Dude, like I was goofing with my monitor and reading the VRefresh and HRefresh frequencies over and over, trying to calculate the number of scanlines being transmitted (Because I figured that it wasn't the 1024 that my video was supposed to be rendering due to some additional marking and the like, much as you would find in a TV) and I noticed something. The numbers where in flux. Now we're talking 1% per every 10 samples, but by just enough for me to notice it. And I got to thinking... If the the monitor thinks the refresh rates are in flux, that would suggest that it's signal locking circuits would have to be wavering. This would mean that the screen would have to be sliding left and right. Now Theoretically it could move up and down, but I know there's enough spacing in the timing of the scan lines that that's really a non issue. So I figured I'd do a quicky experiment to try and support the "wobble" hypothesis. I figured the amount would be small, especially since I've NEVER noticed it in past. So I engineered an experiment that put a single black point on a white background on my screen, and then held the tip of an unwound #1 paperclip up to it. Now although my hands move, I can get the end of the paperclip to stick on the glass at the point of contact. Well, I watched carefully, knowing that the amounts of variation I should see should be somewhere near the dot pitch of the screen. And low and behold. There's my little black dot, covered up by the paperclip poking out from the left side of the paper clip sometimes, and on the right other times. Now this isn't conclusive, but definitively suggests that it's clearly possible that the screen is wobbling by a factor of about 2.5X the dot pitch. We're taking a variation of approximately .55mm in this case, but .55 is still much larger than zero. Also, the fact that the dot never appeared above or below the paperclip shows that it's highly likely that the wobble is not coming from the motion of me or my paperclip, as that motion should be chaotic, and not reserved to just left and right motion.

Just one of those thoughts. Oh, and regarding the original inquiry, it seems that for my display, at 1024 video scan lines, the monitor is receiving 1065 scan lines. This would leave 41 scanlines for syncing and other things. This is quite odd though, since I thought that the VGA standard didn't have any black space beyond the syncing time. But then again who know, maybe VGA needs 41 scanlines worth of time to finish a vertical sync. Hmm

*Messes with his monitor and drops it down to 800x600*

Ok, now at 800x600 I get 46.7 KHz HRefresh I get 625 scan lines... Now if I relate the number of true scan lines (1024 to 600) and then solve for the expected number of "Extra" scanlines, assuming I know that there's 25 extras at 600, I get 40.95, which is 41 based on my rounding. Therefore, since both modes where at 75Hz VRefresh, then I could assume that the amount of time spent is exactly the same (And I'm almost right, they're within 10% or so,) So It's probably part of the VGA standard (I'm gonna call it vertical blanking time, and guess that I'm likely right) and I just need to find a web resource that talks about the VGA standard to prove it. ^^




- posted by Jim The Cactus @ 6/04/2004 04:31:31 PM | Permalink | (0) Comments. Add your own!