Starting at the beginning with Mirror Screws

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Postby Viewmaster » Sat Oct 01, 2011 5:33 pm

Just an interesting extra in that the direction of scan is in the opposite direction for a convex slat compared with a convex one.
To prove it, get a flexible bendable surface ( perspex or cover of a CD).
Bend it to concavity and watching a lamp reflected in it rock it to and fro noticing the direction of the 'reflected light.'

Do the same with a convex bent surface. Notice how the direction of the'scanned light' is reversed.

BTW, my rig could have easily been adapted for convex slats,but the viewing position would have to be at 2 different points to set the swivelling mirrors, but I see, Steve, that you are going down the concave road, as convexity is too VEXing ! :-)
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby Viewmaster » Sat Oct 01, 2011 6:23 pm

Panrock wrote:1) They'll have to be made of something other than stainless steel! Although I have now obtained a good polish on the reflecting edges, there still seems to be a 'grain' effect disturbing the quality of the image. A mirror screw needs to exhibit a perfect, smooth, mirror finish.

Steve O


Quotes from Peter Yanzer's site........
http://www.televisionexperimenters.com/ ... ology.html

"two 32-line stainless steel mirror screws have recently been sold to separate customers for $850 each."

"The stainless steel screw produced an excellent 120-line picture"
........end quotes.

A grain effect maybe caused by insufficient very, very fine polishing with the very finest of grits.
Non stainless may tarnish in time ???

PS.......I'll do your scans later on.
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby Panrock » Sat Oct 01, 2011 7:50 pm

Viewmaster wrote:Just an interesting extra in that the direction of scan is in the opposite direction for a convex slat compared with a convex one.
To prove it, get a flexible bendable surface ( perspex or cover of a CD).
Bend it to concavity and watching a lamp reflected in it rock it to and fro noticing the direction of the 'reflected light.'

Do the same with a convex bent surface. Notice how the direction of the'scanned light' is reversed.


This is true only when we are viewing the concave surface outside its focal plane, when it gives a diminished, inverted virtual image. However in this application we will need to view it inside its focal plane, so that it will magnify the reflected view, thereby extending the width of the picture. Try flexing the CD a little less... :)

Viewmaster wrote:Quotes from Peter Yanzer's site........
http://www.televisionexperimenters.com/ ... ology.html


A handy link to have. Thank you.

Viewmaster wrote:"The stainless steel screw produced an excellent 120-line picture"
........end quotes.

A grain effect maybe caused by insufficient very, very fine polishing with the very finest of grits.
Non stainless may tarnish in time ???


Yes, both could be the case.

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby Viewmaster » Sun Oct 02, 2011 6:59 pm

Steve, I cannot scan full pages but here are all the relevant items extra to what you have already seen about curved mirror screws.....

The last 2 pics show a means of shining the illuminant onto each slat in turn to intensify the light. Not really applicable with modern means of lighting and complex to make anyway.

There are also references to small screw to lock each slat but other than what I have shown there is no further explanation as to what this means in practice.

A rig improvement........
1. put a small vertical marker on each swiveling mirror on which to sight to, the slat width apart.
2. Arrange for a fixed sight or hairline in which to look at the mirrors when setting their angle, rather than trying to put your head against a chair or door. Non movement is important.

I wonder if all this will work ? ...! :-)
Attachments
Picture 868.jpg
screw adjustments ??
Picture 868.jpg (221.55 KiB) Viewed 13014 times
Picture 870.jpg
Picture 870.jpg (191.47 KiB) Viewed 13014 times
Picture 871.jpg
curved mirror screw
Picture 871.jpg (213.77 KiB) Viewed 13014 times
Picture 873.jpg
drum for light intensification
Picture 873.jpg (237.61 KiB) Viewed 13014 times
Picture 874.jpg
light intensification drum
Picture 874.jpg (228.42 KiB) Viewed 13014 times
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Magnification role of mirror screw

Postby kareno » Mon Oct 03, 2011 8:54 pm

I don't want to rain on anybody's parade and I applaud any efforts to realise a mirror screw (because it is very difficult).

I feel I should point out, however, that I came to the conclusion a while ago that the mirror screw is an ingenious magnification device, and so it has all of the unpleasant attributes of a magnifier, foremost being narrow viewing angle.

