Wavy horizontals

[Viewmaster]

I have 'lifted' this photo from The 'NBTVA 2012 convention thread' links so I hope the author will forgive me.

But it does illustrate what I too am suffering from at present only even worse !
What causes the horizontal lines to be modulated by a kinda bent sine wave I wonder, whose frequency appears to be the frame rate ?

I have tried power supply smoothing etc, but to no avail.

Can the disc speed be rising and falling at the frame rate ? If so, why, as the frame lock here and on my set up seems OK.

Varying disk speed seems to be the only answer but I have put on an extra flywheel for even more inertia and it makes no difference.

Has anyone else suffered from this and did they cure it ?

Thanks.

[Harry Dalek]

I have 'lifted' this photo from The 'NBTVA 2012 convention thread' links so I hope the author will forgive me.

But it does illustrate what I too am suffering from at present only even worse !
What causes the horizontal lines to be modulated by a kinda bent sine wave I wonder, whose frequency appears to be the frame rate ?

I have tried power supply smoothing etc, but to no avail.

Can the disc speed be rising and falling at the frame rate ? If so, why, as the frame lock here and on my set up seems OK.

Varying disk speed seems to be the only answer but I have put on an extra flywheel for even more inertia and it makes no difference.

Has anyone else suffered from this and did they cure it ?

Thanks.

Hi i don't think its any thing to do with the video at all as you may have seen on my drum project its wobble vibration ,it doesn't take much ...no wobble the thing has to be 100% right centered and the thing that its made of is thick enough to over come the forces at speed and thats hard unless you have a machine shop and years know how....that pictures pretty good hard to get a mechanical television to work like a crt .

[Viewmaster]

Thanks for reply, Harry. Well, there is no wobble on my system as I have ball race bearings and all runs very smoothly. I agree that it cannot be the video.

I added a big out of balance mass to the flywheel to see if it worsened but it made no difference, so it cannot be an out of balance problem.

The wavy effect varies on various pictures, sometimes worse than others, but always there.

I think it must be mechanical but so far, cannot see where.

Harry, these things are sent to try us.
But who the hell sends 'em?

[Klaas Robers]

This has to do in most of the cases with unprecisiness of the disc or the drum. If you have an aperture drum where the distance of the slits or holes varies somewhat, then you get the waviness in horizontal lines.

The same effect occurs if the drum (it looks to be that) is not running in the same cilinder. If you touch the running drum then you should feel it against your finger for the whole revolution, not only at one point to begin with. Yes that is very difficult to obtain, but if the spindle is not exactly in the center of the drum you will observe the deviations as wavy horizontal lines.

The same occurs with a Nipkov disc if the spindle is not exactly in the center. It even occurs if it wobbles and you are watching the picture from a standpoint somewhat higher or lower than the vertical height of the spindle.

Making the mechanical scanning is a very precise job and not that easy. That is the reason that computer steered drilling or lasering is needed to get a more or less perfect disc.

If this waviness is stable from frame to frame, then Karen Orton made a PIC-solution in which she delays the video signal of the different lines over a line by line adjustable time delay. Then the video signal is been distorted such that this compensates for the errors in the disc or drum.

[Viewmaster]

I daresay I am overlooking something here, but if one looks at an NBTV disk on the right hand side (the normal viewing position) and the disk is off centre, any scanning hole will only move left or right from correct position as one sees it.
Although this will give side expansion and cramping of the picture I don't think it will cause vertical motion. ie wavy horizontals.

Of course, if one viewed the picture at the top, then the picture would be where vertical motion was seen and this would give full horizontally wavyness to the picture.

There's more to this NBTV business than meets the eye !

[Steve Anderson]

I would try and split this down the middle and find out where the imprecision is, is it my monitor or is it in the signal source?

Try using a purely electronic signal source if you have access to one. Klass, Karen and others have published simple designs that are suitable. If your current signal source is a recording, where did it come from? Is the error in the original mechanical camera done by others? Or is it in the recording media? The original may have been recorded on tape before transfer to a PC/CD file format.

