AVS Forum banner
Status
Not open for further replies.
1 - 8 of 8 Posts

·
Registered
Joined
·
124 Posts
Discussion Starter · #1 ·
Here is a topic that may be of interest to those of us considering high scan rate direct view sets such as the new 720p capable Princeton Ai3.6HD 36" 4:3 (with built-in iScan Pro chipset) or the Sampo 34". I looked hard at the Sampo (in person) but it has a lot of well documented quirks (though I love the flat screen and fine pitch, etc.)


OPTIMIZATION OF BEAM SPOT SIZE--OR, WILL THIS SET REALLY RESOLVE 1080i IN A 16:9 WINDOW?

Ignoring for a moment the whole video bandwidth issue (which these sets certainly seems to have enough of), we all know that maximum lateral resolution can be reasonably estimated by width divided by stripe pitch. But looking at how many lines the set can paint across the screen in a given area (full screen or in a letterbox), and actually have resolved (not just jammed together and overlapped to end up soft and fuzzy), it seems to me that herein lay the REAL difference between a 720p 16:9 and a 720p 4:3.


To date while there have been a handful of direct views with vertical image squeeze for DVD, none have offered 720p capability and most don't even claim to be able to do 1080i (basically 540p) in the squeezed window. And none of even the 34" 16:9 sets offered 720p (30" Princeton AF3.0 a smaller screen exception). Now along comes the 34" 16:9 Sampo and this new Princeton Ai3.6HD 4:3 and we have a horse race to think about. (Please let's check our 16:9 vs. 4:3 philosophy at the door and just focus on the technical question I am posing).


Bandwidth-wise this new Princeton should be capable of 1080i in a 16:9 window because the thing claims to go to 1024h x 768p full screen, so 540p (1080i) in a 30% shorter window should not be a problem. Now Princeton recommends 800x600 for sharp VGA input and here is where I start to wonder. 480p DVD in the 16:9 area should look great (600 full screen being about 33% larger than 480).

Sorry for not getting to my point, which is: if a set (suppose this applies to all CRT based RPTV and FPTVs too) is truly 1080i capable with fine enough phosphors to resolve it--and is not overlapping the beam at 1080i--then aren't we seeing gaps between the lines (small I know) when watching 480p? (Even more so if the set is 720p capable). Of course that is why folks use quadruple scalers with the big 9" CRT projectors. I mean, these sets do not change beam spot size based on horizontal scan rate or, in the case of sets doing the squeeze, based on scanned area.


So how does one know what a tube and its beam and electronics are really optimized for? Obviously a lot of RPTVs upconvert everything to 1080i so that is a good clue that they are optimized for that (with varying results of course). But I have yet to see this issue talked about in earnest regarding direct views (even Joe Kane's and Greg Rogers tech articles are light on this topic, in part because there have not been any sets for consumer use where this could be an issue).


Just so you don't think I am a lunatic contemplating something that won't be see at (or maybe you will when I tell you this), I intend to sit 4 feet or less from this 36" set to have it fill my field of view.

Perhaps users of 720p optimized sets (not that this really is one, I do not know, hence the long post) should have 720p scalers or DVD players which do a decent job upconverting to that.


For those that want to see the info on the set that I am looking at, check:
http://www.princetonhdtv.com/Product.../Ai36HDp1.html

And there is another thread in this forum with more info about the exact Silicon Image chipset being used and the units other capabilities. It is:
http://www.avsforum.com/ubb/Forum11/HTML/013415.html


Lastly, for those that are dying to turn this into another one of the many 16:9 vs. 4:3 debates, I plead with you not to. Let's stay on topic. (And if you care to hear MY own feelings for the short term,they are that: a) the Princeton 4:3 is the best big direct view I know of and the 16:9 window will be about as tall and wide as the 34 16:9s); b) since my critical viewing will of course be with 16:9 media, differential phosphor burn in will only degrade the 4:3 image and not the other way as when widescreen users watch 4:3 windowboxed (besides, I run the temp and picture levels low enough to make the tube outlast the period in which the thing will be around).


Looking forward to the discourse.

AJC

P.S. I am not a totally new poster here. For some reason the AVS board purged my prior registration (under same user name that I reregistered on). I've posted about 30 times--but I am a senior lurker!:)
 

·
Registered
Joined
·
1,128 Posts
Interesting post. I thought the limiting factor for resolution for direct view sets was more connected with the design of the "shadow mask" or whatever proprietary name is given to the grid matrix upon which the CRTs focus? Hence, regardless of spot size, resolution was limited by that grid? Maybe this is not the whole story. Does anyone know?

