Film-Tech Forum ARCHIVE: How do you know if there is noise reduction on a VA track?

This is topic How do you know if there is noise reduction on a VA track? 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=007245

Posted by Kenneth Wuepper (Member # 1174) on 06-04-2005, 07:14 PM:

We screen many restored films that have "restored" sound tracks. The original film was MONO sound but the new release has VA or SVA tracks printed on it. How do you determine if the sound has been"stereo-enhanced" and if there is any noise reduction added?

Some tracks have directions printed on them like SR. Can we assume that any SVA track will have at least Dolby A reduction applied to it? All of the films in quesion predate varaible area soundtracks and stereo as we know it so the SVA tracks are new.

Your help is appreciated.

KEN

Posted by Sam D. Chavez (Member # 1841) on 06-04-2005, 08:27 PM:

Ken,

Variable area goes back a long way so this is not an indicator that it has been re-recorded.

You can safely assume any stereo track has at least A-type NR. I'm sure you are familiar with looking for stereo info on the track in the non dialog passages.

There are many mono recordings where the tracks appear to be recorded all the way to the edge. These are generally mono recordings.

The A or SR markings on the leaders are a good start but certainly not reliable indicators.

You have to listen and make a judgement call often times.

Posted by Louis Bornwasser (Member # 3063) on 06-04-2005, 09:13 PM:

The only ones that are surely mono and no noise reduction are density tracks!!

Louis

Posted by Frank Angel (Member # 248) on 06-05-2005, 02:16 PM:

We run into the same problem many times. Why labs don't put NR format information in the soundtrack neg the way SMPTE recommends and why it is not as routine as a putting in a countdown leader is beyond me, but many times they don't, especially prints made overseas.

Sam is right; sometimes it comes down to just listening (is there such an expression as "earballing it?"). We have a small mixer in the booth that will take an earphone jack. When in doubt, I listening with earphones to the left and right channels. Having the L/R sound pumped directly into your ears will tell you a lot. Stereo vs mono is almost instantaniously identifable with phones and many times gives you a much quicker answer as opposed to hunting through footage looking at the track with a loupe trying to find differences visually.

SR vs A while impossible to distinguish by the look-see method, it can be discovered with the earphones, although more difficult than hearing stereo. If you are not decoding correctly, you will be able to detect a more strident, raspy sound in the high end of the spectrum, especially if you using SR on an A track. Just switch back and forth between both on a loud passage and you should be able to tell which one sounds more natural. Switch back and forth on quite passages and you should hear some pumping if you are not using the correct NR format.

Posted by Richard Fowler (Member # 893) on 06-06-2005, 09:10 AM:

Mono tracks by the early 1930's had noise reduction and stereo tracks where demostrated in the late 1930's in the U.K.....30 years before Dolby / UltraStereo / DTS Stereo tracks of today [Razz]

Posted by Sam D. Chavez (Member # 1841) on 06-06-2005, 10:05 AM:

There were any number of experimental stereo soundtracks as Richard states, but we are talking release prints here.

Posted by Frank Angel (Member # 248) on 06-06-2005, 01:05 PM:

Also, there was dynamic gating used on most mono tracks before Dolby. This consisted mainly of shutting down the track when there was no audio present rather than leaving a big white track with no modulation on it in the middle of the soundtrack area. The light valve would reduce the actual track to a very narrow single line so as not to allow an opened white area that the cell could read as noise, especially if any dirt were to accumulate on it. This was an open ended noise reduction in that it required no complimentary decoding on the playback end. I am not sure, but I think this still may be applied on modern analog tracks regardless of the encoded Dolby A or SR NR.

Then there are those odd instances of tracks that are purposely recorded in mono (Woody Allan's MANHATTAN) where the release print could conceivably be recorded in Dolby SR or even SRD and obviously needs to be played back in that format no matter if the actual audio is mono. Again, none of this would be an issue of labs would simply mark the negatives properly as per SMPTE recommended practices. I mean, how hard is it to write a freakin "SRD" in the soundtrack area on the leaders?

Posted by Leo Enticknap (Member # 534) on 06-06-2005, 05:13 PM:

quote: Frank Angel
Also, there was dynamic gating used on most mono tracks before Dolby.

