|
| Author | Thread |
|
|
09/25/2005 02:06:47 AM · #1 |
Greetings everyone,
After wasting $700 on a COOLPIX, I would now like to get some technical info on what factors contribute to a digital camera's light gathering ability. I was taking pics in a dimly lit church and even at f2.6 I needed impossibly slow shutter speeds. The gentleman next to me, however, had no such problem. His monitor was brightly lit. I don't know what camera it was but I did notice that the lens was about 2" in diameter. Upon closer inspection of my camera, I noticed that what I thought was the lens is just a 1", flat piece of coated glass. The physical opening (lens?) into the the camera body is probably 2mm.
So I ask: What should I look for in a camera to allow shooting in low light?
Just an afterthought...does the size of the CCD matter?
Thanks,
Ken |
|
|
|
09/25/2005 02:57:42 AM · #2 |
Originally posted by gibbs:
Just an afterthought...does the size of the CCD matter?
Thanks,
Ken |
Yes. |
|
|
|
09/25/2005 03:02:24 AM · #3 |
| The larger the diameter of the lens the more the possiblity of getting more light into the camera. That's why smallish pocketable/pursable camera are not as good when shooting indoors. |
|
|
|
09/25/2005 03:22:00 AM · #4 |
Originally posted by faidoi:
The larger the diameter of the lens the more the possiblity of getting more light into the camera. That's why smallish pocketable/pursable camera are not as good when shooting indoors. |
This is not strictly true. In theory a given f/stop will trasmit exactly the same amount of light to the sensor regardless of what camera is carrying that f/stop.
F/stop is a ratio of physical diameter of the aperture to focallength of the lens. A 25mm aperture on a 50mm lens is f/2.0. A 25mm aperture on a 100mm lens is 5/4.0. On a 200mm lens it's f/8.0, and so forth. The intensity of light falling on an object (the sensor in this case) decreases witht he square of the distance between source and object. In a lens the "source" is the physical location of the aperture.
In a view camera stype system, where the lens is mounted on a rail, the longer the focal length of the lens the further it must be racked out from the film plane to attain focus, and this equation is quite precise. On cameras like the ones we use, where the location of both the lens mount and the sensor are fixed in relation to each other, focus is achieved optically by the complex glass elements of the lens rather than simnply by physically extending the lens, but nevertheless the f/stop equation is taken into account in how the different apertures are named.
You may be sure that regard;ess of what camera is being used, the same intensity of light is reaching the sensor at the same f/stop, or metering systems and ISO numbers would be meaningless.
The fact that his monitor was brightly lit is meaningless, though. I don't know about the 4500, but my coolpix 5700 and, indeed, every digital camera I've owned, has an adjustment for LCD (monitor) brightness. He probably just had his dialed way up. It's an electronic image, not an optical one, and it can be adjusted for viewing brightness in exactly the same way you can adjust the brightness on your computer monitor or television.
Just because you dial the lCD up bright doesn't mean the captured image is any brighter. Indeed, if you dial the LCD too bright you'll be VERY disappointed at how much darker the actual captured image is.
As for sensor size, it has no effect on the f/stop equation or how much light is being transmitted to the film plane by the lens. Larger sensors ARE better in low light, but that's not why; what makes them better is that they have less "noise" at higher ISO speeds, so you can shoot at, say, ISO 800 instead of ISO 100 without concommitant image-wrecking increases in noise.
My 20D has less noise at ISO 1600 than my Coolpix 5700 does at ISO 100.
Robt.
|
|
|
|
09/25/2005 09:44:03 AM · #5 |
Originally posted by bear_music:
As for sensor size, it has no effect on the f/stop equation |
What I've read agrees with what you said.
But I think I have also read that the sensor size does impact on the equation in one respect: The larger the sensor, the larger the lens has to be in order to provide the same focal length and f/stop.
That's why P&S cameras get away with such incredibly small lenses. Their lenses don't need to capture as much light because they illuminate a much smaller area.
Well, it's what I've read anyway ... and it sounds logical to me. :-)
|
|
|
|
09/25/2005 08:53:42 PM · #6 |
I've heard that the sensor size does matter.
In theory, the f-stop is determined by the focal length / apreture. A 50mm lens on a DSLR that has an f-stop of 1 will have an aperature of 50mm, and will have an area defined as 50mm * pi that will let some given amount of light through. This amount depends on how much light is available (i.e. how bright the subject is)
The light you can use is the light that hits the sensor, so a full-frame sensor will have an area that gets 36mm x 24mm = 864 square mm worth of photons. A 1.6 crop factor sensor will get (IIRC) 24mm x 16mm = 384 square mm worth of photons, or about 44% of what the full frame sensor would.
Complicating this is the number of photosites (pixels). The larger the photosite, the greater the number of photons available. But naturally, for a given size sensor, increasing the number of photosites gives a smaller size for each photosite.
