70mm 8-Perf ?

[Mark Gulbrandsen]

Why not just put marks with a permanent felt tip pen on the face of each sprocket. Next thing Tony is gonna tell you that a still working Cray Model 2 computer is required to do the conversion math... Geez.... Make life easy, it'll still works for 8 perf too, because you know that Allen Barnes is gonna come across a used Iwerks 8 perf projector.

[Ryan Gallagher]

Fun little rabbit hole, now I want an absolute encoder disc designed on a T-Shirt.

[Tony Bandiera Jr]

Why not just put marks with a permanent felt tip pen on the face of each sprocket. Next thing Tony is gonna tell you that a still working Cray Model 2 computer is required to do the conversion math... Geez.... Make life easy, it'll still works for 8 perf too, because you know that Allen Barnes is gonna come across a used Iwerks 8 perf projector.

Boy someone sure has a bad case of the grumps...did you forget your meds?
No need to be a butthead.

A marker will wear off very fast, that is why I'd do the notches.

Now go take a nap Gramps.

[Randy Stankey]

... A marker will wear off very fast, that is why I'd do the notches...

... and probably transfer to the film.

[Mark Gulbrandsen]

... and probably transfer to the film.

From the front face of the sprockets. W@hat kind of permanent markers are you using whose ink creeps around??

[Ryan Gallagher]

Too bad you couldn't put in a rotary encoder or even just an IR sensor to count clicks as the sprockets turn. Then you just make a divide-by-n counter and a digital display. Build in a control so that you can change "n" to the number of sprockets per frame. You'd be able to use it on any format of film that people can throw at you.

The electronic parts should be obtainable from the usual suspects. The only part I can't really figure out is how to put on the encoder/sensor. I'd have to see the device in order to figure it out but I'm sure that there are others who could do it just as well as I could...even better.

If one were to really go nuts with the universal encoder idea, it seems like there is enough shaft out where the knob is mounted to make room for an encoder disc and optical or magnetic incremental encoder wheel, even if you had to shorten the collar of the knob a bit. Then your reader/display etc could mount to the frame left, perhaps even replacing the analog counter if you were so bold.

Or failing that, the fact it is a 2-gang, you could sacrifice it's usefulness for editing by putting all your encoder mods in place of one of the gangs, making it into a single-gang film counter, (which were a thing in this form factor too), I have a couple 16mm single gang style.

You'd have to zero the rotary "DRO" before use, but even with the existing notches for 8perf, that would be easy with the indent handle engaged and cranking it over to the zero frame line.

[Ryan Gallagher]

With Cinemeccanica in theory still in production, I wonder if what remains of the industry as a whole would be better served by making a cost-effective diy 3d printable universal counter design... with perhaps only the film adjacent parts being traditional metal or plastic rollers/sprockets sourced from Cinemeccanica? One could make it a 2 gang with a 35mm sprocket and a 70mm sprocket. Silly, tiny market, but fun project.

Or better than two gang, make it a single but the sprocket and rollers swappable... so it can service 16mm too? Throw in a clever rewind bench mounting solution for film path alignment etc.

[Randy Stankey]

It doesn't matter what kind of ink is used. If it doesn't stay in the place where you put it, where does it go? It doesn't just evaporate. It's getting worn away. What is it getting worn away by? Film! If film rubs on the ink and wears it away, it only stands to reason that the ink is getting onto film.

It might not be a lot of ink, only a few dots. It probably doesn't happen all at once. It takes a certain amount of time. It gets worn away, little by little. Maybe even small flakes but, still, it is ink being transferred from one place to another. Even a small amount isn't something I'd want on my (somebody else's) film.

It's like those idiots who use white China marker to draw rings around the edges of the film on a platter, as they build, to mark the reel ends. Then people complain when the wax flakes off the film and gets all over the inside of the projector. Then that dust gets transferred back to the film and makes black specks all over the picture and people complain.

Maybe, back in the day, when people were running a proverbial grind house, those kinds of tricks might fly but, today, when film isn't being made the same way, you have to be a lot more careful.

In our case, with just a few dots of ink, maybe it doesn't matter so much but you still have to think about it and make an informed decision.

[Mark Gulbrandsen]

There are felt pens that leave permanent marks on alumnium, some steels, and stainless. I can tell you that from my experience as a machinist. I used them to mark high spots on stuff I was machining.

[Randy Stankey]

That's fine! My point is that you are thinking about the problem. Most people don't.

Thinking about problems is what makes the difference between a professional and a wanker.

[Randy Stankey]

If one were to really go nuts with the universal encoder idea, it seems like there is enough shaft out where the knob is mounted to make room for an encoder disc and optical or magnetic incremental encoder wheel, even if you had to shorten the collar of the knob a bit.

Hmm... Do you mean like a reluctor wheel? (e.g. Like the way the anti-lock brakes on cars work.) There is a toothed wheel and a magnetic sensor that detects the passing of each tooth as the wheel rotates. In a car, that information is used to detect one or more wheels slipping on the road in order to activate the anti-lock brakes.

Yes, it does look like there is enough room to squeeze in a reluctor wheel and a magnetic sensor. Now, you've got me thinking!

With a reluctor, versus a stepped encoder, you can detect speed of rotation and use that information to deduce film speed. If we use an Arduino or a Raspberry Pie or some other microprocessor, we can also detect changes in film speed over time. We could set a maximum film speed and set off an alarm if the speed goes too fast. There could be an alarm if the film starts or stops, unexpectedly. I wonder whether it would be possible to tell the difference between the film tailing out and a film break.

