Forum: Troubleshooting and Beginner Help
10-02-2012, 09:01 AM
|
|
Why with a grain of salt? You are perhaps too modest Sir! :) Or do you have reason to discount Takinami's data?
Dave in Iowa
|
Forum: Lens Clubs
10-01-2012, 09:08 AM
|
|
The 150 is superb with the DA55-300 in my experience. One seldom noted advantage is that the built-in flash works for almost all the magnification range (to a bit above 1:1). I also have the Tamron 28-75 and have seldom used it with the Raynox 150.
Dave in Iowa
|
Forum: Pentax DSLR Discussion
09-30-2012, 03:23 PM
|
|
It seems to me a 100% strength birefringent AA filter atop a Bayer sensor array must add at least one 50% bright pixel at the edge of a step function illumination pattern.
That's the result I get thinking it thru & if I simulate it numerically by averaging four adjacent pixels. I think maybe this edge broadening should add according to the Kodak rule with p=2.
Dave
|
Forum: Pentax DSLR Discussion
09-29-2012, 03:55 PM
|
|
|
Forum: Pentax DSLR Discussion
09-29-2012, 09:44 AM
|
|
Are there any tabulated comparisons of AA filter strengths (like gaussian, sinc equivalent, etc) for various Pentax cameras?
Conversely, how might one test for a rough measure?
Like: I have a K-x camera - can I expect edge blur at best to be like that resulting from a Gaussian(1) filter, Gaussian(2) filter, ....?
Dave in Iowa
|
Forum: General Talk
09-28-2012, 08:41 AM
|
|
If she was informed of her Cherokee ancestry by her family, she'd be a lair if she denied it. It is reasonable that such a topic would arise in casual conversation during the process of interviewing for a faculty position. It is also likely that a university would like to demonstrate its diversity by publicizing it.
The real question is whether or not she used the info to unwarranted advantage and did not properly qualify it as being family lore.
|
Forum: General Talk
09-26-2012, 05:26 PM
|
|
I think it more probable that partisan attempts to limit/expand access at the polls will result in legal challenges after the election or violence during the election.
I won't be surprised to see conservative activists at polling places challenging voters to prove their eligibility and liberal activists trying to prevent such challenges. This has a high probability of creating real time conflicts.
|
Forum: Post Your Photos!
09-26-2012, 09:54 AM
|
|
I like the image and appreciated the tutorial in how to do it in PP.
I think the resulting image would be MUCH better if the burned out highlights on the boat's stern (and the top of the last pole) were toned down to match the surrounding slightly greyer levels. As it is, my attention is strongly drawn to those overexposed areas.
Dave in Iowa
|
Forum: Troubleshooting and Beginner Help
09-24-2012, 09:22 AM
|
|
How so? The theory is well established & my application of it was conservative.
The K-5's high pixel count means that a 100% crop maps a very small sensor area to each display pixel.
Here's an illustration from Luminous Landscape of a Gaussian distribution (very similar to the Airy distribution expected from a lens) on a square grid of pixels. The Gaussian "spot" is about six pixels wide, just like op's case.
From LuminousLandscape - Deconvolution sharpening revisited
Clearly, what idealy be a single bright pixel is smeared out over quite a few of its neighbors. The following is an attempt to give a reasonably useful approximation to what's going on in theory.. It ignores AA filters, etc..
Following Airy disk - Wikipedia, the free encyclopedia the Gaussian approximation to the Airy Disk is:
Relative brightness = exp(-(x^2+y^2)/(N/2)^2)/[(N/2)^2(Pi)] where x, y are in microns and N is the f-stop number.
The center brightness of a multi-pixel spot varies with 1/N^2, so an F:11 spot's center is about 4 times brighter than an F:22 spot's center.
Dave
PS I note your signature "Knowledge does not equal understanding." - true, but Ignorance is certainly a worse predictor of understanding.
|
Forum: Troubleshooting and Beginner Help
09-23-2012, 05:34 PM
|
|
Your K-5's specifications are 4928 x 3264 (16.3 MP) with a 23.4 x 15.6 mm CMOS sensor. That's 4.7 microns pixel spacing. The diffraction spot diameter for f:22 is:
d=2.44*N*(.55 microns)= 2.44*22*0.55=30 microns!
That means when you shot the photo, each infinitesimal spot of light was spread over a disk more than 5 pixels in diameter. When displayed at 100%, each point of light in the original scene is spread over 5 pixels on the display (almost 1mm). No wonder it looks fuzzy!
To prevent strong diffraction effects at 100% display, your f-number should be such that the diffraction spot radius is about the same as the pixel spacing. In your case,
N ~ 4.7(microns)/1.22*0.55(microns) ~ 8
Or, avoid high enlargements at high f-numbers.
Dave in Iowa
|
Forum: Pentax SLR Lens Discussion
09-22-2012, 07:44 AM
|
|
Good point; the error you point out is in our sloppy use of assigning the name "2.8" to the number 2*SquareRoot(2) = 2.828427.....
