Megapixel vs. Zoom

"Randy Berbaum" <rberbaum RemoveThis @bluestem.prairienet.org> wrote in message
news:ejrfbs$kv1$1@wildfire.prairienet.org...
> Neil Harrington <not RemoveThis @home.today> wrote:
>
> : "Mueen Nawaz" <m.nawaz RemoveThis @ieee.org> wrote in message
>
> : > I meant if you're looking at it on the screen at 100% size (1:1 zoom).
>
> : I'm afraid I still don't get your meaning, unless you mean for example
> that
> : an object 10 inches high in the real world would be 10 inches high on
> the
> : monitor. But if that's what you mean I don't see how it has any bearing
> on
> : this.
>
> I think he ie refering to a relationship of one screen pixel representing
> one pixel from the image data. When using the various PS products this is
> called a 100% zoom or "actual pixels".

As mentioned, I have seen "100% zoom" used to mean that, but that wasn't my
impression from what he wrote. I think "actual pixels" is the more
descriptive term. I don't use PS so wasn't aware that they used "100% zoom"
in that way. Perhaps that's where the usage comes from.


> The size on the screen has nothing
> to do with the size of the object in real life, or even on the resulting
> print. Such a RL to screen relationship would be nearly meaningless since
> the size of the screen and the resolution of the screen will determine the
> screen dimension of an object of a specific dimensions in pixels.

Right.


> And when
> you include the relationship of RL object to pixel dimension with a
> particular lens length and sensor resolution the number of variables can
> be daunting. And this becomes even more crazy when you throw in the
> relationship of screen size to print size. Different screen
> resolutions/zoom sizes translated through the print driver and printer to
> some highly variable choices of print dimensions compounds the problem.
> True, if you have all the variables and math it should be easy to predict
> RL to Screen to print dimension relationships, but there is very little
> need (short of a CSI investigation trying to measure the length of a
> weapon from a fuzzy photo) for most of us to worry about or care about
> such equations.
>
> If I am displaying an image at 300% so I can do some very fine editing, I
> will not expect the image to be perfectly sharp. I expect that the image
> to look as if it is made up of individual squares. On the other hand if my
> image shows strong pixelation when it is being displayed at 50% resolution
> I know there is a problem either with the display program/device or with
> the image data.

You're using these percentages in a sense consistent with "100%" meaning
nothing more or less than "all the original image pixels," correct? So your
"50% resolution" would mean what, one pixel on screen for every two pixels
*linearly* in the original image?


>
> : The other issue is how you're looking at the final image. As you
> indicate,
> : an 8x10 print is likely to look much better than the same image on a
> : computer screen. That's chiefly because you can actually get *all* the
> : pixels in the image on an 8x10 print, but you can't on a computer
> screen.
> : With a typical 17" or 19" LCD monitor, the maximum resolution is
> : 1280x1024 -- which is only 1.3 megapixels. Even a monitor that will do
> : 1600x1200 is only putting 2 Mp on the screen. So with a 6, 8 or 10Mp
> camera
> : we're going to be throwing away a lot of those pixels as long as we're
> : viewing the image on a computer screen, regardless of the camera's
> sensor
> : size, and there's not much we can do about it. That's really a monitor
> : issue, not a camera issue.
>
> True,as far as you go. But since an image can be displayed much larger
> than the size of the screen, a 10mp displayed at a 1 screen pixel=1 image
> pixel shown on a 1.3 mp screen will still work. The screen will just be a
> small window onto the much larger image that can be shifted around the
> image to reveal different parts of the image.

Sure. That is in fact exactly how I look at some of the images published on
the Web from different cameras/lenses, when the page allows such viewing.
Looking at the full image that's been made to fit the screen is practically
meaningless for purposes of comparison, so much information having been
thrown away.


> Display on a screen is not
> limited to an image that can be shown on the screen all at once. But a
> single dot on the screen will be much larger in RL physical dimensions
> that a single dot on the print. If a print was made at the same resolution
> at the same physical dimensions as the screen very few of us would count
> it acceptable. What looks good and is acceptable on a screen is nowhere
> near good enough for a print.
>
> How many of us would look at a printed image that has just over 1000 lines
> of resolution in a print 2 feet tall and call it sharp and clear (an
> equivalent of about 41 dpi). But if you get a medium sized HD Plasma
> Screen with a 2' tall picture people will be OOoo'ing and wow'ing at how
> sharp and spectacular the image is.

Sure. Mostly because of the usual difference in the viewing distance.