We've got some examples of shuttered unilateral VA tracks in the archive - I'll try and dig one out and scan it at some point this week.

The bottom line is that unless the track is twin bilateral, it almost certainly doesn't have any sort of NR or processing encoded which needs a dedicated decoder in the theatre (with some rare exceptions, e.g. Perspecta tones). Play it on 01, and if your processor supports it, there's nothing to lose by experimenting with 02. I used to find that with thin VD tracks (i.e. ones on which the contrast wasn't very good, thereby reducing the s/n), it would sometimes clean them up very well without any noticeable clipping of the actual signal.

quote: Richard Fowler
...and stereo tracks where demostrated in the late 1930's in the U.K...

That would have been the experiments of Alan Blumlein, I guess. He is also believed by some to have 'invented' stereo consumer audio. He was heavily involved in developing the 405-line, 50 hertz TV standard that was the forerunner of PAL and was used in Britain between 1938 and 1984. He also worked on early radar technology, too - a busy chap!

Posted by Sam D. Chavez (Member # 1841) on 06-06-2005, 05:40 PM:

Good point about the Woody Allen soundtracks, essentially Dolby Mono with SR. There is always the exception that proves the rule.

The narrowing of the bias line is usually called ground noise reduction or GNR. I forget exactly how it works but it needs to open a precise moment before and close right after the passage is complete so you are not aware of "here comes the big one".

Posted by Gordon McLeod (Member # 33) on 06-06-2005, 07:00 PM:

Also there were some DBX encoded optical tracks as well

Posted by Sam D. Chavez (Member # 1841) on 06-06-2005, 08:06 PM:

There was no mistaking DBX encoded tracks. If there was no decoder, the pumping made you feel like you were on a roller coaster ride.

Posted by Stephen Furley (Member # 1277) on 06-07-2005, 01:39 AM:

quote: Sam D. Chavez
The narrowing of the bias line is usually called ground noise reduction or GNR. I forget exactly how it works but it needs to open a precise moment before and close right after the passage is complete so you are not aware of "here comes the big one".

There were various ways of doing it; the main galvanometer mirror or light valve shutters could be biased to narrow the track in quiet passeges, a separate shutter could be brought in from the opposite edge with unilateral and duplex tracks, and there was the strange Marconi Visatone track, where first one, and then a second, extra black line would be inserted in the white area of a unilateral track as the modulation of the track decreased, and then removed as it increased again.

There were at least two methods with VD tracks as well. The light valve could be biased to darken the track, (I don't think GNR was done with glow lamp recorders, but the sound from those tracks was so bad anyway that I doubt if you'd notice), or the actual width of the track could be reduced in quiet passeges; the 'density squeeze' tack. I've never seen one of theese, but there's a picture on one on the BKSTS sound formats wallchart. The darkening type can be reconised by watching the track on a rewind bench; a track without GNR well appear as a constant shade of grey, while one which has it will be seen to lighten and darken.

There are pictures of the older types of VA tracks in the post which I made on page 2 of the 'What are the sound formats on 16mm film' thread in this forum. I can't find a way of linking to it; the URL is just shown as 'http://www.film-tech.com/' wherever I am on the site, so I can't copy and paste it. I know other people have linked to previous posts; how do you do it?

Posted by Leo Enticknap (Member # 534) on 06-07-2005, 02:48 AM:

If you're using Firefox, right-click on the link to the thread from the threads menu of whatever forum it's in, select 'copy link location', and then paste it into the UBB code link in your message.

Posted by Michael Schaffer (Member # 1204) on 06-07-2005, 03:24 AM:

quote: Stephen Furley
There were at least two methods with VD tracks as well.

I am confused. I thought biasing was done only on VD tracks, as the principle of the VA is to narrow and widen the translucent part of the track anyway. Please explain for the slow-brained. Pictures welcome (or working links!).

Posted by Stephen Furley (Member # 1277) on 06-07-2005, 03:40 AM:

Thanks Leo.

quote: Leo Enticknap
The bottom line is that unless the track is twin bilateral, it almost certainly doesn't have any sort of NR or processing encoded which needs a dedicated decoder in the theatre (with some rare exceptions, e.g. Perspecta tones).

