I guess in order to fill the field of view of a larger sensor, you need a faster lens, which then results in a smaller depth of focus. But the reason for the smaller depth of focus is still the lens, even if the selection of the lens is based on field-of-view (sensor) considerations.
a) at 100 pixels/mm, a 24 x 35mm image is only 8.4 Mpixel, not 34M.
I already posted a explanation of the 34 million number in a different place, but I'll repeat it here because this point needs to be clarified. You need to multiply your pixel count by 4 because the measure of resolution is being able to resolve a pattern of alternating black and white lines. The line count is for either just the black lines or just the white lines, so you actually need two pixels for each line you want to resolve (one for the black line, one for the white line). Then the math is [ (24 mm) x (2 pixels/line) x (100 lines/mm) ] x [ (35 mm) x (2 pixels/line) x (100 lines/mm) ] = 33,600,000 pixels.
If you don't account for this and try to take a picture of a 100 line/mm object, you will get an image of a flat grey field instead of a pattern of alternating black and white lines.
You need to multiply your number by 4 because the measure of resolution is being able to resolve a pattern of alternating black and white lines. The line count is for either just the black lines or just the white lines, so you actually need two pixels for each line you want to resolve (one for the black line, one for the white line). Then the math is [ (24 mm) x (2 pixels/line) x (100 lines/mm) ] x [ (35 mm) x (2 pixels/line) x (100 lines/mm) ] = 33,600,000 pixels.
If you don't account for this and try to take a picture of a 100 line/mm object, you will get an image of a flat grey field instead of a pattern of alternating black and white lines.
Large DOF has nothing to do with digital vs. analog and everything to do with the fact that the focal length in most consumer digitals is so short that the hyperfocal distance becomes very short.
Come on, what kind of confused reasoning is that? Digital cameras have shorter focal lengths and larger DOF because digital sensors are smaller than 35mm film.
Well, the focal length and depth of focus are both properties of the camera lens, and have nothing to do with the detector being used (film or CCD/CMOS). The focal length is directly determined by the geometry and composition of the lens, and the depth of focus is a byproduct of the focal length and the aperture of the lens.
I suppose if I sat down and thought about it, I could come up with a more detailed analysis of the effect of smaller detectors on camera design, but I'm lazy...;-)
There are probably things that you can do while taking photographs on film that cannont be reproduced by Photoshop--maybe some combination of lighting, exposure, film response, and focus. Photoshop is an amazing tool, but at the same time, I'm sure there are some things that it can't do.
Actually, the random distribution of silver halide crystals in film probably avoid aliasing artifacts that can occur with the periodic distribution of CCD/CMOS elements on a detector array.
That's when you compile from source and install into/usr/local or/opt...
On my old Debian installation, I didn't have any problems doing things this way. Granted, I had only a few things installed this way--Mozilla, JDK and a couple of other things that weren't available via apt-get or whose versions in the apt-get tree were too old--but it worked fine for me.
Re:What Edison would say if alive today...
on
Engineer in a Box?
·
· Score: 2, Funny
Old lady: "Tell me, Dr. Hathaway, what's Professor Einstein really like?"
Dr. Hathaway: "Dead."
Re:Clean look and good international support
on
Google Does the News
·
· Score: 2
if you think slashdotting is cool, wait till google news points at your community newspaper.
It's a good thing Google provides its own cache...;-)
I think the main reason I'm bitter is I would like to have put something on the dome, but $100,000+ over four years to put something on the dome is a little much. Besides that, I have no regrets of completing my undergraduate studies at a public institution, debt-free. In fact, I think it is smarter to save on one's undergraduate expenses by going to a public institution than to go to a private institution, no matter how prestigious, and incur a mountain of debt. And I know many of my college friends agree with me, because a lot of them turned down offers of admission for MIT, Caltech, and the like. After all, you can always go to those places for graduate school, and in computer science or engineering, they will usually fully subsidize the cost, regardless of financial need. Which begs the question: why not do that for undergraduate studies, too?
Well unfortunately, whether or not military personnel are overspecialized often doesn't matter. Perception is reality--many people believe that military occupations are overspecialized, and that can be a barrier to joining the non-military workforce.
