photo

What is this?

It’s a low cost DIY timecode generator built from off-the-shelf open source hardware (OSHWA) components. It outputs a custom audio timecode based on GPS signal and can be used like any TC box plugged in MIC camera input for dual system sound and multicam sync.

This site proposes a cheaper and better mouse trap: a novel syncing method and apparatus that focuses on time of day record starting and ending determination rather than frame count. Here I give instructions for building a global navigation satellite system (GNSS) open source and open hardware dongle + software combo to:

  • sync camera video with sound from a dedicated audio recorder
  • sync multiple camera takes and show them aligned on your timeline
  • sync multiple audio recorders

As with SMPTE TC, each device has to be connected to a TicTacSync box (43$ a pop!).

New video clips (with synced audio) are created before importing into your NLE with the open source accompanying software tictacsync (that took waay more time to code than the SAMD21…).

Also, the remergemix command can assist in (re)conforming with an easy workflow:

  • post-process cameras and recorders files using the command tictacsync --isos YourMediaDir
  • the --isos flag results in cut & synced ISO files being written in the process, see details.
  • import clips into your NLE, do few edits -> oops… some audio needs rework!
  • remix the ISO files or add SFX, foley, etc…
  • call remergemix YourMediaDir to re-join the new mixdown to the camera file
  • refresh clip in NLE.

And voilà: new sound, no EDL, no pricey field recorder PT workflow.

The Grand Plan

Building up a community of devs and users, I want to incite camera manufacturers to implement Global Navigation Satellite System (GNSS) time of day (TOD) based head-tail microsecond time stamping of their recordings. Where needed, cheap RF 1PPS repeater could be deployed if GNSS signal is too weak.

And I want to break up the gate keeping that emerged among users due to the high cost proprietary hardware and software industry ecosystem: “We’re using those expensive tools because we are pros… if you can’t afford the tools of the trade, too bad…”

Tech is now (or always has been?) used by capitalists, VC investors and private equity firms to lock in creators: let’s Seize the Means of Computation!

Simpler, cheaper, faster : pick three

Yes, other well established solutions exist but this one is 5x cheaper: 45$ for each synced device VS 225$. And it is simpler on many aspects:

  • no jam sync necessary: it’s GNSS based
  • no frame rate to mess with: it’s frame rate agnostic
  • syncing at both ends: automated audio drift correction

Power up, plug, shoot, sync in post.

TicTacSync: syncing for the DIY adventurous

This cheap hardware/software combo will appeal to videographers who have encountered some limitations of their editing software “waveform analysis syncing” and should also interest scientists who need to timestamp data recordings within sub-millisecond precision.

Want to sync dual-system sound without breaking the bank? Doing multicam? Multisound? Need clock drift correction? tictacsync does it all!

  • TicTacCode is a new audio timing track format to timestamp both the start and the end of a recording. It is not SMPTE LTC and does not identifies individual frames (hence by design it is frame rate agnostic).

  • TicTacSync (note the case) is the hardware dongle generating TicTacCode. It fetches UTC time from GNSS signals. You need one for each recording device if your use case is syncing media files. Low cost, less than 45 USD, it is assembled from off the shelve components . Here’s a sample BOM on Adafruit store.

  • For syncing media files, tictacsync is the post-production CLI program that processes recordings shot with TicTacSync dongles, see repo. This program does the syncing before putting your clips into your NLE of choice (see demo on the left). As an added bonus it does time stretching to correct excessive clock drift between your devices, if any (see code).

Plays well with SMPTE LTC too

Another alternative use for it is drift calibration of prosumer cameras with a TC out connector (like the EOS R5 C or LUMIX DC-GH6): once a calibration measure is done in the morning, the camera can run on his own timecode generator in free run mode and before powering off the camera at day’s end another calibration sample is taken: and voilà! automated drif correction in post . No more extra $$ and clutter for a camera that already has internal TC!

Make it, don’t buy it

This won’t be a commercial product (not from me anyway). I’m sharing information and code to build your own devices: some assembly required. The post processing software is CLI only… I reckon a lot of videographers don’t know how to pip install modules or download external dependencies and tools (ffmpeg, sox, python, the Arduino IDE, etc…) so If someone want to code a GUI application and wrap dependencies into an installer for users with near-zero IT knowledge, go ahead!

Flashed Sale

For the inexperienced, flashing the SAMD21 board is the most adventurous task so I’m willing to do this part for you: a pre-flashed (and tested) board is available on Tindie.

Soon I’ll offer hand assembled prototypes too… hoping the project get some traction. Maybe some Shenzhen hacker will be pleased by the idea and whip up a one board PCB, driving the cost down even more.