With all that, you might be able to set things up so that the device keeps track of reels for you. You could know their length. You could press a button to set a marker, of sorts, in the data so that you can record splices, defects or even catalogue scenes in the film, all down to the reel, foot and frame.

Who knows? We might even be able to detect the slight bump as a splice or defect passes through. Then the system could even find and count splices for you.

(Of course, this is all speculation. )

[Ryan Gallagher]

Hmm... Do you mean like a reluctor wheel? (e.g. Like the way the anti-lock brakes on cars work.) There is a toothed wheel and a magnetic sensor that detects the passing of each tooth as the wheel rotates. In a car, that information is used to detect one or more wheels slipping on the road in order to activate the anti-lock brakes.

Similar in concept i imagine, except it's a solved problem... there are both optical and magnetic variants of incremental or absolute rotary encoders disc solutions for precise motion tracking already:

https://photo-solutions.com/the-diff...ry-disc-types/

https://www.usdigital.com/blog/encod...ncoders-intro/

Your mind was headed similar fun places as mine. I was imagining a film counter that could run while inspecting, with user buttons for "splice", "scratch", "repair" etc etc. It could generate an inspection file with exact distance from picture start or first frame etc, measure each reel, yada yada. You were going even further to imagine it alternatively attached to the projector and do more data logging there, and/or fire automations based on reel number and footage, breaks, etc.

[Thomas Piccione]

Hmm... Do you mean like a reluctor wheel? (e.g. Like the way the anti-lock brakes on cars work.) There is a toothed wheel and a magnetic sensor that detects the passing of each tooth as the wheel rotates. In a car, that information is used to detect one or more wheels slipping on the road in order to activate the anti-lock brakes.

Yes, it does look like there is enough room to squeeze in a reluctor wheel and a magnetic sensor. Now, you've got me thinking!

With a reluctor, versus a stepped encoder, you can detect speed of rotation and use that information to deduce film speed. If we use an Arduino or a Raspberry Pie or some other microprocessor, we can also detect changes in film speed over time. We could set a maximum film speed and set off an alarm if the speed goes too fast. There could be an alarm if the film starts or stops, unexpectedly. I wonder whether it would be possible to tell the difference between the film tailing out and a film break.

With all that, you might be able to set things up so that the device keeps track of reels for you. You could know their length. You could press a button to set a marker, of sorts, in the data so that you can record splices, defects or even catalogue scenes in the film, all down to the reel, foot and frame.

Who knows? We might even be able to detect the slight bump as a splice or defect passes through. Then the system could even find and count splices for you.

(Of course, this is all speculation. )

Randy if you're actually looking to try this, compact and integrated incremental encoders are something you can buy commercially. CUI (among other companies) makes various models, but I think the AM10 would be well suited to this task. I've attached the dimensions page from the datasheet, they're quite small. Only 9mm thick!

The arguably much cooler encoders Ryan shared a picture of are for absolute rotational reference, which we don't really need here, and tend to be bigger and more expensive.

I'm not quite sure what you mean when you say that a reluctor can detect speed of rotation and an encoder cannot. They both work on the same principle; there are x number of pulses per rotation, and depending on how many pulses you count per second, you can figure out how fast it's spinning.

I suspect any sort of splice detection would be difficult to make reliable, as tension will vary naturally across the length of a reel, changes in alignment of everything would probably mess with it, and all sorts of seemingly mundane things would make it hard. Storing footage counts would of course be trivial.

I do find all this discussion pretty funny as there were electronic film counters, but the only one I've ever used was utterly terrible! This one from Big Sky would stop counting if you went too fast, and you didn't need to go all that fast for that to happen...

[Ryan Gallagher]

Randy if you're actually looking to try this, compact and integrated incremental encoders are something you can buy commercially. CUI (among other companies) makes various models, but I think the AM10 would be well suited to this task. I've attached the dimensions page from the datasheet, they're quite small. Only 9mm thick!

The arguably much cooler encoders Ryan shared a picture of are for absolute rotational reference, which we don't really need here, and tend to be bigger and more expensive.

I'm not quite sure what you mean when you say that a reluctor can detect speed of rotation and an encoder cannot. They both work on the same principle; there are x number of pulses per rotation, and depending on how many pulses you count per second, you can figure out how fast it's spinning.

I suspect any sort of splice detection would be difficult to make reliable, as tension will vary naturally across the length of a reel, changes in alignment of everything would probably mess with it, and all sorts of seemingly mundane things would make it hard. Storing footage counts would of course be trivial.

I do find all this discussion pretty funny as there were electronic film counters, but the only one I've ever used was utterly terrible! This one from Big Sky would stop counting if you went too fast, and you didn't need to go all that fast for that to happen...

Hah, I figured some version of these concepts had already been tried before. Cool device (if it worked reliably). The optical disc encoders can be made for either absolute (the pretty ones) or incremental/relative measuring (pretty vanilla single track jobbies). And yeah, absolute is not needed here... but they are indeed pretty. Encoder type really depends on the precision required... I was just drawn to the disc ones because they seemed like an easy retro-fit on an existing shaft per the original modding topic, but the universal inductive ones you linked seem appropriate too.

[Josh Jones]

A theater I worked at had one of the !weks systems. It was a Strong built Pro-35 running 70mm 8 perf. Platter was a souped-up stock platter as well as I can remember. I've got a lab can in storage with some trailers and stuff in 8-70.

JJ