The issue of MIT becoming tuition-free is not relevant to the topic of OpenCourseWare, but the high tuition of MIT and the Ivy League universities has been a pet peeve of mine for some time. In truth, MIT, Harvard, Princeton, Yale, and many other elite private universities would still be in perfect financial health if they did not charge tuition, as most of their funding comes from the growth of their endowments or from federal or private research funds. Greenspun's article raises a valid point that the high tuition of MIT and the Ivy League universities is equivalent to price-fixing, and I believe there was actually an anti-trust finding against MIT and the Ivy League over this issue.
I guess I'm really sore about this issue because I had the opportunity to go to MIT or Stanford, but because I didn't qualify for financial aid, I decided to go to a public university that did provide financial support. Ideally, one shouldn't be forced into decision regarding education by financial considerations, but in this educational system, it happens all too often.
Except that military service only demonstrates competence in a narrow specialization (i.e. how to operate specific equipment or something along those lines). I agree that military service is good for developing discipline (as a general rule my college friends who had been in the military or were doing ROTC were much more disciplined than I), but college allows one to demonstrate more general competence in a field more (or less) relevant to the business world.
But the advantage of having completed university coursework for a topic like algorithm design or algorithm analysis is that the grade for that course will represent competence in that field as measured over the course of a quarter or semester. I find it difficult to believe that someone's competence in such an open-ended and abstract field can be assessed by a score on an exam taken in the span of a few hours.
Holography requires sufficient film resolution to record the information content of the object modulated on a high spatial frequency carrier. In simplistic terms, lots of images of the object from different perspectives are recorded on film as a hologram, which means the film resolution requirement for making a hologram of the object is much higher than for taking a photograph of the object. The problem here is that the object is so detailed that you could not find film with sufficient resolution to record the hologram.
The original Science article cites an Applied Optics article from 1984, which I'm would guess basically says what I've said in the previous paragraph.
"Um.... hi... this is Jane Doe. My Commodore 64 started smoking earlier, and I shot it with a fire estinguisher. Um... do you think it's safe to turn it back on?"
Well, it depends on if it was a water extinguisher or a chemical extinguisher. If it's a water entinguisher, maybe if you give it enough time to dry, and if it's a chemical extinguisher, probably (depending on the chemicals--I figure they don't fill them with mercury or TNT).
Re:American servicemembers in Germany
on
When Users Attack
·
· Score: 1
Also 50 Hz vs. 60 Hz. When I was at a study abroad program at Oxford, people couldn't figure out why their US-bought clocks were losing an hour every 6 hours. I thought about it for a while, then I realized that the clocks were probably using the line current as a frequency standard. So while the power adapters change 220V to 110V, they don't change 50 Hz to 60 Hz, and voila, you have clocks running at 5/6ths of their normal speed.
I think CPU heat tolerance is a function of temperature over time. The reason high temperatures are bad is that the dopants that make regions in the silicon p-doped or n-doped start diffusing out at a rate proportional to the temperature. So if the chip stays cool, you have a computer, but if it stays hot for a long time (or gets very very hot for a short time), you get up with a very dirty silicon crystal.
Slashdot Mirror
I guess in order to fill the field of view of a larger sensor, you need a faster lens, which then results in a smaller depth of focus. But the reason for the smaller depth of focus is still the lens, even if the selection of the lens is based on field-of-view (sensor) considerations.
If you don't account for this and try to take a picture of a 100 line/mm object, you will get an image of a flat grey field instead of a pattern of alternating black and white lines.
You need to multiply your number by 4 because the measure of resolution is being able to resolve a pattern of alternating black and white lines. The line count is for either just the black lines or just the white lines, so you actually need two pixels for each line you want to resolve (one for the black line, one for the white line). Then the math is [ (24 mm) x (2 pixels/line) x (100 lines/mm) ] x [ (35 mm) x (2 pixels/line) x (100 lines/mm) ] = 33,600,000 pixels.
If you don't account for this and try to take a picture of a 100 line/mm object, you will get an image of a flat grey field instead of a pattern of alternating black and white lines.
I suppose if I sat down and thought about it, I could come up with a more detailed analysis of the effect of smaller detectors on camera design, but I'm lazy...
There are probably things that you can do while taking photographs on film that cannont be reproduced by Photoshop--maybe some combination of lighting, exposure, film response, and focus. Photoshop is an amazing tool, but at the same time, I'm sure there are some things that it can't do.
