This is topic SDDS backup track defeat 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=002455

Posted by Brad Miller (Member # 2) on 07-05-2001, 03:15 PM:
 
I am setting up a sound rack with SRD/DTS/SDDS and am configuring them to be digital failsafes, should one format drop out during a show. The room is set up for 8 channels, so I wanted to make SDDS the primary audio source, defaulting to the other two digital formats as a backup (instead of analog). There is just one problem. SDDS incorporates their own "backup" track, so that when the primary "high quality" track has a dropout, instead of falling back to SR it falls to the SDDS backup "low quality" track.

This is a really stupid idea and quite honestly, it sucks. I don't know how others feel, but the backup SDDS track in my opinion sounds WORSE than just falling to SR. To make matters worse, I sometimes hear some of the channels dropping to the backup track, while the others that are not experiencing a dropout continue playing in the primary track. This is horribly distracting during a show!

Question #1:
How can this backup SDDS track be 100% completely and absolutely defeated at all times? I do not ever want to hear it play again. Like I said before, analog SR sounds better...but in this particular installation, if the SDDS falls back to it's backup track, there are two more digital systems just waiting to "jump in and take over", but that won't happen until the SDDS drops through it's low quality track. Thus, the SDDS will have to be placed last in line of the 3 digital units (pretty much defeating the 8 channel) if this can not be defeated.

Question #2:
Is there a way to modify an SDDS player so that it will function like a SRD unit, in the idea that it will not attempt to resume digital playback until "X" number of seconds after good data is received?

Anyone?


 


Posted by John Pytlak (Member # 331) on 07-05-2001, 03:26 PM:
 
Dumb question: What would happen if the SDDS reader just couldn't "see" the backup track? For example, what if that side of the optics was blocked? Does the reader need to "see" both tracks to play SDDS, or will it play the primary track in full SDDS digital even without the backup track being readable at all?

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Cell: 716-781-4036 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion
 


Posted by Brad Miller (Member # 2) on 07-05-2001, 03:27 PM:
 
You mean like putting a piece of tape over one lens, right? If I remember the configuration correctly, each side of the film has 4 primary channels and 4 backup channels (to backup the 4 primary channels of the opposite side). So unfortunately, that wouldn't work (unless I have info that screwed up in my head).


 


Posted by John Pytlak (Member # 331) on 07-05-2001, 03:41 PM:
 
Here is what I found on the Sony SDDS website:
http://www.sdds.com/tech/prodspec/dfr2000.html

_____________________________________________________________________

"Eight Discrete Digital Audio Channel Capability
Up to eight discrete digital audio channels (L, LC, C, RC, R, SW,
SL, SR) are recorded on the two edges (S&P sides) of the 35mm
film negative stock. The SDDS Recorder System is also capable of
creating six channel soundtracks on the film negative as well.

Advanced Error Correction
The data patterns are Reed-Solomon encoded for error
detection/correction. In addition to this powerful error correction
scheme (termed Blocked Cross-Interleave Reed-Solomon Code,
BCIRC) a unique feature of the SDDS encoding system is the
recording of "back-up" channels of audio data.

SDDS soundtrack information is placed redundantly on both sides of the film to keep the film
playing in digital should damage occur to the film print. The process of switching to the back-up
information is known as digital concealment. To assist in concealment when switching to the
digital "back-up" channels, discrete level codes for the original three channels are encoded along
with the audio data."

_____________________________________________________________________

Somehow, I recall that they did not "split" the data between the P and S tracks. Any film (e.g., damaged edge), lab (e.g., safelight fog) or reader (e.g., tracking, focus) problems would likely be confined to one side of the film, so it would seem more logical to have a complete set of data on each side of the print.

Should be simple enough to cover each track separately and hear/see what you get.

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Cell: 716-781-4036 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion


 


Posted by Brad Miller (Member # 2) on 07-05-2001, 03:50 PM:
 
Ok, then perhaps I did have that backwards in my head. Regardless, as I've been "told" previously, SDDS has a lower quality backup track to fall back to. If this system doesn't have such a track and both are full quality, then why do I hear the extreme high frequencies on random channels cut in and out on sustained audio? It was explained to me a few years back that the cause was "falling back to the backup digital track". Perhaps that was bad information from another tech?

