Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 

   


   


   















      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[filmscanners] Re: Density vs Dynamic range



----- Original Message -----
From: "Laurie Solomon" <laurie@advancenet.net>
To: <dickbo@btopenworld.com>
Sent: Tuesday, June 11, 2002 5:34 AM
Subject: [filmscanners] RE: Density vs Dynamic range

> I further understood the number of discrete shades to be
> represented by "bit depth" not "resolution."

That is correct, bit depth means the number of grey levels available to play
with between maximum and minimum image density in any given tonal area of a
scanned image.

Actual real world final resolution of detail is controlled by the viewing
methodology. i.e. paper or film which as far as I understand the issue is
more and more affected by inkjet printing characteristics.

In my working life - I am now considered to be an old age pensioner - it was
affected by screen rulings and printing press/paper characteristics.

In traditional photography ye olde bromide paper decided all but than i
expect you know all that anyway.

> If the term, "resolution," is not designated the same thing as indicated
by
> ppi, dpi, line pairs, and numbers of pixels but  is being used differently
> or to designate something entirely different then knowing this is the case
> would reduce my confusion and clarify the comments for this layman.

On drum scanners, resolution is set by input scanning line pitch which is a
function normally related to degrees of original image enlargement and is
usually greater than output line pitch because resolving all the information
contained in an original is ultimately in the eye of the beholder so that no
matter what objective measurements may say, the final decision is made by
the viewer, who of course uses an analogue instrument...i.e. the eye/brain
combination.

Most people I have workd with consider that with CCD devices, resolution is
directly related to the megapixel number which in itself is influenced, in
terms of final visual appearance, by the scanning devices optical qualities
and also the designated gradiation charactistics which in turn would suggest
that once digitised, pixel grey levels will also contribute to the apparant
reproduced original detail, as interpreted by the viewer at final output
stage.

If one were to accept that position then it would also be necessary to take
account of the so called dynamic range althought working photographers will
have no time for that approach prefering instead to adopt straight original
density range as their method used to judge the quality of a scanned image.

Either way pixel depth, which we might describe as a vertical representation
of tonal change and pixel depth which we might also describe as a horizontal
representation of density range will have some influence on the rendition of
fine detail specifically relating to the grey level representation of image
contrast changes between adjacent areas of tone - think USM here my man and
all may well become clear.

I would always take all that stuff as read myself, and look for megapixel
count over any given unit area of a scanned image as the deciding factor
when making a product purchasing decision.

A design engineer, on the other hand, will argue that the whole chain of
components between scaning an image in and digital file out, influences the
result and that CCD pixel rate on it's own is not an acceptable way of
looking at the issue.

That of course is largely theoritical because in a real world manufacturing
situation the company who designs and makes the scanning product has to work
to a final cost calculation from which a profit must to be generated in what
will be a highly competitive and therefore price sensitive market.

I do believe it is generally considered the larger the profit the more
successful the company is performing, so it is not at all unusual to find
one contributing element to image quality may actually result in another
component receiving a qualititive image contribution somewhat lower in
potential quality than it can actually process, in other words a component
of the inage processing chain might be under speciefied in order to keep
manufacturing costs down.

In practical terms this means there will be compromises in qualititive
standards somewhere in the design process, so it is quite possible to find a
product that has top quality excellence at one stage of the scanning process
and qualititive/cost compromises somewhere else.

What actually matter in practice is "will it sell"?

Most photographers or repro men therefore will tell you that resolution
measured by the number of pixels across a given image area, is the true
measure of detail rendition because they consider the CCD to be the key
component in quality generation.

A design engineer, I should point out, may not.



-----Original Message-----
From: filmscanners_owner@halftone.co.uk
[mailto:filmscanners_owner@halftone.co.uk]On Behalf Of Jonathan King
Sent: Monday, June 10, 2002 10:45 AM
To: laurie@advancenet.net
Subject: [filmscanners] Re: Density vs Dynamic range


Ok, I bitten my tongue long enough.  It hurts!


On Mon, 10 Jun 2002 10:32:30 -0400, Austin Franklin wrote:
>
>
>>However my point is that if you can reduce the noise level then you can
>>increase the number of steps (by halving the step size) with real
>>benefit, but without altering the range.
>
>Hi Peter,
>
>Correct, but that INCREASES the dynamic range.
>



No it doesn't.  You are confusing dynamic range and resolution.  Doubling
the number of steps & halving the step size will keep the same dynamic
range, but it will double the resolution.  Resolution here is the number
of discrete shades of R, G or B.  Peter's point is that if the noise of
the system is greater than the step size, decreasing the step size will
just digitize noise - Not Good, unless you play with digital processing
techniques, as alluded to.




>>My principle argument was that a 5000:1 ratio does not specifically
>>define
>>that 5000 steps are requires
>
>Well, yes it does...that's what it means, with respect to what we are
>talking about.  Go download a few linear CCD specs (or if you want, I can
>mail you some), and you'll see they talk about it in exactly the same way
>I
>do.
>
Try this one, unless there is a better example.  It looks like a nice CCD
for a 4000dpi, 120 film scanner?

http://www.kodak.com/US/plugins/acrobat/en/digital/ccd/kli8023Long.pdf

FYI Kodak defines dynamic range pretty much the way engineering schools
I've attended in the U.S., and apparently Ireland, do:  Max. Output level
divided by the dark noise level.

>>Just a quick question - do CCDs really use a +/- voltage swing? I'd have
>>thought that would have introduce noise problems around 0.

It looks like the Kodak chip uses only positive voltages, but the output
has a DC offset that the buffer amplifier and A/D have to deal with.


If you want to split technical hairs, please at least be willing to state
and consistently use terms.  First Nyquist, and now this.  It has the
appearance of trying to baffle the non-technical( Wow, photographers don't
need engineering degrees? ;), but doesn't provide any clarity.

Best Regards,

Jon








----------------------------------------------------------------------------
------------
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



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



 




Copyright © Lexa Software, 1996-2009.