SR separation

[Frank Raia]

You guys probably don't care but..... My friend who used to own a theater was sick of me rambling on about this and rummaged through the last bit of stuff he has and found a small piece of buzz track, enough to make a loop. I was way off! Told you so moment for you guys! After adjusting that alignment the sound improved. So now I am happy to at least have that adjustment done and under my belt. Onward to find at least the pink noise test film.

Again, thanks everyone for your input, and Tony go ahead and say it... I told you so. Lol IMG_20250628_094508.jpg

[Ryan Gallagher]

Looks like you found quite a bit, some people consider themselves lucky to have enough to make a small loop only for threading the reader itself!

[Ryan Gallagher]

Tone loops are probably less critical in a home cinema setup, unless you are dual changeover. But yeah Pink would be a nice to have addition.

[Frank Raia]

Ryan no dual changeover my wife would have a fit if I tried to cram in another set up! I am rebuilding a platter though.

[Tony Bandiera Jr]

Thank you all for your insights, until I find the proper test films I will have to settle for the sound I have. Its not horrible but not correct.

Tony, when you did your alignments by 'eye and ear' was there a huge difference between the sound at that point to the point of it being 'correct'?

Often times, yes. The main thing very hard to get by ear is tight focus and azimuth. The smallest of errors there really affected sound quality. There was only one time I got it so close that it could have waited until the next PM call. It was at The Century Orange cinemas in the mall..they had a JJ projector and had just closed out a 70mm run. The first film they got in 35mm was "Runaway Train" (A good but very dark film) and the call was: "sound is absolutely horrible". (Or as Joe Redifer would say, "It sounds like ass.") When I got there the show had just started and was about at the end of the first reel. I went into the auditorium and it was, ....well a train wreck. EVERYTHING was coming out of the surrounds, it was badly muffled and distorted.

I went back to the booth and found the problem: someone had dicked with the optics (old style lamp and solar cell setup.) The slit was three times normal size and skewed almost 45 degrees. Since I had nothing to lose, I loosened the slit lens holder and adjusted it by ear as the film was running. I got it so good the crowd went crazy and the manager came tearing upstairs asking what I did, "It sounds great."

I explained about the optics and told him I'd come back the next morning (This was a Sunday emergency call, they had put up with the crap sound since Friday.) When I returned with the loops and test equipment, the azimuth was spot on and the focus was only slightly off. (There was an imbalance between LT and RT that I adjusted with 69T).

As for why it was so far off? Someone had bent the "guide rails" that direct 70mm away from the end of the slit lens assy. They were afraid it would scratch the film so some genius loosened the optics and slid them back. (They did at least try to get it back right, there was a scratch mark they made on the barrel of the lens). I fixed the guides and told them to never touch that lens again.

So generally, I can get them by with the "by the ear" tuning, but it was always far enough off that it had to be done right shortly afterwards.

As to WHY that was so....adjustments for focus, Dolby level, etc. use fixed tones (or pink noise) recorded at very specific levels for setting up the processor. A film soundtrack is all over the place so doing what I did for "Runaway Train was like trying to stick a firecracker up a running bobcat's rear end and hoping to not get ripped apart in the process.

Even setting by ear with the loops (other than Dolby Tone levels, where you have the meters , LEDs or display to show when you got it right) will have errors depending on ear fatigue, background noise, etc. The ONLY loop designed to be set by ear is Buzz Track, and that one is pretty fool proof...again, as long as you can hear it clearly. I ALWAYS had my oscilloscope connected and looking at the signals, even with buzz track. (It helped that my test loops were always in the same road case as my 'scope and RTA)

EDITED TO ADD: Frank, I told you so.

[Tony Bandiera Jr]

Tone loops are probably less critical in a home cinema setup, unless you are dual changeover. But yeah Pink would be a nice to have addition.

I disagree. optical sound (for Dolby) MUST have tone loops regardless of the setup.

Now, for a single projector mono only setup, you CAN live without tone loops and could end up with reasonable results just setting focus by ear. But Buzz and Azimuth must be still set up with loops.

[Ryan Gallagher]

I disagree. optical sound (for Dolby) MUST have tone loops regardless of the setup.

Now, for a single projector mono only setup, you CAN live without tone loops and could end up with reasonable results just setting focus by ear. But Buzz and Azimuth must be still set up with loops.

That make sense, I’m coming at this not having had to do a full A chain yet, and my primary experience with Dolby tone is for doing the Auto cal on the 650 between two projectors.

[Frank Raia]

Tony, thank you for the response, I always like to read stories like that. I don't want to be the guy that messes with shit.... Oh wait I am! Lol. But now I have one test film and will continue to find more. Oh and thank you for the edit, I was waiting for it! Lol

[Tony Bandiera Jr]

Tony, thank you for the response, I always like to read stories like that. I don't want to be the guy that messes with shit.... Oh wait I am! Lol. But now I have one test film and will continue to find more. Oh and thank you for the edit, I was waiting for it! Lol

You're welcome. The best part of learning how to do the A chain is that once you get good at it, it takes less time each time, and the results often make the client very happy, as they never realized how far off it can get and still sound sort of ok. But when they hear the difference...suddenly that annual service call is worth it.