Actually, the random distribution of silver halide crystals in film probably avoid aliasing artifacts that can occur with the periodic distribution of CCD/CMOS elements on a detector array.
The lenses that are well-corrected over a 8"x10" field must also cost a fortune to manufacture and purchase...
That's when you compile from source and install into /usr/local or /opt ...
On my old Debian installation, I didn't have any problems doing things this way. Granted, I had only a few things installed this way--Mozilla, JDK and a couple of other things that weren't available via apt-get or whose versions in the apt-get tree were too old--but it worked fine for me.
Old lady: "Tell me, Dr. Hathaway, what's Professor Einstein really like?"
Dr. Hathaway: "Dead."
I think the main reason I'm bitter is I would like to have put something on the dome, but $100,000+ over four years to put something on the dome is a little much. Besides that, I have no regrets of completing my undergraduate studies at a public institution, debt-free. In fact, I think it is smarter to save on one's undergraduate expenses by going to a public institution than to go to a private institution, no matter how prestigious, and incur a mountain of debt. And I know many of my college friends agree with me, because a lot of them turned down offers of admission for MIT, Caltech, and the like. After all, you can always go to those places for graduate school, and in computer science or engineering, they will usually fully subsidize the cost, regardless of financial need. Which begs the question: why not do that for undergraduate studies, too?
What do you think this is, K5? ;-)
Well unfortunately, whether or not military personnel are overspecialized often doesn't matter. Perception is reality--many people believe that military occupations are overspecialized, and that can be a barrier to joining the non-military workforce.
The issue of MIT becoming tuition-free is not relevant to the topic of OpenCourseWare, but the high tuition of MIT and the Ivy League universities has been a pet peeve of mine for some time. In truth, MIT, Harvard, Princeton, Yale, and many other elite private universities would still be in perfect financial health if they did not charge tuition, as most of their funding comes from the growth of their endowments or from federal or private research funds. Greenspun's article raises a valid point that the high tuition of MIT and the Ivy League universities is equivalent to price-fixing, and I believe there was actually an anti-trust finding against MIT and the Ivy League over this issue.
I guess I'm really sore about this issue because I had the opportunity to go to MIT or Stanford, but because I didn't qualify for financial aid, I decided to go to a public university that did provide financial support. Ideally, one shouldn't be forced into decision regarding education by financial considerations, but in this educational system, it happens all too often.
Except that military service only demonstrates competence in a narrow specialization (i.e. how to operate specific equipment or something along those lines). I agree that military service is good for developing discipline (as a general rule my college friends who had been in the military or were doing ROTC were much more disciplined than I), but college allows one to demonstrate more general competence in a field more (or less) relevant to the business world.
But the advantage of having completed university coursework for a topic like algorithm design or algorithm analysis is that the grade for that course will represent competence in that field as measured over the course of a quarter or semester. I find it difficult to believe that someone's competence in such an open-ended and abstract field can be assessed by a score on an exam taken in the span of a few hours.
My $0.02
Holography requires sufficient film resolution to record the information content of the object modulated on a high spatial frequency carrier. In simplistic terms, lots of images of the object from different perspectives are recorded on film as a hologram, which means the film resolution requirement for making a hologram of the object is much higher than for taking a photograph of the object. The problem here is that the object is so detailed that you could not find film with sufficient resolution to record the hologram.
The original Science article cites an Applied Optics article from 1984, which I'm would guess basically says what I've said in the previous paragraph.
What I want to know is who was the guy that went around licking frogs to try to get high...
I thought that not eating cheese was more of a vegan thing...
True...
Also 50 Hz vs. 60 Hz. When I was at a study abroad program at Oxford, people couldn't figure out why their US-bought clocks were losing an hour every 6 hours. I thought about it for a while, then I realized that the clocks were probably using the line current as a frequency standard. So while the power adapters change 220V to 110V, they don't change 50 Hz to 60 Hz, and voila, you have clocks running at 5/6ths of their normal speed.
I think CPU heat tolerance is a function of temperature over time. The reason high temperatures are bad is that the dopants that make regions in the silicon p-doped or n-doped start diffusing out at a rate proportional to the temperature. So if the chip stays cool, you have a computer, but if it stays hot for a long time (or gets very very hot for a short time), you get up with a very dirty silicon crystal.