For example, on a 32 line NBTV mirror screw you can perform framing adjustment by moving your head up and down! Equally though, you lose framing if you move your head for the purpose of, say, sitting down.

The pictures you see on the internet give the impression of a picture 'screen' much like a CRT with wide viewing angle, etc. It is likely that the cameramen chose their standing positions very carefully to get those shots.

Imagine you took a more conventional, say, drum monitor and put it on a table facing away from you. Imagine then you put a larger concave (magnifying) mirror behind the table but facing you, so that you could see the magnified reflection of the monitor screen. That is what a high line count mirror screw image will look like.

In fact, the mirror screw is worse because of the different apparent depth of the image, depending on whether you are eyes are level or tilted. At least the big concave mirror doesn't have that awful atigmatism.

In addition, the aspect ratio of the picture changes as you move closer or further from the screw. At least a concave mirror affects both axes equally.

I just thought I'd mention this in case anyone has unrealistic expectations of the mirror screw. It is a remarkable and ingenious mechaincal scanner but its principle of operation is still magnification with all the attendant problems that brings.
kareno
 

Postby AncientBrit » Mon Oct 03, 2011 11:29 pm

@Karen,

Is the mirror screw exclusively a direct viewing device or can an image be projected onto a ground glass screen?

If so do the variable geometry effects, depending on the observer's view point, no longer become a problem?

Regards,

Graham
AncientBrit
Green padded cells are quite homely.
 
Posts: 858
Joined: Mon Mar 26, 2007 10:15 pm
Location: Billericay, UK

Postby Panrock » Mon Oct 03, 2011 11:50 pm

Hi Karen,

Thanks for the warning. Don't worry... no unrealistic expectations - at least from me. From past experience I know how often NBTV displays can prove disappointing for one reason or another... and the constructional accuracy required in this case will be formidable. That's one reason this design can be switched to 60-lines if necessary - still demanding, but less so.

In general, past experience has shown me that NBTV gives much more convincing results when you don't try to be over ambitious or 'push things'. There's no 'free lunch'. Again, I want to clearly state this rig may never get built. Still. mirror screws fascinate me at the moment and I'm going to try jumping in at the deep end! :D

This morning I've emailed a new file to the laser cutters for a couple of concave trial pieces, radius of curvature 4 metres. This figure hasn't been plucked out of thin air, it's the result of some thought. I'll then see the effect and work from there, with further trials likely. It's simpler and less time-consuming for me to do it this way than construct a test rig, and it doesn't cost too much either.

A question regarding the 'differential focus' (for want of a better term):

Is it true that if the line of light is fairly close to the m/screw but the observer is fairly far away, this effect would be minimised, but at the cost of some horizontal linearity ?

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby kareno » Tue Oct 04, 2011 1:12 am

Hi Graham, Steve,

Yes, I believe the mirror screw has been used for projection, Graham. I suspect it is not efficient because you illuminate the whole screw yet only one mirror is ever reflecting onto the screen at any one time. So its close cousin the mirror drum would probably be a better choice for projection.

Steve, you are right - viewing distance can minimise the astigmatism. You'll find that if you move your head up and down, the light appears to be coming from the screw. But if you move your head side-to-side, then the light will appear to be as far behind the screw as the light source is from the screw, possibly further if curved mirrors are used.

Given that you probably want your viewing position to be no more than 8-10 feet away, and assuming the light source is 3 feet away (only a guess) then I suspect the parallax difference won't be concealed that well.

The test will be when you tilt your head a little!
kareno
 

Postby dominicbeesley » Tue Oct 04, 2011 2:57 am

Hello all,

The viewing angle is indeed tight, though this can be mitigate a fair bit by adding extra lines above and below the picture and by making each line a bit too long.

I tried a bit with the projection idea and lasers and would be possible. An even better idea would be projection along with the lens drum in the articles above

D
User avatar
dominicbeesley
Anyone have a spare straightjacket?
 
Posts: 685
Joined: Sat Sep 13, 2008 11:32 pm
Location: Hebden Bridge

Postby Panrock » Tue Oct 04, 2011 3:17 am

dominicbeesley wrote:The viewing angle is indeed tight, though this can be mitigate a fair bit by adding extra lines above and below the picture and by making each line a bit too long.


