This is topic SR decoding symptom 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=001562

Posted by Paul Mayer (Member # 355) on 10-14-2000, 05:21 PM:
 
Caught a screening of the 1997 Japanese film "Hanabi" (Fireworks) last night and I have a question for the group regarding something I was hearing. AFAIK this was an SR presentation. What would cause the the extreme hi's of the center-channel to show up in the surrounds?

 
Posted by Randy Stankey (Member # 64) on 10-14-2000, 05:45 PM:
 
Off the top of my head, I'd say it's one of two things...

Bad alignment in the sound head. (Lateral alignment or azimuth)

OR

Poorly alignment of the preamp levels.

Either way you need to have your tech do a complete A-Chain. Not something you can do yourself unless you have an oscilloscope and test films.
 


Posted by Antonio Marcheselli (Member # 373) on 10-14-2000, 06:13 PM:
 
I know that Dolby matrix decoding has that "problem": to hear part of the voice (hi's) in surround channels.
I notice that it depends much from the quality of the soundtrack: some film I've shown had voices almost equally divided between center and surround but other ones were like Dolby Digital...

Bye
Antonio

P.S. OF course, if you have too much voices in the surround perhaps your soundhead needs to be adjusted!!
 


Posted by Michael Cunningham (Member # 210) on 10-14-2000, 06:23 PM:
 
From my limited understanding of how Dolby matrixing works, I would suggest you might have a poor recording. As you know, Dolby stereo (both A and SR) has only two recorded channels. Anything in the left and right channels that matches exactly to each other is sent to the center instead. To get the mono surround channel, a super high frequency (inaudible to humans) is attached to the sound you want sent to the surrounds. The processor recognizes this "parasite tone" and diverts those sounds. If the high frequency tone had "bled" a little into other frequencies during printing, you might get the effect described.

-Mike
 


Posted by Randy Stankey (Member # 64) on 10-14-2000, 07:43 PM:
 
If I'm not mistaken, all modern forms of surround sound take the LEFT and RIGHT and sum them to get CENTER, then they subtract LEFT from RIGHT to get the surround. To get sterro surround, they take LEFT from RIGHT and send signals with a positive phase-shift to the LEFT SURR. and signals with a negative phase-shift to the RIGHT SURR.

Thus, if your sound head's lateral alignment is off, you'll get some of the right channel's signal comming into the left or vice versa. If your azimuth is off you'll get a similar effect. If your left and right preamplifier levels aren't exactly equal, all that summing and subtracting will get thrown off.

No matter how you slice it, there's no sure way of solving the problem until you've at LEAST verified your A-Chain.
 


Posted by Russ Kress (Member # 461) on 10-15-2000, 12:16 AM:
 
This is not an "SR" decoding symptom.

"Spectral Recording" is a noise reduction process that shoves the noise level into the basement while providing for extra "head room" on the recording media that allows for impressive peak levels (lots of room on the track for bass and loud signals).

Signal leakage is most likely caused by an incorrect azimuth on the slit.

High frequencies are the most "directional" of sounds. They will be the ones that the brain will notice as being in the wrong place.

You can place a good subwoofer in the men's room. So long as it fires at the same time that the surrounds do, you will swear that it came from the surrounds.

I'd say check the A-chain.

Russ

 


Posted by Richard Hamilton (Member # 321) on 10-15-2000, 06:26 PM:
 
Delay could also be wrong.

Rick
 


Posted by Frank Prete (Member # 627) on 10-16-2000, 04:24 AM:
 
Even the slightest movement in the azimuth alignment will cause high frequencies to go out of phase. The high frequencies are extremly short so it only takes a little bit for them to be out.

Have you had any brain wraps recently in the theatre that may have moved the cell/lens?
 


Posted by Oscar Neundorfer (Member # 458) on 10-16-2000, 06:07 AM:
 
Michael,

You may be confusing the "parasite tone" method of steering surrounds with the 35mm Mag system. 35mm mag had a 12 kHz tone to turn the suround channel on and off as a rudimentary form of noise reduction since the surround track was smaller and therefore noisier than the other tracks.

Matrix decoding depends only on phase and level information in the 2 tracks. High frequencies leaking into the surrounds are almost always a sign of improper soundhead alignment (lens azimuth misaligned) or improper slit loss correction adjustment in the preamp (causing high frequency response to be unequal between channels). It could also be caused by a NR channel being defective and not handling high frequencies correctly, but this would be rare.

Hope this helps.

------------------
Oscar Neundorfer
Chief Engineer
SMART Devices, Inc.

oscar@smartdev.com
 


Posted by Gordon McLeod (Member # 33) on 11-04-2000, 06:56 PM:
 
Randy said"o get sterro surround, they take LEFT from RIGHT and send signals with a
positive phase-shift to the LEFT SURR. and signals with a negative phase-shift to the RIGHT SURR."
Optical stereo either SR or A type has only mono surrounds encoded on them

The matrix has two parts to it
to derive the centre channel it takes first the sum of Lt and Rt (the nr outputs)
If the l and R are equal in phase and magnitude a gain riding circuit effectivly turns up the gain of the centre matrix output and turns down the gain of the L and Right matrix outputs

The surround channel is a little more complex as the Lt and Rt are first passed through a pair of 90 degree phase shifters then the differenece of the summed pairs is the surround channel and the L and R are left behind
A similar logic circuit then does the gain riding magic in the smae fashion as the centre channel but uses the If the LandR are equal in magnitude but out of phase the Surround channel is boosted and the L and R are attinuated.
If either L or R are the only signal present there component that is found in both the C and S channels are attinuated and L and R are boosted
Most simple matrixs omit parts of this and that is where errors occur
The process started with a derivation of the QS matrix (cat116) and then went to useing the chipsets of the SQ system with the tate enhancements (cat 150a-d)
The cat 150e introduced the latest of Wilcox's patents on matrix decoding and the 150F put it onto a single chip
hope that helps
 






Powered by Infopop Corporation
UBB.classicTM 6.3.1.2