CurtPalme.com Home Theater Forum

[Person99]

Well, I was bitten by the constant image height bug (since I am sick of the tiny letterbox 2.35:1 movies—they are supposed to be bigger not smaller god damn it!), but since I watch too much 1.85 and 1.78 content, I did not want to do a 2.35:1 screen because 1.78 suffers more than I would like and it would be too dim (not to mention the PJ would be so far back it would render my elevated 2nd row unusable). (A 2.35:1 screen of decent size with a single CRT is too dim—I’ve noticed in the craze for CIH, many people [both the digital and few CRT guys] that are doing this tend have pathetically short 2.35:1 screens because their room/projector can’t really handle a 2.35:1 screen of the “right size”, so their 16:9 content is not bigger than an RPTV—no wonder they hate watching anything but 2.35:1 content!)

So, I started looking at (gasp) bright digitals with anamorphic lenses. Although they have come a long way and would make anyone who has not seen a well set up top end 8” LC or 9” machine very happy, I have a problem. I have seen those, in fact, I own one of those! Although the JVC RS1 is right now the only digital that really competes with top CRTs for image quality, it has issues including cartoon-ish colors and it just doesn’t have that “magical” film-like look of a CRT.

Further, the anamorphic lens is not all it is cracked up to be. Since digitals don’t have memory blocks, the 2.35:1 image is noticeably dimmer than the 16:9 image and there is no easy way to recall the settings for different ARs on many dPJs. Further, most of the lenses introduce changes in color temp (so 2.35 with the lens will have a different color temp then 16:9 without the lens—or you can leave the lens in place and reduce the resolution of you < 2.35:1 content—yuk). So, to compensate for these two issues, you’ll need to use a video processor to set up the different memories and hope you have one with enough inputs and memories for all your sources. Also, they all have chromatic aberrations that manifest as color fringing at the edges of the screen--I like perfect convergence damn it!

I toyed with the idea of anamorphic lenses for CRTs and played around with that a bit, but couldn’t get it to work because in a 2 prism lens, for the second prism to be large enough, there is not enough room due to the angle of the R and B guns (their 2nd prism is bumping into the green’s second prism). Their may be a creative way to get this to work and I may play around a bit more before giving up, but there will still be the downsides of anamorphic lenses if I get this to work. Besides, making 3 sets of coated lenses is not going to be cheap (around $1800).

So, in the quest for a wider 2.35:1 image, I started playing around with the idea of a pseudo-constant area set up based around a 2.05:1 screen. I’ve done some testing and can even accomplish this pretty well without a processor. So, I have to wonder why the hell more people are not doing this? I’d love a discussion about this. Let me outline where I’m at and the negatives I see. There have been advocates of constant area for years (for instance Bjoern http://www.videophile.info/Screen/Page_01.htm and http://archive.avsforum.com/avs-vb/showthread.php?s=&threadid=15649), but I didn’t really want a “true constant area”. I think the 2.05:1 may be a better option.

The CRT PJ is really a marvelously flexible display device. Kind of sad that they have not been able to capture even a fraction of this flexibility in digitals, but its flexibility will allow this type of set up with no problem.

Let’s start with my goals:
1) A 2.35:1 image that is noticeably wider than a 1.85/1.78:1.
2) No loss of quality or resolution on 1.78:1 content.
3) A 1.78:1 Image that is the same or almost the same size as I currently have.

The equipment:
1 Zenith Pro 1200X 8” LC Projector (2001 build—basically the latest and greatest Barco Cine series)
2 Toshiba HD-A1
3 Motorola QIP 6416 Cable Box (Verizon FiOS service)
4 JVC 30K D-VHS machine (DTheater and MPEG2 decoding)
5 Mitsu 2K D-VHS machine (recorded HD)
6 Crescendo Systems RTC2200 (aka kimcoder)
7 HDFury (connected to kimcoder)
8 Extron 202Rxi (very late build)

Note: thanks to the flexibility of the CRT, I’m doing this with no video processor.

