Understanding & maintaining accurate audio and video synchronisation is a complex topic - but in Pro Tools 11 things have just got a whole load easier and a whole load cheaper too!
In Pro Tools 11 (and therefore 12) Avid now use the AVE (Avid Video Engine) from its flagship video editing and grading platform, Media Composer. In Pro Tools 10 and below, accurate sync was only EVER achievable with Avid media and an Avid Mojo. The second one stepped off piste by using QuickTime files (i.e. non Avid media) or with any other peripheral then it was game over for sync. Desktop video in Pro Tools 10 and below was like a drunk fish on a bike as I have previously explained and demonstrated. A total no go area.
The Arrival Of HD Video
When HD video started rearing its head in the days of Pro Tools 10, Avid sanctioned the use of BlackMagic cards to output the HD video, as Mojos were SD devices only. The Sync Avid peripheral was discontinued and the Sync HD replaced it as it could cope with tri-level syncs and the more esoteric HD frame rates we all know and love today.
Avid MXF was still the media of choice on the BM cards for accurate sync as they would use the quicktime engine to playback quicktimes - which were a bit rubbery!
Conventional wisdom was that if you wanted accurate sync between picture and audio, your Pro Tools must be clocked to the same video reference (via the Sync peripheral) as your video I/O interface (Mojo or Blackmagic). This practice would reduce two important things:
- The randomness between audio starting and video starting (as audio would be timed to start on the frame edge boundaries) between each play pass.
- Any drift or meandering during a single play pass (like desktop video in Pro Tools 10) as the audio and video are both resolved to frame pulses.
This is good basic engineering practice and is still the perceived wisdom of the day.
The AVE in Pro Tools 11
When Pro Tools 11 arrived, the new AVE promised to do serval exciting new things:
- To release the dependancy upon Avid media (MXF) for accurate sync - as the QT engine was no longer in use.
- To release the need to transcode media - any media was playable.
- Have more accurate desktop playback.
- Playback on desktop AS WELL AS out the video peripheral.
Avid also qualified AJA video peripherals. Now the fun starts….
My Pro Tools 11 HDX Rig
When I went to PT 11 HDX from HD3, I went with a Blackmagic Decklink Studio (a pcie card with a video reference input) to put in my designed Pro Tools PC, and a Sync HD to be able to feed my PT rig video syncs.
Due to having some driver issues and problems with jogging audio and scrubbing picture, I tested the AJA IOXT thunderbolt video peripheral. It was like driving an Aston Martin compared to the Blackmagic Card, fast, responsive and never crashes - so I got my cheque book out and bought the AJA IOXT. Happy days.
Upon hearing about the AJA T-Tap, AJA's diminutive palm sized I contacted Andy Killinger at Digi-Box, AJA's UK distributor for a review unit.
The T-Tap really is palm sized! It is a small box barely bigger than a Smint dispenser or a box of matches. It connects to your computer via Thunderbolt, and offers SDI and HDMI outs. It's a "portable professional monitoring solution for your pocket" as the AJA box says, "because it matters". Not much else to say about it really.
As I already had the AJA IOXT (its big brother) it made sense to use that as a benchmark to measure and compare the T-Tap against.
The test rig: A 3.5Ghz quad i7 Pro Tools PC with 32GB ram running Thunderbolt 1, running OSX 10.8.5: AJA IOXT and AJA T-Tap, Pro Tools HDX 11.3.1 with an Avid Omni, an Avid Sync HD, and a Rosendahl Nanosyncs for clocking duties.
I am still running AJA software 10.5.2 (the 10.8.5 version), so I just had install the driver for the T-Tap. Current versions for the AJA software on Yosemite include the drivers for all devices. How easy is that? After installing the driver, I unplugged the IOXT and attached the T-Tap. The T-Tap was immediately recognised and the software reported that it needed to change the firmware in the T-Tap. It tells you how long the update will take and when it is complete, and then then to power cycle the T-Tap (i.e. unplug the Thunderbolt). I then re-booted for good practice sake.
The T-Tap was immediately recognised when I restarted the AJA control panel
The shot above was taken once Pro Tools had booted - hence the "in use by Avid" in the top right hand side in red.
The session was one video track and one audio track, going straight out to my interface (Avid Omni). The videos were 1080i/50 DNxHD QuickTime videos from Jeremy Curtis at Harkwood that came with my fabulous Sync-One2 AV sync measurer. If you need to know about sync? Then get one of these boxes - we awarded it a Pro Tools Expert Editors Choice!
The IOXT Result
Here's the SyncOne-2 in front of the screen.
Looking at the two quick iPhone grab videos below you can see the IOXT results...
The first video is evenly spaced flash and blips, which is good for assessing the difference between audio and video (i.e. screen delay). The second video is a multi-spaced random pattern, which stresses the system more and would show up any perturbations on a single play pass (i,e if the av sync is wobbling and drunk). There were no elephants in the room.
The heartbeat shows the "happy measurement" of both A and V. The figure on the left is difference between audio and video in frames (to 1/100 of a frame) for each reading, the middle reading is the difference in milliseconds (remember 1 frame at 25F = 40 milliseconds), and the right hand reading is the average difference in frames (to 1/100 of a frame).
Here's the session window for the IOXT. The Sync HD is clocked from the Nanosyncs and the IOXT is also referenced to the same Nanosyncs.
The sharp-eyed amongst you will note that the composite output was also attached (this is how my booth is fed) and the IOXT was doing a simultaneous down-res on the fly, not used in this test. I was using the 3rd output (the HDMI) to feed my Panasonic Veira screen.
The T-Tap Session
The T-Tap has no video reference input, so I ran the Pro tools session on its own internal reference.
The T-Tap Result
Here's are some quick iPhone grab videos so that you can see the T-Tap results. There are 4 videos: the first with the regular spaced pattern and the next three showing the multi-spaced pattern.
The IOXT is rock solid to 1/100 of a frame accuracy. Outstandingly tight sync indeed. Never wavering - as you would expect from a totally referenced system.
The T-Tap was very close behind with only 8 ms on average difference (less than 1/4 of a frame) or about 27cm extra distance from the speakers in terms of time of arrival difference!. I took three play passes and achieved virtually constant consistent results each time. Please note that the T-Tap does run quite warm to touch, almost hot.
It would appear that the combination of the Avid Video Engine in PT11 and the ability of the AJA T-Tap to shift digits from your drive, up the Thunderbolt bus, through the T-Tap and into your digital display are leading to sync results that were virtually unimaginable a few years ago unless you spent some serious cash and bought some serious hardware. It is now possible to achieve professional results for around £200 which I have to say is outstanding! The T-Tap is a very accurate but affordable video peripheral and I can heartily recommend it.
However I suspect people who do video with Pro Tools without a Thunderbolt equipped computer may be feeling rather "left out in the cold" by now!