This phenomenon is most noticeable on items like the THX Broadway logo. It sounds as if someone is A/B switching between two EQs where one eq has everything on the highest frequencies completely cut...and only on certain channels! It doesn't happen on every film, but when it does, it is very obvious and highly annoying. (And yes, I've done this test where the film was not even threaded through the analog reader.)

The problem is at it's worst when the reader (2000 series) is threaded "to the red line", as the manual says. When the reader is threaded one perforation less tension, it does not happen near as often and the sprockets of the film are not under so much stress. Actual dropouts to analog are also less frequent (if at all) when threaded one perforation less. Of course, try and demo this to hard headed people and it's like talking to a brick wall.


 


Posted by John Pytlak (Member # 331) on 07-05-2001, 03:56 PM:
 
The "backup" is probably lower in quality. But I recall that one side is the primary, and the other is the backup, and the data is not split between the two. I have a SDDS brochure where the S-track (analog soundtrack side) is clearly labeled "SDDS Track", and the P-track (opposite edge of the print) is clearly labeled "SDDS Back-Up Track".

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Cell: 716-781-4036 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion
 


Posted by Joe Redifer (Member # 3) on 07-05-2001, 04:29 PM:
 
I say block one side and see what happens.
 
Posted by Gordon McLeod (Member # 33) on 07-05-2001, 04:32 PM:
 
Doesn't SDDS work best with both sides blocked
 
Posted by Nic Margherio (Member # 273) on 07-05-2001, 07:07 PM:
 
Blocking one track would only partially solve the problem, but there would be a big trade-off.

The P-side carries the "full quality" information for the L, LC, LS, and SW channels along with a "back-up quality" track for the R, RC, RS, and C channels. The S-side is opposite of this. (Full quality R, RC, RS, and C, backup quality L, LC, LS, and SW) Continuously blocking one or the other tracks would provide for switching to analog if digital info is invalid, but you would be playing four entire channels in the lower quality back-up track at all times. I would not want to watch (listen) to a movie this way. Not to mention that the system would still (if you had a bad reel) jump between analog and digital continuosly, regardless of the length of time valid data is present, unlike SRD.

Are you installing a DFP-2000 or a DFP-3000? If your automation is flexible enough and it is a DFP-3000, you could define a format for SDDS-8, (with DTS & SRD fallback) add an extra cue for 8 channel features, and run the rest of the time in DTS or SRD. This way you would retain SDDS-8 capability and eliminate many potential problems with SDDS-6 tracks. If it is a 2000 (or 2500) you could probably use your little trick you describe in the Tips section to bypass the unit unless you have an 8 channel print. Just an idea.

A little bit off the subject, but I have rejected many SDDS reels that cut between the back-up and full quality tracks, only to have Technicolor bill us for the shipping cost of the reels. Apparently when they get the reels back for inspection, they don't have the ability to monitor the individual tracks, only whether it plays in digital or not. Thus, they deemed that the reels were fine and that the theatre should pay for the shipping costs. ("uneccessary reel replacement") Needless to say I would not allow that to happen.
 


Posted by Brad Miller (Member # 2) on 07-05-2001, 07:11 PM:
 
Thank you Nic...so I'm not crazy then about the track configuration. (The unit is a DFP2000.) That backup track is pretty bad quality, isn't it?

I've been told that there is a way around this, because the labs supposedly have set up their units in this manner to check the printing. Someone out there must know how this is done!


 


Posted by Joe Redifer (Member # 3) on 07-05-2001, 09:06 PM:
 
I wonder... what do people who work at Sony actually think of SDDS? They don't seem to support the format very well (you don't see any Sony reps who post here or anywhere else).
 
Posted by Mike Judge (Member # 558) on 07-06-2001, 04:42 AM:
 
If I were them I'd be afraid to post on here.
 
