This is topic Sync Check DCP? in forum Digital Cinema Forum at Film-Tech Forum ARCHIVE.
To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=16;t=001037
Posted by Dave Macaulay (Member # 813) on 02-28-2012, 09:41 PM:
Is there a "sound sync check" DCP available anywhere? This would be like the Dolby SR-D reader setup film with a flash frame and synchronized beep or pop, so that the digital sound delay could be set correctly for different D/A converters.
If anyone knows of one, or if anyone can make one, please let me (us?) know.
Doremi servers (for example) come set up to have correct sync with the Doremi D-A converter feeding an analog cinema processor. Fully digital processors have a but of latency and the lipsync is close but the sound is slightly late. Sending the AES digital signals to a digital cinema processor like the CP750 gives lipsync that's obviously wrong. Finding reliably synced content has me stumped. I have a few trailers that are "fair" for setting up sound sync, but having a real sync check DCP would be great.
Posted by Sam D. Chavez (Member # 1841) on 02-29-2012, 07:08 PM:
This was an idea originally brought to me by Lonny Jennings some time ago.
The DCP sync file is easy enough to make but measuring the absolute sync pop is another matter.
I'm happy to report that we are working on incorporating it into an existing product.
Stay tuned.
Posted by Dave Macaulay (Member # 813) on 02-29-2012, 09:18 PM:
As usual, Lonny is several steps ahead of me. I look forward to whatever lets me set sound sync relatively easily.
I was with Chuck not long ago, he says hi if I ever talk to you...
Posted by Carsten Kurz (Member # 5396) on 03-02-2012, 01:47 AM:
Making a sync DCP is relatively easy with todays free tools.
To actually measure the delay, a very cheap way would be a simpl audio/light sensor connected to a stereo line-in of a laptop or portable audio recorder. One channel with a microphone, the other with a very basic photo-transistor setup. Start Record, beep-flash the sensor, and you will see the delay in waveform view of the recording.
If you have a decent digital camera with movie function, you can do the same by just capturing the sync DCP from any 'mean' seat position.
Then you just need a basic movie editor that allows you to see picture and audio waveform to analyse the audio shift from a flash frame.
- Carsten
- Carsten
Posted by Sam D. Chavez (Member # 1841) on 03-02-2012, 09:59 AM:
Carsten says,
"Making a sync DCP is relatively easy with todays free tools."
I just said that two posts back.
Posted by Steve Guttag (Member # 268) on 03-02-2012, 08:18 PM:
He signed his name twice...perhaps he was just making sure the message go across?
Posted by Carsten Kurz (Member # 5396) on 03-03-2012, 08:15 AM:
quote: Sam D. Chavez
I just said that two posts back.
No Sam, you didn't mention free tools ;-)
Aside from that, Studio Six Digital are currently working on an A/V Sync module for their iPhone tool-suite http://studiosixdigital.com/audiotools/
using the iPhone video recording function.
- Carsten
- Carsten
- Carsten
Posted by Paul H. Rayton (Member # 1817) on 03-07-2012, 04:23 PM:
I'm not sure if this would apply to ALL digital still cameras, but if I were to use mine for such a visual "sync check" comparison (as described by Carsten), there might be an issue because there is a slight latency in my camera's video output, due to image stabilization. So the visual pop might be offset a bit from the audible pop.
All we really want and need is a sync "pop" sound of some sort, that is precise and well-defined, and an absolute visual reference -- and have it loop. Then, we can set it to running, go out and listen, see if it's in sync or not, and then dial up (or down) the system.
(And then, **hope** that when we get festivals in, with tapes and other video content from all manner of diverse sources, that the "kids" made their movies properly!)
Posted by Carsten Kurz (Member # 5396) on 03-07-2012, 06:00 PM:
There's all sorts of strange gear around, but still even cheap digicams should be able keep video and sound in sync to within one frame or so. And that should be much more accurate than 'guessing' on a loop. If there is an inherent delay in a cameras video processing pipeline, the developers should know this and compensate for it.
I'm not talking about live video anyway, but looking at the recording.
There are sync check videos with a visual pendulum and the sync pop at the center transition, like a musical metronome. Or more elaborate ones like this with multiple visual sync cues:
http://www.youtube.com/watch?v=bCPEidaVzQU
They are certainly good to align a system by direct sensing that is way out of sync, but around +/- one or two frames, it gets difficult. Not that it is really THAT important to be precisely on the spot, but if a cheap gizmo can help you, why not use it.
You can check your cameras sync capabilties easy with a clapperboard. Or simply two sticks or a clap of your hands.
- Carsten
Posted by Jussi Siponen (Member # 5572) on 03-20-2012, 01:32 PM:
Inspired by the BBC sync test video linked by Carsten.
I took the sound from the BBC video and adapted the scale for 24fps.

This is frame 7 (upper) and frame 12. The "tick" sound occurs half way through frame 12 being visible.
