| Author |
Message |
HiTracey
Joined: 07 Apr 2011
Posts: 11
Location: United Kingdom
|
| Posted: Thu Apr 07, 2011 10:42 am Post subject: Stacking three crt pjs |
|
|
Hi folks
I'm looking for help and advice. I'm non-technical, I'm not blessed with a large bankroll for my hobbies, and have a fairly restricted amount of time. Within those parameters I'm trying to improve my home cinema experience in a reasonably quick and straightforward manner and, as everyone will know, this is a rocky path. I'm blundering down it blind.
I've done everything practical in terms of a darkened room, with blackouts and dark colours/dark non reflective surfaces. I'm now trying to create a neat installation for three pjs (Barco 808s) and have run into a problem. I've hung all three on the ceiling with the outer pair toed in, and I simply can't converge properly on the l/r 10% of the screen. The inner 80% is acceptable; any loss in sharpness is countered by a great improvement in brightness and contrast. I can't seem to adjust the horizontal linearity sufficiently at the edge. It appears clear that I'm exceeding the ability of the pj to adjust to such an extreme angle to the screen.
Any ideas?
I've other options:
Plan B - cut it down to two pjs which, my calculations suggest, will be fine for angles - shame to lose a pj;
Plan C - As above plus third pj slung below - difficult engineering and not as aesthetically pleasing;
Plan D - blending three pjs - a real pain from the sound of it;
Plan E - blending two outer pjs over a full screen pj - this sounds best.
My concern about blending is that it appears expensive, time consuming and I can't really divine whether there is something I can buy, plug in, twiddle around with and I'm there. That is, I've paid someone else to do the technical work and make it simple for me, as I can't spend weeks and months fiddling around with it. Is there a reasoanbly cost effective product that fulfills this demand that anyone can recommend?
My problem is, having jury rigged the projectors vertically (aesthetically horrible) and been blown away by the resultant image, it's simply not possible for me to revert to one pj. The difference is simply night and day.
Thanks in advance.
|
|
| Back to top |
|
 |
Boilermaker
Joined: 21 May 2006
Posts: 527
|
| Posted: Thu Apr 07, 2011 11:26 am Post subject: |
|
|
How about Plan F - Blending two of the projectors and use the third as a spare. Your light output will be more than the three stacked, and your ANSI contrast will go way up because of the greatly reduced bandwidth required. You should be able to get a pair of TVOne's for about $2,500 or so.
BTW - there is absolutely no advantage to blending three for video use.
Bob
|
|
| Back to top |
|
|
HiTracey
Joined: 07 Apr 2011
Posts: 11
Location: United Kingdom
|
| Posted: Thu Apr 07, 2011 11:41 am Post subject: |
|
|
| Boilermaker wrote: | How about Plan F - Blending two of the projectors and use the third as a spare. Your light output will be more than the three stacked, and your ANSI contrast will go way up because of the greatly reduced bandwidth required. You should be able to get a pair of TVOne's for about $2,500 or so.
BTW - there is absolutely no advantage to blending three for video use.
Bob |
Bob
Thanks for your input. That is certainly viable for me. The triple blend seemed expensive and overly complex. The use of the third pj as a full screen overlay seemed a way of giving a further kick to the contrast/brightness of a basic blend.
Having experienced improvements in these parameters by firstly shrinking my screen to roughly half the size (14ft down to 10ft), then darkening the room, then stacking the pjs, I'm like a junky looking for my next fix. To be fair, I had the big screen/light room for many, many years before starting this journey, therefore I'm looking for sensible improvements, not some unrealisable elusive 'perfection'.
I've tweaked the sound/LFE system to where I want it and I'm done (and in fairness that's where 90% of the budget went), so I'm capable of drawing my line in the sand and saying "It's enough".
We all recognise that lines in sand can disappear fairly quickly when the wind of change blows.
Regards, Tim
|
|
| Back to top |
|
|
Boilermaker
Joined: 21 May 2006
Posts: 527
|
| Posted: Thu Apr 07, 2011 1:27 pm Post subject: |
|
|
| Quote: | | The triple blend seemed expensive and overly complex. The use of the third pj as a full screen overlay seemed a way of giving a further kick to the contrast/brightness of a basic blend. |
As long as your screen AR is 1.78 - 2.4 range, there is absolutely no advantage to using a triple blend as you would greatly reduce the use of each projector's available raster area.
I also would not use the third projector at all:
1 - The total raster area would only increase by about 25% because the third projector would use only about 1/2 that of the blended pair. Especially so if you use a 2.4 AR screen.
