>> Back to Black Level
Tip by Dean Roddey: Dean used black
velvet/felt to improve the black image on his screen. Here is the initial post he
made at digitaltheater forum.
"Ok, so I've been on a serious software binge lately, but got a little burned out
and decided to 'relax' this weekend by getting back to the TW65H80 again, to see if I
could deal with my internal reflection problem (and improve it in general as well.)
As a quick recap, for those not following the problem, I got a diffusion screen for my
Toshiba RPTV. Because this screen (which otherwise looks gorgeous) is not very
directional, it reflects quite a bit of light back into the set. This backsplashed light
causes shadown images on the bottom half of the screen when small bright objects are in
the top half on an otherwise dark scene.
So I decided to line the interior with black velvet as an experimental step, which at
worst would improve the set overall even if it did nothing for the reflection. First about
the lining adventure.
Well, I'm so sore I can barely stand up. I whacked my head on the inside top of the set
and have a huge gash on my head. I have carpet burns on my knees, and I sat down on an
X-acto knife! But, it was worth it. I've almost completely lined the interior (just a
couple spots on the top yet to get), and the contrast has gone up very noticeably. It now
is much closer to that 'direct view' contrastiness that people miss with an RPTV.
I basically used standard Elmer's wood glue, and very black, black velvet from the
local fabric store. It took about 3.5 yards from a standard... ummm... what are those
things called? Bolters or something like that? Its a lot of work at
uncomfortable angles, but its not terribly difficult. Like most such projects, by the time
you finally really work out a good system, you are about done.
It's well worth doing because as soon as you lay down that first piece, its obvious
that the velvet is at least a couple notches blacker than the interior paint on my set
anyway. By the time is completely lined, it so much blacker in there than before that its
not even funny.
Other than the standard contrast and brightness patterns from Avia, I've been using
Mystery Men as a means of gauging increased contrast and ability to hold black. And the
amount of low IRE detail I can see now is vastly increased. On patterns like the pluge
against a half white field, the amount of bleedover onto the pluge side is vastly reduced.
So anyway, do it if you have the back for it. Now on to the reflection problem...
I completely analyzed the reflection problem this time. As I covered more and stuff and
the reflections persisted, I became more and more convinced that my original theory was
correct, i.e. that it was bouncing off the mirror itself, not off some piece of exposed
metal or something. This turned out to be almost right. There were actually a couple of
smaller reflections off the screws in the lense assembly, which is why I couldn't seem to
find 'the' source previously, since there were a couple of sources, though the primary
problem is what I thought it was.
So, what is happening is that the light comes up from the guns, it hits the mirror up
towards the top and bounces off at a pretty good angle to hit the screen. The fresnel
layer straightens it up so that it comes out the front pretty much straight out. It then
bounces off the back of the diffusion screen and goes back through the fresnel layer.
However, the fresnel layer is obviously not symmetric, so the light goes back the other
way pretty much straight through. It hits the mirror again, above the point where the
original incoming beam hit, and bounces back down onto the screen lower down.
The reflections cut off exactly at the mid point because, at that point, the light path
works the other way and it starts bouncing down into the bottom of the set, i.e. it works
for me once the half way point is passed. So a bright object at the top of the screen has
a reflection just below the halfway point. A bright object just above the halfway point
has a reflection at the bottom of the screen.
In the process of figuring this out, I discovered that about the top 3 inches of the
mirror is not used, i.e. the guns never shoot up that high. At least this is true on my
set. By putting something over this part of the mirror, I found I could limit the
reflections to the bottom 1/4th of the screen. This is because the light reflecting back
into the mirror from the upper parts of the image is actually bouncing back onto the
unused part of the mirror at the top.
So, I very carefully masked off the top 2 and 3/4" of the mirror (left a little
for paranoia's sake) with velvet. This has improved the reflection by pushing it further
down, and also contributed more to the contrast by cutting down on the light bouncing
around in the top of the set.
But, that's the best I can do. I cannot get rid of this reflection totally unless I can
prevent the backsplash. I assume that the only way to do this is a one one layer in the
screen stack. Its not that it totally sucks now, since its pushed it pretty far down now.
But, on a dark film like "Mystery Men" or "Devil in a Blue Dress", it
gets on my nerves because the rest of it is looking so completely righteous at this point.
But, overall, its looking very, very good now. Probably I'm the only one who would
notice the ghost images for the most part, because I'm completely obsessed with them now
:-) But, I guess I'll have to continue forward to look into some sort of more tinting
layer or one way layer or something like that. But the effort so far has been worth it. I
have to do a little more lining in the top, then I'm going to put it back together until I
can do some more research.
So here is a question... One of the things I did was to make a 'shroud' for the lense
assembly. I took a piece of velvet, cut holes in it for the three lense caps, and draped
it loosely over the lense assembly. So it completely covers all of that exposed metal and
screws and whatnot. This made a big difference as well, and cut out a couple of the
previously mentioned ancillary reflections. Does anyone think that this would cause a heat
problem? How much is the heat dispersion of the set dependent upon heat being able to come
up into the mirror cavity? I'd think not much, since it has no venting to the outside, but
I want to make sure I'm not going to fry it by not allowing the air to freely flow up
around the lense assemblies.
So that's where it stands at this point. Things are continuing to improve, and other
than this one lingering problem in the lower quarter of the screen, it looks very, very