| Author |
Message |
cmjohnson
Joined: 03 Apr 2006
Posts: 5180
Location: Buried under G90s
|
| Posted: Mon Jun 28, 2010 3:20 am Post subject: |
|
|
| Spanky Ham wrote: |
All this sets the stage for my disagreement with your post CJ. The fact is 8" pjs can resolve 1080p. Scott proved this years ago. Will it be sharp in the corners? Probably not, but people may want to give that up for the higher res/refresh. Even if the center of an 8" has a lower MTF than a 9", it still may be good enough for some. IIRC, MTF didn't budge a great deal from 1080p (it may have been higher than that) down to some lower resolutions. It seems like MTF really spiked at like 800x600. THis was four years ago in Scott's lab with Darin, Scott and myself playing around with the MTF device. Everything was on the fly, so I didn't have any tests planned out. It would be interesting to test MTF on a variety of res/refresh with both a 9" Marquee and an 8" Marquee (with an Ampro thrown in for good measure  ).
|
Well, we really don't HAVE any disagreement here. I've also seen 8" projectors resolve 1080p and do a pretty good job at it...but the MTF achieved by 9" CRTs under otherwise identical test conditions will be a higher value.
Interestingly enough, while Scott was doing his tests with the high res 8" tubes (made by Lexel Imaging) for the Marquee 8500HR, that was about the same time I was doing tests with a slightly upgraded 8000 chassis and a P19LNQ green tube in brand new condition, aside from a nasty central spot burn.  This was the basis for my orginal FrankenYoke experiments. I was using a Sencore CM2125 color monitor analyzer as my generator source.
I was also using healthy, recapped, refurbished, late model neck cards and a 2035-02P VIM in the machine. It was really an 8500 without astig. Late model 8000, with all 8500 parts. A software change and an added astig board and it would have been an 8500. But I was only using ONE tube most of the time. I had cheater plugs connected to the neck card connector cables so I could do that. Makes for a simpler test setup.
Maybe you're going to believe me. Maybe you won't. But as far as I can determine, NOT having an MTF meter on hand
to prove it, I was able to essentially match what Scott got out of his HR prototype. He used custom tubes and stock yokes. I used a higher resolution tube class (The LNQ type has a finer beam spot size than a 180DMB) and FrankenYokes.
I cranked the resolution all the way up as high as the Sencore would go, 2048x2048. I was easily able to look right into the lens face (or to the tube face with the lens off) and count the one on, one off scan lines in the central region of the picture. And I don't mean, I could tell that there were separate scan lines, I mean, there were BLACK lines between the bright green lines. I can't give a number but the MTF was rocking, I'm sure of that! It literally brought a tear to my eye, so I turned the contrast level down to non-painful levels. 
I could count the scan lines all the way to the corners of the picture with the lens off. With the lens on, the corners got a bit fuzzy at that level.
The lens was of course an HD-8B. I found that with that lens I was not able to get sharp enough focus to resolve on the screen what was as clear as can be on the CRT face. Reportedly, the newer HD-8Bs are better. Maybe they could
have helped.
The math works out such that 2048x2048 resolution in 4:3 mode is the same line pitch as 2048x1536 (QXGA) compressed
to a 16:9 raster. FAR beyond 1080p requirements. But I couldn't get that incredible sharpness on the screen due to lens limitations.
If you doubt I did this, I've given you the exact formula I used to make it happen. You could duplicate the test if you want, and I'll even provide a free set of optimized FrankenYokes for an 8000 or 8110. (No astig windings.) Go for it. Find yourself a P19LNQ green tube and a signal source (A PC with a reasonably modern graphics card will do) and an 8500 chassis and give it a shot.
But, obviously this isn't a stock projector, is it? Non-stock tubes, non-stock focus coils. A stock projector might display the image but it would not be able to rock the MTF values the way this sucker could.
HD145s would have helped. They ARE better lenses. But I didn't have any.
I don't dispute that an 8" machine can display 1080p and even do a good job on it. But as for whether or not it can
really fully resolve it is a question I can't answer with any confidence. Certainly they can look good at 1080p, and
that's what's most important and I'm happy to leave it at that.
CJ
|
|
| Back to top |
|
 |
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Mon Jun 28, 2010 3:44 am Post subject: |
|
|
I think that x2 switch was suppose to be on x1.......
Just kidding CJ.
|
|
| Back to top |
|
|
benareeno
Joined: 22 Mar 2006
Posts: 1614
Location: ottawa, canada
|
| Posted: Mon Jun 28, 2010 5:03 am Post subject: |
|
|
you really think you can discern 2000 lines with an 8 inch tube? Roughly 250 lines per actual inch...adding the scanlines in between that actual picture data would seeming put the size of the lines too small for human vision...wouldn't it?
