This is topic Audible effect of high SRD error rate? in forum Film Handlers' 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=1;t=003096
Posted by David Stambaugh (Member # 1102) on 02-16-2002, 11:06 AM:
If the SRD reader is experiencing a high error rate, but still getting enough data to work with to avoid a hard dropout, will the high error rate cause audible "defects"? I'm talking about tiny, split-second "blips" in the sound. This is especially noticeable during music containing sustained notes from string instruments. There's just the slightest audible blip in the sound. Might occur as often as once every couple of seconds.
------------------
- dave
Look at this! His chin strap has been cut!
Posted by Joe Redifer (Member # 3) on 02-16-2002, 11:56 AM:
Yes. Absolutely. I have heard this many times before. You will hear a teeny tiny "blip" like effect at super high error rates, even if it never drops to analog. You will hear a much harder blip when it actually drops out even for a split second. This is one way that I can always tell a film is playing in Dolby Digital. Of course if it is getting good readings, then you won't hear the blip and all will be fine.Good ears! You are the only person besides myself and Erika (another member on the forum from long ago) who has seemed to notice this. Once again I'd like to stress that what we are hearing is not SR-D dropping out of digital. This has been confirmed by Erika staying in the booth watching the processor and me listening to the blips in the auditorium. Usually the Dolby processor just reads "F" briefly when the blip occurs. Never goes to analog. I always like to say "I can hear the Dolby Digital F'ing!"
Posted by Jesse Skeen (Member # 586) on 02-16-2002, 12:33 PM:
Yes, I've noticed this too. The LED on the reader was turned up and that seemed to take care of it. I also always notice that funny "buzzing" sound it sometimes makes when a splice goes through.
Posted by David Stambaugh (Member # 1102) on 02-16-2002, 12:58 PM:
I've noticed this in at least 2 of the local Cinemark's SR-D auditoriums. It's as if the error correction system is momentarily overwhelmed and has to throw out a very small sample of sound, and can't conceal it. Kind of like a break in continuity.------------------
- dave
Look at this! His chin strap has been cut!
Posted by Antonio Marcheselli (Member # 373) on 02-16-2002, 02:06 PM:
Yes, I noted too. When the "blip" occurs, the "block error" red led into the CP500 will light on for a moment.I was in a theater that had continuing "blip" during the show because they had oil on the reader.
I wonder why CP500 allow that kind of stupid thing: I believe that after few "blip" it is better to drop to SR: if the "blips" cannot be recognized during dialogue, they are very annoying during music scenes.
Bye
Antonio
Posted by Ray Derrick (Member # 1001) on 02-16-2002, 04:01 PM:
How many block errors can be tolerated before you take the rather drastic and far more audible step of reverting to analog would have been a software design consideration for the Dolby engineers. It is likely that they did not regard a rate of block errors below a certain level as bad enough to warrant reversion.Has anyone noticed if this problem is software version related? If it is only on say, later versions, there is always the possibility that it is a bug and Dolby don't know about it.
It is also very important for projectors to run as close as possible to correct speed because, unlike the optical soundtrack which just shifts up or down in pitch, a digital system is designed around a constant data rate and so must use some method such as software "splicing" to correct for the speed difference. This can cause audible artefacts if the speed variation is too great.
------------------
Ray Derrick
President/Chief Engineer
Panalogic Corporation Sydney, Australia
Phone: 61 (0)2 9894 6655 Fax: 61 (0)2 9894 6935
Posted by David Stambaugh (Member # 1102) on 02-16-2002, 04:15 PM:
Ray, "splicing" is a good description of the artifact. It sounds kind of like a bad analog tape splice. But if it were a digital problem caused by the projector running too fast or slow, wouldn't these blips be more likely to repeat at a constant interval? What I'm hearing is basically random.