2 - Your ANSI contrast would actually be reduced slightly because the video signal to the 3rd projector would be seeing a video signal almost twice the frequency of each of the blended pair.
3 - Besides being more difficult, convergence would not be as successful because of normal thermal drifting.
Your choice of a 10' (assuming width) screen is right on for an 8" blend. My screen is 10'-6" wide at 2.40 AR and it is VERY bright. Make sure you use a screen that does not exceed unity gain.
Bob
|
|
| Back to top |
|
|
garyfritz
Joined: 08 Apr 2006
Posts: 12088
Location: Fort Collins, CO
|
| Posted: Thu Apr 07, 2011 2:27 pm Post subject: |
|
|
| Boilermaker wrote: | | Your light output will be more than the three stacked, and your ANSI contrast will go way up because of the greatly reduced bandwidth required. |
!? How does bandwidth affect ANSI contrast!? That's going to be determined by light scatter (especially since his 808s aren't LC), not the input signal...
|
|
| Back to top |
|
|
Boilermaker
Joined: 21 May 2006
Posts: 527
|
| Posted: Thu Apr 07, 2011 2:51 pm Post subject: |
|
|
| Quote: | | !? How does bandwidth affect ANSI contrast!? That's going to be determined by light scatter (especially since his 808s aren't LC), not the input signal... |
If I'm not mistaken, ANSI contrast is defined as the ratio from one pixel at 100IRE to the next pixel at 0IRE. Digitals have an easy time with this which is why they appear sharper.
With the analog equivalent, this signal would go from full voltage (.7 volts) to zero volts at the rate of the video signal to emulate the pixels. The bandwidth capability of the video curcuitry determines how well the signal actually follows this ideal.
Obviously it doesn't quite get all the way to .7volts and it doesn't get back down to 0 in the limited time, therefore the full white pixel isn't quite white and the black pixel isn't quite black.
Supposedly, if a specification of bandwidth is listed as XMHz, then it is actually 3db down at that frequency, which is only 1/2 the voltage.
Bob
|
|
| Back to top |
|
|
zaphod
Joined: 16 Jun 2006
Posts: 2002
Location: Cloverdale
|
| Posted: Thu Apr 07, 2011 3:08 pm Post subject: |
|
|
blend two, stack the third over those ... call it a "stend" or a "black". copyright the idea. make millions. 
_________________
walk gently. leave a good impression.
|
|
| Back to top |
|
|
Boilermaker
Joined: 21 May 2006
Posts: 527
|
| Posted: Thu Apr 07, 2011 3:14 pm Post subject: |
|
|
| Quote: | | blend two, stack the third over those ... call it a "stend" or a "black". |
Better yet - get a fourth projector for a double "Stend" and run it with 3D and call it a "4X3Stend"
Bob
|
|
| Back to top |
|
|
zaphod
Joined: 16 Jun 2006
Posts: 2002
Location: Cloverdale
|
| Posted: Thu Apr 07, 2011 7:01 pm Post subject: |
|
|
| Boilermaker wrote: | | Quote: | | blend two, stack the third over those ... call it a "stend" or a "black". |
Better yet - get a fourth projector for a double "Stend" and run it with 3D and call it a "4X3Stend"
Bob |
that would be the "X-stend"
_________________
walk gently. leave a good impression.
|
|
| Back to top |
|
|
Boilermaker
Joined: 21 May 2006
Posts: 527
|
| Posted: Thu Apr 07, 2011 7:18 pm Post subject: |
|
|
| Quote: | | that would be the "X-stend" |
Would that turn an 8" into a 9" and make it last longer???
|
|
| Back to top |
|
|
overclkr
Joined: 08 Mar 2006
Posts: 4227
|
|
| Back to top |
|
|
mc86
Joined: 20 Sep 2008
Posts: 767
Location: pittsburgh, pa
TV/Projector: ECP 4500 (Vidikron box), ECP4500+, wanting 07MS/07MTS, evaluating pc soft-blend
|
| Posted: Thu Apr 07, 2011 9:55 pm Post subject: |
|
|
One of the best things about this forum is that I can show videos like Cliff's above to my wife...while that doesn't buy me any real political capital, it does remind her that my fledgling abilities could always be "worse"... 
I'd have to say that having a machine as a backup would have to be the best route if you plan on keeping the setup for a while...I've got 150 hours on my ECP (very few more on the chassis, just old) and I'm already getting some funky start-up behaviors and other intermittent stuff going on. Will be nice to just swap inexpensive replacement parts on-hand (IF those work!).
