Yes, I did search for this topic, and received no joy.
My question is this: how is it possible for DTS audio to slowly drift out of sync with the picture? We are showing EP2. The distributor insists that we play it in DTS, despite the fact that our largest house also has SDDS and SRD. The technicians come and set the sync, and it's perfect. When they leave, we get complaints. I don't see any problem for quite some time, until I walk into the cinema and the sound is around 1 second out of sync. I rush up to the booth to switch to SR, and by the time I get there the sync is perfect again.
I know that if the time-code is lost for more than a few seconds, the switch back to SR is automatic, so I'm assuming it's not that.
What we have is a DTS-6D and a Dolby DA20 hooked up to a Sony 3000 SDDS unit. All the readers are on the penthouse.
Any tips on this? Even if you can just give me a decent way to trouble-shoot this? I have one idea - monitor the green LED on the reader, using a PC. Is there a way to check the status of this LED, perhaps via the 9-way D-Sub connector? I don't want to have to open the reader up and connect directly to the LED.
------------------
"One man can make a difference."
Steve
------------------
"Old projectionists never die, they just changeover!"
--jhawk
Frank
What would they rather do, play it in SRD while you work out the DTS problem, or give back their 90% of every ticket that is given back for a refund?
I agree with Steve, probably your rotary switch is full of dust and/or oxidized. Try clean it.
I had a similar problem for that.
Bye
Antonio
> I only noticed after writing this post that you had already written that you were bypassing the other readers, so what I just wrote does not apply to your problem, but you should check anyway!
What model of DTS player do you have? And, it's serial number?
Is the whole film out of sync or some reels?
Do other films you play also have the sync drift problem? If not, it is a print problem and you should request a new print.
If other films do have the same problem, write down the OFFSET setting. Then, get a small flat bladed screwdriver and rotate each (of the two) rotarty switches two full revolutions. Once that's done, return to the OFFSET setting that you wrote down.
Sync is determined by counting the number of frames from the DTS reader's lens to the picture aperture and mutiply by 1.25. Subtract one from the result and that is your OFFSET setting. Be sure that is what is selected by the rotary switches on the player.
When threading, make the same size loops as you did when measuring the OFFSET.
Sync drift is not caused by a faulty reader or bad tension. It could be the rotary switches or the timecode board itself. Please send me an email for more information.
Karen at DTS
khultgren@dtsonline.com
What I intend doing tonight is hooking up two processors simultaneously - DTS-6D and DTS-6AD - to the same reader, then comparing their center channels while running through the film. I've borrowed discs from another site. Results tomorrow.
By the way, is it true that the DTS disc may have a backup soundtrack? I sincerely doubt this, but this seems to be the understanding of the manager here, who spoke to the technician.
The lip sync problem has been apparent with other prints (E.T.), but it was simply 'out', then it was set 'in'. What we have confirmed now, only with Star Wars, is that it is, in fact, drifting.
I would test the whole thing out and find no problems....
A few days later they would phone and say they had wow
and out of sync.
DTS suggested I upgrade the firmware to the latest version as this was a known problem, which I've done and no complints since.
I've also done the same at my own theatre where we have two DTS 6 units. NO problems of any kind EVER!
I can't rememebr excatly what the firmware version I upgraded to
is (possibly 1.24 from memory?). It may be worth checking that out.
The IC is normally clearly labeled with it's version number so
check it out. DTS are VERY helpful in evey way so give them a call and they will keep you right and supply the required IC's.
Regards Ken.
The DTS backup track is the analog track. When there is a problem reproducing the sound the player switch automatically in Dolby SR. This will happen if timecode is lost for more than 4.5 sec, CD rom problem or hardware failure. If your offset switches are bad, the player will see a different offset than what you set but it is not considered a fault! I cleaned the switch with a spray contact cleaner. I had the sync changin continuously during a show... 
Do the A/B comparison and let us know!
Karen,
I wish add to Michael question: should count the frame when shutter has illuminated the print the first pass or the second? 
Bye
Antonio
I expected that there must be a processor fault, because:
* We had run the print on several occasions in other houses, using the 6-ADs, and it was flawless.
* We had reports of similar problems from another site that has a DTS-6.
* The troubleshooting guide from DTS mentioned this 'drift' problem and suggested replacing the timecode board.
So, naturally, if it were a processor fault, I would expect that the faulty one would drift way out of sync with the working one. What really happened?
Both processors drifted out of sync, together! I don't even remotely know how that's possible. I mean, if the timecode is lost for over 4 seconds, the sound will cut out. How can even an intermittent time-code cause the sync to wander, slowly?
