Re: Where is digital camera resolution going?

In article <443DCCFA.6060703@xxxxxxxxx>, "Roger N. Clark (change username to rnclark)" <username@xxxxxxxxx> writes

So, Kennedy, how many cycles/degree can you see? With your better
eyes, you should be able to do about 50 cycles/degree or even
better. Try it in different lighting conditions.

I'll have to leave that test to run later, Roger. I never expected this discussion to run as long as it has and, due to the holiday and business, I am now going to be offline for a couple of weeks. I'll follow-up when I get back.

So what? Just because no photo printer can currently print at 500 lppi
has nothing to do with problems with sampling and what the eye can resolve.
If the eye resolves more detail, then we need better printers.

Surprising that none exist is there is a "need" for them.

The same argument used to be made when digital cameras reached 2
megapixels, then again 3, then 4, and so on.

That isn't the same at all - as I stated in the part you didn't quote, there isn't a technology restriction in providing such resolution in printers. When people were making such comments about 2, 3 & 4 megapixel cameras it was at a time when 2, 3 & 4 megapixel sensors didn't exist at a commercially viable price. That is a pretty hefty technology barrier.

I print photos in highest quality mode using 1200x2400 dpi.
That is the dot density - not the pixel density!
Your HP printer resamples everything sent to it to 600ppi *before* dithering ink dots at the higher resolutions you are quoting!

No it does not resample.

Technically you are correct - it isn't the printer that resamples, it is the printer driver. There is a Windows API that you can call to interrogate the printer for its resolution - for HP inkjets that *is* 600ppi. I am sure you can look into this while I am away.


The only
flaw in the above data is that the MTF numbers are too low because
I didn't take into account the MTF of the optical system recording
the data from the print. But that should be a small effect.

There is obviously a flaw you don't understand in either your methodology or interpretation since it is impossible to obtain any modulation, let alone resolution, at 600lppi on a printer which resamples images to 600ppi prior to dithering ink dots.

The flaw is in your resampling assumption.

That isn't a flaw. It is something that has been tested by HP users and reported. It is also something that has been documented by third party printing software - eg. Qimage.

I guess it is possible that you have written your own non-standard driver, but I doubt it.

http://www.ddisoftware.com/qimage/quality/

Note it says "almost always." Apparently, in my printer driver,
that is not the case.

"Almost always", because for large format printers it is less than this - 300ppi in the case of HP and Canon and 360ppi in the case of Epson.

Just like I predicted. Care to comment on the data in Figure 24?
I thought not.

Actually I did comment - and you even quoted my comment in the first line above! Nevertheless, here is further comment - it doesn't appear to deviate much from the 1.7cy/mrad figure! 20% modulation threshold at 30cy/deg except when using retinal illumination beyond the normal pupil accommodation levels.

For those jumping in, the reference is:
http://webvision.med.utah.edu/KallSpatial.html
The figure is:
http://webvision.med.utah.edu/imageswv/KallSpat24.jpg

The maximum brightness is for 900 trolands.
A troland is:
http://www.photonics.com/directory/dictionary/lookup.asp?url=lookup&entr
ynum=5436&letter=t
"That level of retinal illuminance resulting when a surface
with a luminance of 1 candela/m^2 is viewed through a pupil
with an area of 1 mm^2."

That is 1 troland, the full definition is at
http://encyclopedia.thefreedictionary.com/Troland
"The Troland typically refers to the ordinary or photopic Troland, which is defined in terms of the photopic luminance:
T = L · p
where L is the photopic luminance in cd / m2 and p is pupil size in mm2"

So the maximum light is 900 candelas/sq meter with a
pupil of 1 square mm. That is the equivalent of an
overcast day. Overhead sunlight would be over 50
times brighter.

Last time I checked, my pupil was a lot larger on an overcast day - and I am not sure it ever gets as small as 1mm^2. You need to take account of *actual* pupil size to determine the scene illumination - and it is much brighter than your estimate.
--
Kennedy
Yes, Socrates himself is particularly missed;
A lovely little thinker, but a bugger when he's pissed.
Python Philosophers (replace 'nospam' with 'kennedym' when replying)
.

Relevant Pages

  • Dark adaption and pupil size - an experiment (longish)
    ... when I was about six years old looking though a neighbors reflector ... The 11x80 bincos were very impressive from a dark ... I'd do an interesting experiment to see if increasing my pupil size ... often used in eye drop form by eye doctors to dilate patients ...
    (sci.astro.amateur)
  • Re: DOF preview in OVFs of DSLRs is crippled
    ... lets one stop down the lens. ... that the sensor accepts a ray bundle with a larger angle than the viewfinder. ... if instead of a ground 'glass' focusing screen, if you had an opaque white projection surface and looked at that from the other side, that should look correct but you see that image combined with a clear view from your eye and your eye has a much smaller sensor than a DSLR. ... the entrance pupil of the viewer's eye must be aligned with and be of similar size to the instrument's exit pupil. ...
    (rec.photo.digital)
  • Re: Image Luminosity vs magnification (long explanation)
    ... >the human eye pupil, or about 7.62 mm at its maximal dilated pupil. ... I'm assuming "G" is the relative light grasp of an aperture in relation to the ... aperture equals 7-mm, magnification equals 1X ...
    (sci.astro.amateur)
  • Re: Where is digital camera resolution going?
    ... Your HP printer resamples everything sent to it to 600ppi *before* dithering ink dots at the higher resolutions you are quoting! ... That is 1 troland, ... where L is the photopic luminance in cd / m2 and p is pupil size in mm2" ...
    (rec.photo.digital)
  • Re: Accommodation in Aphakial Eyes -- Bates has not been disproven on this.
    ... noticeable to the naked eye. ... about the normal eye is how much the pupil is constantly adjusting for ... I think it is true that the amount of sun energy entering a healthy ... And importantly the retina has the ability to adjust its light ...
    (sci.med.vision)