Some SRD titles showing "reverted" (but not all titles?)

[Ryan Gallagher]

We have a potential SRD snag with two upcoming 35mm screenings.

Our prints of "X" and "Pearl" are not being picked up as SRD by the 702s+CP650. 702 LEDs are in the green, have run other prints just fine.

Last time I hit this it was with a personal print I snagged off ebay and assumed it was a festival print issue. Confirmed our trailer reels still triggered SRD and ignored it.

I'm in again tomorrow and I'll re-test a trailer reel to make sure it is not something deeper. But anyone have a suggestion as to what might be the cause for certain prints that definitely have SRD, especially modern prints, to immediately revert? X and Pearl are 2022!

Note we have not checked alignment with a scope yet, it's been on my todo list. But we've seen prints run with 0-3 error rates, so I wasn't too worried about it previously.

Thanks.

[Edit to note, the Screenings are Saturday 10/14, I have a little time]

[Mark Gulbrandsen]

How long has it been since the 702 was last used? I'd do a complete setup procedure, especially LED level and waveform alignment... I did have a number of Dolby readers where the lateral guide wore on one side only, letting the film drift over and the very edge of the digital info was almost out of scan range. Also check all the bearings. Check those printed in 2022 SRD tracks to see if there is any microscopic damage to them. If they look good... and the reader is in perfect alignment... but still doesn't play back the 2022 prints reliably but all other film does, then the tracks are likely defective.

[Ryan Gallagher]

Thanks Mark.

Last SRD feature we ran was 7/16, and only a handful of dolbySR and A-Type features since then. But my test with a trailer reel was last month after experiencing the same glitch on my personal print. About a year ago when we booked our first 35mm post pandemic, we discovered our LEDs were aged beyond the point they could be adjusted to green, we swapped and tuned the LED output levels (not scoped/DRAS... just the trim pot).

I'll inspect for wear etc and I'll give the waveform alignment and voltage level tuning process a go if I manage time in the schedule, especially if a test with a trailer reel exhibits the problem now too. I brought my scope to the theatre for a reason, guess it is about time to use it.

[Marco Giustini]

Did you swap the 702's LED without doing a full alignment?

As Mark says, if you have access to a scope and DRAS, don't waste time testing reels, just play something good and do a full calibration. DRAS will probably give you a good pointer but if you also have a scope you'll fix that in half a hour.

Also make sure the drum is free to spin, free from dirt, no stuck rollers etc - sorry, you probably know this but I thought I'd mention!

[Ryan Gallagher]

Thanks Marco. At the time we swapped led modules we did not have a scope handy. We called it a win when we had digital audio below an error rate of 3. Time to go back and do the full alignment it seems

[Randy Stankey]

One thing I check when debugging SRD is film tension through the reader.

Sometimes, if there isn't enough back tension on the film as it goes through the reader, it won't track consistently. Even though the system plays at a low error count, it can suddenly seem to drop in and out, throughout the show. Adding an extra roller in the film path to apply just a bit of tension often helps the system play better.

It's easy to tell whether this is the problem if you gently squeeze the film by the edges, just before it enters the top of the reader. The problem will go away while you are putting tension on the film and it will come back when you let go.

In my experience, putting an extra roller in the path, just before the reader, so that the film has to make an extra zig-zag often adds just the right amount of tension.

When I was running Dolby Digital at Mercyhurst, I almost always had a Kelmar Film cleaner at the top of the stack and that applied the amount of back tension that was needed. If I ran a film without the cleaner, the Dolby Digital sometimes dropped in and out. There were a few films where I wasn't able to use the Kelmar (with Film-Guard) because putting any extra substances on the film wasn't allowed. That's when I added that extra roller and the problem stopped, pretty much, all together.

I don't know if this is your problem but it's an easy fix that's worth trying. If it works, your problem is solved and you didn't have to go through a lot of trouble to do it. If it doesn't work, you haven't spent anything but a few minutes time.

[Marcel Birgelen]

I can only underline the importance of doing a recalibration using a scope or DRAS. Years ago, I encountered a similar issue after LED failure. Didn't have the time to do a recalibration as it happened "in flight", SRD wouldn't reliably engage anymore or drop out regularly. Some prints had less issues than others. Don't know exactly why, but probably has something to do with the slightly different translucency of the different film emulsions. Some of the more enlightened Dolby techs around here may have some insights in that.

[Ryan Gallagher]

Thanks for the tips all.
We did find time this afternoon while a live show was onstage rehearsing to go through the motions.

With a scope image we were able to get closer to the recommended led output. We started getting non-reverted DD at that stage. We adjusted the modules slightly to try to improve the waveform and reduced errors further.

One module is basically out of trim headroom now but giving us a solid 2 rate on the prints we have handy. The other module has headroom still but we could not get it any better than 6s on these prints.

