Originally Posted by markmon1
This is the part that's confusing me. The image is scaled not stretched, so the aspect doesnt change. If your screen is 16:9 and you scale to 16:9 res, then you reduce the image position so that the overflow for 17:9 is off the sides of the screen. I think we agree. Then if you want to go full panel you reduce the overall image so that the 17:9 image width fits the 16:9 screen. I think we agree here, too. When you display a 16:9 image on that, the top and bottom will show up as bars because the panel is not capable of displaying the full image. So some of the image is cut off. In this case 17:9 panel gives 16:9 with black bars on top and bottom because the panel cant display those.
I think we're saying the same things here. So first, for 16:9 content:
With a 16:9 screen, and using the 'Auto' aspect ratio (meaning 3840x2160 pixels): When you say "reduce the image position" if you mean you use lens zoom so that the content portion of the full 3840x2169 pixels of the chip, Aspect Ratio 16:9, fills the screen, then yes. When you do this, the unused sides of the chip (4096-3840 = 256, so 128 pixel wide bar on each side) are projecting black bars, which go into the screen frame.
With a 16:9 screen, and using the 'Zoom' aspect ratio (meaning 4096x2160 pixels), the original 3840x2160 pixels of the content gets scaled up to 4096x2304 pixels. Since the chip only has 2160 pixels, then 144 pixels (2304 - 2160 = 144) are cropped off, 72 pixels on the top, 72 pixels on the bottom. The resulting 4096x2160 output from the full chip is 17:9 (or about 1.9:1). You would have to adjust the lens zoom to fit this image to the full width of the screen, and since the screen's aspect ratio is 16:9, and the content is 17:9, there are narrow bars on the top and bottom.
Originally Posted by markmon1
Ok on 2.35:1 image when you go 16:9 scaling, the width of the screen is filled and the top is black bars for the remainder of the aspect.
If you scale to 17:9 then you reduce the image to fit your screen. The width fits and the black bars on top are the same since you are scaling not zooming.
You do gain ~100 lumens in brightness but its likely not to be noticeable.
Ok, let me work through this, with the math (and let me use 2.40:1, as that's what I've been working with for Anamorphic lenses, but it should still be close for 2.35:1):
If you have a 2.40:1 source, the actual content is 3840x1600, with the upper and lower black bars 280 pixels tall each (2160 - 1600 = 560, divided by 2 is 280). This is with the 'Auto' aspect ratio.
If you take that 3840x1600 source, and choose 'Zoom' in aspect ratio, it gets scaled up to 4096x1707 (slight rounding error). Since you still have the original 2160 pixel height, and the content now occupies 1707 pixels, you now have narrower bands (2160 - 1707 = 453, divided by 2 is about 227). The aspect ratio is unchanged, you have full use of the chip's width, and 27 more pixel rows vertically (480 - 453 = 27). The pixels you've 'lost' on the top and bottom were black anyway, so there is no content loss. And some improved light output, which actually worked out to more than I might have thought; hadn't done these calculations before:
Auto Aspect Ratio: 3840 x 1600 = 6,144,000 pixels in use
Zoom Aspect Ratio: 4096 x 1707 = 6,991,872 pixels in use
Difference is 847,872 more pixels in use. Divided that by the original 6,144,000 pixels, and you get about 13.8% more pixels active in displaying content, when choosing 'Zoom' over 'Auto'.
Well, that's a lot more than you were asking for, but once I got started, I couldn't help myself!!!