Matt
|
|
| Back to top |
|
|
ChrisWiggles
Opinionated SOB
Joined: 12 Mar 2006
Posts: 2529
Location: Seattle
|
| Posted: Fri Apr 08, 2011 12:26 am Post subject: |
|
|
| Boilermaker wrote: | | Quote: | | !? How does bandwidth affect ANSI contrast!? That's going to be determined by light scatter (especially since his 808s aren't LC), not the input signal... |
If I'm not mistaken, ANSI contrast is defined as the ratio from one pixel at 100IRE to the next pixel at 0IRE. Digitals have an easy time with this which is why they appear sharper.
With the analog equivalent, this signal would go from full voltage (.7 volts) to zero volts at the rate of the video signal to emulate the pixels. The bandwidth capability of the video curcuitry determines how well the signal actually follows this ideal.
Obviously it doesn't quite get all the way to .7volts and it doesn't get back down to 0 in the limited time, therefore the full white pixel isn't quite white and the black pixel isn't quite black.
Supposedly, if a specification of bandwidth is listed as XMHz, then it is actually 3db down at that frequency, which is only 1/2 the voltage.
Bob |
No, ANSI Contrast is simultaneous contrast averaged over 4x4 checkerboards.
Measurements of one-pixel ON and one pixel OFF is entirely different, you're most likely thinking of MTF (which is a little more complicated than that, but is part of what MTF measurements capture), not ANSI Contrast.
|
|
| Back to top |
|
|
km987654
Joined: 25 Jul 2007
Posts: 2874
Location: Australia
TV/Projector: Barco BG809s
|
| Posted: Fri Apr 08, 2011 12:31 am Post subject: |
|
|
| ChrisWiggles wrote: | | Boilermaker wrote: | | Quote: | | !? How does bandwidth affect ANSI contrast!? That's going to be determined by light scatter (especially since his 808s aren't LC), not the input signal... |
If I'm not mistaken, ANSI contrast is defined as the ratio from one pixel at 100IRE to the next pixel at 0IRE. Digitals have an easy time with this which is why they appear sharper.
With the analog equivalent, this signal would go from full voltage (.7 volts) to zero volts at the rate of the video signal to emulate the pixels. The bandwidth capability of the video curcuitry determines how well the signal actually follows this ideal.
Obviously it doesn't quite get all the way to .7volts and it doesn't get back down to 0 in the limited time, therefore the full white pixel isn't quite white and the black pixel isn't quite black.
Supposedly, if a specification of bandwidth is listed as XMHz, then it is actually 3db down at that frequency, which is only 1/2 the voltage.
Bob |
No, ANSI Contrast is simultaneous contrast averaged over 4x4 checkerboards.
Measurements of one-pixel ON and one pixel OFF is entirely different, you're most likely thinking of MTF (which is a little more complicated than that, but is part of what MTF measurements capture), not ANSI Contrast. |
I understand that ANSI contrast is a measurment between two adjacent sections of screen one with light and one without. I don't see that it matters if its one pixel or a group of pixels when it comes down to it the end result will be the same and CRTs are not good at this because light easily bleeds from one area of light to an area of no light.
|
|
| Back to top |
|
|
ChrisWiggles
Opinionated SOB
Joined: 12 Mar 2006
Posts: 2529
Location: Seattle
|
| Posted: Fri Apr 08, 2011 6:15 am Post subject: |
|
|
| km987654 wrote: | | ChrisWiggles wrote: | | Boilermaker wrote: | | Quote: | | !? How does bandwidth affect ANSI contrast!? That's going to be determined by light scatter (especially since his 808s aren't LC), not the input signal... |
If I'm not mistaken, ANSI contrast is defined as the ratio from one pixel at 100IRE to the next pixel at 0IRE. Digitals have an easy time with this which is why they appear sharper.
With the analog equivalent, this signal would go from full voltage (.7 volts) to zero volts at the rate of the video signal to emulate the pixels. The bandwidth capability of the video curcuitry determines how well the signal actually follows this ideal.
Obviously it doesn't quite get all the way to .7volts and it doesn't get back down to 0 in the limited time, therefore the full white pixel isn't quite white and the black pixel isn't quite black.
Supposedly, if a specification of bandwidth is listed as XMHz, then it is actually 3db down at that frequency, which is only 1/2 the voltage.
Bob |
No, ANSI Contrast is simultaneous contrast averaged over 4x4 checkerboards.