Electronically or software generated audio files played on a PC or CD player are just as good as 'live' generated electronic signals...not being 'funny', but you need to eliminate any irregularities in the source signal. Gary may have some suggestions here on this matter...

Try and establish any relationship between the position of the 'wave' within one line and the average (perhaps nearby) video levels, i.e. does a bright line preceding and/or following line shift the 'wave' up? Or the opposite? Likewise dark lines. This could point to an electronic problem rather than a mechanical one. DC restorers and sync separators would be the prime suspects in this case.

The common occurrence of this on monitors with very different manners of construction, mechanical inertia, skill of the constructor, access to the tools and facilities required, all manner of things, makes me wonder if at times we are blaming the mechanics unduly. I'm not saying they're blameless however.

In essence what I'm asking is does the picture content alter the position, amplitude shape/nature of this 'wave'? If it does I doubt it's (at least mainly) mechanical. When replaying the same video later on, is the 'wave' the same? i.e. is it consistent, dependent on the video signal?

I might mention hum finding its way onto the source waveform. In this case the 'wave' won't be locked/stationary and the 'wave' will tend to wander around....again the hum could be in the original recording...

Steve A.

[gary]

IGary may have some suggestions here on this matter...

I can't see how this can be caused by a variation in the signal in such a way as it to be NOT apparent when displayed digitally (i.e. by TBP), for it not to be apparent digitally it would mean the variation would have to be less than a pixel in duration and the variation we see in the example is much more than that.

I am 99% sure this is always introduced mechanically and almost certainly by eccentricity.

Albert, as the variation is over 360 degrees it won't matter where you are viewing it you will see the same thing - look at it this way, at some point in a revolution the (say) disk will be at it's zenith or highest point - at that time the aperture that is in the gate will be higher than it should be by that amount. When the disk is at it's nadir or lowest point the opposite aperture will be in the gate and lower than it should be by that amount - and so on in a perfectly sinusoidal fashion. Across the lines you would just see a some widening or narrowing of the picture I would have thought.

Another common(ish) way of this happening is in the layout of the (say) disk when it is done electronically (computer) as any rounding off errors tend to sinusoidal in nature. This may also happen in some mechanical construction techniques no doubt.

[Steve Anderson]

I can't see how this can be caused by a variation in the signal in such a way as it to be NOT apparent when displayed digitally (i.e. by TBP).

Ah! That's data I don't have...does the same file play OK in TBP? That I don't know. I don't know where the source of Alberts files is/are.

What I was more driving at above (r.e. suggestions) was a known good source of test files such that without confirmation using a scope, the output from the PC or CD player should be as perfect as expected from a test signal. Files that could be downloaded, posted here or e-mailed.

Steve A.

[gary]

Ah! That data I don't have...does the same file play OK in TBP? That I don't know. I don't know where the source of Alberts files is/are.

What I was more driving at above (r.e. suggestions) was a known good source of test files such that without confirmation using a scope, the output from the PC or CD player should be as perfect as expected from a test signal.

Steve A.

Well I guess I was really implying:

a) Anything generated by Video2NBTV is a good test source, and

b) The validity of any signal can be check by displaying with TBP (or with NBTVAnalyser if you want to get fancy).


But I should also point out that the club CDs are meant to be THE source of test signals, but even a 12.5kHz square wave (variable duty cycle preferably) from a good quality signal generator or similar would do the trick.

There have, of course, from time to time been newsletter articles on the construction of crystal locked test signals as well.

Edit: whoops - I meant, of course, 400 Hz not 12.5Hz - I had to get out of the bath to post this - can't let mistakes like that go uncorrected.

[Viewmaster]

IGary may have some suggestions here on this matter...

I am 99% sure this is always introduced mechanically and almost certainly by eccentricity.