Posted by Joe Redifer (Member # 3) on 02-16-2002, 08:05 PM:
This problem seemed to vary from print to print with me. I'd start checking out a new print, smooth music would come on and then I'd hear "blip" "blip" "blip" and I would go insane! I then ran the movie in SR since it sounded so much better that way until I could move it to a DTS house. Realigning the readers seemed to minimize or eliminate the problem.I can sometimes hear a similar effect when I am watching UPN. Not sure if it's just my station or the network itself, but it was very prominent during Star Trek Voyager. Drove me up the wall because there was no "SR" button I could press. I'm sure it was a digital error in whatever low-end system UPN uses.
Posted by David Stambaugh (Member # 1102) on 02-16-2002, 08:35 PM:
The severity of the SR-D problem seems to vary from one reel to another. If the reader is marginally out of whack, maybe normal lab processing variations between reels would make the problem better or worse? I've noticed the same thing with UPN and a couple other cable networks. My theory there is they are running the show through some kind of "time compressor" so they can squeeze in more commercials. 
Posted by Ken Jacquart (Member # 369) on 02-17-2002, 12:55 AM:
What you guys are hearing are "block repeats". If a Dolby Digital block cannot be read, even after all error detection/correction, then the previous block will be repeated in its place. A Dolby Digital block occurs every 1/96 per second (or every 10 ms). A repeated block will consume 10 ms of time and can barely be heard if you are paying attention to it. If two blocks are repeated in a row, then it is a little bit easier to discern (albeit very slight). If the processor cannot read 3 blocks in a row, it will revert to SR (3 strikes and you're out!). So, whenever you see that brief "F" on the Cat 673 card, you will hear a block repeat (or two). If this occurs on a regular basis, then you probably have a problem with alignment... namely, most likely, video amplitude. Drag out the scope and check it... good ears on those block repeats!
ken
Posted by Steve Guttag (Member # 268) on 02-17-2002, 01:49 AM:
Like everything else in software, it seems that there is always a senerio that beats the system...I once had a bad reader/print situation (a penthouse reader at that!!!) where the block repeat was just enough to be annoying yet not enough to have the DA-10 drop to analog. So it sounded like one was virtually under water for the last reel of Harry Potter. Using DRAS...looking at r-9 the azimuth jumped from say -1 to about -8 and stayed there! One reader did ok with it, the other, did not!
I've recently taken a liking to the skinny splicing tape either 16mm or 19mm...no drops and no seemingly ill affect regarding print longevity.
Steve
Posted by Brad Miller (Member # 2) on 02-17-2002, 01:57 AM:
I still fail to see why you need to use the skinny clear splicing tape to prevent your SRD from dropping out when I have used full opaque tape before with no dropouts at all.
Posted by Steve Guttag (Member # 268) on 02-17-2002, 03:18 AM:
Yes, you can use opaque tape and not revert but the fewer blocks you cover, the fewer repeats, the splice also has less hop on the screen. Personally, I was sad when cement spliced stopped working (read polyester)...between perf splices worked for me!Steve
------------------
"Old projectionists never die, they just changeover!"
Posted by Brad Miller (Member # 2) on 02-17-2002, 03:47 AM:
Hey, just get an ultrasonic splicer for those ultrasonic prints. It's even cheaper than a Kinoton OR a 2nd projector!!! 
Posted by Antonio Marcheselli (Member # 373) on 02-17-2002, 05:20 AM:
The projector that had the block repeat had some dirty on the sound drum and the print jumped regularly on the digital reader. On the CP500 I had 5-f-5-f-5-f.
I was surprised because 5 was not a bad number and I was wondering how can the reader travel between 5 and f.
Then, I found the dirty and I resolved the problem.Pls note: there were not my projectors, there are no dirty on my projectors...
Bye
Antonio
Posted by Gordon McLeod (Member # 33) on 02-17-2002, 10:56 AM:
The problem with both cement (aka wet splice) and UltraSonic is the loss of 2 frames every time the print is made up and broken down
Posted by Mark Gulbrandsen (Member # 72) on 02-17-2002, 02:25 PM:
As a dealer with alot of customers in remote areas I'd also like to express my concerns of what seems like more than normal the amount of prints with high SRD error rates, or even plain defective prints that are bad for sections then go on playing fine. We are considering selling DTS players exclusively for this very reason. We feel that the possibnility of not getting a disk for a day or two is way less hassle for any of our customers than the SRD stuff has been.