Current screen:
1) 52”x92” 1.2 gain (they call it 1.3 but it is 1.2) Dalite Cinema Vision (not as nice as the Stewart but for the price has a nice “glow” to it).
2) Projected Image sizes:
a) 1.33:1 – 52”x69” (VERY rarely used)
b) 1.78:1 – 52”x92”
c) 1.85:1 – 49.75”x92”
d) 2.35:1 – 39”x92”

Target Screen:
1) 50”x102.5” Wilsonart (probably but might do Hurley or Dalite material), maybe curved???
2) Projected Image Sizes:
a) 1.33:1 – 50”x66.5”
b) 1.78:1 – 50”x89”
c) 1.85:1 – 50”x92.5”
d) 2.35:1 – 43.5” x 102.5”

Notice that my 1.78:1 is barely smaller and my 1.85:1 is actually a pinch larger. Further, the 2.35:1 is getting closer to a respectable size gaining 10.5” in width and 4.5” in height over the current constant width set up.

The masking system is also relatively straightforward as one of the masks (top/bottom or left/right) is always open (at the edge of the screen). Forgetting about 1.33:1 masking because I almost never use that AR, there is only 2 positions for the top/bottom mask and 3 for the side masks (I’m not worrying about “odd ball” ARs at the moment):
1) 1.78:1
a. Top/Bottom – position 0 (edge of screen)
b. Right/Left – position 2 (narrowest)
2) 1.85:1
a. Top/Bottom – position 0 (edge of screen)
b. Right/Left – position 1
3) 2.35:1
a. Top/Bottom – position 1 (narrowest)
b. Right/Left – position 0

Given this, the masks do not have to be fancy and never are stacked (one on top of the other).

So, now to the most important part: how does this work out on the tube face? Not too bad actually. If you do the math, you will find that assuming my active image width stays the same as it is now (at a maximized 16:9 shape), the new set up would put 2.35:1 using that much phosphor width, but 1.78:1 using only 87% of that width. But, the tube sides are not straight, they curve, so a mazimized 2.35:1 image area can actually use a very small amount more of the tube width putting my 16:9 image (for the 2.05:1 screen) at closer to 90% of its current width. So, how did it do with HD content at 90% width—very well, from 1.45x projected image width, I could tell no difference in quality (including fine details like hair) between the current raster width and the image produced using 90% of the width. Indeed, 1 on 1 off patterns showed I was still seeing it all.

Ah, but the height of the active image area is also reduced to 90% of its former size—what about that? Again, no problem at 1080i. It *may* have slightly reduced scan line visibility, but again, no visible loss of detail from the front row. Frankly, even at 90% of its former size, it is actually close to the size it is if you follow the set up manual (and bigger in size than if you follow the NEC set up manual).

So, what are the downsides I see to this approach?
1) The 2.35:1 project image has about 23% more area than the current 2.35:1 image (because 102.5” is pretty darn wide for an 8” machine). That means the contrast will be have to be bumped up 23% for the same brightness. Well, a 23% bump is contrast up to 77—yipes. So maybe split the difference and run a 70 for 2.35:1.
2) The higher contrast of 2.35:1 images will cause faster wear—faster than the other ARs will wear, so I could have brighter parts of the image in a few inches at the top and bottom of all other ARs.
3) Even 16:9 will have to be driven about 11% harder than currently because it is only 90% of the current size on the tube face.
4) A curved screen would be best since with the lenses on my PJ, at 102.5” wide, the focus on the edge is about 5” out from the flat screen plane. How would I build a curved top/bottom mask because a mask that is 5” in front of the screen in the center might look odd?


So, my biggest concern comes down to having to drive the tubes harder. I like a bright image and my current one is bright enough for me, but I’m near the top of my comfort range with how hard I’m driving them. This set up puts me out of that range. If tubes were $200 each, I could deal with it, but at $2000 to retube the baby, I want them to last longer than 1000 hours! This is really the main thing stopping me from charging head long in.

Anyway, thoughts, opinions, etc?

[CZ Eddie]

Dave,

First let me comment that this was a well thought out post you've written above. A bit long to read for someone like myself who has the attention span of a chichuaua.

We are actually looking at the same setup, you and I. My screen will be 1.3 gain and about 105" wide. And a 2.05:1 as well.

