On the flip side, I mostly shoot digital -- I've shot almost a hundred thousand frames of it in the last five years since I got into photography. But I'm now wanting to borrow my dad's old mothballed OM-1 and get some Velvia, just to try it.
The loss hit is minimal, too, if you use a few extra bits.
I have a 10MP camera that spits out 5.5MB JPEGs, and unless I have something that I know will require extreme post-processing, I don't mess with the raw files. I have the thing set to raw+jpeg, but the raws just take up space most of the time -- the jpegs are just fine.
(And I *am* an image-quality fiend most of the time. I am picky, but I just don't get anything extra out of the raws. It helps that Olympus has a good jpeg converter, unlike Canon where they look like mush.)
Quality at what light level? (I assume you're comparing medium-format film to 35mm digital).
A Nikon D700 can make very good images at ISO 1600. They don't even *make* non-crappy color film at that speed. (Also, good luck finding a f/2 lens for your Hasselblad.)
Digital utterly demolishes film at high sensitivities. I have some great shots at ISO 800, and I shoot *quarter-frame*. No color shift or loss of saturation, just slight graininess.
It's not that bad. I have a friend who shoots it on an old twin-lens reflex camera that works pretty well. He's a grad student like me and can afford it, so...
Granted, I spend nothing per image. This lets me spend my photo budget on nice lenses.:)
There are two arguments here: what happens at the same *angle of view*, and what happens at the same *focal length*.
I use an Olympus camera, which has a sensor half the size of film.
On 50mm film, a 400mm f/7 lens would have a certain amount of DOF.
Shot on my Olympus, it would produce a more magnified image, since the sensor is smaller. But it would have *less* DOF. Since the image has to be magnified more (from the sensor to print), smaller amounts of blurring are acceptable.
If I want something that acts like a 400mm lens on film, I want a 200mm lens on Olympus. A 200mm f/7 (shot on Olympus) gives me *more* DOF than a 400mm f/7 on film. A 200mm f/3.5, which is what I use most of the time, gives me the same angle of view and the same DOF as the 400mm f/7 on film.
Wrong way on the depth of field. Larger formats have LESS depth of field, not more. Actually, if you want the same amount of depth of field, you get the same amount of shot noise on any format.
It depends. Yes, there is softening due to diffraction, and softening due to shot noise and noise reduction.
But there is useful detail in those extra pixels. I used to have a Panasonic FZ50, which is a consumer-grade 1/1.8" sensor with a pretty nice lens on the front (35-420 equivalent superzoom), and I've taken pictures where all that resolution *does* improve the image. Looking at one right now, actually -- printed 16x20 on my wall, looks great. It's of a butterfly with spread wings, and you can make out every scale. Can't do that with 4MP.
In the sciences you put a huge effort into quantifying error. A result might be quoted as:
60 +- 2 due to limited sampling in a Monte Carlo experiment (statistical error) +- 0.5 due to uncertainties in a previous result that this one relies on +- 0.2 due to using an approximation in our math +- 0.8 due to uncertainties in how we corrected for a bias (systematic error)
The presidential pollsters do this: they'd quote some number as "58% for Obama, with a 2 percent statistical margin of error, and an additional 1 percent error coming from the fact that we're not quite sure if we're over- or under-sampling cellphone-only voters."
If your estimates aren't *precise*, that's okay. You can still give an honest estimate with a large error bar. Do it, and honestly quantify your uncertainty.
nobody can explain how it works (there aren't even any good theories*)
Quantum electrodynamics produces results that agree with experiment to thirteen significant digits. It is probably the most accurate, successful theory ever devised.
Slashdot Mirror
I prefer the hot-librarian-looking women to actually be literate. Kind of ruins the fantasy if they're not.
In the case of humans the color comes from pigment, and pigmentation is subtractive.
On the flip side, I mostly shoot digital -- I've shot almost a hundred thousand frames of it in the last five years since I got into photography. But I'm now wanting to borrow my dad's old mothballed OM-1 and get some Velvia, just to try it.
You can print those digital files before the robopocalypse.
The loss hit is minimal, too, if you use a few extra bits.
I have a 10MP camera that spits out 5.5MB JPEGs, and unless I have something that I know will require extreme post-processing, I don't mess with the raw files. I have the thing set to raw+jpeg, but the raws just take up space most of the time -- the jpegs are just fine.
