My latest tool for playing cylinders

[Brad]

Norm,

If you have the appropriate motor speed control you could add the ability to detect the changes in the diameter of a warped cylinder by the changing angle of the tone arm and dynamically adjust the motor speed to remove the Wow.

[richardh]

Norm,

If you have the appropriate motor speed control you could add the ability to detect the changes in the diameter of a warped cylinder by the changing angle of the tone arm and dynamically adjust the motor speed to remove the Wow.

Now that really would be a "smart" record player. I wonder if that kind of system would work.

RJ

[Viva-Tonal]

Imagine the calibration nightmare involved. I can imagine an opaque leaf on top of the pickup arm, modulating the amount of light trained on a photocell, having the changes in the photocell's resistance regulate the voltage fed to the motor, changing its speed accordingly. Mucho complicated!

[WDC]

Thank you all for the nice feedback here and on youtube!

Brad, that's a nice idea but I am sure that the motor would be incapable of adjusting fast enough to do this anywhere near realtime. For this appliance one would certainly require a micro controller that operates a step-motor (which is usually very noisy). A fixed value for height-speed ratio would not work as every record can have a different thickness. The player would have to be calibrated for a certain speed and pickup height and then would have to detect any offset from this value. Indeed, very, very, very tricky!

I think the 3D scanning technique could (or already does?) cope with the problem of an uneven surface through image processing software.

We would face at least two different problems a machine would have to distinguish. First, a real warped cylinder with a kind of deformed surface and second, a cylinder that is just off-centered. A wobbling cylinder with an otherwise even surface would be easier but just technically assuming a steady surface distance while any uneven surface would additionally require a compensator to compress or expand the certain area with flawless increasing/decreasing values.

I actually had thought of a scanning possibility for cylinders about 5 years ago by using the optical unit of an ordinary flatbed scanner.

The record would be scanned for just a little more than one single turn and a software would process an audio signal from the picture.

However, after doing some calculations I proofed to myself that a current scanner would not supply a sufficient resolution to pick up any real high frequencies. As far as I remember it would only go up to something of 3 KHz which is only about 7% of what a regular pickup does. Even if most of the recorded content stays in that range, any declicker program would have serious problems to mask the clicks properly. And, by now others have realized this idea by using confocal probe scanners which none of us could ever effort these days.

I really wonder what will be possible in 20-30 years from now. Maybe replicating techniques will be precise enough then to reproduce your own box full of vintage North American cylinders.

[Brad]

Brad, that's a nice idea but I am sure that the motor would be incapable of adjusting fast enough to do this anywhere near realtime. For this appliance one would certainly require a micro controller that operates a step-motor (which is usually very noisy). A fixed value for height-speed ratio would not work as every record can have a different thickness. The player would have to be calibrated for a certain speed and pickup height and then would have to detect any offset from this value. Indeed, very, very, very tricky!

Tricky, but fun. At the slow speeds of the mandrel, it would not be difficult to detect the deviation in the cylinder circumference and generate an error signal that could be used to control motor speed. It all falls to the ability to control the motor speed.

Another approach would be to implement a similar detector on the tone arm and record the "error" synchronized in time with the recording, then via software use the error to interpolate the samples during the wow's which would in essense recover a clean reproduction. Again, fun!

[richardh]

Norman,

just thinking about your scanning solution....I wonder what optical / software tricks they used to bring those phonautograph recordings back to life last year....but even so I am sure the equipment required would be very expensive.

RJ

[bostonmike1]

I am really delighted how the experimental work of the last three has turned out. This weekend was absolutely great. I just got my brand new electric cylinder player and have been using it all day. It features all I need and does work with all standard cylinders, including longer 3-minute and dictation cylinders. The arm can be moved backwards to use an adapter for concert cylinders too.

The transmission allows to set anything between 40-210 rpm at a steady speed. Just for fun, I connected it with the 1938 radio in the living room and it came out quite clear. The previous player I had did work but always had to be re-adjusted because the plastic base would not allow the top assembly to stay in place. The passive tracking tone arm was very delicate to construct. The wire is silver-shielded with a teflon-coating. Once a S-shape is there, the tracking works fluently and I don't have to worry about different groove threads and shrunken celluloid cylinders.

Sir--- What is the motor you used in the picture shown in your post? I have a need for one not related to your device, but for a flex-circuit I am working on. Any help you can give would be appreciated. MICHAEL P.S Great work!!!

[JohnM]

What if you could design a system where there was a rubber roller say ½" diameter that went inside the cylinder bore and up against the plaster and then a rod on the outside of the cylinder with two rubber rollers that just gripped the cylinder at the ends by pressing the cylinder between the inner roller and the two outer rollers. That way the cylinder could turn eccentrically but still maintain the same surface plane between the outer rod and the inner roller. The stylus could track under the rod. Does this make sense?

John M

[Brad]

What if you could design a system where there was a rubber roller say ½" diameter that went inside the cylinder bore and up against the plaster and then a rod on the outside of the cylinder with two rubber rollers that just gripped the cylinder at the ends by pressing the cylinder between the inner roller and the two outer rollers. That way the cylinder could turn eccentrically but still maintain the same surface plane between the outer rod and the inner roller. The stylus could track under the rod. Does this make sense?

John M

Interesting idea John. In theory it should work. Challenge is that many cylinders have ribs or ridges inside which would interfere. Also, if a Blue Amberol cylinder has been reamed, then the plaster may be smooth and circular whereas the surface may not be circular or even concentric to the inside.

[Aaron]

What if you could design a system where there was a rubber roller say ½" diameter that went inside the cylinder bore and up against the plaster and then a rod on the outside of the cylinder with two rubber rollers that just gripped the cylinder at the ends by pressing the cylinder between the inner roller and the two outer rollers. That way the cylinder could turn eccentrically but still maintain the same surface plane between the outer rod and the inner roller. The stylus could track under the rod. Does this make sense?

John M

Interesting idea John. In theory it should work. Challenge is that many cylinders have ribs or ridges inside which would interfere. Also, if a Blue Amberol cylinder has been reamed, then the plaster may be smooth and circular whereas the surface may not be circular or even concentric to the inside.

Yes John that in theory would work, but as Brad said the ribs would cause a problem. If you went so far as to try this, a way of trying to overcome the problem would be a hollow mandrel which the cylinder would slide onto and then the wheels follow the inside of the mandrel but then the cylinder would wobble again once you began to play it. The only way that would definitely work would be to smooth out the inside of the cylinder completely but this would render the cylinder unusable on a normal mandrel....

Aaron