| |
Parody
Registered: Jul 2013
Posts: 7 |
A question about resolution and refresh rate
I have a question, in the form of a set of assumptions, and I hope you will correct me if I'm wrong. :)
The old PAL TV-system had 576 visible lines, each drawn every other 1/50th second (576i 50Hz). The PAL C64 has 200 lines inside the border, so I assume that every C64-line covers two TV-lines. Then, on the TV the two lines that make up the C64-line will be drawn up every other, and opposite, 1/50th of a second. That is, one line is drawn one 1/50th of a second and the other line is drawn up the next 1/50th of a second and so one. The C64 will give you 50 raster interrupts on, let say, line 60 each second (or close enough). So, would this mean that by changing the contents of a line every 1/50th of a second in a raster interrupt you could effectively double the vertical resolution?
Have I understood it correctly? Is what I'm describing FLI? |
|
| |
Oswald
Registered: Apr 2002
Posts: 5094 |
to my understanding c64 always "instructs" the TV to draw the same set of those lines. so every other line remains always blank.
fli is a different thing, it increases nr of colors possible to use within character cells in gfx modes. |
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
what oswald said, the image is displayed always on the same lines, every other line is left blank (which creates the "scanline" type of appearance).
you can verify it yourself easily, just create two images with a single vertical line at the same place and switch between them every frame. |
| |
Parody
Registered: Jul 2013
Posts: 7 |
Thank you for the replies! :)
So, if the C64 only uses one field and then effectively only updates the screen every 1/25th second, does that mean that you only get 25 raster interrupts on a line every second?
...or, do you get 50 raster interrupts, but that every other raster interrupt has no immediate effect on the TV? |
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
every 50th of second a full frame is shown, and you get 50 interrupts per second |
| |
Parody
Registered: Jul 2013
Posts: 7 |
Good! So, you get 50 interrupts every second. :)
...but, I don't understand how you can show a full "frame" every 50th of a second if you only use every other line. As I understand it old TVs were interlace and showed a set of every other line one 50th of a second, and the other set of every other line the next 50th of a second, and so on. If you then only use one of these sets of every other line then you should get an effective screen update every 25th of a second. Surely old TVs were not capable of progressive scan...? |
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
15khz monitors always show 288 lines per frame (50 times a second), with every other line left blank. whether every other frame is offset by one line depends on the video signal, its not a feature of the monitor. |
| |
Parody
Registered: Jul 2013
Posts: 7 |
Ah, ok, I think I understand now. So, the C64 could via the video signal tell the TV or monitor to display the same field every 1/50th of a second, instead of the every other field pattern of normal TV at the time. You have enlightened me! :) |
| |
lft
Registered: Jul 2007
Posts: 369 |
What groepaz said. The television set needs a way to distinguish odd fields from even fields, otherwise it could end up displaying them in the wrong order, and that would look like a really badly de-interlaced dvd rip. Thus, every field is marked with an odd/even identifier.
However! That explanation makes sense to us in the digital era, but television is much older than that.
Here's what really happens:
The electron beam is guided to a position on the screen using the horizontal and vertical components of a magnetic field. The important insight is that these components are handled completely independently. Horizontally, the beam sweeps at a constant rate from left to right, and is reset by horizontal sync pulses. Vertically, the beam sweeps at a slower constant rate from top to bottom, and is reset by vertical sync pulses. As a consequence, the rasterlines are actually slightly diagonal, at least in terms of the independent horizontal/vertical axes of the beam director. But in many tv models, the beam guiding apparatus is physically tilted to compensate (although this creates a slight amount of shearing instead).
It follows that by changing the time delay between the vertical sync pulse and the following horizontal sync pulse, it is possible to adjust the vertical position of the image. This is indeed how interlace is implemented. The c64 (and Amiga in non-interlaced modes, and many consoles) maintains a constant time delay between the vertical sync and the first horizontal sync, and this effectively results in a progressive display with visible gaps between the scanlines.
The trouble nowadays is that monitors try to be clever and make assumptions about the signal in order to clean it up. Partly, this is necessary in order to de-interlace the signal and resample it for the internal frame rate of the LCD. But depending on the algorithm used, this may not work well (or at all) with a signal consisting of only odd fields. |
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
Quote:with a signal consisting of only odd fields
you mean, with a signal where field = frame. there is no odd/even frames at all :) |
| |
AmiDog
Registered: Mar 2003
Posts: 97 |
I suggest googling for 240p. For example this page does offer some insight into the progressive signals of older game consoles and computers. |
| |
lft
Registered: Jul 2007
Posts: 369 |
Quoting GroepazQuote:with a signal consisting of only odd fields
you mean, with a signal where field = frame. there is no odd/even frames at all :)
For somebody who is accustomed to the standard PAL signal, I'm sure that the fields generated by the VIC are quite odd. |
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
indeed =D |