Anyway, a nice direct view has a great deal going for it, such as brighter image, better contrast ratio, and potentially more saturated color, when compared with RPTV, and I hope these sets continue to improve and get larger. The big problem with direct views for me is their HD image is so small, it requires sitting right in front of it!

John in VA
 

·
Registered
Joined
·
398 Posts
Superdad.


Have you seen the 37" monitors from Mitsubishi and NEC? They will do 1920 X 1080p 60 Hz and "squeeze" DVD and HDTV to 16:9 too. I too wonder about the effective resolution though.


The specs are better but I haven't actually seen a side by side comparison with the new Princeton.


Mike
 

·
AVS Forum Special Member
Joined
·
11,139 Posts
An interesting point for discussion; not sure myself if it's ever been pinned down here. Asked similar questions myself about my new RPTV last summer after reading the manufacturer's claim it would resolve a 1080i image; believe they say display 2 million plus pixels.


The starting point for me, and I hoped to see this at the P.G. site, too, was a screen image photo (cropped, high-contrast) with a test pattern clearly showing its 1000+ full screen width resolution. Never could find one for my RPTV. You'd think this would be an important techie selling point for set makers making such claims. Also, for a direct view, I'd want to pin down just what the screen and phosphor stripe/dot pitch is, then do the math and see how close you come to 1000+ full screen width, taking beam widths into account. You'd think, with the growing numbers of moderately priced HDTV test-pattern generators, those concerned could snap high-resolution screen images and publish them on the web. Like you, hope the commentary doesn't diverge to a 4:3 vs 16:9 debate. Similarly, hope it doesn't evolve into a discussion of what is actually being delivered by broadcasters and how much of that picture resolution remains, after filtering, on any screen surface. That's a fat thread by itself. -- John


------------------

STOP DVI/HDCP AND DFAST
 

·
AVS Forum Special Member
Joined
·
11,139 Posts
Believe another question to ask is what is the crossover point as far as resolution capability and brightness between computer monitors and consumer-oriented monitors such as Princeton Graphics'. Too much beam current to deliver extra brightness with a computer monitor and the beams diverge, spoiling resolution--likely damaging the ultra-fine grille or mask needed for high resolution, too. So, in an 'HDTV' consumer-oriented, direct-view CRT monitor, if you want enough brightness for group viewing at a distance, how do you build in an ultra-fine grille or mask? If someone claims they have, let's see the test patterns to prove it.


When you say looking to resolve 720p "across them", I'm assuming you mean the 1200+ full-width horizontal resolution of that HDTV format. Of course, that format seems to be rarely transmitted these days. And filters in sets whack about 20% off what's delivered, meaning the set only has to display roughly 1000 lines horizontal resolution (full width) with 720p, assuming the maximum detail was ever broadcast. Also, since prerecorded 1080i broadcasts appear to be limited to 1000-1400 horizontal resolution, minus the 20% set filtering, this format's resolution demand isn't significantly higher (for now, at least). Still, to be repetitious, wouldn't mind seeing the proof that direct-view CRTs provide even this much detail. -- John


------------------

STOP DVI/HDCP AND DFAST
 

·
Registered
Joined
·
124 Posts
Discussion Starter · #6 ·
As gnosys pointed out, in my original post I sort of combined the factors of phosphor and shadow mask by refering to "stripe pitch," when, for any proper exploration of this topic we need to speak of them separately.

That is:

a)the red, green, and blue phosphor matrix that is actually painted onto the back side of the glass; and the pitch and pattern of same, which I believe are often full height-of-glass vertical stripes perhaps with tiny black horizontal tick marks to improve contrast (not sure about this--is anyone else?).


b) the type and configuration of the shadow mask, a metal grid (often made of the heat resistant material trade named Invar, which I recall Zenith originally patented many years ago in the early PC monitor days) suspended close to the inside fac of the tube, which acts in part to narrow the focus of the 3 color beams as they are driven across the width and height of the glass.

When shadow masks were perforated with round holes in dot-triads, the max. number of horizontal lines that could be painted ("vertical resolution") was limited by pitch (the way it is for horizonal res across the screen even with "aperture grill type masks).

Nowadays, there seem to be two prevalent types of so called shadow masks in use on all CRTs:

1)Sony's Trinitron aperature grill which consists of hundreds of thin vertical wires and none horizontally, save for a couple of support wires across (which used to be visable on my old Sony computer display);

2)Everyone else's grills which I guess are Invar metal sheets with tiny vertical rectangles punched in them (does anyone know about the type and variation of the dominant "stripe pitch" masks used these days? I'd really like to know).


