Lost Chord

(Note: If you wish to refer to the individual projectors mentioned, the best route is via the link at the top or bottom of this page to the Projector Index)




Well, not really. Actually, in search of the sound quality hidden in 9.5mm optical sound tracks, but which can be so difficult to extract. The story starts somewhere around 1965, with my first Super Vox projector and continues thru many of the machines produced that are capable of reproducing 9.5 sound. In search of the ultimate in 9.5 sound, I have tried at various times, as well as the Super Vox, the standard Vox, the Pax, the Son (ugh!), the sound version of the Bolex tri-gauge, two versions of the Buckingham 9.5, the Cinegel Royale 9.5/16  the Heurtier Universal, a converted Comète and a converted B&H. I looked at, but did not buy, the French Eiki conversion that Roger Spence imported. I have, but have more sense than to use, a sound unit for a Specto. There was also sound heads for the Gem, 200B and H. I also do 8, 16, 17.5 and 35 sound, so feel I have some qualification to comment.

There are so many factors that I propose to work my way thru the entire supply chain, from the film to the point where the sound reaches your ears. 

The Film

9.5 sound tracks are notoriously variable, from the faint and wandering like Edge of the World, to the clarity and quality of such as Land Without Music and many of the Popeye and Betty Boop cartoons. Not a lot that can be done about the film, except to ensure it is kept clean and in good condition a lot of background noise can come from dirt and scratches.

The Light Source

I have often read that exciter lamps need to be low voltage and maybe DC rather than AC to avoid hum etc. Not sure how that applies to the likes of the Vox, 16mm Debrie or GBL516, using a 110v lamp for both picture and sound, but let that pass. The point is that the illumination of the sound track can have a significant impact. A blackened lamp, or one that is slightly out of alignment, can reduce sound output. I found on the Pax that the precise position of the exciter relative to the sound optic could make quite a difference. I have used Vox and Super Vox converted to horizontal burning "peanut" QI bulbs. A slight tilt of the filament can greatly affect the amount of light reaching the optic. I have also tried a dichroic mirror lamp in these machines, with a bit of the silvering scraped off to allow light thru to the optic. It is in fact possible to have too much light, leading to very sharp and distorted sound.

The Sound Optic

On 16mm these are usually factory set and rarely need adjustment. Such precision is rare on 9.5 so we have to know how to tweak. There is a lot that can affect sound quality:-

The alignment of the optic must be precise. The light from the slit must hit the film at 90 degrees to the vertical. It must be square to the sound track, not crossing it on the slant. And it must be focussed just so. Usually the only way to do this is by ear with a film running thru. The best point of focus for sound may not be the point that appears best to the eye.

The imprecision in the printing of 9.5 tracks means they are not always in precisely the same place. Most 9.5 machines have a built-in sideways adjustment, but I had to add this feature on the Heurtier.

The linear size (ie how much of the track it covers along the length of the film) of the sound slit is important the narrower it is, the better the treble response (usually a vital consideration with 9.5), but the lower the volume. On the Vox/Super Vox, it is possible to dismantle the sound optic and adjust the slit. Never tried it on other machines tho.

Width is also important. Earlier Buckingham machines had the original 16mm sound optic masked down to match the slightly narrower 9.5 track (the Heurtier has a built-in adjustment for this). Vox users will know that if you move the sound optic you can pick up all sorts of noise and distortion at the edges of the track. I have heard that a common 35mm practice was to narrow the sound slit by masking to read only the cleanest central portion of the track on well-worn films in second-run houses.

If the optic isn't clean, you will not get the best sound. As with all lenses, clean carefully to avoid scratching, which will definitely make things worse.

Picking up the Light

By which I mean the photocell or equivalent. Trust Pathé to do it the hard way, with the 17.5mm Home Talkie  using a mirror to send the sound to a cell, a practice perpetuated in the Vox. The alignment of the mirror then becomes another vital link in the chain. This approach was brought to its apogee of idiocy in the Pax, with a tiny twisted mirror of polished metal sticking out to where it could easily be knocked. I could never get mine tightened up enough to stay put, and the slightest misalignment lost the sound altogether. Partly this is a function of having the sound track on the outside - tricky to get a big fat photocell in position without making threading next to impossible (tho' 35mm always seems to manage). An interesting solution to this was the 17.5 Super Rural, with a large hollow bearing about which the sound drum rotates (and the projector pivots for vertical alignment). The photocell pokes right thru this from the back of the machine. An added problem, particularly on the Vox, is that the silvering on the tiny mirror decays and corrodes over time.

Another rather odd problem I have come across is the effect of stray light from the lamp or from room illumination. This can induce a decided hum under the sound. On my Heurtier Universal I have had to put a mask in place to shield the solar cell from scattered light from the back of the gate. The Heurtier has a 1 inch air gap between the condenser and the back of the gate!

Whilst it is of course perfectly possible to get good sound from a photocell, they do tend to fade with age so the solution is to fit a modern photo-diode or solar cell. These are small enough to fit directly under the film and read the sound track direct, obviously far more effective than a mirror (but they do occasionally need a very careful dusting). I have used the BPY 10, MS4B, OCP 71 (photo-transistor and not a success) and solar cells intended for 35mm. Nowadays, it is hard to get hold of such devices and one is reduced to cannibalising 16mm machines (If anyone knows where diodes or solar cells can be obtained it would be a great service to share that knowledge). It takes a little ingenuity to lead the wires away so as not to interfere with threading, but I have done it on the Vox, Pax and 17.5 Home Talkie. A screened wire is of course essential.