BTW, is this the right thing to do -- should it be timed to the "leading edge" of frame 12?
If the timing is determined to be correct and I do a QC pass on a real D-Cinema system (in stead of a software DCP player) I can post a download link later this week if there are any takers.
Posted by Paul H. Rayton (Member # 1817) on 03-20-2012, 02:40 PM:
Hi, Jussi -- Yes, great. Please consider this at least one "vote of encouragement" to proceed with your plan. I'd be happy to try it (when you finish it), utilizing my somewhat "antique" Quvis server.
Posted by Ian Freer (Member # 1940) on 03-20-2012, 04:02 PM:
I have a DCP clip that is almost identical to this that must have come included on some other content I received. I don't recall ingesting in directly, but came as part of something for a film festival so I was ingesting 'everything' to make sure I had the correct versions of DCP's available. I haven't looked at it for a few months but from memory it has the logo of a London, UK post house or something on it. I'll have a look how big it is and see if I can upload it or send it to Brad to put up on the site somewhere. Saves everyone having to re-invent the wheel as it were...
Cheers,
Ian
Posted by Jussi Siponen (Member # 5572) on 03-21-2012, 01:22 PM:
DropBox is still uploading as I write this. The download link will start to work as soon as the upload is complete (shouldn't take more than 30 minutes):
SyncTest DCP (340 MB, runtime 1:30)
The delay scale now shows also milliseconds.
Re-inventing the wheel is kind of fun, but also potentially embarrassing, given the audience. I hope I didn't mess up somewhere
Posted by Marco Giustini (Member # 4544) on 03-22-2012, 01:34 PM:
Jussi, this looks very good, thanks!
Posted by Dave Macaulay (Member # 813) on 03-28-2012, 08:34 AM:
Thanks a bunch! I haven't run this yet (waiting for projector power) but the DCP did ingest OK into the Doremi server.
Posted by Carsten Kurz (Member # 5396) on 03-31-2012, 04:32 PM:
Jussi - I just though over it, and it might be useful to have a DCP with the same sync beep at exactly the same time in ALL channels. You might even want to compile a 7.1 version. As all tracks are the same, this could be used on 5.1, 7.1 and SDDS layouts just as well.
I guess I could rewrap it myself from the existing file, but as you have posted a DropBox link, you might just replace the file with a surround version. All that's needed is to place the same mono wave file into all surround channels instead of using only a stereo file for L/R.
I know usually sync is most important for center speaker, but having the sync beep on all channels will enable people to also check various L/R and surround delay issues e.g. by turning down individual channels/amps.
- Carsten
Posted by Jussi Siponen (Member # 5572) on 04-01-2012, 04:57 AM:
Hi Carsten,
Good point with the 7.1 - and why didn't I think of that myself?!?
*headdesk*
I'm going to be busy with work for the next couple of days, I'll get to it later this week.
Posted by Jussi Siponen (Member # 5572) on 04-01-2012, 01:31 PM:
Had some free time in the evening after all
SyncTest DCP, 7.1 channel version (337 MB, 1:30)
DropBox is still uploading - if the link doesn't work, wait 30 minutes and try again.
Posted by Carsten Kurz (Member # 5396) on 04-01-2012, 07:17 PM:
Great Jussi, thanks again.
- Carsten
Posted by Wolfgang Woehl (Member # 6901) on 09-08-2012, 09:22 AM:
Jussi, great effort and thanks for sharing.
Here's a conversion to Bluray from your SYNC-TEST_TST_F_20_2K_IND_20120317_DCP_OV package: SYNC-TEST_TST_F_20_2K_IND_20120317_DCP_OV.udf. Get the image and burn it to disc.
We used this in combination with a Syncheck II to debug some issues. So I've stared a lot at this thing in the past days:
I'd probably go for less visual clutter like dropping the upper left title, dropping the "Adjust" note.
The Syncheck II is a straightforward sync checker, has a light sensor and a mic. It goes for white flash so the bigger the white region is on sync flash the easier to catch. Their test material is actually black with 1 full white frame. But you can make do with the flash circle.
We've noted the off-boundary position of the sound pip and, like you, I'm asking myself if that is correct sync. Film sound editors would know. Anyway, we were not chasing half a frame but more so that was not a biggie.
Btw, there is a short and low-volume organ kind of sound just before the first pip.
Posted by Wolfgang Woehl (Member # 6901) on 09-17-2012, 11:19 AM:
So I've asked a film sound editor: He says perfect sync sits at the 0 position (my words) of a frame, the opening or starting boundary of a frame.
Duh, right? Well, always good to talk to experts
Posted by Jussi Siponen (Member # 5572) on 09-27-2012, 12:43 PM:
Looks like I'll be re-visiting this project next weekend.
The sync test is off by 0.5 frames / 21 ms -- hardly the end of the world, but I'd rather have the DCP right *on* than "close enough for government work"...
Gentlemen, stop your downloads for now. I think I can swap the files under the links with corrected versions and then make a post saying I've done so.
