Originally Posted by Dominic Chan
What’s “reasonable” depends on familiarity and feasibility. One could say it’s reasonable to expect the picture brightness spec to apply to the whole screen, but in actual fact the corners are always less bright, by 30, 40 or 50%, and no one would claim that’s a defect.
Judging by the recent posts in the JVC forums, most people seem to have accepted that it’s “reasonable” to have the paper test showing brighter corners.
It seems less "reasonable" if you already had a £6K projector devoid of brighter corners, only to have it replaced with one with them...
. The difference is not slight between the units. And while many folk understand lens vinetting etc - those tend to be much more constant, design specific matters which do not vary to such a degree from unit to unit. We're talking about 5-6x less native contrast at that corner area than in the lens centre at worst case iris settings.
Of course my "good" unit has an erratic lens iris, but I doubt very much that is much implicated as no matter the iris setting on the "not so good" unit it doesn't make any useful improvement to the contrast of those areas.
At the end of the day once the screen goes up in my room I'll be able to see it in all it's glory. It looks like the contrast in at least one corner is around 20,000:1 regardless of iris; if I set the iris to get my 56cd/m2 this means the black level I end up seeing at the edges (they're clearly not just corners) will probably be something like 0.004-0.006nits instead of <0.001 nits in the centre of the screen. Most of the time it won't be visible due to eye bias, but it will be visible to the naked eye easily. I'll be interested to see what starfields make of it. Probably the seemingly better convergence will make more difference to the image quality for me than the black field uniformity.
I'm not aware that I'd seen anyone measure the contrast in those areas before; but it does sound like we're setting ourselves up as apologists for poorer than expected contrast before the new 4K units have even arrived...
. I mean, how far do poor contrast areas have to extend into the frame before you start questioning what is the contrast ratio really achieved? And I really do think that we should start talking about what is a reasonable way to talk about contrast on product.
Interesting thoughts and questions which could only really be discussed in the presence of either user data or an open discussion policy by JVC on the details of the phenomenon:
Are the 4xx units likely to have better black field uniformity due to less pushing of the envelope for contrast?
Are the 6xx units likely to have better black field uniformity due to hand picked parts (maybe I got panels that could have gone in a 6xx in my first unit, but they didn't need them so they went in a 5xx).
How likely am I to have corners with these issues vs not? (I've got a 50% hit rate at the moment).
I'm a bit disappointed that the attitude is generally "it is what it is". For my own part looking at this with my "production test" hat on - I think clearly I've demonstrated there are good (at least 1) and not-so-good units out there. It's completely measurable. It would be possible to devise simple tests (took me minutes) to measure these optical blocks for the phenomenon.
The testing is either being done at the factory already (and these units are passes - in which case we could be told what the test limits are); or the units passed some stricter criteria in the factory and something has happened to them since (misalignment due to shipping /packaging, rapid degeneration, insert random theory here); or they simply aren't tested in this way. In all cases it should be possible to have an open understanding of what exactly it is you are buying into at 130,000:1.