(And I *am* an image-quality fiend most of the time. I am picky, but I just don't get anything extra out of the raws. It helps that Olympus has a good jpeg converter, unlike Canon where they look like mush.)
Quality at what light level? (I assume you're comparing medium-format film to 35mm digital).
A Nikon D700 can make very good images at ISO 1600. They don't even *make* non-crappy color film at that speed. (Also, good luck finding a f/2 lens for your Hasselblad.)
Digital utterly demolishes film at high sensitivities. I have some great shots at ISO 800, and I shoot *quarter-frame*. No color shift or loss of saturation, just slight graininess.
It's not that bad. I have a friend who shoots it on an old twin-lens reflex camera that works pretty well. He's a grad student like me and can afford it, so...
Granted, I spend nothing per image. This lets me spend my photo budget on nice lenses. :)
There are two arguments here: what happens at the same *angle of view*, and what happens at the same *focal length*.
I use an Olympus camera, which has a sensor half the size of film.
On 50mm film, a 400mm f/7 lens would have a certain amount of DOF.
Shot on my Olympus, it would produce a more magnified image, since the sensor is smaller. But it would have *less* DOF. Since the image has to be magnified more (from the sensor to print), smaller amounts of blurring are acceptable.
If I want something that acts like a 400mm lens on film, I want a 200mm lens on Olympus. A 200mm f/7 (shot on Olympus) gives me *more* DOF than a 400mm f/7 on film. A 200mm f/3.5, which is what I use most of the time, gives me the same angle of view and the same DOF as the 400mm f/7 on film.
Wrong way on the depth of field. Larger formats have LESS depth of field, not more. Actually, if you want the same amount of depth of field, you get the same amount of shot noise on any format.
It depends. Yes, there is softening due to diffraction, and softening due to shot noise and noise reduction.
But there is useful detail in those extra pixels. I used to have a Panasonic FZ50, which is a consumer-grade 1/1.8" sensor with a pretty nice lens on the front (35-420 equivalent superzoom), and I've taken pictures where all that resolution *does* improve the image. Looking at one right now, actually -- printed 16x20 on my wall, looks great. It's of a butterfly with spread wings, and you can make out every scale. Can't do that with 4MP.
And they say that the rest of the people who call themselves Christian aren't the real Christians -- they are.
When it comes to political Christianity, it's the bible-thumpers that have the clout, sadly.
Christianity. Ever been to the American South?
(Disclaimer: I am Alabamian.)
In the sciences you put a huge effort into quantifying error. A result might be quoted as:
60
+- 2 due to limited sampling in a Monte Carlo experiment (statistical error)
+- 0.5 due to uncertainties in a previous result that this one relies on
+- 0.2 due to using an approximation in our math
+- 0.8 due to uncertainties in how we corrected for a bias (systematic error)
The presidential pollsters do this: they'd quote some number as "58% for Obama, with a 2 percent statistical margin of error, and an additional 1 percent error coming from the fact that we're not quite sure if we're over- or under-sampling cellphone-only voters."
If your estimates aren't *precise*, that's okay. You can still give an honest estimate with a large error bar. Do it, and honestly quantify your uncertainty.
And they're not planets.
So we are to accept that just because lots of Americans are too dumb to understand how misplaced our priorities are, that we shouldn't worry about it?
How does that change the fact that he started as an ordinary dude and is now president?
Yes, he's half white too. White people can have "living the American dream" stories too, y'know.
Do you really want private companies going to the Moon and commercializing it?
The military budget is not being cut (significantly). US military spending, regardless of how it is classified, is discretionary in reality.
You could fund a manned Mars mission (pessimistic estimated total cost: $100 billion) with a 3% cut in the US military budget for ten years.
Uncontrolled willfully ignorant are less harmful than the willfully ignorant under control of assholes.
Because religion is the cause of a great deal of dumbassery in the USA.
nobody can explain how it works (there aren't even any good theories*)
Quantum electrodynamics produces results that agree with experiment to thirteen significant digits. It is probably the most accurate, successful theory ever devised.
That was my first guess, too.
Q: What's the difference between Sarah Palin's mouth and her vagina?
A: Only 20% of what comes out of her vagina is retarded.
It is patently absurd for two sixteen-year-olds to rape each other.
Running code on 4096 Opteron cores on a supercomputer atm, will get back to you when it finishes.
Code that I tested on my computer, and didn't have to register with anybody to get a copy of gcc.