I would prefer a 2.35 or even 2.40 screen but it's simply impractical IMO. A couple of years ago I ran my Marquee 9000 on a 118" wide 2.35 screen at ~0.8-0.9 gain. It was tons of fun of course (though not a videophile's dream of quality). But in only 1000 hours I already was seeing the slightest of 2.35 raster wear on my tube faces. Not bad at all of course. But noticeably quicker wear than if I'd run a 16:9 screen.

This means that at 1/2 life, you are not going to get anyone to purchase those tubes if you sell them used. So that is one cost to factor in. The person I sold my 9000 to commented on this and said it was "OK", but under certain conditions he would notice the wear onscreen. That was the 2nd PJ I'd ever sold and I've been way upfront about tube wear to potential buyers since then.

So yes, the tubes will wear more quickly with 2.35. But it is not a drastic difference at all!

With that said, I too would prefer a medium. So 2.05:1 is my goal for the theater I'm building now. I *must* have my 2.35 material be wider than anything else. Or it's psychologically "not as cool as 16:9".

With 2.05:1 your 2.35 is wider than 16:9, and yet your 16:9 is taller than 2.35. So it's the best of all worlds! Both screen setups will have something better than the other.

But the one thing I have not looked into yet, and will eventually be forced to research... is an easy to build, inexpensive solution to masking. Because you don't spend hours & hours of setup and spend tons of money on a great h/t and not get any masking for it! Masking is so important to the final image quality!

But... how to do it? I have no idea. Because I would want to be able to do it with a remote control.

If I have to get up out of my chair to adjust it everytime, then I could simply use velcro to make things simple.

[Person99]

Eddie, thanks for the comments. I have a mechanical engineer friend of mine looking into a DIY masking system that is affordable. I'll let you know what I find.

A 9" on a 118" wide 2.35 is going to be about the same phosphor wear as my 8" at 102.5" wide, so yep, you confirmed my fear about tube wear. Decisions, decisions.

A true CA freak would find one fault with the 2.05:1 based screen. Although 16:9 and 2.35:1 content is about the same size (area), 1.85:1 will actually have a bit more area.

I was not going to try to do much of a sophisticated masking system as you see from my post. I was not going to try to support 1.66:1 or 1.33:1 which helps make the masking system simpler since there is no overlap. If that is what you are also planning, maybe we can share notes (when I get some).

[kal]

Curved would be definitely the way to go if you're worried about tube wear (assuming you don't have too many people sitting too far off axis, or don't care where your friends sit as long as you have the best seat in the house).

I doubt you'd like the hotspotting of going any higher gain than 1.2/1.3 on a flat screen.

It's a good idea Dave.

If nothing else, you've actually depressed me a bit as I didn't realize some of the pitfalls of CIH on digitals. *sigh*. I really like the idea of CIH too. There's something so wrong with how 2.35:1 movies look on my setup vs. 1.78:1 movies (since my screen's 1.78:1 like yours). I don't have room to go any wider though so unless I change houses, I'm not going to be able to do what you're doing. There's just no room (and my L/R speakers are just too high).

Kal

[kal]

Curved would be definitely the way to go if you're worried about tube wear (assuming you don't have too many people sitting too far off axis, or don't care where your friends sit as long as you have the best seat in the house).

I doubt you'd like the hotspotting of going any higher gain than 1.2/1.3 on a flat screen.

It's a good idea Dave.

If nothing else, you've actually depressed me a bit as I didn't realize some of the pitfalls of CIH on digitals. *sigh*. I really like the idea of CIH too. There's something so wrong with how 2.35:1 movies look on my setup vs. 1.78:1 movies (since my screen's 1.78:1 like yours). I don't have room to go any wider though so unless I change houses, I'm not going to be able to do what you're doing. There's just no room (and my L/R speakers are just too high).

Kal

[CZ Eddie]

Kal, you could always go with acoustical screen.

Dave, like I said.. yes it does wear noticeably quicker, but not drastically! I figure I got the same wear in 1000 hours as a 16:9 would have given in 1300 or 1500.

