16mm OPtical Soundtrack specs.

[Robert Houllahan]

Does anyone know what the specs are for 16mm optical sound? Is there a Kodak paper?

 

I was thinking a frequency response in the 200hz to 6khz range and 40db to 50db S/N is this about correct?

 

Mixing my soundtrack and making a print :)

 

-Rob-

[Phil Rhodes]

Hi,

 

I'd encourage you to avoid 16mm optical sound like the plague - those specs sound about right, and it sounds really poor.

 

Phil

[Robert Houllahan]

Hi,

 

I'd encourage you to avoid 16mm optical sound like the plague - those specs sound about right, and it sounds really poor.

 

Phil

 

 

Well the film is a std.16 original which I just cut on a steenbeck and I had planned to make a print from the beginning so there is no avoiding the sound.

 

We make allot of 16mm answer prints here at Cinelab and while I agree that 16mm optical is fairly lo-fi it can be tailored to be acceptable in it's own way.

 

-Rob-

[Bernhard Zitz]

and it sounds really poor.

I second that.

 

16mm optical sound has it's own aesthetic and can be appreciated for this, in a certain way... I wouldn't ad to much ambient sound and music that covers the dialogue. Everything that goes further than dialogue and short distinguishable noises might disorientate the ear... I'd edit and mix the sound in a very heterogeneous way...

 

Can be hard to find a movie-theater or a festival that still screens std.16

 

cheers, Bernhard

[Clive Tobin]

Does anyone know what the specs are for 16mm optical sound? ...

Most projectors are sort of flat (maybe -6 dB) from 100 to 5000 Hz. Figure the s/n is about 40 dB. If the audio will be heard through a projector that is clattering away in the same room as the audience, make any important audio be within a very narrow volume range. Volume compression of speech is recommended. Don't boost frequencies below 100 or above 5000 to try to overcome projector limitations as this really won't help and it will just reduce the volume.

 

16 optical is not really hi-fi but it can sound pleasant like a good AM table radio. Avoid trying to use a Poulenc Pipe Organ Concerto for background music. :-)

[Robert Houllahan]

Most projectors are sort of flat (maybe -6 dB) from 100 to 5000 Hz. Figure the s/n is about 40 dB. If the audio will be heard through a projector that is clattering away in the same room as the audience, make any important audio be within a very narrow volume range. Volume compression of speech is recommended. Don't boost frequencies below 100 or above 5000 to try to overcome projector limitations as this really won't help and it will just reduce the volume.

 

16 optical is not really hi-fi but it can sound pleasant like a good AM table radio. Avoid trying to use a Poulenc Pipe Organ Concerto for background music. :-)

 

 

Thanks Clive!

 

This is about what I figured from what I hear in the lab but wanted a bit of a confirmation for working with my sound mixer.

 

No Pipe Organs! Damn, well there is always the DVD! :lol:

 

-Rob-

[K Borowski]

Forgive the ignorance, but how do those specs given here for 16mm optical compare to those for say a CD, or analog tape, or a 35mm analog track?

 

Regards,

 

~Karl

[Clive Tobin]

Forgive the ignorance, but how do those specs given here for 16mm optical compare to those for say a CD, or analog tape, or a 35mm analog track?...

CD: 20-20,000 Hz, s/n claimed to be 80-90-? dB but has quantization noise instead of the soft hiss of analog. Very low level signals can be distorted. Overload results in hard clipping.

 

1/4" analog tape: depends on equipment and speed, 30-16,000 Hz, 50-60 dB. Overload gives less objectionable distortion, reducing the extreme waveform spikes of speech or singing to make it sound louder on average.

 

35mm Academy mono optical track: 45-8,000 Hz, 48 dB. Normally recorded at a high level so there is frequent or even constant clipping distortion, which is made somewhat less objectionable by the Academy high frequency playback rolloff, offset by boosting highs in the mix.

 

35mm Dolby Stereo optical track: 30-12,000 Hz maybe more, depends on projector sound optics, 48 dB. Slight occasional clipping, nowhere near that of the old mono tracks, is considered acceptable.

[Robert Houllahan]

CD: 20-20,000 Hz, s/n claimed to be 80-90-? dB but has quantization noise instead of the soft hiss of analog. Very low level signals can be distorted. Overload results in hard clipping.

 

1/4" analog tape: depends on equipment and speed, 30-16,000 Hz, 50-60 dB. Overload gives less objectionable distortion, reducing the extreme waveform spikes of speech or singing to make it sound louder on average.

 

 

I have recently heard some 1/4" tape which when played back at my friends sound studio really struck me. The recording was of a piano concerto and I thought it was better than any such recording I have heard on CD.

 

I have heard a lot of 16mm print soundtrack over the last 4 or 5 years here at Cinelab and I think that there is allot of sound you can stuff into that little squiggle.

 

-rob-

[Guest Zulkifli Yusof]

So let's say I'm mixing down a film project which will be projected on 35mm Dolby Stereo, should I take the specs of the optical track into account? Like EQ'ing away anything that's below 30 Hz and above 12,000 Hz since it's not gonna make it into the print anyway?

Edited February 7, 2007 by Zulkifli Yusof

[Robert Houllahan]

So let's say I'm mixing down a film project which will be projected on 35mm Dolby Stereo, should I take the specs of the optical track into account? Like EQ'ing away anything that's below 30 Hz and above 12,000 Hz since it's not gonna make it into the print anyway?