Using the names of the f-stops in calculations is usually good enough; what the actually aperture opening a manufacturer calls F:2.8 on a particular 100mm lens is highly unlikely to be precisely 28.28427mm.
I suppose a reasonable thing to do for quick estimates is to use f-stop names in calculations then round up or down to the next common f-stop name; in this case f:3.9 rounds to f:4 (as would f:4.1).
Best wishes,
Dave
|
Forum: Pentax SLR Lens Discussion
09-21-2012, 10:29 AM
|
|
No.
1mm x 11/2.8 = 3.9mm
Same magnification on the sensor.
Dave
|
Forum: Pentax DSLR and Camera Articles
09-18-2012, 11:06 AM
|
|
I think the statement must be modified to emphasize the stochastic nature of the beast. In aggregate we collectively get what we pay for, but may individually get better (or worse) than we paid for.
Dave in Iowa
|
Forum: Troubleshooting and Beginner Help
09-18-2012, 09:10 AM
|
|
For any lens system there is at least one optimal combination of parameters. There is an aberration limit at the wide open end of the aperture range (if you open the aperture more the image gets worse) and a diffraction limit as the aperture gets smaller and smaller.
When you add auxiliary optics or reverse the lens (or go beyond the minimum focus distance) I don't think there is a way to predict how aperture changes affect the image. Experiment is the best way to find out.
Dave in Iowa
|
Forum: General Talk
09-09-2012, 12:30 PM
|
|
The first few comments after the article were more disturbing than the article itself.
Dave in Iowa
|
Forum: Pentax SLR Lens Discussion
09-09-2012, 10:19 AM
|
|
|
Forum: General Talk
09-05-2012, 01:23 PM
|
|
You could say that you understand that women have different (and often more discriminating) views on color, composition, lighting, etc than men so you'd like to see that in action. Maybe a mixed gender team will be better than (or at least as good as) single gender; ie. you may learn from her & vis-versa.
Partnering with her would be a good way to see if there's any truth to that. Maybe your outlooks will be complementary - you won't know 'til you try. Treat her like a colleague or partner not a date.
Old Dave in Iowa
|
Forum: Pentax SLR Lens Discussion
09-05-2012, 09:08 AM
|
|
There should be no risk of electronic damage to the camera by short circuiting any of the camera mount pins. That's because the camera is designed to receive all metal mount K and M type lenses which always short circuit the camera pins.
Dave in Iowa
|
Forum: Pentax Camera and Field Accessories
09-03-2012, 09:48 AM
|
|
No problem! Someday maybe.
Dave in Iowa
|
Forum: Pentax SLR Lens Discussion
09-02-2012, 08:52 PM
|
|
It is very close to exact if you are far from hyperfocal conditions (ie. the far background is noticeably out-of focus).
|
Forum: Do-It-Yourself
09-01-2012, 04:06 PM
|
|
Awwwww, thanks [Blush].
Actually, you only need to put a tube /box on the lens & cut it back 'till it doesn't show any more. (As you point out!)
Dave in Iowa
|
Forum: Do-It-Yourself
09-01-2012, 10:23 AM
|
|
In the equation Hood.diameter=Lens.diameter(Hood.depth/focal.length +1) the term in the parentheses must be bigger than 1 (because 1 is added to another positive number, so hood diameter must be greater than lens diameter - probably you used (hood.depth/(focal.length+1)) instead of
(hood.depth/(focal.length)+1).
Here's a better equation and a description of where it came from.
using your numbers and the above equation
Hood diameter > Hood.depth*sensor.width/focal.length + lens diameter = 21*24/28+20 =38
With film, sensor.width is 36, not 24 so the hood diameter should be greater than 21*36/28+20=47, so I'd expect vignetting at the corners.
Dave in Iowa
|
Forum: Pentax Camera and Field Accessories
08-31-2012, 07:55 AM
|
|
Cropping the original image may be a good idea for comparison.
The best way to make sure the effective f-stop is the same is to use the same lighting, ISO, and exposure time - adjust the f-stop (not the time) so the exposures are the same. At first this seems to make no sense, but if +/-EV, ISO and exposure time, lighting, etc are the same, effective f-stop must be the same.
This method is better than calculating effective f-number I think.
Dave
|
Forum: Pentax Camera and Field Accessories
08-31-2012, 07:02 AM
|
|
Thanks Lowell, you are right & the situation is clearly going to vary lens type to lens type.
Here's an image comparing the corners of good diopter lenses on the end of extended bellows. They looknabout the same to me and pretty good!
Dave in Iowa
|
Forum: Pentax Camera and Field Accessories
08-31-2012, 06:52 AM
|
|
I think what you and Lowell say are both true. I have seen cases where tubes are better at the edges and where diopters are better at the edges.
However, I'm still seeking example photos of either case to discern which is more common and how to predict.
At this point I don't know what to tell someone who asks "Which is best; Tubes or Achromats?" Especially if edges are a concern.
Dave in Iowa
|