I'm not interested in masking for anything other than 2.35 and 16:9. My Lumagen will fit everything else within those two. A tiny bit of unmasked screen is not a big deal. I just have issues with say, watching 2.35 on a 16:9 screen. The unmasked part is just too big IMO. As it would be watching 16:9 on a 2.35 screen.

2.05:1 makes things easier but I have a feeling I'll still want some masking! At the very least, for when I'm watching 2.35 movies. 16:9 and 1.85 just aren't as critical as often.

[kal]

[kal]

[Person99]

[Person99]

[VideoGrabber]

Dave,

this is a very interesting topic, and a well-written analysis of some of the options and impact thereof. I pondered for a while how I wanted to respond to this. My first thought was to start with your question...

> So, I have to wonder why the hell more people are not doing this? <

One reason might be that whenever it's been suggested, the experts completely rip it to shreds. Some examples from "Anyone running a 2.05:1 screen?":

> for a CRT PJ, there are only two aspect ratios that make sense, 4:3 and 16:9. Anything else makes no sense at all. <

That's pretty unambiguous. and...

> big issue--image quality... When you try to run the 1920 on a smaller than full width raster (as you will if your screen is an AR bigger than 16:9), you will loose horizontal resolution. Also, trying to fit the 720 or 1080 lines in the small vertical height of a 2.35:1 shape raster can't be done for most PJs, so again, you loose resolution here. <

and...

> I'm amazed that already there are 7 people running ARs that significantly reduce the quality of their image <

I could go on, with more examples of the same, but you get the point. People get discouraged when they're lambasted for trying to find ways to get "outside the box" of a CW setup mandated exclusively by phosphor considerations on 4:3-ish (or 5:4-ish) CRT tubes. I'll avoid pointing out who those comments came from. Don't want to embarrass anybody.

But I decided to take a different, more positive approach instead.

Rats! I'm out of time. Gotta go. I'll have to follow up on this later.

[Person99]

I want to try this on a 180 tube machine also. From my experiences, I'm pretty sure that you won't get the full 1920 in 89-90% phosphor width with one of those. I'm thinking that more people don't do it and I was opposed to anything not using the full phosphor because of all the 180 tube machines out there. Assuming my hypothesis that you need a P16 tube 8" to do this well on an 8" machine means just a few 8" machines.

Also, there is still the tube wear considerations and as Eddie point out, they are very real. He had noticible wear within a timeframe that I've never seen wear on a 16:9 set up.

But you are correct, I lumped the 2:05:1 screen in with 2.35:1 screens which do significantly compromise the 16:9 image. Should have done more calculations on the 2:05:1 before. Ooops.

[Jesse S]

Gary Murrell is doing 1920x800p or so for 2.35:1 on an XG. 1920x800 should be just as difficult for the tubes to resolve as 1920x1080, the only difference being lower bandwidth usage. How well can an XG (P16's) resolve these resolutions?

[jask]

quote
But... how to do it? I have no idea. Because I would want to be able to do it with a remote control.
quote
I keep thinking that the answer is to use a couple of stepper motors and and an X/Y drive board that could be controlled off the HTPC but doing the actual work and building it....number 128 on my project list (and that does not include the Honey-do list!!)

[Person99]

[Person99]

[madpoet]

Very interesting thread, but you COMPLETELY lost me on this one:

"Further, the anamorphic lens is not all it is cracked up to be. Since digitals don’t have memory blocks, the 2.35:1 image is noticeably dimmer than the 16:9 image and there is no easy way to recall the settings for different ARs on many dPJs."

The purpose of the anamorphic lens is to allow you to use the full panel of a 16x9 digital projector on a 2.35:1 resolution. I don't understand where you are getting the noticeably dimmer from. Most anamorphic lenses add significantly to the brightness because of the full panel usage.

[Person99]

[CPJr]

In this day of discount-price CRT PJs (<g>), has anyone considered or tried using two identical PJs, one further away from the screen?

I've been thinking about this issue lately as I too don't like the idea that 2.35:1 movies are smaller than 1.78:1 on my 90" screen.

Something I would not want, at least for HDTV movies, is a digital projector which rescales letterbox format movies to a higher number of scanlines.

Ideally, we nead a hi-def disc format which uses all 1080 lines for ultra-wide movies.

Charles

[Person99]