Measurements of one-pixel ON and one pixel OFF is entirely different, you're most likely thinking of MTF (which is a little more complicated than that, but is part of what MTF measurements capture), not ANSI Contrast. |
I understand that ANSI contrast is a measurment between two adjacent sections of screen one with light and one without. I don't see that it matters if its one pixel or a group of pixels when it comes down to it the end result will be the same and CRTs are not good at this because light easily bleeds from one area of light to an area of no light. |
Oh it makes a massive difference.
When you're measuring checkerboards, you will still get pretty decent contrast ratios. When you're measuring down to the per 'pixel' level, there is so much more local spill happening, beam spillage, lens resolution, focusing etc. You're really measuring resolution, not contrast ratios.
They are radically different measurements.
As far as ANSI contrast per se, ANSI Contrast is actually a very tightly defined measurement method on a specific pattern. Most people (like myself) use the term loosely to refer to any 4x4 checkerboard-type measurement. ANSI CR is not just any random simultaneous CR.
|
|
| Back to top |
|
|
Boilermaker
Joined: 21 May 2006
Posts: 527
|
| Posted: Fri Apr 08, 2011 9:49 am Post subject: |
|
|
| Quote: | No, ANSI Contrast is simultaneous contrast averaged over 4x4 checkerboards.
Measurements of one-pixel ON and one pixel OFF is entirely different, you're most likely thinking of MTF (which is a little more complicated than that, but is part of what MTF measurements capture), not ANSI Contrast. |
Chris - Thanks for the clarification and correction.
My intention is to explain the fundamental advantages of a blend over a conventional setup and specifically how it effects the apparent sharpness of the picture.
There are obviously two factors that effect contrast, however measured: Electrical performance in the video chain and physical/optical performance.
Electrical - In a conventional setup running 1080P@72Hz, the pixel clock frequency is 211.750Mhz. In a blend setup at 2.4AR with the same effective resolution the frequency goes down to 85.240Mhz. The difference in the projectors ability to properly excite the pixel to pixel at the proper level is quite large.
Physical/Optical - In comparison to a conventional setup, each pixel excited by the beam is almost twice the distance, both horizontally and vertically, from each other. Whatever anomallies occur both with the phosphor excitation and the light path to the screen has to be substantially reduced due to the "spreading out" of these "pixels" on the raster.
This is how I explain the difference in "sharpness" that I see in my blend. It is not subtle!
Bob
|
|
| Back to top |
|
|
Tim in Phoenix
Joined: 21 Oct 2006
Posts: 4409
Location: Phoenix
|
| Posted: Fri Apr 08, 2011 1:31 pm Post subject: |
|
|
Let us not forget that blending lets you light up a lot more phosphor, 80 or 90 percent of the tube face. Displaying 2.35 on one projector only lights up 56% of the available phosphor! You would not run a V12 engine on seven cylinders would you?
.
|
|
| Back to top |
|
|
Spanky Ham
Joined: 22 Mar 2006
Posts: 5643
Location: Comedy Central
|
| Posted: Fri Apr 08, 2011 3:02 pm Post subject: |
|
|
| Boilermaker wrote: |
There are obviously two factors that effect contrast, however measured: Electrical performance in the video chain and physical/optical performance.
Electrical - In a conventional setup running 1080P@72Hz, the pixel clock frequency is 211.750Mhz. In a blend setup at 2.4AR with the same effective resolution the frequency goes down to 85.240Mhz. The difference in the projectors ability to properly excite the pixel to pixel at the proper level is quite large.
Bob |
Could you explain the formula you used to determine the pixel clock?
|
|
| Back to top |
|
|
noos@xp37+
Joined: 17 Jun 2008
Posts: 464
Location: Berlin/Munich
|
| Posted: Fri Apr 08, 2011 3:16 pm Post subject: |
|
|
Hello,
try this before you change any hardware: http://nthusim.com/
Use the multi projector trial. Full geometric control plus new edge blending. Just give the free trial a try....
Marc
|
|
| Back to top |
|
|
Boilermaker
Joined: 21 May 2006
Posts: 527
|
| Posted: Fri Apr 08, 2011 4:23 pm Post subject: |
|
|
| Quote: | | Could you explain the formula you used to determine the pixel clock? |
I cheated! TvOne has a real nice little calculator that does it for you. I tried to send it, but could not do it. Just go to tvone's website and look for the Corio2 resolution creator software.
The nice thing is that it calculates the proper timing values and porches so that the figures you get are actual values. Let me know if you have trouble finding it!
Bob
|
|
| Back to top |
|
|
|
|
