Do you think those of use without microtonic will be able to transfer data to the PO-32?
If the data format was documented, that might help us.
Yes, in the interviews they mention saving the data tranfer sound as a wav file and emailing it to a friend.
I don’t think TE will provide the data format beacuse it could hurt the uTonic VST sales. But anyone could analyse the transfer data, try to figure it out and create an editor/generator.
1 Like
Note that there are at least two things to figure out: the sound encoding protocol (modem) and the data transfer format. For the later, it may help using the uTonic VST to explore how each individual control affects the tranfer data.
Things may get quite a bit harder in case they added an encryption layer to it…
1 Like
@punji if this can be done on the PO-32 data, it can be done on the OP-1 patches as well ain’t it?
since the metadata can be seen and it can be saved both as patch (.aif) and sound data, it might help to decipher the PO-32 ones…
1 Like
has anyone got a sound recording of the noise?
1 Like
@bichuelo I guess the PO-32 data tranfer format is totally unrelated to the OP-1 patch format. It is most likely binary data, not text metadata, in order to be more efficient when transfering over the low speed modem protocol.
@eesn Not sure if this is pure enough:
1 Like
@punji one sound = 0.8 sec, whole bank = 6.9 sec (from another video), edit: bank+patterns = 17sec /edit. strongest carrier seems to be at 2700hz, bulk of bands seems to sit between 2500 and 5000 hz. I tried a variety of protocols (RTTY, Wideband, Dual, PSK, FSK, etc) but without knowing what to expect, specifically the type of data, and with the number of options available (type of sideband, modulation, polarity, baud), it's pretty difficult to know. cocoaModem seems to decode all sorts of gibberish, but binary is to be expected. µTonic 3.2 isn't out yet, so no way to look inside either...
2 Likes
@punji one sound = 0.8 sec, whole bank = 6.9 sec (from another video). strongest carrier seems to be at 2700hz, bulk of bands seems to sit between 2500 and 5000 hz. I tried a variety of protocols (RTTY, Wideband, Dual, PSK, FSK, etc) but without knowing what to expect, specifically the type of data, and with the number of options available (type of sideband, modulation, polarity, baud), it's pretty difficult to know. cocoaModem seems to decode all sorts of gibberish, but binary is to be expected. µTonic 3.2 isn't out yet, so no way to look inside either...
I love you guys and gals. Seriously, between this post and the custom firmware thread, the amount of dedication and know-how here is just stellar. 
2 Likes
Looking at an old (v2) microtonic preset (mtdp, if anyone has a v3 preset, i'd love if you sent me one). If all the numbers inside were 32-bit floats, a patch would fit inside about 80 bytes. *Could be* 1200bps but that would be a little too slow to transfer 16 sounds and patterns as in the other video. 2400? Compressed?
Anyway, below is a crop of TE's photo of Magnus Lidström's desktop, showing "DPSKModem" and "FSKModem" .h/.cpp and some "PocketRhinoEngine" parameters..
2 Likes
@eesn That picture is a great finding! I though the PO was coded in C, not C++… Looking at the modulation signal, it seems some kind of FSK modulation, but the spectrum plot doesn’t fit exactly, maybe we need a better recording. Anywa, there is a preamble with a little over 3KHz tone amplitude modulated at 600 Hz. That could a synchronization phase with two carriers separated by 600 Hz. If those same carriers are used for MSK modulation, it would give a 1200 bps rate, as you suggested.
1 Like
@punji the little plot at the start is a click of him pressing a button. I did not hear any sort of handshake. a quick google found a bunch of docs and 2400bps software modem working in that spectrum. I guess we find out when the first units are sent out and we get more recordings.
I’d still love a v3 microtonic preset, if anyone is willing to send me one.
ps not sure if this is the Microtonic or PO project. edit: it says “po_tonic” there at the end in the window title
2 Likes
carrier frequencies/harmonics in the spectrogram: 300 (weak), 900 (weak), 1500 (weak), 2100 (weak) 2700, 3300, 3900, 4500, 5100, 5700, 6300, 6900 (weak), 8100, 8700 Hz.
2 Likes
There is 0.8s on each side, the part to the left is what I called a preamble because it seems very constant, but it may just be a sequence of the same symbol:
The peaks there are at 300 (weak), 2700 (2nd), 3300 (1st), 3900 (3rd), 6300 (weak) and 8700 (weak).
1 Like
Yeah I assume left is just the carrier(s), what I mean by handshake is whether the PO responds in any way, which I don’t think it does. There’s fsk-decode.py on the intertubes, which takes a mono 16 bit wav and spits out bytes. Hard to tell what those bytes mean without getting hold of the original message or an example of one.
That said, the sources show 2-byte hex numbers in use. Finding out the struct we’re after would be interesting.
2 Likes
Hopefully TE will release some sort of SDK for the PO-32, as korg did for the Volca Sample
Volca Sample SDK
2 Likes
You guys are so smart. If this was a real conversation, I’d be looking verious and trying hard to act like I knew what the hell was going on.
1 Like
Mine shipped today! Can’t wait. Happy to make some recordings if you guys tell me what you need?
I’m perplexed yet fascinated. I’m going to follow just to see where the heck this ends up!
Mine shipped today! Can't wait. Happy to make some recordings if you guys tell me what you need?Definitely could do with a line-out recording from the computer output, e.g. instead of sending it to the PO just record it into a recorder. Then save that same sound as a native microtonic preset (i.e. not DAW preset). Take a screenshot of microtonic. Pack all files together as a zip =)
p.s. waiting on soniccharge.com/tonicnow