KBK said: I'm not a fan of the filters because they ruin or damage a ANSI CR that is already cutting close to the limit. "Just a little bit more flatness' is a large price to pay for the small increase in other areas, in my opinion. One may play with the filters, but one may also find after much experimentation, that the image looks better without.
You're not necessarily correct there, KBK. To my mind, the whole point of using the filter is to increase contrast.
Here's why: the no-signal black (RGB all = "0") is the darkest "black" the projector can produce. Yet, there is a haze over it, consisting of mostly excess B and G. This haze does nothing for detail. All it serves to accomplish is to reduce contrast, by making no-signal blacks not quite "black". The idea of the filter (Light Salmon is magentaish-yellowish) is to provide a basis for cutting back on the B and G haze in the blacks, but not to interfere with the R part of the "black" signal. Hence, after filtration, no-signal blacks should have:
(a) less haze (and be thus closer to achromatic, "no-color" black) and
(b) more "blackness" because the B and G haze elements have been reduced.
This leaves the problem of the mids and highlights, which take on a decidedly "pinkish" appearance. I use the individual color contrast and brightness controls (in the ADVANCE menu) to remove the pink by reducing R, rather than increasing B and G.
If you increase any of the brightness sliders above "0" then this will add brightness to all levels, including the no-signal black level. Conversely, reducing any brightness level cannot darken the no-signal black, since no-signal black is, by definition, at its blackest already. Reducing brightness sliders will, however, reduce the particular color in the mids and highlights, which is what I needed to do to remove the pink cast.
In effect, using the filter outside the internal optical system of the projector (i.e. screwed onto the front of the lens after the haze has been added internally) "anchors" the no-signal black at its blackest and most achromatic state, allowing you to "rubber-band" the brightness curves of individual colors to remove color cast from the mids and highlights.
Potential for dimming of the picture is as follows:
(a) the filter itself removes about 2/3rds of a stop of light.
(b) reflection by the filter back into the light path removes more.
(c) reflection from the rear face of the clear glass filter removes more.
All in all, my estimate is that about a whole stop is lost by my process. The light is halved. In order to get back some of this lost light I use the projector in high-bulb mode, coupled with DYNAMIC picture mode. "Taming" DYNAMIC mode is an exacting process but, as I proposed above, it can be done.
The resulting pictures keep all but the very brightest "near white" highlights in the most contrasty movies. Reducing the CONTRAST slider can very often reduce the (already minimal) loss of these highlights. A good telecine transfer leaves some headroom in the whites and many movies do not require this extra muting of CONTRAST. Likewise, the MID gamma slider can be reduced back to +1 or even 0, depending (once again) on the quality and nature of the transfer (using a projector amplifies flaws and enhancements in transfers much more than a relatively small CRT tube will ever do).
By contrast (forgive the pun), the three CINEMA modes and NATURAL look awfully muddy and dim. NORMAL and VIDEO come closer, but don't quite have the punch that DYNAMIC has. Their colors may be OK, but they're just that little bit too dim (quite dim for the CINEMA and NATURAL modes) for the large size of image I prefer (about 110 inches across), especially with Cinemascope format titles which require even more enlargement.
On Telecine Transfers Generally
I know that, in theory, telecine transfers destined for DVD are supposed to be the highest quality possible. Certainly, if they are originated in high definition (as most recent movies have been) and then interpolated down to PAL and NTSC resolution, they would probably have been created as flat and unenhanced as possible. But I suspect that somewhere in the interpolation process instructions are given from the producers or wonks in the movie company management chain to "add a bit of sparkle" to the lower resolution images by means of extra unsharp masking, more contrast and brighter colors in particular. DVD transfers are destined for consumption by the masses, most of whom (despite our little perfectionist cadre here) have relatively small, traditional CRT screens to watch their movies on. They take little pride in adjusting their set-ups to achieve 6500K nirvana. They have the contrast wound out too far and colors too high. They like bright pictures and plenty of them. So the movie companies provide what their public wants.... not in all cases, and not as much nowadays, but regularly enough.
Result? A nice, sharp, punchy image on a 76cm CRT screen. But when the same title is projected 110 inches across the long axis, with a wider, closer viewing angle (because projector aficianados like BIG cinema-type screen dimensions) the artificial sharpening halos around light-to-dark transitions are about an inch wide and stick out like a sore thumb. MPEG artefacts in static scenes sparkle... and distract. Detailless shadow detail, only part of an eyeful on a TV ten feet away, becomes the whole eyeful when you're looking at a projected image five or six times the width of your TV image... and then you notice for the first time that there's just no meaningful information there under this actor's chin or that actor's eyelids... at all.
So we tweak to try and defeat the "helpful enhancements" that the movie companies have so thoughtfully provided us. We reduce sharpness, contrast and color (the COLOR slider reduction is about the only tweak universally agreed upon by tweak advisers here).
I'm not certain that being able to present a movie exactly as it was projected in a cinema, from film, isn't a Holy Grail that may never be found until DVD authoring companies and production designers realise that their "enhancements" show up poorly on the increasing numbers of big screens - projection and plasma - that are being introduced into domestic homes. I guess what I'm saying is that it's all very well and good to make your projector conform to a test signal, but it's a pity that DVD companies don't seem to have used the same test signal (or have wilfully ignored it) when they author their titles.
A well scanned, 35mm film-originated, high bit-rate transfer is truly a joy to behold on the AE-700, almost indistinguishable from what we remember the actual movie looked like at the cinema complex when we saw it a year or so ago (even more convincing if we never saw the cinema version). Suspension of disbelief and immersion in the program material (rather than VB and white-flash complaints) overcomes any residual resistance to the fact that we're watching an image - the product of an imaging system designed in the 1950s before even video tape existed - that was never intended to be blown-up to more than about 24 inches across. That suspension of disbelief is as good as it will get. Unfortunately, much of the enhancements that we are dished-up by the movie corporations, seeking to serve the lowest common denominator, serve to diminish this ideal state of viewer appreciation.
I've found my own tweaks and can finally rest easy with almost any DVD transfer I care to throw up on the screen without having to endlessly adjust the display parameters to suit the particular, individual program. A tweak here, a tweak there if I'm feeling picky, but not much more than that.
That, to me, is bliss. The Holy Grail of perfection will have to wait for another day.