Mark @ Home
Posted by Gordon McLeod (Member # 33) on 02-17-2002, 03:09 PM:
Mark a while ago your policy was SRD only because you couldn't get disks and bad drives what gives
Posted by Mark Gulbrandsen (Member # 72) on 02-17-2002, 03:34 PM:
Gord,
This area has received an onslaught of bad SRD prints lately. Its a come and go thing and U never know when its going to come. My change in stance is the result of all the problems that I have had with SRD Digital only based processors and the manufacturers lack of cures for the problems at hand. SRD units are THE only equipment that I sell that I get frequent complaints about from our customers. I had no probs. with CP-500's till a few years ago when all of a sudden they started having alot of problems. Some you know about, and some of the problems have been experienced by yourself as well. I guess that I'd have to say that I too got caught up in the this SRD digital package is the best bang for the buck thing too. But all they've done is to go bang and in all the wrong places! Yes, I still have alot of DTS units come in for upgrade too, some with other problems as well. And some customers still do not receive disks with the films. I have an upgrade kit on order right now, and a cabinet fuull of drives, boards and cables that I've replaced from other units. Someday I'll weld them all together for a nice artsy display. However, once the DTS units are upgraded I never see them again, or get complaints from our customers about them. Having been a DTS Authorized Service Center since I came to GTS I can also say that DTS is also way ahead of the game with dealer/customer service and they seem to be more into into rectifying a problem when it comes into being. There is ALWAYS someone at the other end of the phone that can help solve a technical issue! Other cinema equipment manufacturers could learn a lesson or two from them and I commend them on their service and support to their dealers. On another note....Would I ever sell a DTS 6AD? NO! Never!
Mark @ Home
Posted by Steve Guttag (Member # 268) on 02-17-2002, 06:22 PM:
>>"Would I ever sell a DTS 6AD? NO! Never!"<<And yet amazingly enough, I would! It actually has many rather nifty features in it.
Steve
------------------
"Old projectionists never die, they just changeover!"
Posted by Ray Derrick (Member # 1001) on 02-18-2002, 07:28 AM:
Ken said "If the processor cannot read 3 blocks in a row, it will revert to SR (3 strikes and you're out!)". Maybe Dolby should revise their software by also including "3 block repeats and you're out"!
------------------
Ray Derrick
President/Chief Engineer
Panalogic Corporation Sydney, Australia
Phone: 61 (0)2 9894 6655 Fax: 61 (0)2 9894 6935
Posted by Steve Guttag (Member # 268) on 02-18-2002, 08:36 AM:
I'm not sure Ken's statement is entirely accurate. Since a 1" opaque tape can be used without a reversion, more than 3 blocks are obstructed, in theory, every splice should cause a reversion in the 3-strikes and you are out scheme.Steve
------------------
"Old projectionists never die, they just changeover!"
Posted by Ken Jacquart (Member # 369) on 02-19-2002, 05:05 PM:
You guys don't miss a thing!
You are correct, I mispoke or should I say, mistyped...
What I should have said was, The processor will revert if it cannot read 4 blocks in a row... The three strikes and your out rule referrs to another reversion scenario where if there have been three reversions in a short period of time, the processor will remain reverted for a time before it attempts to come back to Dobly Digital (it assumes there may be a bad section of film and doesn't want to irritate the audience with many reversions). Sorry about the confusion.
Here is another one for you... Many films today have "splice caching". Throughout part of a given reel, audio data relating to several feet of film around the splice, is read into memory. Upon reel change, the processor will revert to internal memory backup of these few feet (if need be... if it would have reverted) instead of reverting to SR. So, if a given film has this feature enabled, then we don't care how bad of a splice you make, it will playback good data from its memory.ken
Posted by Brad Miller (Member # 2) on 02-19-2002, 05:31 PM:
Ken, is that done only at reel changes, or continuously throughout the reel?