There always has to be an exception; the sample of Dolby Stereo on the BKSTS wallchart is a strange duo-unilateral format that I've not seen elsewhere.

To return to a point raised by Sam,

quote: Sam D. Chavez
Variable area goes back a long way so this is not an indicator that it has been re-recorded.

VA certainly does go back a long way; it was used by Eugene Lauste, and you can't go back much further than that with optical sound, he was demonstrating sound on film to audiences before the First World War, so it was certainly in use long before sound reached normal cinemas. On the other hand, VD tracks were still being recorded into the '60s, and duo-bilateral ones, as we have today, were in use in the '30s.
During the '30s and 40s a new film might be released with just about any type of track, and VD and VA tracks were both in use on 'normal' release prints for a period of over 30 years.

web page

The main area of confusion today is likely to be a 'A' type print mis-labeled as SR. I've even seen quite a few Academy mono tracks labeled as 'SR', but at least in this case it's obvious that it's wrong. It's not uncommon to find 'SR' used to refer to any analogue track, in the same way that anything that's not anamorphic tends to get labeled as '1.85'

Posted by Kenneth Wuepper (Member # 1174) on 06-07-2005, 06:06 AM:

Gentlemen,

Thanks for all of this wonderful information. I knew that the experienced people here could provide not only answers but also insight into this perplexing question.

Although the idea of trying the sound to see what you like (if you have the time) is functional, it is in no way as satiafying as doing it right.

Where can one obtain one of the sound track wall charts as mentioned?

KEN

Posted by Stephen Furley (Member # 1277) on 06-07-2005, 06:22 AM:

From the BKSTS. web page

They're expensive, but they are nicely produced. I've seen cinemas that have framed copies on the wall for the public to look at.

I think somebody was distributing them in the US, possibly Boston Light & Sound, but I may be wrong about that.

Posted by Gordon McLeod (Member # 33) on 06-07-2005, 09:44 AM:

Sam said "There was no mistaking DBX encoded tracks. If there was no decoder, the pumping made you feel like you were on a roller coaster ride."
Maybe that is why I saw it used on some old Porn Flicks Also I believe Kintek encoded some NR on some special venue stuff a long time ago
I remember on the old Yellow Kintek boxs a Kintek NR function

Posted by Sam D. Chavez (Member # 1841) on 06-07-2005, 10:30 AM:

GNR was used on VA tracks before Dolby and still is used today. The idea is to narrow he bias line to a minimum in the absence of modulation. The trick is not to let it get so small the bias line pinches shut, as you get all sorts of unpleasant artifacts when that happens.

Posted by Stephen Furley (Member # 1277) on 06-07-2005, 12:40 PM:

quote: Sam D. Chavez
The trick is not to let it get so small the bias line pinches shut, as you get all sorts of unpleasant artifacts when that happens.

What sort of artifacts? When there's no line there would be no light, and therefore no sound. Where there is a line of constant width there would be light, but unmodulated, therefore no sound other than noise caused by film grain and photocell, therefore I assume that the problems would occur at the point where the line disappears and re-appears; is this the case?

Posted by Martin Brooks (Member # 1269) on 06-07-2005, 02:45 PM:

quote: Frank Angel
We have a small mixer in the booth that will take an earphone jack. When in doubt, I listening with earphones to the left and right channels. Having the L/R sound pumped directly into your ears will tell you a lot. Stereo vs mono is almost instantaniously identifable with phones and many times gives you a much quicker answer as opposed to hunting through footage looking at the track with a loupe trying to find differences visually.

Another technique is to wire up a Left minus Right switch. (Combine one channel out of phase with the other channel.) If you hear nothing, it's mono. If you hear something, it's stereo. Won't tell you whether there is NR though. I still use an old Apt-Holman preamp that has this built in. It's great for all kinds of diagnostics (tape head azimuth, FM antenna tuning, etc.)