So from this, am I to understand the way to go will be to add (say) 10% to the length of each of my slats and then expect that the line of light will be seen simultaneously at the opposite extremities of adjacent slates, that is - when the observer is at the distance which gives the correct aspect ratio...?
Phew - a bit of a mouthful that !

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby kareno » Tue Oct 04, 2011 8:38 am

Hi Steve,

Yes, a little overscan is a good idea to cope with sync drift.

Also, if you choose the slat width in strict accordance with the aspect ratio for the height of the stack (i.e. slat width = 4/3 * stack height for a 4:3 aspect ratio) then you will be forced to place the light source right next to you.

If you want the light source to be to one side, i.e. some degrees around to the left or right, then the active mirror (that which is showing a light spot) will be tilted away from you by half that angle, and the apparent width of the mirror will diminish. You'll need to compensate for this.

This brings to mind a key to understanding mirror screws. We all know that if you were to illuminate an entire Nipkow disk, you'd see 32 little images in a circular band, all at various states of synchronisation. Well a similar things happens with mirror screws.

The world reflected in the mirror screw has a circular 'fence' of images - 120 of them! The radius of this circular fence is, for the case of flat mirrors, the same as the distance of the light source from the screw. The width of the slats has to be sufficient to see one of the images on this imaginary fence - preferably the one that is correctly synced!
kareno
 

Postby Panrock » Tue Oct 04, 2011 6:11 pm

Karen, it seems to me you have encapsulated all the essential wisdom about mirror screws in just one post.

kareno wrote:Yes, a little overscan is a good idea to cope with sync drift.
I'll definitely include this now.


kareno wrote:Also, if you choose the slat width in strict accordance with the aspect ratio for the height of the stack (i.e. slat width = 4/3 * stack height for a 4:3 aspect ratio) then you will be forced to place the light source right next to you.
This bit I still have to get my head around.


kareno wrote:If you want the light source to be to one side, i.e. some degrees around to the left or right, then the active mirror (that which is showing a light spot) will be tilted away from you by half that angle, and the apparent width of the mirror will diminish. You'll need to compensate for this.
Could this be done by simply stepping back a bit?


kareno wrote:The world reflected in the mirror screw has a circular 'fence' of images - 120 of them! The radius of this circular fence is, for the case of flat mirrors, the same as the distance of the light source from the screw. The width of the slats has to be sufficient to see one of the images on this imaginary fence - preferably the one that is correctly synced!
Very useful. Thank you.

Steve O
Panrock
Green padded cells are quite homely.
 
Posts: 870
Joined: Mon Feb 05, 2007 8:25 am
Location: Sedgeberrow, England

Postby Viewmaster » Tue Oct 04, 2011 6:27 pm

Here's a 60 line colour mirror screw built to perfection.
Run the videos. Interestingly the camera moves arounds a little but the picture is still perfect.
There is also an intersting mirror angle adjusting pic.

http://la-radiovision.fr/projetmirror-suite.htm

Do a Google translation.
User avatar
Viewmaster
Frankenstein was my uncle.
 
Posts: 1306
Joined: Fri Apr 06, 2007 4:50 am
Location: UK Midlands

Postby kareno » Tue Oct 04, 2011 8:16 pm

Could this be done by simply stepping back a bit?


I don't think so, Steve. The extreme case is when the light source is ninety degrees away from you, i.e. to one side of the set while you're in front. In that situation the mirror showing a reflection of the light source will not be flat on to you but rotated 45 degrees. The mirror will therefore only appear 71% (one over root two) of its width.

If I were building a screw and planned to put the light source at 90 degrees to me, then I'd make the slats 41% wider than aspect ratio would demand to compensate for this.

A more realistic angle for the light source is perhaps 60 degrees max, 45 degrees being more typical. These will demand 15% and 8% over-sized slats respectively. Formula: 1/cos(0.5*angle)
kareno
 

Postby dominicbeesley » Tue Oct 04, 2011 8:53 pm

Plus add some more so that the person sitting next to you can see the image too. You need to calculate in how wide a viewing angle you'd like or just add 15% to the length of each screw anyway.

You have to remember that moving towards or away will make the percieved width change but not the height...

Dom
User avatar
dominicbeesley
Anyone have a spare straightjacket?
 
Posts: 685
Joined: Sat Sep 13, 2008 11:32 pm
Location: Hebden Bridge

PreviousNext

Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 31 guests