The technician who came by yesterday says that he aligned the reader. I asked if he aligned it with an oscilloscope, and he said that, well, the reader is something that either works or it doesn't. Since he was unable to correct the problem, and can't explain it, he now says that we don't know how to lace the projector correctly. That an incorrect upper loop size will cause a constant sync error (not a drift) is logic that seems to have eluded him. That the other three DTS houses work perfectly also doesn't phase him.
Karen, this is a mystery to me - do you have any ideas? We're talking two different processors out-syncing simultaneously!
Are you saying that the SAME print in a different house is ok and on that houses with two different processors has a sync problem?
Strange...
Wait for Karen's reply... 
Bye
Antonio
Antonio, yes, that's what I'm saying. And I'm saying that a processor with no problems at all, when moved to the troubled house, drifts out of sync too.
>Wait for Karen's reply... 
That's a given! 
Michael:
>You also checked if the unit was still playing dts when the drifting
>occurred?
Absolutely - there was NO analogue sound available. I hooked up ONLY the time-code to the two DTS processors, and monitored their outputs.
>Remember that analogue can also be a little bit out of sync when the
>loop below the intermittent sprocket is very big or very small!
Again, analogue wasn't an issue - it was non-existent.
>If
>you are dealing with a very large audtorium, that can also add to
>the problem. In that case, you cannot check sync while listening to
>the monitor and watching the movie from the booth window because in >the back of the auditorium, the sound will be late.
I agree, but the sound shouldn't drift that far out on the monitor.
Adam Martin:
>Is the timecode cable firmly seated and screwed down in the reader
>and in the DTS unit? Have you tried grounding the reader to the
>projector chassis with the wire provided on the timecode cable? Are
>the timecode cable connectors corroded? Have you tried a different
>timecode cable?
Adam, I made a completely new cable for the test, because I needed to run two processors simultaneously. So the old cable is not a suspect.
Very scientific, your responses were. But a scientist, I am. Already addressed in my test, your concerns have been. Meditate on this, I will.

Damn - I've watched that film a couple times too many! 
Does it always fail in the same place by the same amount? What's the magnitude of the drift, by the way? Are we talking 500ms or 5 seconds, or...?
Have you tried cleaning the optics on the timecode reader?
--jhawk
Brian, do you have another reader? If so, try to swap the "faulty" one and see if the problem disappear.
More: does the "Timecode" green led on the DTS6 is blinking all the time or is it steady?
Let us know!
Antonio
"I'm so cool; when I go to sleep, sheep count me..." I could even hook the time-code output to a microcontroller with a 7-segment display. This would display the time-code being output. But, because the time-code is not going to vary with one print (unless you slice the film down the middle and shift it a couple of frames, which, with some projectionists, you never know
), I don't see this as a lucrative avenue. I might try it just to learn something.What I'd need to know is the data format of the time-code. Anyone know this? Is it standard SMPTE?
Answer to your question: I don't know if it always fails in the same place, because it doesn't always fail. But when it does, goes up to more than 500ms.
Antonio: I would have swapped readers, but the technicians have bolted a film roller to the top of the reader. This won't suit any of our other projectors, because the reader in this cinema is uniquely on top of both the Dolby and SDDS readers. The LED does flash on occasion, and the sound does drop out, despite the tech-head assuring me that he'd adjusted it.
Gordon: The problem is [b]not[/i] likely to be an intermittent time-code board, because it affected both the DTS-6D, and DTS-6AD equally. Unless there is a board in the reader to which you're referring.
Thanks for the useful responses. I await Karen's response.
Rollers on top of the reader shouldn't be a problem. If I remember correctly (it's been a while since I've worked with DTS), the reader itself is attatched to a bracket that the rollers are bolted to. Just remove the two readers from their brackets and swap them.
Gordon, the problem is not the board in the processor, because both processors were affected equally. The second processor was from another house, and is a known working.
So I would say that if your reader has big problems perhaps you can reach also sync problem...
Before aligning the reader be sure that the reader itself is aligned with the projector and with film path. I would not recommend to correct the alignment of the reader/projector (physically) with the lateral screw, use that to correct the alignment of the reader.
Bye
Antonio
So, if the DTS signal is at RS-232 levels (at least 6V, and up to 24V difference between positive and negative levels), then you should be able to get a very decent run. My test cable was about 5m long, but the installed cable may be up to 10m. (16' / 32')
I'd really shocked if it wasn't within the RS-232 voltage spec, though (you could easily enough measure to be safe).
Sure, ideally you'd know the format. If anything's clear, we know it's not SMPTE timecode, since there's a special jumper for that. But you don't even really need to know the format (even if it is not RS-232, you don't really care). All you have to do is timestamp the data from the reader and look at the differences, and see if the rate of timecode coming from the reader changes. If they're opaque blobs or if your UART shows them up as "~~{{{{", as long they're regular, then you can make a useful measurement if you look for variations in that regularity.
