This one is about is about CRT displays, not projectors but maybe someone will find it interesting. I was reading recently about how with digital audio a technique called "dithering" is used whereby a small amount of noise is added that increases the perceived resolution. I had a listen to a *6* bit audio file with and without dithering and the first sounded terrible and distorted, the second sounded quited hissy but otherwise completely understandable; a big improvement.
Anyway, I thought about it for a while and then forgot about it but wheels must have been still turning inside my head because yesterday I got this idea you see... With a shadow mask or vertical stripe CRT the resolution is fundamentally limited by the spot size and the stripe or dot pitch. Have a close look at an old mono green screen - with no mask and continuous phosphor - the image can be razor sharp. My idea then is to perceptually overcome the limiting factor of the stripes or dots.
For a stripe tube the resolution limitation is in the horizontal plane so what you do is (seriously now) adjust the raster so it finishes a little in from the edges, and hold the monitor or at least the crt in a mechanical fixture that can move the crt side to side maybe several centimetres in a sinusoidal motion , and at the same time an ac current proportional to and in sync with the lateral movement of the crt is applied to either the horizontal deflection yoke or the horizontal shift cct. The result should be that the crt moves side to side at several Hz (depending on it's mass) while the actual image remains in the one apparent position.
The point of all this is that the verttical edges of any image are now not dominated by the stripe pitch at all because the image edges are continually sweeping across many stripes that effectively blur into one. This is just the same as driving down the street in your car and being able to see clearly "through" the palings of a "moving" picket fence. The spacing of the palings doesn't affect the details of what you can see provided they are moving!
A long persistence phosphor would not work very well in this regard, you would have to try it and see. Also, a shadow mask type crt would have to be moved in the vertical plane as well.
/Circlotron - pushing the ideas envelope so hard people will think he is nuts.
Anyway, I thought about it for a while and then forgot about it but wheels must have been still turning inside my head because yesterday I got this idea you see... With a shadow mask or vertical stripe CRT the resolution is fundamentally limited by the spot size and the stripe or dot pitch. Have a close look at an old mono green screen - with no mask and continuous phosphor - the image can be razor sharp. My idea then is to perceptually overcome the limiting factor of the stripes or dots.
For a stripe tube the resolution limitation is in the horizontal plane so what you do is (seriously now) adjust the raster so it finishes a little in from the edges, and hold the monitor or at least the crt in a mechanical fixture that can move the crt side to side maybe several centimetres in a sinusoidal motion , and at the same time an ac current proportional to and in sync with the lateral movement of the crt is applied to either the horizontal deflection yoke or the horizontal shift cct. The result should be that the crt moves side to side at several Hz (depending on it's mass) while the actual image remains in the one apparent position.
The point of all this is that the verttical edges of any image are now not dominated by the stripe pitch at all because the image edges are continually sweeping across many stripes that effectively blur into one. This is just the same as driving down the street in your car and being able to see clearly "through" the palings of a "moving" picket fence. The spacing of the palings doesn't affect the details of what you can see provided they are moving!
A long persistence phosphor would not work very well in this regard, you would have to try it and see. Also, a shadow mask type crt would have to be moved in the vertical plane as well.
/Circlotron - pushing the ideas envelope so hard people will think he is nuts.
Circlotron said:/Circlotron - pushing the ideas envelope so hard people will think he is nuts.
Just wanted to agree with you on that point.. 😀
Organic Dithering.
Video CRT's use electromagnetic deflection as a means of making them affordable for televisions. Electrostatic CRT's (like an O-scope) cost a lot more to make.
Magnetic deflection limits the precision of the beam landing.
Electron beams also exhibit a degree of gaussian distribution when tricking a target. There is the center of the target surrounded by a zone of decreasing probability of electron collision.
The mask in a color CRT shapes this blurring, but does not eliminate it.
Then there is the persistence of the phosphors. The beam scans the phosphors and they continue to glow for some period after they've been hit. This also aids our persistence of vision.
All of these account for the gaussian blur that help make a CRT work. I don’t know that throwing some dithering on top of that is going to aid the picture much.
Charles Poynton has written some good books on digital imaging where he explains CRT operations quite well. He used to also have some good web articles.
Dithering has however been used for years to aid longitudinally scanned FM modulated recordings of both analog and digital signals. The pickup head's have piezoelectric or magnetic deflection coils attached that have a dithering signal applied.
As the head rotates across the moving tape path the RF level of the FM signal is monitored. The dithered head wiggling across the tape path can then be steered into the optimum position. Matsushita used to sell a consumer VHS machine that did this, though it is more commonly found in analogue and digital recording devices costing between $10k and $500k.
8 bit digital video signals can be tough to combined or sometimes even look at without applying a 1-bit dither to the LSB.
16 bit digital audio workstations sometimes have a dithering option to make the -lowlights- sound less grainy. Back in my studio days I typically kept the digital tape hiss turned off.
Ikegami has experimented with using monochrome CRTs with LCD chroma overlays with some striking results, and some strikingly prohibitive prices. The gaussian blur of the light source beneath the LCD chroma mask may be very close to achieving what you’ve suggested. Hughes-JVC took this principle to the extremes with the light amplifier video projectors.
-Dave
Video CRT's use electromagnetic deflection as a means of making them affordable for televisions. Electrostatic CRT's (like an O-scope) cost a lot more to make.
Magnetic deflection limits the precision of the beam landing.
Electron beams also exhibit a degree of gaussian distribution when tricking a target. There is the center of the target surrounded by a zone of decreasing probability of electron collision.
The mask in a color CRT shapes this blurring, but does not eliminate it.
Then there is the persistence of the phosphors. The beam scans the phosphors and they continue to glow for some period after they've been hit. This also aids our persistence of vision.
All of these account for the gaussian blur that help make a CRT work. I don’t know that throwing some dithering on top of that is going to aid the picture much.
Charles Poynton has written some good books on digital imaging where he explains CRT operations quite well. He used to also have some good web articles.
Dithering has however been used for years to aid longitudinally scanned FM modulated recordings of both analog and digital signals. The pickup head's have piezoelectric or magnetic deflection coils attached that have a dithering signal applied.
As the head rotates across the moving tape path the RF level of the FM signal is monitored. The dithered head wiggling across the tape path can then be steered into the optimum position. Matsushita used to sell a consumer VHS machine that did this, though it is more commonly found in analogue and digital recording devices costing between $10k and $500k.
8 bit digital video signals can be tough to combined or sometimes even look at without applying a 1-bit dither to the LSB.
16 bit digital audio workstations sometimes have a dithering option to make the -lowlights- sound less grainy. Back in my studio days I typically kept the digital tape hiss turned off.
Ikegami has experimented with using monochrome CRTs with LCD chroma overlays with some striking results, and some strikingly prohibitive prices. The gaussian blur of the light source beneath the LCD chroma mask may be very close to achieving what you’ve suggested. Hughes-JVC took this principle to the extremes with the light amplifier video projectors.
-Dave
- Status
- Not open for further replies.