Originally Posted by rossandwendy
On the Whole Adjust I did H 1 and V -1. That was all I could do as a compromise, because almost every zone outside the center is off by different colors and amounts. So I then went into zone and had to adjust every single block, most blocks needed a minimum H -10 and V 10, and the worst blocks near edge needed a whopping H -20 and V 24.
That is a bit much on the edges. I'm assuming your pixel grid isn't well defined, go into MPC enhancement and turn off E-shift and set all other enhancement sliders to 0. Load up something on your PC (some text preferably, like the Lazy Brown Fox pattern or even just LONG sentences of text in a browser) and expand the browser to fill the projected image. If you can and have a capable receiver with (2)+ HDMI outputs, compare it to your hw45es in real-time (not by memory) and put the images side by side (zoom down all the way if needed).
Let me see if I can put this into some perspective thought here.
I have (2) JVC's here, an RS-45 that falls somewhere near exceptional, but not perfect convergence, but let's say top 20% easily of all JVC units I've seen. Then, I also have an RS-420 (not much different than your RS-540), and it is what I would call 'average' convergence.
The RS-420 is off by red by about 1/3rd to a 1/2 of a pixel (about 0.6+ pixels on red), and about 1/3rd (0.5) on blue. Now even if I use absolute max Lens shift and even shoot the projected image at a slight angle increasing chromatic abberation, the RS-420 only shows a total error difference from edge to edge of approximately 0.8 (meaning if I were to use zone correction, the bottom might need -2, top +6).
You are telling me -24 to -10, that is a variance of 14 vs. my variance of 8.
So I would say this all depends on the focus of the pixels themselves, more than convergence alone, depends. See if you can compare text and look for the actual pixel delineation (the ability to make out the square pixels that make up the text or image).