From what I've heard, noise does not increase proportionatly with sensor size, so the S/N ration (i.e. "good" photons to noise) is greater with larger photosites.
In general, digicams have much smaller sensors (and much smaller photosites) than DSLRs.
|
|
|
|
09/25/2005 09:54:19 PM · #7 |
Originally posted by hankk:
I've heard that the sensor size does matter.
In theory, the f-stop is determined by the focal length / apreture. A 50mm lens on a DSLR that has an f-stop of 1 will have an aperature of 50mm, and will have an area defined as 50mm * pi that will let some given amount of light through. This amount depends on how much light is available (i.e. how bright the subject is)
The light you can use is the light that hits the sensor, so a full-frame sensor will have an area that gets 36mm x 24mm = 864 square mm worth of photons. A 1.6 crop factor sensor will get (IIRC) 24mm x 16mm = 384 square mm worth of photons, or about 44% of what the full frame sensor would.
Complicating this is the number of photosites (pixels). The larger the photosite, the greater the number of photons available. But naturally, for a given size sensor, increasing the number of photosites gives a smaller size for each photosite.
From what I've heard, noise does not increase proportionatly with sensor size, so the S/N ration (i.e. "good" photons to noise) is greater with larger photosites.
In general, digicams have much smaller sensors (and much smaller photosites) than DSLRs. |
Think of "correct exposure" as "thickness of light on the sensor". Your aperture area x pi ratio is accurate, but it describes the "thickness" of light that the lens projects, not the amount of it. If this were not the case, then a 50mm lens that covers a full 35mm frame area would be FASTER on a smaller sensor than on a full-frame sensor and this isn't so.
It's the same principle that allows you to measure rainfall by how much of it is collected by a small tube in your backyard; that more rain has fallen (in absolute quantities) on your backyard is irrelevant for determining howe much rain has fallen on a given AREA of the backyard.
By your reasoning a given 50mm lens mounted on a full frame camera would be slower than the same lens mounted on a 1.6 crop Canon 20D, and this isn't true.
Robt.
Message edited by author 2005-09-25 21:57:37.
|
|
|
|
09/25/2005 10:02:18 PM · #8 |
| wasting money on a coolpix? i resent that |
|
|
|
09/26/2005 02:25:03 PM · #9 |
Originally posted by bear_music:
[quote=hankk]
Think of "correct exposure" as "thickness of light on the sensor". Your aperture area x pi ratio is accurate, but it describes the "thickness" of light that the lens projects, not the amount of it. If this were not the case, then a 50mm lens that covers a full 35mm frame area would be FASTER on a smaller sensor than on a full-frame sensor and this isn't so.
It's the same principle that allows you to measure rainfall by how much of it is collected by a small tube in your backyard; that more rain has fallen (in absolute quantities) on your backyard is irrelevant for determining howe much rain has fallen on a given AREA of the backyard.
By your reasoning a given 50mm lens mounted on a full frame camera would be slower than the same lens mounted on a 1.6 crop Canon 20D, and this isn't true.
Robt. |
Ok, I've read your post a couple of times and think I understand what you're saying. I can see that I clearly wasn't clear enough. Seems what you read wasn't what I meant, probably because what I wrote wasn't exactly what I meant. You are correct in that the f/stop number is a good measure of how much light hits the focal plane per square mm.
I should have pointed out that the number photons per square mm would be constant regardless of the size of the sensor for any given lens at a given f/stop, but the number of photons per pixel would vary depending on the size of the pixel. But this just proves that ISO maters, and larger sensors (or larger photosites) will have higher max ISO settings, and you referenced this in your first post.
As far as the full frame vs 1.6 crop factor goes, if you had a 1/1.6 (0.625) extender that would get the full image circle to shrink to the size of the 1.6 crop factor sensor, you would get more light on the sensor (just as a 2x extender makes the image dimmer). But based on the sizes of the Canon 1.4x and 2x extenders, you'd have to build it with a negative length :-)
Note that one other advantage from a larger sensor is that larger sensors may use a higher percent of the sensor for the photosite. As the photosite gets larger, the seperation from other sensors will remain constant. (I'm assuming that the seperation between sensors is generally the same within each generation of sensors, regardless of sensor or photosite size because the interconnecting circuit area is based on existing CMOS design rules.) Thus, if you have an 8MP sensor on a DSLR that is 36x24mm, and an 8MP sensor on a digicam that is half that size, the 36x24mm sensor will have proportionally more area available for photosites becasue the circut area will be the same on both sensors.
(See //web.canon.jp/Imaging/cmos/technology-e/light_gathering.html for a picture of old vs new technology).
Message edited by author 2005-09-26 22:39:00. |
|
|
|
Current Server Time: 08/13/2025 10:40:33 PM  |
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 08/13/2025 10:40:33 PM EDT.
|