Posted by Mark Gulbrandsen (Member # 72) on 02-19-2002, 06:12 PM:
Steve,
You can go ahead and sell all the 6AD's you want to. They are all over the Edwards chain out here and its the ONLY piece of equipment in their theatres that regularly suffers from the hickups. As I stated before, in the interests and well being of GTS's customers, I will stick to an analog processor with outboard digital whatever they want. That way their show will always go on.
Steve, All processors out there have at least some "nifty features", even a CP-50 does for that matter, but its not those extra nifty features that sell the tickets.
Mark @ Old GTS
Posted by Joe Redifer (Member # 3) on 02-19-2002, 08:13 PM:
I would assume that the splice cache feature is only for reel changes as it would take quite a bit of memory to buffer everything in a particular reel. Though with some tricky programming on the encoder's part it COULD be done, I am betting.
Posted by Steve Kraus (Member # 476) on 02-19-2002, 09:08 PM:
Joe espoused:
>I would assume that the splice cache feature is only for reel
>changes as it would take quite a bit of memory to buffer everything
>in a particular reel. Though with some tricky programming on the >encoder's part it COULD be done, I am betting.No because you could not send that much data in background. By the same means that data can be delivered to update the Dolby Digital playback firmware, they will slowly build up in memory the data necessary to cache the entire reel change. I've forgotten the bitrate for Dolby Digital on film...320kbps comes to mind. They'd probably want to cache a few seconds on either side of the splice. Whether that data is sent in background slowly over the entire course of a reel (along with updates and other goodies) or just in the minutes before someone else will have to explain.
At least that is how I interpret it. As always, reality may be at variance with anything I say.
Posted by Joe Redifer (Member # 3) on 02-20-2002, 01:15 AM:
Dolby Digital needs huge hard drives installed in every unit that you can load the digital sound on from CDs so it can revert to that if it fails. Oh wait.....
Posted by Phil Hill (Member # 371) on 02-20-2002, 02:33 AM:
Joe, Get ready for the new DTS systems. It will have hard-drives loaded from CD's. Trust me, I know. Of course I now will be killed!
>>> Phil
Posted by Steve Kraus (Member # 476) on 02-20-2002, 09:31 AM:
No reason the other two digital systems could not incorporate HD's as well. Continue to read the track for synch purposes and if subsequent reads produce a block of data with less errors then swap it for the one already stored but presuming that, in general, the best read is the first one when the print is brand new you'd be mostly playing from the stored data. While we're at it let's add a LAN so that when prints get moved the system can check if any other CP at the site has the data and can import it. All possible of course but at a significantly increased price.Now, more seriously, a couple of questions for Ken that I forgot to ask: When a film with the caching feature comes along, is this feature invoked immediately on the first run or does the initial run just set up the software so that it then happens on subsequent runs? And how does this technique apply to changeover situations? I'm curious whether my own setup benefits given that it's a screening room and generally a film does not screen more than once in a row.
Posted by Karen Hultgren (Member # 395) on 02-21-2002, 11:52 AM:
Ah-oh! tisk-tisk Phil ! Giving away DTS secrets! ;-) Yes, a new DTS player is due out this summer, and yes, it will have a hard drive. When its info makes it to our site, I'll let you guys know.
Karen at DTS
Posted by Pete Naples (Member # 761) on 02-21-2002, 12:54 PM:
I have got 4 of hese DTS 6AD's under my care up here, these things were installed and other than routine tweaking, they have never skipped a beat, which is more than can be said for the 4 CP500 in the same building. Steve's right, there are some clever wee touches in there, copying eq from channel to channel for example, that can save a heck of a lot of time, and result in a better overall eq. I notice the the CP650 software can do this too now though. It's a shame that the theatre can't get discs, they are intending to replace the DTS's with CP650 as soon as funds allow, because they advertise that they have digital sound in all theatres, which is true, but if your in one of the DTS theatres, there's a 99.9% chance that you are hearing analogue, which leaves the owners open to all sorts of flak, trades descriptions etc.
Posted by Brad Miller (Member # 2) on 02-21-2002, 03:19 PM:
Oh Karen, you've REALLY screwed up now! Now you simply must tell us the one nagging question...Will the new dts unit have enough hard drive space to hold at least TWO full length movies and a couple of trailer discs worth of trailer audio files so that double-ups in one auditorium do not require constant handling of the discs between every show?