Call me crazy, but I just doubt that very much...
why do people obsess about 1080p? Can you even discern it at 10 or 15 feet?
masking a 2.35:1 picture will do a lot more than sqeezing another 10% of resolution out of an image...and how many of you are doing that?
High resolution is for test pattern viewers...not movie watchers. Luckily this forum is full of test pattern viewers.
|
|
| Back to top |
|
|
Mark_A_W
Joined: 15 Mar 2006
Posts: 3068
Location: Sunny Melbourne Australia
|
| Posted: Mon Jun 28, 2010 6:33 am Post subject: |
|
|
| macgyver655 wrote: | | Mark_A_W wrote: |
(And despite implied suggestions to the contrary, the only real difference between a top end 8"er and a 9"er is the tubes. . |
And that alone gives the 9"er the ability to more easily resolve a higher resolution. |
That's what I meant.
The electronics make some difference - a G90's electronics make about 30% difference compared to a G70, the tubes make the rest.
|
|
| Back to top |
|
|
huggy
Joined: 02 Aug 2008
Posts: 927
Location: Melbourne,Australia
|
| Posted: Mon Jun 28, 2010 7:13 am Post subject: |
|
|
| benareeno wrote: | you really think you can discern 2000 lines with an 8 inch tube? Roughly 250 lines per actual inch...adding the scanlines in between that actual picture data would seeming put the size of the lines too small for human vision...wouldn't it?
Call me crazy, but I just doubt that very much...
why do people obsess about 1080p? Can you even discern it at 10 or 15 feet?
masking a 2.35:1 picture will do a lot more than sqeezing another 10% of resolution out of an image...and how many of you are doing that?
High resolution is for test pattern viewers...not movie watchers. Luckily this forum is full of test pattern viewers. |
Ben
While I'm in agreement with you,I think the point here and on many other tweaks is more or less the association with the hobby side of Crt rather than actual image enhancement.I think many here have gotten the most out of their display's and the rest is just academic for the most part.
I mean, if it wasn't for these threads,we'd have nothing else but Ron's political rants and Wan's.........well......rants!
Dave
|
|
| Back to top |
|
|
CasetheCorvetteman
Joined: 09 Nov 2008
Posts: 6326
Location: Australia
|
| Posted: Mon Jun 28, 2010 9:52 am Post subject: |
|
|
| Mark_A_W wrote: | Back to the topic....
Case
I played around with 1920x804p at 72hz on my XG.
And I could not tell any difference from 1920x1080i at 96hz.
So I just stuck with the single res, 1920x1080i at 96hz (called 48hz in powerstrip).
(And despite implied suggestions to the contrary, the only real difference between a top end 8"er and a 9"er is the tubes. There's no huge difference (as in orders of magnitude) in the performance of the electronics..well, until you get to a 909!). |
Cheers Mark I tried Powerstrip a while ago but wasnt too keen on it, with the nVIDIA drivers on Windows 7 ( not sure about others ) these days you can adjust the resolution in dot for dot sizes, you can set the frequency and all that sort of stuff, this pic will explain it better:
Seems to work quite well. Dont know how it compares to powerstrip, i didnt really get a chance to test it as well as i have with this. Havent tried any of those smaller resolutions ( height wise ) yet myself, but i was keen to hear what others had to say about it. I figured it would be less effort for the projector if nothing else, didnt seem too beneficial drawing 400 lines of black!!
If i had the cash last time i saw a Reality 909 in the products for sale section, it wouldve been hanging from my ceiling right now I bought the solar electric system and the solar hot water with that money though, so it was well spent.
| benareeno wrote: | you really think you can discern 2000 lines with an 8 inch tube? Roughly 250 lines per actual inch...adding the scanlines in between that actual picture data would seeming put the size of the lines too small for human vision...wouldn't it?
Call me crazy, but I just doubt that very much...
why do people obsess about 1080p? Can you even discern it at 10 or 15 feet?
masking a 2.35:1 picture will do a lot more than sqeezing another 10% of resolution out of an image...and how many of you are doing that?
High resolution is for test pattern viewers...not movie watchers. Luckily this forum is full of test pattern viewers. |
Itll be alot more than that, by the time you take unused phosphor into account itll be more like 600 or 700 lines per inch. I certainly dont think youll see them on the tube face either, but its not the point really.
