberic wrote:
Ledfyr wrote:

I moved parts of the note handling from the Arduinoboy to the Raspberry Pi and it reduced the dropped notes to almost 0. The raspberry now handles the sending of note-off messages.

according to https://github.com/ledfyr/lsdj_mo/blob/master/README
you say that "MIDI notes are often "missed". The Arduino seems more stable, or there are bugs in my code sad"
is it more stable now?

That is my unfinished repo for using the Raspberry's GPIO port. In that case almost half of the notes are missed. I kind of paused this, dont know how to continue right now.

The fix I talk about in the original post (mono mode) requires a special Arduinoboy version + a Raspberry Pi with MIDIcloro. See info in the link in the original post.

38

(18 replies, posted in Nintendo Handhelds)

This is awesome! I'll have to try the Teensy for USB MIDI. Thanks alot for keeping developing this great project!

39

(21 replies, posted in Nintendo Handhelds)

trash80 wrote:

Hey there, neat project.

Ledfyr wrote:

* Mono mode - improves/solves the problem with lost or dropped MIDI notes from the Gameboy (use with custom Arduinoboy software, only tested with GBC, see README).

This really shouldn't be a problem, or at least solvable outside of Arduinoboy, it's actually a pretty lightweight task to handle the data stream. The issue may come from the Gameboy itself, or that is the code on the GB, since it is kind of a hack. What version of LSDJ were you using? I am wondering if this was a wrapping buffer issue that was solve awhile back.

Let me know if you need any clarification on my end, feel free to email me.

Thanks for all the info and advice in the mail conversation!

I thought I should clarify the problem with note dropouts some more in this thread. I cleaned the original post a bit and put some info about note dropouts below:

I wanted MIDI sequencing from the GB to be super stable. My goal is to use the GB as my main sequencer. I had noticed some occasional note dropouts, about 1 note every 30-50 notes was dropped). This is not a big problem at all, but I still wanted to try and make it better.

I moved parts of the note handling from the Arduinoboy to the Raspberry Pi and it reduced the dropped notes to almost 0. The raspberry now handles the sending of note-off messages.

It is probably possible to solve this on the Arduino itself, and I think that would be a better solution.

Hi!

I thought I could mention my project for the Raspberry Pi:
http://chipmusic.org/forums/topic/17404 … pberry-pi/

It's more of a merge box, but it can do some of the things you've discussed, like channel switching and tweaking of velocity (there's even a "random velocity" mode). The program could be changed to send on multiple outputs.

If you want to, feel free to expand on this or use ideas/parts of the code for your project.


Another product related to this:
http://www.audiothingies.com/product/midibro/

Are you using alkaline batteries?
My frontlit gbc was dim on rechargeable (nimh). It took a while for me to figure out I had to use alkaline.
Seems like 1.5 V is needed.

automatonsubsys wrote:

i really want to get one of these. I have been experimenting with MIDI sequencing in Python in Linux, so I am sure it will work fine on this device. I have also been getting into other trackers in Linux. seems like a good handheld choice.

has anyone tried this on a raspberry PI? seems like it would be similar

found this, but i only really want MIDI out

http://www.instructables.com/id/PiMiDi- … earned-to/

I have worked with MIDI in/out on the Raspberry Pi using a C++ lib called RtMidi (I think its available for Python too).

No need to follow that instructables page unless you really want to, since most USB MIDI devices works directly when plugging them in to the RPi (MIDI interfaces, keyboards etc).

Not sure if it works like this on the C.H.I.P, but I cant see why it shouldnt.

My project if you're interested:
http://chipmusic.org/forums/topic/17404 … pberry-pi/

43

(21 replies, posted in Nintendo Handhelds)

Thanks for the advice. I did mail with zii.hrs who is the author of PiBoy and he did help me by spotting a bug, and we talked a bit about the problem. But we couldn't solve it.

I suspect sending sync signals out of the RPi might not be as timing critical as reading bytes from the GB (LSDJ MIDI out mode).

44

(21 replies, posted in Nintendo Handhelds)

Regarding connecting the GB to GPIO directly:
I got MIDI out mode working this way. Sadly it's very unstable - a lot of notes are dropped. I've given up on this for now, thinking that the RPi is not well suited for this (though I might be wrong). Possible solutions could be to try with a RPi 3 and/or try to optimize the real time build (requires RT kernel).

Here is the code: https://github.com/ledfyr/lsdj_mo

45

(21 replies, posted in Nintendo Handhelds)

New functions are added. See updated original post and README on github!

nanode wrote:

It looks cool but tbh I'd rather go with something a bit more fleshed out like Goat tracker.

There are Linux versions of GoatTracker and Little GP Tracker. Would be really cool to use them on the Pocket Chip.

Pocket Chip runs Linux ... so it should work! Not sure if any adaptations are necessary, or if it works out of the box though.

Not sure about the differences in sound quality, but you can use the existing headphone jack.
A.k.a. Internal Prosound mod.
Then the speaker will turn off when you plug something in the jack.

Note that some headphones will be too quiet, some will work fine.

If the LEDs blink as usual, I think that the Arduinoboy works.
If not, something might have gone wrong when uploading the software.


Regarding the bug ... sad
I can't figure out why channel switching doesn't work in PU2/WAV/NOI, but I guess channel switching in PU1 is better than nothing.
If anyone can spot the error please let me know.