Posted by Joe Redifer (Member # 3) on 07-06-2001, 05:13 AM:
 
It's easy to attack an enemy who won't defend itself.

To be honest, I think most people would be impressed that a Sony rep was posting. They'd be fools if they wasted their time attacking the person instead of discussing solutions.
 


Posted by John Pytlak (Member # 331) on 07-06-2001, 08:38 AM:
 
Nic said: "The P-side carries the "full quality" information for the L, LC, LS, and SW channels along with a "back-up quality" track for the R, RC, RS, and C channels. The S-side is opposite of this. (Full quality R, RC, RS, and C, backup quality L, LC, LS, and SW)"

If you're correct, I guess my SDDS brochure that identifies the S-track as "SDDS Track" and the P-track as "SDDS Back-Up Track" was written by a marketing person, rather than someone who knew the details of the format. As I said, it would be more logical NOT to split the channels between the two tracks, as any problem would tend to affect one side of the print.


------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Cell: 716-781-4036 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion


 


Posted by Michael Brown (Member # 890) on 07-06-2001, 09:06 AM:
 
Yup, I was lead to believe that one side was the primary and the other a backup track. With both tracks pretty much identical.
 
Posted by Jason Burroughs (Member # 68) on 07-06-2001, 12:54 PM:
 
The arrangements of the digital channels is such:
P=picture S=soundtrack

P side: C L LC SL - Rm [backup] (mixed R RC SR - compressed SW)
S side: R RC SR SWC - Lm [backup] (mixed L LC SL - compressed C)

If for example the P side is damaged and unreadable, it will pull the C L LC SL from the backup contained on the S side.


This way if one side of the film is completely unreadable, is still has a backup of the channels that were on that side. The backup channels are of very high compression and lower quality taking up less bandwidth. I may be mistaken, but some of the channels may be even drived via matrix such as R + LC = C but of that I'm not sure of. Blocking 1 side of the film would only force the backup channels of the affected side. There has to be a software swith that would disable the backup tracks however Sony probalby has those locked in a deep dark corner of the coding, then again it may be hard coded that way as well.


 


Posted by Michael Brown (Member # 890) on 07-06-2001, 01:03 PM:
 
Could sombody please clarify what the SW channel is?

I am assuming that:
L=Left LC=Left center R=Right RC=left center C=Center LS=Left surround and RS=Right Surround.

 


Posted by Jason Burroughs (Member # 68) on 07-06-2001, 01:07 PM:
 
SW = Sub Woofer
 
Posted by Michael Brown (Member # 890) on 07-06-2001, 01:54 PM:
 
Thanks,

I found this on www.sdds.com
-------------------------------------------------------------------
"With SDDS, both edges are used to provide two continuous streams of data interleaved using a cross-redundant error correction technique to further prevent drop-outs from film damage or scratches."
----------------------------------------------------------------------
It suggests that both sides are used and that one side does not function just as a back-up.

So I am slightly confused now.


 


Posted by Jesse Skeen (Member # 586) on 07-07-2001, 01:02 PM:
 
Is there any display on the system that tells whether it's playing the primary or backup SDDS track?
 
Posted by Steve Guttag (Member # 268) on 07-07-2001, 03:08 PM:
 
Where to begin...so much misinformation!

Ok...on SDDS there is no back up side or track in the conventional sense. Both sides in are in use at all times. The only redundant channels are C and SW, that is if either track is covered up, both Center and Subwoofer will play unaffected....but what about the other channels, I hear you cry!

SDDS has what they call "Digital Concealment Mode" or DCM for short (in our current world, why use a single word when mutiple words for describing backup can be used...also the neat abbrieviation DCM)...