I like the test patterns, it reminds me of days gone by when we used to have LANs and play lots of video games, we would do all sorts of bench marking to see who had the best rig at the LAN. It was a bragging right, and it was good fun. Test paterns and resolving higher resolutions is like bench marking a bunch of gaming computers at a LAN. Saying these test paterns dont matter would have been like turning up to one of our gaming nights years ago and telling us bench marking doesnt matter. It matters allright, especially when youve got the most powerful rig in the room with 16 fans in the full tower case keeping it cool.
| huggy wrote: |
Ben
While I'm in agreement with you,I think the point here and on many other tweaks is more or less the association with the hobby side of Crt rather than actual image enhancement.I think many here have gotten the most out of their display's and the rest is just academic for the most part.
I mean, if it wasn't for these threads,we'd have nothing else but Ron's political rants and Wan's.........well......rants!
Dave |
Exactly right, but there is nothing wrong with doing some testing!!
|
|
| Back to top |
|
|
virusc
Joined: 11 Apr 2007
Posts: 358
Location: Massachusetts
|
| Posted: Mon Jun 28, 2010 1:26 pm Post subject: |
|
|
most 1080p sources do not even offer full horizontal resolution anyway. 1500 or so. Also we do not perceive horizontal resolution as much as vertical anyway. Joe Kane I think spoke about this when talking about 1080i vs. 720p. the 1280 vs. 1920 did not make that much of a difference horizontal anyway.
I have seen nearly what CM is speaking of but with using edge to edge on a 8" tube at very low drive(contrast) and perfect astig and such. at usable contrast the slight bloom took this away. Also you do have to look straight into the lense as CM mentioned. Non of the 8" lenses are that good. at center some of the very rare all glass are at center but not at edges, especially so when you get the extreme edge to edge which is out of the sweet spot of the lense.
|
|
| Back to top |
|
|
cmjohnson
Joined: 03 Apr 2006
Posts: 5180
Location: Buried under G90s
|
| Posted: Mon Jun 28, 2010 5:43 pm Post subject: |
|
|
| benareeno wrote: | you really think you can discern 2000 lines with an 8 inch tube? Roughly 250 lines per actual inch...adding the scanlines in between that actual picture data would seeming put the size of the lines too small for human vision...wouldn't it?
Call me crazy, but I just doubt that very much...
why do people obsess about 1080p? Can you even discern it at 10 or 15 feet?
masking a 2.35:1 picture will do a lot more than sqeezing another 10% of resolution out of an image...and how many of you are doing that?
High resolution is for test pattern viewers...not movie watchers. Luckily this forum is full of test pattern viewers. |
I clearly outlined the equipment required to duplicate my procedure and results. You are welcome to follow this line of exploration if you want to, but don't say it can't be done if you're not willing to put forth the effort yourself.
Tse achieved similar results with the Marquee 8500HR. You can order one from VDC right this minute if you really want to,
and due to its better lenses, it'll actually display that resolution capacity on the screen.
Do you doubt HIM?
Let me fry your mind with this: AmPro's infamous light valve projectors used a CRT to address the liquid crystal light valve.
That CRT has a 3 inch diagonal faceplate. Its primary emission is in the near IR but you can see a visible dark red trace
on the CRT if it's not attached to the light valve.
That 3" CRT delivers 2500x2000 resolution as specified in the AmPro specifications in the manual.
There are some CRTs made for special applications that have even finer beam spot sizes than that.
CJ
|
|
| Back to top |
|
|
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Mon Jun 28, 2010 5:51 pm Post subject: |
|
|
| cmjohnson wrote: |
That 3" CRT delivers 2500x2000 resolution as specified in the AmPro specifications in the manual.
There are some CRTs made for special applications that have even finer beam spot sizes than that.
CJ |
But the CRT itself is limited in light output. Super high resolutions are easily capable at low light output. Its when you try to get a watchable light output the resolution suffers.
|
|
| Back to top |
|
|
virusc
Joined: 11 Apr 2007
Posts: 358
Location: Massachusetts
|
| Posted: Mon Jun 28, 2010 6:43 pm Post subject: |
|
|
|
CJ, The Ampro CRT in the light valve is really not a CRT as you know. It really is a deflected electron beam and is not a light emitting device so it should not be compared to FP CRT's which are the source of the illumination. Resolutions much above what LUGs can do is not really possible as the tradeoffs with danger/cost/service life are not feasible.
|
|
| Back to top |
|
|
cmjohnson
Joined: 03 Apr 2006
Posts: 5180
Location: Buried under G90s
|
| Posted: Mon Jun 28, 2010 7:27 pm Post subject: |
|
|
Sorry, you're wrong about that. I know that system pretty darned well. It's a 3" high resolution CRT with an IR phosphor, with versions made by Thomas Electronics or IST (Imaging and Sensor Technology) with a glass fiber optic faceplate.
Like any projection CRT, it is subject to phosphor wear and burns and has a typical quality lifespan of 10,000 operating hours.