(By the way, the dot-triad shadow mask technique is not dead and I feel that for computer displays it can have a great advantage. Fine dot triads generally render text at small point sizes more smoothly in my experience than do equivalent stripe pitch tubes (note that it is not fair to directly compare dot pitch and stripe pitch center-to-center figures). Hitachi has really done a neat thing with their dot array computer displays in two ways: they make the vertical pitch almost the same as the horizontal pitch; and they mazimize the hole size/minimize the between dot metal area, which I think allows them to have a brighter image before dreaded "blooming" of the beam. Check out http://www.hitachidisplays.com/how_monitors/page3.htm My Hitachi CM771 while lacking the "punch" of many aperature grill monitors, renders fine lines and tiny text with extraordinary focus.)


So anyway, for CRT-based HDTV, it seems that the important three parameters (after metting basic requirements to scan high, have wideband video amplifiers, and decent geometry) are: 1)phosphor pitch/pattern, 2)shadow mask pitch and construction, 3)beam spot size and focus.


One advantage I can see to an aperature grill (at least one that has no horizontal wires--such as Sony--how does Toshiba do it?) is that vertical resolution (horizontal scan lines) is not limited by the mask the way it would be with a dot-triad mask.

Does anyone think that the phosphor and shadow mask pitches (again I guess there isn't one with Sony tubes, but how about the others) are already much finer than the 720p we are looking to resolve across them?


Can someone do a little research on these issues and report back some facts? Boy, I would love to take a tour of a broadcast or computer monitor manufacturing plant to see how they make these amazing tubes.


Regards,

AJC
 

·
Registered
Joined
·
292 Posts
I asked a similar question to Superdad's in a Sampo-related post:


"The Sampo has been praised for having high resolution compared with other HDTVs of similar size. If the Sampo screen has a dot pitch of 0.82 mm (per the review), then it is similar to if not higher than the dot pitch on other 34 inch flat screen direct view HDTV monitors. Surely the Sampo could not have higher resolution than what is implied by the dot pitch. If the 34" diagonal screen is about 29.6 inches wide, that is about 753 mm. With that dot pitch, the screen can show about 918 lines across its width, maximum. Isn't that pretty much the same as other monitors with the same dot pitch? How could the Sampo be so superior?"


The aperture grille, beam spot size, signal up/down conversion, etc., may also conspire to reduce resolution to some extent. But it seems to me that the direct view set can't produce better resolution than limitations in the phosphor pitch. And the phosphor pitch is externally measurable. So regardless of the HD format (1080i or 720p) I doubt that the 34" direct view sets can resolve more than 900 - 1000 lines across. (Not that that is bad!) Right?
 

·
Registered
Joined
·
124 Posts
Discussion Starter · #8 ·
Well, I am happy to be seeing some action off my original post, but if you scroll back up to it you will see that my interest for the moment revolves around vertical resolution (how many horizontal lines), and what these sets are optimized for (in a 16:9 window).

While yes, horizontal res (lateral resolve) is of course ultimately limited by phosphor pitch, that does not seem to be as much the case in the other direction. That is: if the phosphor stripes are continuous top to bottom, and the shadow mask is really a slot mask or aperature grill, then it seems that beam spot size and what it is optimize for is the limiting factor. (Am I wrong about modern tube construction? Particularly Toshiba tubes?)


On a practical basis, what I am asking is this:

1)How much better will DVDs (a 720h x 480 medium as far as encoded content goes) look scaled up to 720p vertical on a direct view that is capable of displaying it?

2)For this Princeton 36" 768p capable (yes 768, as it can sync to 1024x768, but they say it looks shapest at 800x600) 4:3 set, what might be optimum to display in its 16:9 window ('cause the Dish 6000 I plan to buy next month can output 720p)?

3)Do the scalers that go to resolutions of 720p and above do anything with the horizontal content? That is, do they try to interpolate anything additional on each line, or do they just interpolate (or deinterlace in the case of basic 480p 2:3 pulldown chips) additional lines to tighten spacing?


The last question is one that I do not recall seeing addressed anywhere. And while obviously additional horizontal information (and a set's ability to display it) is a key factor in making true HDTV look amazing (that's why the fantasy ATSC specs go up to 1920h), I just do not see how DVDs suffer any on sets limited to around 800-900 lines.

And does anyone have an estimate as to what HD broadcasts are really putting out in terms of horizontal information?


I pose a lot of questions to stimulate dialog, but please don't take this thread too far away from my core questions concerning optimized direct view spot size and shadow mask limitations.


Best to all,

AJC.
 
1 - 8 of 8 Posts
Status
Not open for further replies.
Top