Perhaps the lens assembly needs a focus adjustment still, but we successfully restored function on the problem prints. (it was always functioning on our trailer reel(s)).

I’m attaching our two scoped images if anyone has any advice for further improvements.

1st is the no headroom 2s unit.
2nd is the headroom 6s unit.

[Ryan Gallagher]

Note, I did not have ideal probes. Channel 1 Video was McGuivered from some very sharp multi meter probes and a bnc to banana probe. Clamp channel was a 1x proper probe. Not sure I was getting the best image due to this.

Any recommendations for probes or an insert that will hold itself in the tiny probe points on the cp650?

[Mark Gulbrandsen]

I used these. A set lasts for ever. They have gotten pricy, but are worth it.

[Ryan Gallagher]

Thanks mark. I eventually did see that part referenced in the manual. Ahead of ya and already ordered two from Digikey and some 50ohm RG58 extensions to pair with them.

Also in hindsight, I now understand that the top trace in the waveform is the open sprocket light, and THAT is what they want you to tune to 4-4.5v, not the data trace. I think we overshot a bit and is evidenced by at least one projector diode LED sitting on amber with no film present.

These diodes were working pulls off our 701s from the back room. I have to revisit this calibration soon anyway when we decide when to put on the "new" diodes that I soldered into the old modules. Sooner the better I think so these can remain functional spares with a little life left in them for emergencies.

[Ryan Gallagher]

Does the CP650 have to be in "setup" mode for serial communication (WinDRAS) to function? Tried connecting regular using the correct com port and no luck.
Maybe it is a baud rate/parity thing. I didn't see those references in the docs.

Ryan

[Marco Giustini]

the 650 can be fussy about the serial port! No, it doesn't have to be in a special mode. If your serial adaptor doesn't work, try another one. The Tripplite USA-19HS used to be one of the few working reliably.

The first picture definitely needs a tweak! The top trace - which you rightly say should be at 4 V - is not flat. That means that one side of the soundtrack is getting more light than the other. You should be able to adjust the LED until the level is maxed AND flat. I never worked on a 702 so I'll let you discover how to adjust the LED by reading the manual from page 6

I'd not expect the LED trimmer to be maxed out with a new LED, particularly with a 702 which is one of the best readers you can get.

While you have a scope, also run WinDRAS and check the FOCUS value there. The scope is handy to set the trace to an acceptable position, then you have to look at DRAS. Same with the lateral position and azimuth. But the main adjustment, the light amplitude and flatness, are very important and must be performed before you do anything else.

Glad you're going somewhere! Keep us posting!

[Markus Lemm]

I'm just gonna throw in my 2 cents: I last aligned 2 BACP readers a year ago.

You are not aligning to get the lowest error rate, even new prints can have 3s, 4s 5s and even 6s. You are aligning to get the best chance of reading the Dolby blocks. Each trailer or feature reel is different.

The scope is the best way to get the waveform looking great and getting the DRAS to register. Running both at the same time is best.

Once you have the scope looking good use the DRAS to fine tune.

Jitter, is what I find, causes the highest error rates. Get this as low as you can first. Rollers, back pressure etc.

The focus should be 70+, should be easy to do on most prints.

Azimuth 0 and Lateral position 0.0.

I also found that the voltage should be between 3v and 3.5v for best results overall. As you go over 4v the error rates go down but once you get too bright errors spike and you start getting 7s and Fs.

For led brightness a good place to start is the Histogram, there are 2 lines (white and green). Try to get those close together.


The 650 and the DA20 need 9600 baud 8n1, from what I remember. Try 38400 if that doesn't work.

I had lots of problems with WinDRAS on modern machines (win10) so I have an old XP laptop that runs it great.

Good luck.

[Gordon McLeod]

I'm just gonna throw in my 2 cents: I last aligned 2 BACP readers a year ago.

You are not aligning to get the lowest error rate, even new prints can have 3s, 4s 5s and even 6s. You are aligning to get the best chance of reading the Dolby blocks. Each trailer or feature reel is different.

The scope is the best way to get the waveform looking great and getting the DRAS to register. Running both at the same time is best.

Once you have the scope looking good use the DRAS to fine tune.

Jitter, is what I find, causes the highest error rates. Get this as low as you can first. Rollers, back pressure etc.

The focus should be 70+, should be easy to do on most prints.

Azimuth 0 and Lateral position 0.0.

I also found that the voltage should be between 3v and 3.5v for best results overall. As you go over 4v the error rates go down but once you get too bright errors spike and you start getting 7s and Fs.

For led brightness a good place to start is the Histogram, there are 2 lines (white and green). Try to get those close together.


The 650 and the DA20 need 9600 baud 8n1, from what I remember. Try 38400 if that doesn't work.

I had lots of problems with WinDRAS on modern machines (win10) so I have an old XP laptop that runs it great.

Good luck.

I run legacy software in Oracle virtual box