Neil

"Randy Berbaum" <rberbaum DeleteThis @bluestem.prairienet.org> wrote in message
news:ejrfbs$kv1$1@wildfire.prairienet.org...
> Neil Harrington <not DeleteThis @home.today> wrote:
>
> : "Mueen Nawaz" <m.nawaz DeleteThis @ieee.org> wrote in message
>
> : > I meant if you're looking at it on the screen at 100% size (1:1 zoom).
>
> : I'm afraid I still don't get your meaning, unless you mean for example
> that
> : an object 10 inches high in the real world would be 10 inches high on
> the
> : monitor. But if that's what you mean I don't see how it has any bearing
> on
> : this.
>
> I think he ie refering to a relationship of one screen pixel representing
> one pixel from the image data. When using the various PS products this is
> called a 100% zoom or "actual pixels".

As mentioned, I have seen "100% zoom" used to mean that, but that wasn't my
impression from what he wrote. I think "actual pixels" is the more
descriptive term. I don't use PS so wasn't aware that they used "100% zoom"
in that way. Perhaps that's where the usage comes from.


> The size on the screen has nothing
> to do with the size of the object in real life, or even on the resulting
> print. Such a RL to screen relationship would be nearly meaningless since
> the size of the screen and the resolution of the screen will determine the
> screen dimension of an object of a specific dimensions in pixels.

Right.


> And when
> you include the relationship of RL object to pixel dimension with a
> particular lens length and sensor resolution the number of variables can
> be daunting. And this becomes even more crazy when you throw in the
> relationship of screen size to print size. Different screen
> resolutions/zoom sizes translated through the print driver and printer to
> some highly variable choices of print dimensions compounds the problem.
> True, if you have all the variables and math it should be easy to predict
> RL to Screen to print dimension relationships, but there is very little
> need (short of a CSI investigation trying to measure the length of a
> weapon from a fuzzy photo) for most of us to worry about or care about
> such equations.
>
> If I am displaying an image at 300% so I can do some very fine editing, I
> will not expect the image to be perfectly sharp. I expect that the image
> to look as if it is made up of individual squares. On the other hand if my
> image shows strong pixelation when it is being displayed at 50% resolution
> I know there is a problem either with the display program/device or with
> the image data.

You're using these percentages in a sense consistent with "100%" meaning
nothing more or less than "all the original image pixels," correct? So your
"50% resolution" would mean what, one pixel on screen for every two pixels
*linearly* in the original image?


>
> : The other issue is how you're looking at the final image. As you
> indicate,
> : an 8x10 print is likely to look much better than the same image on a
> : computer screen. That's chiefly because you can actually get *all* the
> : pixels in the image on an 8x10 print, but you can't on a computer
> screen.
> : With a typical 17" or 19" LCD monitor, the maximum resolution is
> : 1280x1024 -- which is only 1.3 megapixels. Even a monitor that will do
> : 1600x1200 is only putting 2 Mp on the screen. So with a 6, 8 or 10Mp
> camera
> : we're going to be throwing away a lot of those pixels as long as we're
> : viewing the image on a computer screen, regardless of the camera's
> sensor
> : size, and there's not much we can do about it. That's really a monitor
> : issue, not a camera issue.
>
> True,as far as you go. But since an image can be displayed much larger
> than the size of the screen, a 10mp displayed at a 1 screen pixel=1 image
> pixel shown on a 1.3 mp screen will still work. The screen will just be a
> small window onto the much larger image that can be shifted around the
> image to reveal different parts of the image.

Sure. That is in fact exactly how I look at some of the images published on
the Web from different cameras/lenses, when the page allows such viewing.
Looking at the full image that's been made to fit the screen is practically
meaningless for purposes of comparison, so much information having been
thrown away.


> Display on a screen is not
> limited to an image that can be shown on the screen all at once. But a
> single dot on the screen will be much larger in RL physical dimensions
> that a single dot on the print. If a print was made at the same resolution
> at the same physical dimensions as the screen very few of us would count
> it acceptable. What looks good and is acceptable on a screen is nowhere
> near good enough for a print.
>
> How many of us would look at a printed image that has just over 1000 lines
> of resolution in a print 2 feet tall and call it sharp and clear (an
> equivalent of about 41 dpi). But if you get a medium sized HD Plasma
> Screen with a 2' tall picture people will be OOoo'ing and wow'ing at how
> sharp and spectacular the image is.

Sure. Mostly because of the usual difference in the viewing distance.

Neil