(Ideally I would want 30gig or larger drives to store EVERYTHING in the average sized complex, with the option to upgrade to larger file storage for those big 30 plexes.)
Posted by Antonio Marcheselli (Member # 373) on 02-21-2002, 03:29 PM:
Brad,I believe that hard drive in the new player will be 20GB minimum. Where do you can found 10gb disk right now?
Bye
Antonio
Posted by Joe Redifer (Member # 3) on 02-21-2002, 03:51 PM:
Antonio is right, it's really hard to find a hard drive that doesn't have the space to do what Brad suggests. But the question is "will the DTS system ALLOW you to do this?" Personally I don't see why not. Without the timecode the data would be useless anyway. Maybe the trailers will be the main thing residing on the hard drive? We'll see. Can't wait until the info is available. Let's get our hopes up now so we can complain later!
Posted by Ken Jacquart (Member # 369) on 02-21-2002, 07:17 PM:
Splice Caching:
Throughout each reel of film, there are "bogus" blocks, which are in essence, extra blocks. With this extra data, we utilize the "dynamic load" feature (which can upgrade the DA20 portion of your processor), and the splice cache feature. It takes only several minutes to build up enough data to encompass 2 to 4 feet around the splice. We repeat the "several minute cycle" throughout the reel to ensure that it all gets read into memory. yours for happy reel splices,
ken
Posted by Phil Hill (Member # 371) on 02-21-2002, 07:32 PM:
"Excuse me while I whip this out..." Sorry, watching Blazing Saddles. It would defeat the purpose of the "other two digital systems" to have to include a disc with each print to load a hard drive. Their claim to fame is they are self-contained on the film. A single system whereas the DTS is a double sound system. Both systems have respective advantages (and disadvantages).
>>> Phil
Posted by Scott Norwood (Member # 30) on 02-21-2002, 08:02 PM:
So, Karen, where is my royalty check for suggesting the "DTS unit with built-in hard disk" concept in the first place? (check the film-tech archives--it's there!) Seriously, though, that sounds like a great idea that should solve many (most?) of the complaints about the DTS hardware. I'm sure that many of us are looking forward to seeing it.
Posted by Steve Guttag (Member # 268) on 02-22-2002, 10:47 AM:
Mark, I can't claim to have a lot of DTS-6ADs in my service but I don't seem to be having problems. Perhaps air in your part of the country doesn't allow for integrated analog/digital processors.
Steve
------------------
"Old projectionists never die, they just changeover!"
Posted by Steve Kraus (Member # 476) on 02-22-2002, 11:21 AM:
Hey, I was suggesting a problem with all that clean mountain air that Mark breathes years ago. I mean as long as we are taking credit for past ideas. hehehehe
Posted by Mark Gulbrandsen (Member # 72) on 02-22-2002, 12:12 PM:
You two should be so lucky as to breath the quality of air that we have out here!
Mark @ Old GTS
Posted by David Stambaugh (Member # 1102) on 07-18-2003, 04:39 PM:
Bumping this up since the subject of high SR-D error rates came up in the SDDS thread. ![[Big Grin]](biggrin.gif)
Posted by Aaron Garman (Member # 1653) on 07-19-2003, 02:08 AM:
I remember sometime last fall, one of our SRD houses was constantly doing this 'blip' and reading F on the processor. Eventually, it got to the point where it was dropping out of digital all the time. Of course, this was the only thing that got anyone to notice. All we did was put in a new bulb inside the Cat. 700 reader and boom: it was reading 1's and 2's. It has been fine ever since.
Recently, our other SRD house is beginning to do this. We are gonna try the same thing again, but I was told we have to order the bulbs. Anyone ever try this?
AJG
Posted by Gordon McLeod (Member # 33) on 07-19-2003, 11:19 AM:
Eventually the lamp degrades the reflector flacks and the fialment sags (just like the obslete excitor lamps for optical sound)
When this happens either the illumination is too uneven to grap the complete block or two low to get sufficient level