Correcting Nipkow Disc Aspect

[Andrew Davie]

It occurred to me that if the Nipkow disk was askew, relative to the viewer (that is, if the image is on the right, and the disc center is closer than the right edge, and the left edge of the disc is closer still), then there would be some visual correction of the "wedge" shape of the typical Nipkow display. Also, the curved scanlines would appear to straighten (slightly?).

Of course, the skewed image would be much narrower than needed, so this would be corrected by making the holes in the Nipkow disc wider and further apart (and hence the image area larger on the disc itself). The holes' width could also be adjusted to remove any apparent line thickness changes. When viewed from directly in front (with an askew disc), the image would look less Nipkow-like. I think.

If the same hole reconfiguration was used, then the disc could be used in normal configuration, and the correction could be done by a pair of mirrors in front of the display -- the first reflecting the display 45 degrees or so to the second, alongside, which reflects the image to the eye. By rotating the mirrors around the vertical axis, one could correct the 'askewness'.

Just thought I'd throw this idea 'out there' for posterity.

Cheers
A

[Stephen]

This is a great idea Andrew! It would be ideal for converting images produced by a source with linear scanning, such as a camera with a drum type scanner or PC, for viewing on a display with a disc type scanner. Of course, you would not want to do this with a signal from a source with curvilinear scanning, such as a camera with a disc type scanner, because the image itself would become distorted in the process of "straightening" the scanning lines.

Having said that, I think that a camera with a disc type scanner could employ your proposed scheme by projecting a focussed image upon the disc at an angle as well, with the disc having appropriate aperture sizes and spacings. This is a very interesting prospect.

[Andrew Davie]

A further possibility: By spacing out the Nipkow holes and making them wider (say, twice as wide) as described above, but not skewing the disk angle, we have now also doubled the brightness of the image. Holes twice the size --> twice the light getting through. Post-correcting the image width with double-mirrors as described will restore it to the correct aspect ratio.

[Stephen]

Interestingly, John Logie Baird had a similar but different means to provide some correction of this rectangular distortion. In his British Patent GB329644, filed 19 February 1929, Mr. Baird suggests that the angle between each aperture increase as the apertures spiral toward the disc centre. In this way, the arcs of all the scanning lines remain of equal length, thereby reconstructing an approximately rectangular picture, even though the scanning lines themselves remain curvilinear.

Another advantage is that because each scanning line has an equal length arc, each pixel is of equal height. With an ordinary scanning disc, the pixels in the inner lines are "squeezed" together so that they overlap, so that the inner lines have slightly less resolution than the outer lines.

This is also a nice approach to experiment with. I have posted this patent at http://www.taswegian.com/NBTV/images/GB329664A.pdf for reference.

[Stephen]

Upon re-reading the Baird patent, I noticed another embodiment with reference to Figures 13 through 15, the description of which starts on page 4, column 1, line 58. Mr. Baird suggests the use of a cupped disc in combination with a magnifying lens. The apertures along the edge would be cupped toward the lens, providing less magnification than the inner apertures that are further away. This embodiment would also approximate a rectangular picture, although the lines would have different widths from one side to the other and the cupped disc would be a nuisance to fabricate. It is clever though, and it is along the lines of what you are proposing with the tilted disc.

[Stephen]

It has further occurred to me that correcting the keystone distortion of the disc display window would allow a multiple spiral disc that provides a much larger display window than would otherwise be possible. John Logie Baird describes a high efficiency multiple spiral disc in his British Patent GB322822, filed 11 July 1928. This disc comprises a group of two spirals of glow lamps, one glow lamp for each scanning line, with a commutator to distribute the modulating signal to each of the glow lamps over two revolutions of the disc.

Normally, such a large display window in a disc would have significant keystone distortion, but with such corrective measures as we are discussing, such as proposed by Andrew or Mr. Baird in his GB329664 patent, this issue largely disappears. Such a disc could use LEDs in place of the glow lamps, or it could comprise an optical fibre disc and a single high power LED with an optical commutator as I have proposed, modified to have two or more spirals.

I have posted a copy of Mr. Baird's GB322822 patent at http://www.taswegian.com/NBTV/images/GB322822A.pdf for reference.