What has been said above is essentially true however...there is a back up track (in the singular form) located on each side...it is known as the "prime" track...a bit confusing since "prime" to theatre people nomrally means main or first choice...not so in SDDS speak. The SDDS prime tracks contain a sum. In the case of Left Prime, written as L', you have Left, Left-Center, and Left-Surround all mixed together. You then also have encoded how loud the channel should play on a channel-by-channel basis. However, with film space getting tight and bits available running low...you don't get too many bits for how loud things should be so, and this is only from memory so I may be a little off on this, there are are only 256 "steps" or levels each channel may be played. So, in theory, if the surrounds were playing at say 82dB and Left was playing at 70dB and Left-Center was playing at 0dB...then the summed L' track will play out of those channels at their respective levels within the resolution of 256 steps. If an effect was originally playing out of Left channel only, the surround channel, in this case, would actually be playing it louder than Left while in DCM (82 vs 70dB).

To try and make the system more bullet proof, the P and S tracks are out of sync with each other (17 frames comes to mind but again, I may be off on that number)

Does the system work? Yes, in limited applications...for instance, during a splice, both tracks are blocked but due to the 17 or so frame offset, each track goes into DCM separately and MOST people won't notice it. It was designed to keep the levels pretty close to actual. In fact, it is more seemless than an analog fall back.

But wait, I hear you scream, why do I keep hearing when the SDDS system spazes out (Brad's term, and accurate). Becuase the prime tracks are ok to fall back to for an instant...not to be played on or repeatedly gone to. It is like a magician that moves so slow you see the trick...it then becomes obvious something is wrong.

SDDS should have a better failure algorythm such that if one track is blocked for an extended period of time (say two frames...allows a splice) then it should release the digital system. Furthermore an algorythm should be set up to keep track of how many DCMs happen over a period of time...so lets say two DCMs over 24 frames would also go analog...then monitor the tracks and if they clear up, come back to digital. I'm not saying these are the ideal algorythms, actual listen tests would need to be done to come up with ideal times to deterimine what is noticable rather than what is not.

What SDDS hasne't figured out (at least on the 2000) is that switching to either DCM or ACM a lot is worse than staying in analog. Don't get me wrong, the DCM method has merit. Unlike Dolby or DTS when fauting to analog is almost always detectable (particularly in the surrounds), a fast and short DCM doesn't have the surrounds jump in level or start hissing at you. The real annoying thing is that this sort of problem should be fixable in software/firmware.

Brad:

I believe you said you have the DFP-2000 on your JM11/JM21 loop. If you are handy with electronics...you could monitor the "Data Present" tally on the "Automation" connector...if it is off or strobing, it has faulted. If it is solid, it has valid digital data to send. I'll have to check to see if the "data present" tally strobes when in DCM. If it doesn't, I'll have to look further. There may be a test point to pick off the DCM (I know there is for an ACM) where by you could use the test point in conjuntion with a DTS relay board to switch the SDDS out whenever it is in ACM or DCM...I'll investigate.

Steve

------------------
"Old projectionists never die, they just changeover!"
 


Posted by Al Barton (Member # 941) on 07-16-2001, 09:04 PM:
 
Brad,

SDDS has an error correction mode called DCM (Digital Concealment Mode). This mode is not a backup mode but error correction if there is minor data loses on one track. Even if the track goes to DCM it cannot be heard unless the DCM is of a long duration. Most DCM's are of very short duration. If a DCM occurs in both tracks then the system will fallback to first fallback selection.

While I am not the clear expert I am assigning John Luttrell to monitor all forums and to respond to technical questions that arise. We will not comment on other equipment or manufacturers.

If you have a beef I will be happy to take your call.

If you have a technical question please contact John through our Technical Assistance Network. (201) 833-5333 the press 8 for Cinema products and then 2 to speak to an engineer.

John will be making a new post that will outline all contact information and our new 800 #.

Again my # is 310-244-7415. If you have questions other than service, parts or emergency technical help I look forward to hearing from you. If I am not in the office I will get back to you ASAP:-)

Al

------------------

 


Posted by Al Barton (Member # 941) on 07-16-2001, 09:17 PM:
 
Brad,

I am posting again with my Signature attached this time.