But if the LED is flashing, then you have even more reason to suspect a problem with the reader. Assuming your print doesn't have a problem with the timecode, you should never have a flashing LED...
--jhawk
Karen at DTS
Karen at DTS
I posted a lot of information for you in the first page of this thread! I also asked you to email me. I prefer to not handle on-going discussions or make lengthy postings on message boards.
Drifting sync is not caused by the length of the cable. It is possible for it to be caused by a bad reader but I have never seen that. If the reader has a problem, its green LED will start blinking. The voltage to the reader's LED should be checked and recalibrated when necessary. Details are posted in our online techcenter at www.dtsonline.com/cinema
It is possible for the player's timecode board, D422, to have a problem. I have seen that but it is not very common.
If sync drifts, it is most likely to be a print problem.
As mentioned in the first posting, if the entire print is out of sync, then either (1) the threading pattern has changed -or- (2) your OFFSET is not set correctly and needs to be recalibrated.
If only part of the print is out of sync, then you have either a print problem or hardware problem. The way to tell is if you move the print and the problem stays with the print, the print is bad and needs to be replaced. If the problem stays with the hardware, then you need to look at that.
The other possibility (sync temporarily off) is that is the way the sound track was intended.
I believe you said that you moved the reader and the problem followed the reader. If that is true, then the reader should be swapped (contact your local dealer for a repair/exchange).
If replacing the reader does not help, then the D422 timecode board inside the player is bad and should be replaced.
For more inforamtion, PLEASE SEND ME AN EMAIL.
Karen at DTS
khultgren@dtsonline.com
One of the other techs here at work was reading this thread and reminded me of another possibility (HI MIKE!). Be sure to verify that the speed of your projector is constant. If the projector speeds up or slows down, it could effect the sync.
Karen at DTS
Have you checked the firmware version yet?
However, the tech was knowledgeable enough to ponder whether it might be a power-phase issue. Bingo! If the Christies use the power phase to determine their speed, then that may well be it.
How to check the speed of the projector? Well, since quartz crystals are very accurate, I could use a quartz crystal-based battery-operated microcontroller circuit to count frames or perforations / sprocket holes and graph the frame-rate throughout the performance. If I graph that, I will post the graph here. Question to the Christie gurus: where can one find a frame/perf timing signal from the P35GPS that we can use to time it?
Sorry, I haven't checked the firmware version yet. Being in South Africa, it would not be a quick fix to change it, unless I write the EPROM myself. Unless, of course, the DTS-6D allows you to link your laptop for a firmware updatte... I will check the version and get back.
At the end of the day... the reader either works or it don't! no if's no but's so that can be counted out.
When you think about it if the projector is running too fast, or for that matter too slow.... or even irregular how can the DTS keep up with it! The time code is fine, it knows where it is in the film. thats the easy bit. But what does the DTS unit do. Does it play back the sound faster or slower.... can it maintain the original pitch of the audio while its trying to keep pace with the position it should be in the film or does it just drop frames to keep sync.
Do your projectors have a speed control on them..... to allow 25fps running etc. if so cut the fingers off anyone playing with it...not very practical but satisfying none the less.
Check that everything is as it should be >> 24fps <<. Until you can be sure you have common ground then you may never find the solution.
But that's irrelevent. There is no requirement for consistent frequency of power between the DTS player and the timecode reader -- the player synchronizes itself to the pulses from the timecode reader. (I have an unfortunate problem of 2 projectors that run at slightly different speeds, and DTS handles it just fine, though in some cases you can hear the pitch bend across the changeover.)
So even if you run your projector at 25fps, that's not going to cause this problem. And ground is even less plausible to be relevent.
And no, the DTS 6D firmware is a socketted EPROM, not an EEPROM. To reprogram it you would need to remove it from the socket and burn it in an EPROM burner, and at that point, you'd be much better off just getting DTS to ship you a new IC.
--jhawk
I had a free firmware upgrade from DTS, contact Karen and she will help you. However, since both player went out of sync in the same moment I do not believe that firmware upgrade will help you in this case.
Perhaps proj speed is a thing to check.
Be sure that print enters the reader freely (it doesn't have violent pull that can cause the reader to see the print changing its velocity).
I had problem for a reader not firmly screwed to the projector.
Bye
------------------
Antonio Marcheselli
Florence, Italy
I did ask all on this forum: How can one count frames on a Christie P35GPS? Is there anywhere I gan get a frame-passing signal, that I can pass to a sub-circuit that does the counting?
Anyone know this? Or do I have to start a new thread?
A suggestion on how to count frames, why not lace up a length of leader through the projector, then mark the appropiate frames, unlace it then count them that way... Just a thought.
------------------
Lewis Bohannon
"The Man Up-Stairs"