Gory

I know this app (and the developers too, they are located in Budapest). For the DVS function you need to buy the app, but it works great. This is my attempt to process dvs signal in my hardware and make MIDI out of it, and send it to the DJPLayer app.

Do you have a button in the ttm that was used for this differentation in vestax compatible cdj. You could use it to activate scratch mode. It is not convenient to alway pushing a button when you would like to scratch or manipulate the vinyl. I use some kind of intelligence instead when the velocity change over time is big enough, it means you are scracthing. When it gets low, it means you stopped. It should be derived from turntable velocity in non scratch mode and obvied in scratch mode due scratching is handled pitch control of the waveform (no?) Yes basically, You should eliminate the fluctuations, and you should be able to activaty the pitch mode if the dj just slightly touches the palatter. Usually in dvs timecode one channel goes backwards and the software use both signal comparation to accurate the analysis. I can uplad some technical documentation from Ms. Pinky Sdk if someone wants to take a look but I will do it this afternoon (now it is 15:06pm) or tomorrow. I think i'm talking avout a different thing. In a DVS time code you have 2 signals 90 degrees shifted thats true. I'm talking about how to send the calculated relative speed to the PC or iPad, instead of sending MIDI messages, which is slow. And the alternative could be data sent over audio channel which is usally used to send the digitized vinyl audio, but now as we process it externally, we can use the audio channel to send the pre-processed data.

For me there were some challanges while making TTM to MIDI jogwheel conversion. 1) There is a normal mode, when the vinyl spins normally. Stutter is eliminated by averaging the right amont of samples to keep the tempo stable. The TTM has some variations in ticks/second by default. And because the resolution is not that fine, It could cause tempo fluctuation otherwise. 2) I have to sense when somebody is touching the vinyl and scratching because I have to send out the jog touch event to be able to switch to scratch mode from pitch bend mode (or switch master tempo on/off) In scratch mode I use a more responsive but less accurate method to convert vinyl moves to MIDI. I use some kind of simple analysis in the time domain. 3) I have to sense when the pitch potmeter on the turntable is moved and the normal mode speed changes, because I have to move the pitch slider of the app with MIDI commands. I derivate this information from tempo change over a period of time, because unfortunatelly it is not possible to directly handle the pitch fader moves without modification on the turntable. 4) Also I have to be able to differentiate when the DJ uses the platter to slighli modify the tempo, when he moves the turntable pitch fader , or when he scratches. I think the most elegant way to pass all these information to a sowftware is using HID or using an audio channel backwards to send digital information with a more higher speed, than MIDI or HID. The software can than handle an incoming mono audio stream as a relative vinyl speed data. It can be done easily on an iPad because it can handle USB Audio Class, while it cant handle a generic HID class device.

In the first place, I would like to make a vinyl interface for my iPad app. So that's why I didn't consider anythng else. I like the app, it is stable, feture rich, contains all I need and ipad is more portable. I don't really like xwax on the other hand. I have some connection with other djay application developers for mobile. My idea was costumising they application to easily work with my hardware and to get the lowest possible latency on the market. (My sound card has 1ms latency)

If you read the whole blog you got the answer I reverse engineered my Reloop Beatpad, and emulated it with the STM32 to the software. It nees the USB VID, PID and maybe the MIDI interface. The optical mouse stuff is an optical mouse chip. I interfaced it throug SPI to my board. But in the end i calculated that it would be too slow for this purpose. 1600dpi with 33 1/3 rpm on the outter edge. But maybe i remember wrong. I never had time to finish it. It does not need any optical marks. It can work on the pure vinyl surface.

In my setup, responsibility or resolution does not really depend on the vinyl or TTM1. The TTM1 generates around 4100 ticks per rotation on the outter edge of the vinyl. And of course it generates less if I place it closer to the center. The timecode signal is 3000 on my Mixvibes vinyl which means around 5400 ticks per rotation (maybe times 4 because of the two phase shifted signals generates twice as much and I can copare to both rising and falling edges). And it has the same resolution towards the center obviously, so it dos not depend on where you place the needle. A MIDI jogwheel has around 800 ticks per rotation, maybe different for a differnet dj controller. And that limits the resolution of the movement. The other problem is with the MIDI. You only can send a packet in every 1ms, but in real life it is around 4-6ms. That limits the responsibility. Unfortunatelly iOS devices cannot handle HID. It would be better because it is easier and works with 1ms at least.

Hi there! The Algoriddim TTM hack is done by me. It is a small STM32 processor based device. I made a 4 channel ultra low latency USB soundcard which is iOS compatible, and a MIDI interface. A smart algorithm translates the vinoy movement from TTM encoder ro MIDI messages (midi controller scratch mode and jog movement) I did the same with directly analyzing the time code vinly audio, so the newer version works with the vinyl only without the TTM too. If you have any questions feel free to ask. I don't check this thread too often, so maybe questions could also be commented in the wordpress blog up there.