Albert, as the variation is over 360 degrees it won't matter where you are viewing it you will see the same thing - look at it this way, at some point in a revolution the (say) disk will be at it's zenith or highest point - at that time the aperture that is in the gate will be higher than it should be by that amount. When the disk is at it's nadir or lowest point the opposite aperture will be in the gate and lower than it should be by that amount - and so on in a perfectly sinusoidal fashion. Across the lines you would just see a some widening or narrowing of the picture I would have thought.

That's a nice piece of reasoning there Gary, which is I can now see is right......well I did say I was probably overlooking something with my sketch and the 360 degree part, is it. Silly ol' me.

I can say from experience that Nipkow NBTV pictures certainly do widen and narrow if the disc is running eccentric......I found this out when adjusting the Nippers film disc to run true.

BUT, if entirely mechanical, why does the wavy line vary in shape and intensity depending on the various pictures shown?
This seems to indicate some variation in a signal property....I have noticed the the o/p from the 4046pll varies somewhat on different pictures, all with good sync pulses, and all frame locking.

Maybe the answer lies there somewhere, but that would mean that a
1/2 kilo flywheel's speed is increasing/decreasing slightly 12 1/2 times per sec, which seems improbable ? But that would explain the wavyness.

Steve mentioned hum, but again that would produce a constant waveform on all pictures I would have thought?
BTW, I have also tried loosening/tightening the drive belt but it makes no difference.

Thanks, guys for your thoughts.

[gary]

BUT, if entirely mechanical, why does the wavy line vary in shape and intensity depending on the various pictures shown?

Whoops I missed that part - where did you see that?

Steve mentioned hum, but again that would produce a constant waveform on all pictures I would have thought?

I agree, although I suppose if the sync pulse itself was sinusoidal then lifting and lowering the signal with hum might give that effect as the sync slice would be triggered earlier and later in time - but I think that's drawing a long bow...

[Viewmaster]

BUT, if entirely mechanical, why does the wavy line vary in shape and intensity depending on the various pictures shown?

Whoops I missed that part - where did you see that?

Gary, I'm sorry not sure what you mean by, " where did you see that?"

?.........I see it on the Nipper's picture where the disc is not more that a pixel in error and it runs true on a ball race bearings shaft.
Sometimes the frame line is pretty straight but rises up or down at the end and at other times it looks like a crooked wavy line across the whole picture.
I am coming to the tentative conclusion that in spite of a heavy flywheel, our old friend the 4046pll is doing it somehow or the other, even when the frame is locked.
Adjusting the 4046 10k speed pot and the 200k pot though doesn't cure the fault.

[gary]

Hi Albert, well I suppose what I meant was, where can I see it?

Ok, well in that case ditch everything I have said because I haven't seen what you describe. What I was referring to was the phenomenon as pertains to the photo in your first post (which is quite common). Since you have made what is probably the smallest Nipkow disk ever made (outside of a confocal microscope) perhaps you are seeing something unique to that, i.e. the result of a miniscule amount of flex based runout or harmonic vibration.

a) is there anyway you can post a video?

b) just exactly what is the signal you are using and can you post that?

BTW a ball race doesn't guarantee no runout, heaven knows if it did I would get you to build me a spindle for my CNC machine, I have tried a number of "off the shelf" solutions and they all have exhibited runout to a more or lesser degree - so much so I can't make any more bead disks until I find a solution (at least one that doesn't cost me thousands of dollars.

[Viewmaster]

Gary, the Nipper is partly down at present but here is an earlier video of it.
Note the wavy frame line at the top which varies in shape and size depending on what picture is running, sometimes being very much worse than this example.

In this case it's track 44 of the club's test CD showing a lady's face.
I hope this link works for you.

Re disk flexing....it is securely sandwiched between two perspex washers.
Maybe the mini video camera is upsetting the apple cart? When back together I shall turn the camera off and see what happens.

[gary]

Well that looks like a fairly static wave to me - sure you are getting some other glitches which would be due to some sort of sync jitter - but I would be astounded if that wave is not due to either eccentricity or periodic hole misalignment. Best I could judge from that short clip anyway.