[Martin Brooks]

Tony, thank you for your response. I do realize that I can not achieve optimal alignment without test film, but this is not in a theater this is in my basement. I do have decent sound at the moment but I notice that it is not very discrete. I am looking for advice on how to make it better with what I have. If I can't than it is what it is, but I have the luxury of playing with it over and over. I guess the next time I go to the last mom and pop theater that is open in my area I could find out who their tech is.

Dolby Optical was not very discrete between adjacent channels. So there's not a lot of separation between left and center and center and right and between right or left and surround, but there is substantial separation between center and surround and between left and right.
On playback:
Left is -15db in the center channel and infinite in the Right channel
Right is -15db in the center channel and infinite in the Left channel
Center is -15db at Left and Right
Center and Surround are exclusive (infinite)

[Steve Guttag]

It's more complicated than that, Martin. Plus Dolby Stereo evolved, over time. The CAT150F, the last in the line before going to software based decoders in the CP500 and CP650 used the Analog Devices SSM series of chips (SSM-2125/2126). From their datasheet, you get a minimum of 35dB of separation between Center and Left or Right on the SSM-2125. In fact, a Center input yields a "typical" 48dB separation in Left or Right. Surround, however, is 35dB of separation to any output.

The very notion of "infinite" separation is absurd. I don't know what cut sheet you got it from but it is impossible. The very reason for adding an additional 20ms to the delay on the surrounds is to mask the crosstalk of the center channel (dialog) in the surrounds by using the Haas effect.

The prior decoder, on the CAT150E was a Sanyo chip that was considered superior to the AD one and used on the SDU4 through to the end of production. The CAT150E was horded by Dolby to be used on studio systems so I would presume its specs are a bit better. The only Sanyo chip datasheet I could find was not in the same DIP package but it boasts 40dB of separation from channel-to-channel.

This is all quite a bit more than the gloomy numbers by Martin above...or is it?

So, with ideal conditions, the analog decoders, were, at least good to 35dB and possibly quite a bit better in their digital representation.

How about at the other end of the spectrum? If one were to look at the CAT110/116 systems as used in the early CP100 and CP50...the stuff used around the time of Star Wars. Well those were lucky to get 6dB of separation on Center. Yes, you had a "draped" dialog where Center was louder but there was plenty of bleed into Left and Right. Anyone that worked with the Eprad Starscopes of the late 70s/early 80s also got to hear that. The surround decoding was by the Sansui QS decoder on the CAT116. The Cat 150 was more than just a "sandwich" card that put both the L/C/R and the S decoders in one module...it brought steering to the system and probably is where Martin's 15dB of separation numbers came from. It was a HUGE improvement over the prior versions.

Okay...so we have a range of decoders (and the movies of their day would have been mixed for the decoders of the day...any of the Star Wars IB prints that are being shown have decoders that are FAR superior to what was heard back in the day and it is possible that some exaggerated channel separation could result).

But what about the input side? How clean is that? If we are on full-coat mag film or some digital audio source...it is probably pretty good with negligible crosstalk. But 35mm optical readers, particularly using forward scan all had crosstalk. Once the slit of light hits the film, it is out of focus heading towards the solar cell. You WILL have crosstalk or you are dragging the cell on the film. Reverse scan readers, particularly late model could again deliver essentially no crosstalk issues. So, you should be able to get whatever the decoder can deliver.

[Frank Raia]

Steve my head is spinning, lol. So I guess I can presume that there is no real separation in the optical sound just lower volume to the surrounding tracks? I have noticed that once I ran the test films and aligned that the center channel was more dominant than the other channels and some odd reels where better than others.

[Marco Giustini]

Frank,
Don't forget that the centre channel can be intentionally added as reverb on all channels. In real life, unless you are outdoor and/or in an anechoic room, if someone talks to you you'll hear the sound from them plus all the little riverberations from the room. So I think it's quite normal to hear "echos" of centre channel from other channels.

To test you can use the cat69T. That has a 1KHz (was it 998Hz?) tone in phase between L and R. If not mistaken, if you run it through Dolby SR, you should only have sound on centre channel. Nothing else. But, once again, I'd expect some crosstalk, nothing is perfect.

[Steve Guttag]

There is separation but it depends on the decoder, the source and how it was originally mixed. Most movies are very "Center Heavy" with dialog locked into Center...like it or not. That is not a function of the decoder but of how it was mixed. Even with today's discrete content, they mix dialog just about exclusively into Center and the mix is often center heavy.

Looking at one of my Dolby Atmos theatres running Superman...just randomly doing a screen grab:

image.png

There is Center nearly 20dB hotter than Left and Right with just trace audio in other channels. No doubt, this was a "talking" scene. There is no "decoding" going on there aside from Atmos...but that would be discrete on the "bed" channels (traditional 5.1/7.1 channels). With Atmos, you only get "decoding" when you have "objects" where they can be located in 3D sound space that the decoder interprets the location based on where the speakers are configured during setup.

Looking at just the 7.1 track (back up track) on the same title:

image.png
It's the same story. Track, 14 is a sync track used for when the server and the Atmos decoder are not the same.

[Frank Raia]

Marco, I have had a hard time finding test film. I did find buzz track and pink noise so that are the only alignments I can do. Running those 2 loops really did help. Tony would be proud!