[filmscanners] Re: HP PhotsSmart - questions

["Arthur Entlich" <artistik@shaw.ca>]

Maybe I can help.

Imagine that grain in film was constant and consistent in position and
size (as in a digital camera CCD).  In fact, let's say there are 1000
"grains" across the film width on a 35mm frame, and 1500 grains in the
length (which is approximately the aspect ratio of 35mm frames).

Forgetting about the grain in the photo paper for a minute, place a film
frame in an enlarger.

Project the film image so it's dimensions are 8" x 12".  Now, count how
many "grains" across the 8" width.  There are... of course. 1000.
However, how many comprise one inch in the projected image?  The answer
is 1000 divided by 8, or 125 grains per inch.

OK, now, we are going to move the enlarger head down and refocus,
reducing the image size to 4" x 6".  Now, count the number of "grains"
across the width.  Still 1000 (although harder to count, at half the
size, eh? ;-)  Now, how many of those "grains" make up a 1" width on
that size enlargement?  Not 125 anymore... now there are 250 points of
grain per inch.  The "resolution" of the enlargement changed from 125
grains per inch, to 250 grains per inch.

In the scanner example, its the same concept, except rather than
enlarging the image from the film source, we are condensing it onto a
CCD sensor.  In a variable resolution scanner, which can refocus the
source image onto the CCD, resolution is a relationship between the
total number of discrete points on the CCD relative to the size of the
physical image source we need to record.  The only other variable is how
much of the CCD unit will be used to record the information.  If less
than the full width is used (for instance just the middle area is used,
because the projective lens cannot accurately focus otherwise, then the
resolution will drop, since less discrete sensor points will be used in
recording the source image area.

Art


Laurie Solomon wrote:

> It is very possible that I have oversimplified things greatly as you
> suggest; but not being an engineer that is the way I tend to grasp the
> basic general gist of what is being said.  :-)
>
> While I think I understand the gist of your argument along with the what
> you are showing in the two animations; I am in the unenviable position
> of not knowing enough of the technical information well enough to
> comfortably disagree with what you are saying or to comfortably agree
> with the position in its entirety.  :-)  I guess my main problem is that
> my point of reference is with scanners that do not have variable optical
> resolutions or variable focusing, which leaves me uncomfortable when
> trying to conceptualize such a scanner and the relationship between
> optical resolution and effective optical resolutions (I am not talking
> at all about interpolation or resampling upward or downward)in such a
> scanner.
>
>
>>Any scanner can downsample to achieve lower resolution than
>>optical, thats why optical resolution is so important.
>>Refocusing to capture a wider strip is NOT the same as downsampling.
>
>
> Yes and any scanner can upsample to achieve higher resolutions than
> optical(but both of these types of resampling are not hardware functions
> but software functions external to the scanner and its operation).  For
> my arguments, I was not in any way referring to any form of resampling
> when I was referring tot effective optical resolution.  What I was
> referring to was to a fixed number of pixels that the sensor can capture
> in terms of its maximum limit in which one varies the size of the strip
> that is being covered by that sensor.  This changes the effective
> optical resolution of the scanner in terms of the number of pixels per
> inch that one can achieve.  I am not sure how this relates to focus or
> depth of field which is what I think your discussion revolves around,
> although I do think it does impact on it as you suggest.
>
> As I have come to realize, we may be in agreement with respect to the
> general thesis, although the specifics are what leave me uncomfortable
> at this stage - maybe because I am having some trouble visualizing and
> comprehending the details when it comes to variable focus and resolution
> scanners.  The camera analogy does not help me; in fact it only confuses
> me. :-)  Unlike the camera one can change the both the field of coverage
> and depth of field by changing lenses from telephoto to normal to wide
> angle as well as by moving closer and further away from the subject and
> in addition to any focusing that one might do by turning the focusing
> ring on the lens which will move the lens elements in relation to each
> other or any manipulation in the lens' aperture.  I do not conceive of
> scanners as allowing for changing sensors from those which can capture x
> number of pixels within a given configuration to those which can capture
> y number of pixels in the same or a different configuration; nor do I
> understand scanners as allowing one to change apertures on either the
> lenses or the sensors (or what would be the equivalent).
>
>
>
> ----Original Message----
> From: filmscanners_owner@halftone.co.uk
> [mailto:filmscanners_owner@halftone.co.uk] On Behalf Of
> wbgilloolyjr@charter.net
> Sent: Monday, August 08, 2005 4:53 PM
> To: laurie@advancenet.net
> Subject: [filmscanners] Re: HP PhotsSmart - questions
>
>
>>You are not wrong, but you have very much oversimplified things.
>>
>>Basically, the optical resolution of a scanner is determined
>>by how many pixels the sensor (CCD,CMOS) captures and how
>>wide a strip the scanner scans.
>>
>>So, a scanner with a 5,000 element (pixel) sensor that
>>covered a strip 1" wide would be a 5,000ppi scanner
>>(5,000pixels/1"=5,000pixels/inch).
>>If the same 5,000 element sensor covered an 8.5" wide
>>(full-bed) strip, you'd have an 588ppi scanner
>>(5,000pixels/8.5"=588.25pixels/inch).
>>
>>There is, of course a lens between the artwork/flatbed and
>>the sensor and the lens-sensor assembly is very much like a
>>camera.  Just as you can move closer or farther away from a
>>subject refocusing the lens to make sure everything is sharp
>>a scanner can move the sensor-lens assembly closer (for
>>higher resolution) or farther away (to cover a wider area).
>>It must refocus to achieve maximum sharpness.  Not all
>>scanners do this (in fact pretty few), but some can.
>>
>>Any scanner can downsample to achieve lower resolution than
>>optical, thats why optical resolution is so important.
>>Refocusing to capture a wider strip is NOT the same as downsampling.
>>
>>My original description is the most accurate way of
>>describing the focusing procedure for any optical assembly (including
>>your camera). Your lens to film (or CCD/CMOS) distance that you set
>>while "focusing" is actually setting the magnification ratio.  Your
>>distance
>>from the subject sets what's in focus.  This isn't obvious
>>for large distances (greater than about 2'), you can move
>>your self then "focus" the lens.
>>But anyone that does a lot of macro photography knows that
>>moving the camera-lens assembly forward and back to achieve
>>focusing after setting the lens is much easier and more
>>straightforward.  This is why a scanner with variable optical
>>resolution sets the lens to sensor distance first (to achieve
>>magnification ratio/strip width/optical resolution - they can
>>be used interchangeably here) then moves the lens-sensor
>>assembly to achieve focus.  Scanners work at macro distances
>>as your reproduction ratio is usually about 1:1.
>>
>>Mr. Bill
>>
>>
>>Laurie Solomon wrote:
>>
>>>>If the optical resolution is variable and YES there are scanners
>>>>that can do it
>>>
>>>
>>>I will take your word for it; but according to my understanding, the
>>>optical resolution has little to do with the distance between the
>>>lens and the sensor and more to do with the size and capacity of the
>>>sensor.  The effective resolution may change with the changes in the
>>>distances between the lens and the sensor; but the actual native
>>>hardware optical resolution remains the same.  But I could be wrong
>>>in my understanding; I am not an engineer.
>>
>>--------------------------------------------------------------
>>--------------------------
>>Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
>>filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate)
>>in the
>>message title or body
>
>
>
>

----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body