Al

------------------
********************************
Al Barton
Sony
GM SDDS Customer Support Group
GM Digital Cinema
10950 West Washington Boulevard
Suite 200
Culver City, CA 90232

Phone: 310-244-7415
Fax: 310-244-2024
Email: al.barton@am.sony.com
********************************
 


Posted by Antonio Marcheselli (Member # 373) on 07-18-2001, 11:57 AM:
 
Welcome to Film-Tech Al!!

I wonder what are the benefits of a "digital backup track" or DCM. Why "waste" space on print with that type of data?
I believe that a if an analog reader is well calibrated you cannot hear the Digital/Analog drop in quite scenes; and in loud scenes you'll probably hear the DCM drop since it is a Mix of channels.
I never heard an SDDS movie; perhaps I'm wrong. But this is my opinion.

So, why use print space for DCM? Why not use ALL the available space for SDDS High Quality track?

Antonio
 


Posted by John Luttrelle (Member # 942) on 07-18-2001, 02:27 PM:
 
In addition to Al Barton's comments I would to add the following contact information
For SDDS Technical Support.
These numbers can direct your call to its appropriate destination.

201.833.5333 Technical Assistance Network (T.A.N.)
press "8" for Cinema Products then "2" to speak with an engineer
866.SONYBPC toll free
techsupport@sdds.sony.com - email technical support <A HREF="http://www.sdds.com" TARGET=_blank>www.sdds.com</A> - sdds website for information on sdds, products, films, and software downloads.

Thank you,

John Luttrelle
Sony Cinema Products
Customer Support Group
10950 W. Washington Blvd
Culver City, Ca. 90232
866.766.9272 press "1" then "8" then "2"
john_luttrelle@scpc.sony.com


 


Posted by Brian Hogan (Member # 940) on 07-20-2001, 03:22 AM:
 
i know i am asking for a beat-down... but whenever i notice the on/off switching sound with sdds, i tend to push the bypass button so that im in analog. perhaps i am totally missing the whole point of this topic... so i think i will take a nap.

-hogan


 


Posted by John Pytlak (Member # 331) on 07-20-2001, 10:17 AM:
 
Brian Hogan said: " whenever i notice the on/off switching sound with sdds, i tend to push the bypass button so that im in analog."

That's the easy way out, but it's cheating the audience who deserve tho hear the movie in digital sound. Has your service tech looked at the problem? When is the last time the sound system was aligned and checked out? Does it happen more on one projector than the others? Does it occur on all prints, on a few prints, or only after prints have been running awhile so projector abrasion of the edges is taking its toll? Problem solving involves gathering data, looking for correlation with the problem, and then taking action to fix the suspected cause, whether it be alignment, marginal track quality, or abnormal projector abrasion.

Now that Sony is participating in Film-Tech, and has given additional contact information, their experts are on hand to help you get the best performance from the system, working with your service engineer.

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Cell: 716-781-4036 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion


 


Posted by Amanda Gullen (Member # 874) on 07-21-2001, 07:56 PM:
 
The tracks are not identical, they both have back up tracks for the opposite channels eg;all left channels on one side with backups for rightand vice versa. Our DFP3000 defaults to SR back-up from digital, which i find quieter than if I selected SR its self.
 
Posted by Joe Redifer (Member # 3) on 07-21-2001, 08:43 PM:
 
Maybe I am misinterpreting, but how can SR be quieter than SR?

Al Barton and John Luttrelle both work for Sony Cinema Products (SDDS).


 


Posted by Richard Topping (Member # 871) on 07-23-2001, 06:34 PM:
 
You can tell by looking at the front panel of a DFP-3000 if it is in DCM because the film title will be flicking on and off. The data present LED will stay solid throughout DCM and only start flickering if it goes into ACM.

You can tell which track is failing in DCM by popping open the front panel and looking at LED's 400 and 401 on the DSP-82 board I think. (Not at a machine to check ) They relate to each side, so say LED 400 is off and 400 relates to P-Side, then the P-Side track is failing.

Using the above, you can usually identify if you have a Processor/Reader fault or a faulty reel.
 






Powered by Infopop Corporation
UBB.classicTM 6.3.1.2