Posted by Jason Raftery (Member # 6291) on 09-27-2012, 03:18 PM:
Jussi, thanks for this helpful DCP.
The sync test DCP ingests and runs fine on the Dolby servers I've tested it on but won't ingest on GDC servers. The GDC doesn't even "see" the DCP on the flash drive and thus won't ingest it.
Posted by Wolfgang Woehl (Member # 6901) on 09-27-2012, 03:28 PM:
Jussi, great and thanks. The same deprecation note goes for the Bluray version of Jussi's Sync Test package. The link I had posted above is dead for now. I'll post a link to the corrected Bluray version once Jussi's done.
Posted by Jussi Siponen (Member # 5572) on 09-29-2012, 06:03 PM:
Here's the latest:
Synctest rev4, combined 2.0 and 7.1 audio versions package (343 MB)
What has changed (both per Wolfgang's feedback):
- corrected the sound timing
- made the white circle on frame 12 larger to make it easier to pick up with measurement tools
The previous URLs now point to a zip file with only a README.txt file with this message:
quote:
Hi!
This package has been superseded by:
https://dl.dropbox.com/u/12106363/SYNC-TEST_TST_F_71_2K_IND_20120929_DCP_OV.zip
Best regards,
Jussi Siponen
@Jason: I don't have access to a GDC server. Could you try the updated package and if it fails, see if your server provides any clues (e.g. in logs)?
When copying the DCP files to your flash drive, the files must be at most one directory deep (i.e. at the drive root or in a folder -- NOT in a folder inside another folder).
Also, will the GDC ingest anything else from the same drive?
Posted by Wolfgang Woehl (Member # 6901) on 09-29-2012, 06:55 PM:
Jussi, great.
2 minor issues:
The zip file
SYNC-TEST_TST_F_71_2K_IND_20120929_DCP_OV.zip unzips to
SYNC-TEST_TST_F_71_2K_IND_20120401_DCP_OV
while the composition's IssueDates are indeed set to 2012-09-29...
And the organ-like sound before the first pip is still there, in both sound versions. Don't know if that is intentional?
For what it's worth: When we had the Bluray version loop over and over it was vexing at first but useful later on. It allowed us to acoustically anchor the loop start which turned out to be crucial.
Posted by Jussi Siponen (Member # 5572) on 09-30-2012, 02:01 AM:
quote: Wolfgang Woehl
And the organ-like sound before the first pip is still there, in both sound versions.
Whisky-Tango-Foxtrot?!?!
The waveform should be a perfectly flat line before the first pip, no DC offset either...
Posted by Wolfgang Woehl (Member # 6901) on 09-30-2012, 07:21 AM:
He, right, guess it should. But
asdcp-test -x 2a111bb8 2a111bb8-fc50-456d-98cc-00718a44b450_pcm.mxf ; sox 2a111bb8_1.wav trim.wav trim 0 0.4 ; sox -v 200 trim.wav trim-v200.wav ; play trim-v200.wav
The original (quite faint) signal is clearly audible over a standard auditorium PA @ 7.0. Well, like I said: It kind of has its use
Posted by Carsten Kurz (Member # 5396) on 09-30-2012, 08:40 AM:
Yes, I can hear the short 'organ' beep just a fraction ahead of the very first sync click, even on my laptop speakers.
No big deal for the purpose of the sync check DCP, but it IS there in the audio.
Should keep no one from downloading and using it, though.
- Carsten
Posted by Carsten Kurz (Member # 5396) on 09-30-2012, 08:15 PM:
(Sorry, can not respond by PM to your PM it seems)
---
Oh, and Jussi, I would propose another test pattern as well - a crop test pattern, a 2k full screen image (2048*1080) with numbered rulers on every edge.
Like e.g. this:
http://www.cine4home.de/tests/projektoren/SanyoPLVZ3/cropping_dig2.jpg
- Carsten
Posted by Jason Raftery (Member # 6291) on 10-06-2012, 09:13 PM:
Thanks for the new DCP.
I have just one location that uses GDCs right now and am there infrequently at best, but I'll try the new DCP next time I'm there.
The files were in the root directory of my flash drive. The server was able to see and ingest the TI DLP logos from here that were on the same flash drive but curiously could not see or ingest Dolby 3D and Dolby 7.1 DCPs that were on the same drive (as well as the sync check DCP).
Posted by Dave Macaulay (Member # 813) on 10-11-2012, 06:53 AM:
I finally downloaded the 7.1 version, and found it won't ingest on the two Doremi servers I have tried. I've been using the original 5.1 version so obviously it ingests fine.
The only difference I can see is that in a directory listing of my usb stick, the original one shows with the usual "UUID" string of a lot of hex numbers. The new one shows up with the DCP name. Should the DCP be in a folder with the UUID? I didin't do anything different between the two, just unzipped the file downloaded from FT.
Powered by Infopop Corporation
UBB.classicTM
6.3.1.2