Anybody figure how big those pixels are?

My math says* about 1mm diagonal on a 65" 16:9 set. I can live with that for a long time.

*I only ran it once - check me!

Diagonal pixels == 1469 (square root of 1280x1280+720x720)

Diagonal inches == 65

65/1469 == ~.044 per pixel, or 22.5 pixels per diagonal inch.


Charlie


Worked a second way - a 16:9 65" screen is 56.7x31.9
1280 divided by 56.7 == 22.6
720 divided by 31.9 == 22.6



[This message has been edited by charlie (edited July 15, 2002).]

Charlie,
I'm no good at math, if you like crunching what would:

3.15 Million
(1366x768x3)
or
1.05 Million
(1366 x 768)each panel

Diagonal Pixels = ?
Diagonal Inches = 60

run out to?

Well....

Assuming a 16:9 60" screen, that would be 52.3 x 29.41...

52.3 / 1366 == .038 or 26.1 pixels per inch
29.4 / 768 == .038 or 26.1 pixels per inch

The 'diagonal pixel' thing was me being stupid - the pixels are square but the 16:9 diagonal is not a 45 degree line - the screen diagonal won't cross at the pixel corners. To compute pixel diagonal multiply by 1.4 (for square pixels)


Charlie

Thanks Charlie
I tried to run the #'s myself after reading your earlier post, but was coming up with mis-mash.

No problem!

I ran those first set since I'm thinknig about a HD2 based RPTV in the 60-65 inch range and there was some concern regarding 1080i getting down converted to the native 720 resolution. As a simple test I printed a pattern of blocks on a 1/20 inch grid on paper and viewed it at different distances. My preliminary conclusion is that (unless the DLP looks very different) the pixel size at real viewing distances is, for me, a non-issue.


Charlie

I can't see pixels on my 65" Mitsu. from ~13' away where I sit (or even standing much closer) and that even when only watching DVD's that are projected at 480P.

1080i from say,... PBS, look much sharper though, but it depends so strongly on the original picture source's res. AND quality.

I doubt you'll have any problem between 720p and 1080i as far as total resolution or downconverting, but since 1080i seems to be more popular I don't dig 720P as much, and once 1080P sets become possible, everyone who's into it like we all are will be upconverting all that 1080i to progressive for a much better than 720P picture.

'Real world speaking' none of this matter right now though IMO.

One thing to note on DLP.
The current DLP RP sets (from Mitsu, Panasonic and/or Pioneer -I forget which) seemed to have a blurry effect on fast motion.
If you look at one in a store that feeds the same HD signals to all their HD sets and you see something fast like a Football game...

Look at how blurry/choppy (hard to describe) the pic looks as it pans across. It's like it can't keep up w/ the picture's speed.

Now look at the CRT-based sets. They'll follow the motion much better.
I haven't gotten a chance to look for this effect on the FP DLP systems I've seen, but it should be the same since it's the same DLP chip.

Not sure if the newer DLP chips move any faster?

Could be 'cuz the original feed is 1080i and the 720P DLP sets are too slow at downconverting? Just a wild thought based on total guessing.

I would doubt the actual DMD was the problem since they can 'switch' at 5kHz. The conversion, maybe motion detection stuff in processing would seem a good bet, as you guessed.

But I'm just guessing too.

Charlie

I read somewhere that this was a problem with DLP technology but latest generation TI chips should be faster. The reason stated had something to do with the electrical charge that attracts the corner of each micro mirror to ‘tilt’ it to display the required light/color. It’s the speed that the charge can tilt each mirror at, - which is the limiting factor. The first generations were supposed to show this ‘choppiness’ on fast action where the mirrors can’t keep up. I forgot watching Samsung displays to ck for that occurrence. Have you looked for this effect on the 2 newest Samsung displays Azryan? Their resolution alone looked sharper than the other competing DLP’s on the floor but I truly did not watch for the ‘fast action’ problem to see if that area had improved.

I don't know if the cause was slow mirrors, slow processors, slow color wheels, or some combination, but first generation DLPs definitely had a fast motion blur problem. I didn't find the problem anywhere near as bad as it is on a slow-refresh LCD monitor, but it was there. It was particularly noticeable on the DLP-based RPTVs (I watched a lot of HDTV football demos on those units). In the limited time I had with the Samsung DLP RPTVs, I didn't notice this problem at all, nor have I seen it on fast color wheel front projectors.

-avi

My whoops Avi,
I retract my prior post just ran into the infor. it DID list the blurred fast action as due to the three devices used for primary colors or the color wheel speed, whichever technology is used.