 

 

I am planning to do the sound mixdown full bandwidth i.e. 24bit 96k and have my sound mixer apply a bandpass filter to it to check for compatibility with the 16mm optical track. That way I have a fuller soundtrack for the dvd and a good idea of the sound from the projector.

 

 

-Rob-

[Philippe Lignieres]

Well the film is a std.16 original which I just cut on a steenbeck and I had planned to make a print from the beginning so there is no avoiding the sound.

 

We make allot of 16mm answer prints here at Cinelab and while I agree that 16mm optical is fairly lo-fi it can be tailored to be acceptable in it's own way.

 

-Rob-

 

One year ago, I have seen Belgium cinematographer Boris Lehman's movie called Babel, a kind of filmed diary.

6 hours, all 16mm, projector inside. It sounds pretty good, I was surprised because of the very short band-width. Boris Lehman is very poor and have only one or two 16mm prints of his movie. Moreover, he always come with his film, to be inside and talk with people after screening.

It is a very good and not very expensive solution. Interesting to try, anyway.

Buddha said : Be low fi !

 

Philippe

[Stephen Baldassarre]

 

I have recently heard some 1/4" tape which when played back at my friends sound studio really struck me. The recording was of a piano concerto and I thought it was better than any such recording I have heard on CD.

 

I know this is an ancient post but most of you guys are still here. A well-maintained, professional 1/4" tape recorder is about 25Hz-25KHz at the standard 15 In/S and about a 70dB dynamic range without noise reduction. Add Dolby SR and you get a perceived dynamic range of about 94dB, which is better than many CDs. Furthermore, you don't have ringing or sub-Nyquist alias distortion like brick-walled digital formats do, so even though CD technically has less inherent noise and distortion, it has other artifacts that can be hard on the ear.

Many of the Mercury "Living Presence" classical albums were recorded on custom 35mm magnetic 3-track, which boasted an 80dB S/N ratio without Dolby! Many of those recordings from the 60s still sound amazing today.

 

Any way, to answer the original question, "it depends". Under ideal circumstances, 16mm optical can range from sub-sonic to about 6KHz, though I know a guy who made some customized optical heads that could do a little better. Most projectors, which were generally budget-oriented for the 16mm market, are about 200H-4KHz. That is largely restricted by cheap audio components and making a compromise between frequency response and noise. The narrower the slit, the better the frequency response (until you hit the diffraction limit) but also more noise since less light passes through the track at any given time. A brighter lamp can help make up the difference but only to an extent, as film grain becomes the limiting factor.

Signal to noise ratio is 30-40dB, depending on who made the optical negative, quality of the positive etc. I just performed an experiment with using Dolby on 16mm with mixed results. Dolby SR is totally useless but Dolby A makes a substantial improvement to noise without sapping out too much high end.

 

Another idea I've thought about using was to put SMPTE on the optical track and have a computer chase it. I'm not sure if a normal 16mm projector is stable enough to maintain good sync though.

 

In the end, I have some 16mm prints that sound fairly good in my living room home theater while others are terrible. One of the things I've noticed is that people who try to compensate for the lack of top end wind up adding too much distortion. The mix should be bright but without excessive energy at any range.

[Simon Wyss]

Something I must get rid off: dynamics. Frequency response, naturally essential. Wow and flutter, perforation influence, slippage with printing, dubbing stages also.

 

Dynamics are not so well understood. Now on the optical tracks field we have, how many, basically two different processes? Variable area and variable density? No, we have at least three. There is an almost unknown third process, namely variable height. It was described in 1941 and consists of a constant width light band whose height varies, thus producing a record that looks like a VD one. In practical use it is not better than any of the two other, so it never caught on.

 

If I may draw your attention, Ladies and Gentlemen, to the fact that with VD tracks the dynamic range is limited by the films’ photographic density range. A contrast of 1:1,000 can easily be accomplished with usual black and white print stocks, with color stocks, too. Expressed as loudness value 1:1,000 is 30 dB. It comes from the contrast between (almost) transparent film and maximum density which can be pushed to log 5 or 1:100,000 or 50 dB. Admittedly, special print stock needs to be employed for such values.

 

A variable area sound track’s dynamic range is limited by the contrast between the thin transparent line at no signal and full track width which is a tenth inch at the best. Geometrically we speak of something between 0,1 mm and 2,5 mm, a contrast of 1:25 or 12.5 dB. Why doesn’t that bother? Because most of what happens on sound tracks is not loud, the range suffices. Gunshots, the hammering of a blacksmith, a train rushing by, we don’t reproduce those sounds as loud as they are in reality.

 

The relation of loudness and frequency is the point. We find optical sound recordings poor because they are simply not loud enough. The means are at hand. I have heard big metal parts pounding from a VD track of a film made in the 1920s, it was a Tri-Ergon recording, that came brilliantly. Loud and clear with high tones. 24 frames a second times 0.3 inches equals 7.2 inches or 182,88 mm per second. With a slit height of one hundredth millimeter the highest reproduced frequency is 182,88 mm / 0,01 × 2 = 9,144 Hz. The factor 2 is there due to the double form of a sine wave. At 0,008 mm or 0.000315" slit height we would reach 11,43 kHz. Just a figure. Sound engineers simply don’t charge as much as is technically possible. In a way I can understand them since the quiet sounds are buried in noise at higher amplification. Here again microfilms can help.

 

The weakest link in the sound chain is not the sound track, it’s the speakers. That was already well known in the twenties. The baffle wall was invented back then.