My first efforts in this direction were in fact as a result of force majeure. I acquired a Super Vox without an amplifier or even the casing for one. As an impecunious schoolboy, that was all I could afford. So I had to fit a BPY10 and connect it direct to a guitar amplifier and crank volume and treble up as far as possible. Still a long way to go at that stage.


The next step is pre-amplification, to increase the range of amplifiers that can be successfully used. Many years ago, 9.5 magazine published an article and circuit pre-ampadiagram for wiring up a little pre-amp from Maplin to do this. Even I managed to build one (see left), and I am no electronics genius. I even fitted a sort of volume control to make it easier to match the output of the pre-amp to the requirements of the main amplifier. For many years, I used an Amstrad hi-(ha!) fi amp, sold cheap in the early 70s with speakers and a record deck. The treble had to be cranked way up and even then it was not good enough. Sadly, Maplin have long discontinued the pre-amp module, but there must be microphone pre-amps out there to experiment with.

Which brings us to one of the most important points. Hi-fi amps are not suited to the reproduction of optical sound tracks. They seem to be OK for magnetic tracks, which of course have a lot in common with ordinary tape recording. I have found that hi-fi amps and, equally important, standard hi-fi speakers, can produce a muddy, muffled and bass-ey sound from optical tracks, tho' a properly designed pre-amp may help overcome this. With 9.5 in particular, where one is constantly struggling to squeeze the higher frequencies from a tiny track, an amplifier and speaker designed to do precisely this is needed. I have found valve amplifiers particularly good (I use one of my Heurtier amps usually), tho I wouldnt regard this as that critical. I use Bell and Howell 16mm, 16 ohm speakers. I used at one time to run my Vox thru a Bell and Howell 600 series projector amplifier via the microphone input. It gave markedly better treble than the hi-fi amp. Interestingly, tho I have very little experience of it, 8mm optical sound doesnt seem to have the same problem, nor is it much of an issue at the other end with 35mm.

The Rest of the Projector

It is a matter for debate just how much, if any, of the wow, flutter etc often associated with 9.5 sound tracks is actually printed in by Pathescope (when you look at the way the picture wanders about on my copy of Glass Mountain, the answer is probably quite a lot) and how much is down to projector problems. Indeed, some problems are inherent to certain projectors - all one can do is take steps to minimise them as far as possible.. To give some examples, the Vox has a fairly ropey system of sound smoothing. This comprises first a flywheel with inevitably rather worn point bearings (I can't quite recall; one side might be a ball bearing in a cup). These can be adjusted to ensure there is neither excess play, causing lumpy rotation, nor excess tightness so there is no real smoothing effect at all.

Second, we have the sound roller, which seems to ride on the flywheel rather than the film itself. This may not be wholly bad, but dirt in the flywheel groove that the inner rim of the roller rides in could be a problem. The roller must of course be clean and free to rotate smoothly on its shaft.

Third, the film is pulled thru the sound head by a sprocket on the end of quite a long shaft from the gears under the motor. Wear in the bearing or a slight bend in the shaft is going to induce a lumpy pull through, causing wow. Not a great deal that can be done about this unless you are more skilled than I in the engineering side or know a man who can. The problem was recognised by Pathé in the Super Vox with an additional bearing on the outside of the bottom sprocket. I suspect that even a lumpy and uneven take-up can feed back to the sound head on a machine like the Vox, where there is little to prevent it.

The Vox does however score over the Pax and Son (ugh) by having a better motor. Governor control is noisy both mechanically and electronically and the sparking can easily break through into the sound. The Pax also has the abominable chute, carried over from the 17.5 Home Talkie (what a lot that machine has to answer for). Together with very small sharp teeth on the sprockets, this seems to me like a recipe for film tearing. The discerning have usually replaced the chute with rollers (they don't use Sons[ugh!] at all). I once dismantled a Gem (the obvious wrong-side-of-the-bedclothes parent of the Son) and an 8mm P8 at the same time. The contrast in the complexity, standard of engineering and sophistication was stark, and the sound conversion did not add quality.

The final point is to keep the noise of your projector away from the audience, with a separate projection room, if at all possible. I once tried a sound insulating phone booth hood over a very noisy Pax, but it weren't a lot of good. I had to get a (quieter) Super Vox instead. One of the snags is that usually you can only tell what your sound is like, and whether all you have tried to do has done any good, when you get to listen to the sound track without a projector blasting in your ear.

Far and away the noisiest narrow gauge machine I have known was the tri-gauge Bolex sound. It was basically a G3 with a very loud mechanical governor added to the mechanism, a lump added for the sound head (with a rather cute little sound gate) and a huge amplifier case underneath. If memory serves, light for the sound was provided by a fuse-shaped/festoon bulb, the image of the filament serving in place of a sound slit, and the result was funnelled down to the photocell in the base. The whole thing weighed a ton. It is in one of Andrew Alden's Bolex books. 

A lot of the fun of tinkering with 9.5 sound has been taken away by the Buckingham conversions, which tend to share the sound virtues of the 16mm original, and are nearly quiet enough to operate in the same room and not notice.

The end result of all this is that I use mostly the Heurtier for 9.5 sound, with the Buckinghams running second, second mainly because they're too new to be really appealing to an old-projector freak like me. I don't like to put a twist in the film, which you normally have to do for 9.5 on the Heurtier. Instead, I let the film run thru the wrong way round and correct the picture with a straight-thru prism. The Heurtier has a really heavy-duty sound smoothing system, with a sprung V-gate affair before the sound drum and a solid sprung wheel which rides directly on the film onto the sound drum itself. They recommend giving the flywheel a manual start to help it get up to speed once you start the machine. There is a pair of smoothing rollers after the sound head too. The whole affair takes a little while to settle down on starting up and I find a longer leader than usual is helpful. But once its going, the sound is about as good as it gets on 9.5.