Digital BW, The Print

Todd Flashner wrote:

> <HUGE SNIP>
> >
> > The concept of increasing the real pixel count by doing 4 exposures (one
> with
> > each of the RGGB filters).  And limitations, in that one cannot do 'live'
> > subjects, as they move (including most landscapes) and would not ''register'
>
> > in
> > the sequential exposures.
> >
> > Harvey Ferdschneider
>
> Okay. See what I'm wondering is with 4 shot backs whether the pixel count
> really increases, or if the pixel information is just more accurate, ie
> captured not interpolated? If pixel count doesn't increase, wouldn't it also
> hold true that with the one shot backs, when used for grayscale images, the
> pixel count doesn't quarter.

For all intents and purposes are not interpolating at this point, but have a
true 5 megapixel (or whatever)  camera.  On the other hand, the manufacturers of
those '4 shot' cameras might claim 4 times the pixel coiunt, and we'd be right
back to where we started this discussion.

>
>
> IOW, pixel count (resolution) is independent from color info. You and Austin
> have maintained that a 6 megapix camera would only give you one quarter of
> that many pixels in grayscale mode (correct me if I'm wrong).

I actually assume that if you shoot in grayscale (not convert in Photoshop) you
will actually have the pixel count that the manufacturers claim (but that is a
guess on my part).  And it would depend on how each camera actually works to
capture the information.

> I am
> questioning that. I'm suggesting that that *IS* the resolution of those
> cameras, and the file sizes get larger from there as the color information
> gets added on top of that. IOW, a 6 megapix camera may yield an 18MB RGB
> TIFF. You don't get 18MB of resolution, but you do get 6 megapixels of
> resolution.

This is such a 'relative ' argument.  Define what makes 'resolution'. The camera
yields whatever file size it says it will, but the question is: Where does the
camera get that information from?  How much is real and how much is
interpolation?

There finally is no way a1 color sensor can yield a real pixel of information
that must contain 3 colors of information.

Harvey Ferdschneider
partner, SKID Photography, NYC




>
>
> So, if you go back to the top of this message, where I came in, Austin says
> for RGB you need four times as many sensors, I'm asking if that is true with
> 4 shot devices, or do the separate exposures work as, or in lieu of, extra
> sensors.
>
> Let me say that I don't know what I suggest to be so, I'm putting that
> argument/question out there somewhat generically. I need Austin to do the
> math... ;-)
>
> Todd
>


[Non-text portions of this message have been removed]