Posted by Stephen Furley (Member # 1277) on 06-07-2005, 03:22 PM:

quote:
I am confused. I thought biasing was done only on VD tracks, as the principle of the VA is to narrow and widen the translucent part of the track anyway. Please explain for the slow-brained. Pictures welcome (or working links!).

Michael, with a VA track without GNR the width of the clear area will vary to follow the waveform of the sound, but on average will be about 50% of the width of the track. Where there is no sound the width will be a constant 50% ish. this large clear area will transmit a lot of light, which will result in noise from the photocell, and any dust, scratches, etc. will also cause unwanted clicks, pops etc. To avoid this the clear area can be reduced by various means. During quiet passages the clear area can also be reduced, but by a lesser amount; as long the modulated edge of the track is not touched the sound will be anaffected, but the noise from the clear film will be reduced. Have a look at the 16mm sound formats thread that I mentioned earlier for some pictures of various methods which have been used.

Posted by Richard Fowler (Member # 893) on 06-07-2005, 03:50 PM:

Some mono optical sound recording galvanometers have a mechanical shutter to close the track during long silent passages in addtional to the closing of the audio track electronically....this creates a noise in itself in operation. One of Ray Dolby's early systems for improving sound recording was to use an audio delay to allow these circuits and hardware to react to as not to clip incoming or outgoing audio being recorded.

Posted by Josh Jones (Member # 447) on 06-07-2005, 05:15 PM:

To me it would seem that they would use some sort of envelope detector circuit which is sniffing the incoming audio. the detector is only sensitive in one direction, so that it cannot cause the valve to 'crash' or completely close down. the output of the detector is summed with a minimum bias voltage that controls the minimun track width, as well as modulation headroom.

It makes sense to me, does it to the rest of you guys?

In reguards to audio artifacts from the valve closing down during silent passages, when the valve opens, it will cause a thump, or pop, depending upon how fast the transition is, in the playback audio.

JJ

Posted by Richard Fowler (Member # 893) on 06-07-2005, 06:12 PM:

More like a wump depending on the audio getting softly clipped [Razz]

I have done plenty of 35mm sound recording when I worked in a film lab that specialized in cinema ads and low budget films...we had little audio tricks, compression and EQ setups depending where the product was shown. On cinema ads for in house production, the sound tracks where recorded on each side of the sound negative and we slugged the picture negative to restore sync; the customers did not own the elements and future reprints where a new order. I think Ray Dolby's set up was like 10ms delay to keep the clipping from happening based on my feeble memory. Josh you should have been there in 1931 to get them to do it right [Big Grin]

Posted by Gordon McLeod (Member # 33) on 06-07-2005, 07:18 PM:

There was a system that used a CCD camera to scan the soundtracks to eliminate the need of noise reduction systems in the galvo But I don't think they ever caught on

Posted by Josh Jones (Member # 447) on 06-07-2005, 07:24 PM:

well, they didnt have op amps in the 1930's, and I dont want to design precision DC amplifiers on tubes [Wink]

Posted by Sam D. Chavez (Member # 1841) on 06-07-2005, 07:29 PM:

It looks like Mr. Furley's questions got answered.

You do want a bias line instead of closing all the way to black not just to prevent wumps, but also because the noise floor will be more constant and even with a small bias line vs. the scratch and low frequency background noise you will get from an unbiased dark film.

Try running black stock with no soundtrack at all and you will hear it immediately.

As to driving the GNR, in the not too distant past before Dolby Labs hit the scene, one method used was a preview head slightly ahead of the playback head on the mag dubber that carried the audio for making the printing master. When modulation was sensed by the preview head, it opened up the bias line in anticipation.

Pretty low tech, eh?

Posted by Louis Bornwasser (Member # 3063) on 06-07-2005, 08:05 PM:

The "wump" was absolutely evident during the slightly-before-Dolby era when I was experimenting with high grade preamps, large power amps, no EQ, and speakers similar to the 4675; all in a small room.

The main complaint was that front row viewers became sick from the constant LF energy from the ground noise reduction gating.

Louis (Now I really feel like an old-timer.)

Posted by Sam D. Chavez (Member # 1841) on 06-07-2005, 09:36 PM:

That's because we are. Oracles of obselete info that is.