I've worked on them. I have one example of one on my shelf of dust collectors. I've seen them in operation without the LV attached.
It IS a CRT. Somewhat specialized, but it is NOT a "deflected electron beam and is not a light emitting device". That is false.
Its light output (not visible to the eye except for a little bit of top-end visible red emission) is the drive signal for the light valve, but
the drive signal required to get peak white out of the system requires the CRT phosphor to be driven to saturation, just like a direct drive
projection system. (Projection CRTs)
The CRT's light output may be limited but the power density on the phosphor is still the same...saturated phosphor at peak.
But, those little CRTs also use very fine deflection/focus systems with an impressive number of poles for fine beam control. I haven't
counted how many poles but it's a bunch.
CJ
|
|
| Back to top |
|
|
benareeno
Joined: 22 Mar 2006
Posts: 1614
Location: ottawa, canada
|
| Posted: Mon Jun 28, 2010 8:33 pm Post subject: |
|
|
I'm saying that you can't see 300+ individual lines in 1 inch of real estate on a tube face with the naked eye...that's what I'm saying.
I'm sure you saw something, but not that many individual lines...
|
|
| Back to top |
|
|
jkruger
Joined: 24 Oct 2007
Posts: 2435
Location: Carlsbad, CA
|
| Posted: Mon Jun 28, 2010 8:36 pm Post subject: |
|
|
|
300 lines in one inch are the width of a human hair. I can see those up close even with bad eyesight.
|
|
| Back to top |
|
|
ecrabb
Forum Moderator
Joined: 13 Mar 2006
Posts: 15909
Location: Utah
TV/Projector: JVC RS40, Epson 5010
|
| Posted: Mon Jun 28, 2010 9:04 pm Post subject: |
|
|
| benareeno wrote: | | I'm saying that you can't see 300+ individual lines in 1 inch of real estate on a tube face with the naked eye...that's what I'm saying. |
Not sure why you'd say that, Ben... Maybe you mean 300 white lines with tiny/no gaps? That would definitely be difficult to see with the naked eye, though I suspect a sharp 9"-er is probably in the ballpark...
To clarify, 300 alternating black/white lines (150 on/150 off) are certainly - and easily - visible to the naked eye.
SC
|
|
| Back to top |
|
|
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Mon Jun 28, 2010 9:25 pm Post subject: |
|
|
I'm implementing a new shovel rating system. It will consist of 1 to 5 shovels. They will be used accordingly with no explanation and no reasoning behind the number of shovels. Enjoy! 
|
|
| Back to top |
|
|
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Mon Jun 28, 2010 9:28 pm Post subject: |
|
|
Would someone please post a picture here of a 1 inch white block with 300 hairs laying side by side so I can see this for myself. Oh, and dont forget to state what color hair you used in your picture.
Thank You!
|
|
| Back to top |
|
|
benareeno
Joined: 22 Mar 2006
Posts: 1614
Location: ottawa, canada
|
| Posted: Mon Jun 28, 2010 9:33 pm Post subject: |
|
|
don't forget the real estate taken up by the actual scan lines...you'd have to add that in too. So 300 human hairs, with the distance of the scan lines in between them....
We need mythbusters here...but I simply don't believe it.
I believe that you believe it and you believe you saw it, but I think you are seeing something other than that exact resolution...at least, to your naked eye you are.
|
|
| Back to top |
|
|
cmjohnson
Joined: 03 Apr 2006
Posts: 5180
Location: Buried under G90s
|
| Posted: Tue Jun 29, 2010 12:16 am Post subject: |
|
|
| benareeno wrote: | I'm saying that you can't see 300+ individual lines in 1 inch of real estate on a tube face with the naked eye...that's what I'm saying.
I'm sure you saw something, but not that many individual lines... |
You are quite wrong about that.
44 gauge magnet wire is .0022 inches in diameter. It is easy to see individual strands of this wire. It's easy to see the individual strands of this gauge wire when they're laid together. At reasonable viewing distances, like at arm's length or closer.
If you have a printer capable of printing 600 DPI, you could make your own print test patterns with one on, one off, or checkerboard patterns, and see for yourself that 300 DPI is very easy to resolve with the naked eye.
CJ
|
|
| Back to top |
|
|
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Tue Jun 29, 2010 1:09 am Post subject: |
|
|
|
|
|
| Back to top |
|
|
macgyver655
Joined: 22 Aug 2007
Posts: 8508
|
| Posted: Tue Jun 29, 2010 1:12 am Post subject: |
|
|
First its scan lines, then its hairs, then its wire and now dots. Dots dont work in the analogy by the way. And you guys are doing the math wrong.
Where's my picture?
|
|
| Back to top |
|
|
|
|
