PDA

View Full Version : Noise with 300d




SpAtZ
Feb 11, 2006, 04:33 PM
My 300d has continued to give me noise issues even when shooting jpegs. The shot that I have attached was at 400 ISO. Is this normal or a well known problem? Is there any way to resolve the issue?

http://homepage.mac.com/cspatz/.cv/cspatz/Sites/.Public/IMG_5129.JPG-zip.zip



penguinman
Feb 11, 2006, 04:47 PM
My 300d has continued to give me noise issues even when shooting jpegs. The shot that I have attached was at 400 ISO. Is this normal or a well known problem? Is there any way to resolve the issue?

http://homepage.mac.com/cspatz/.cv/cspatz/Sites/.Public/IMG_5129.JPG-zip.zip


looks fine to me.. 400 is a pretty noisy ISO still.. shoot that picture at 100ISO and then try to complain.. you're "pushing" the photo at 400ISO in order to freeze the image.. the reason u cant use 100 ISO is bc ur shutter speed will be too long to freeze the motion.. so you're kinda.. "pushing" it.. now if that was a bright day and same ISO you'd have less noise

iGary
Feb 11, 2006, 04:48 PM
What noise?

It is just underexposed.

Considered using flash?

Nice pic, though.

I loves me some Carolina/Black-Capped Chickadees.

rasp
Feb 11, 2006, 04:53 PM
That looks pretty clean to me, maybe I'm less picky.

Here is an interesting article on the subject of noise and high iso

http://www.zenadsl5251.zen.co.uk/photos/highiso.html

mkrishnan
Feb 11, 2006, 04:59 PM
I also think that's within normal limits for noise at ISO 400... Sorry. :(

But what do you mean by "even when shooting jpegs?" JPG will be noisier than RAW, in the sense of artifacting. If you're concerned about noise, then you want to shoot in RAW...not JPG. Although you'll get the biggest gain from tweaking your exposure and finding ways to be able to use a better ISO.

form
Feb 11, 2006, 05:16 PM
I can't view the image. The zip file won't read properly.

That said, are you brightening the image in photoshop? That will drastically increase the visibility of noise because of the uniform result of adjusting brightness on a shot already taken. If you take the same shot at ISO800 instead of ISO400 and you don't have to brighten it afterward, you will get less visible noise overall.

Also, if you use sharpening or increase the saturation, noise will become more visible as well.

One other thing: What quality JPG are you shooting? I'm assuming it's fine because of the zip file's size. If so, then compression artifacts are a moot point.

SpAtZ
Feb 11, 2006, 06:12 PM
I can't view the image. The zip file won't read properly.

That said, are you brightening the image in photoshop? That will drastically increase the visibility of noise because of the uniform result of adjusting brightness on a shot already taken. If you take the same shot at ISO800 instead of ISO400 and you don't have to brighten it afterward, you will get less visible noise overall.

Also, if you use sharpening or increase the saturation, noise will become more visible as well.

One other thing: What quality JPG are you shooting? I'm assuming it's fine because of the zip file's size. If so, then compression artifacts are a moot point.

That may be it, plus I had a large crop. I heard that RAW has more noise though?

form
Feb 11, 2006, 06:27 PM
RAW is typically unprocessed, or less processed, than JPG. Noise isn't much of an issue at ISO400 with a Canon dSLR. I don't believe RAW has any more or less noise than Fine quality JPG, but it does have more versatility as far as post-processing.

jared_kipe
Feb 11, 2006, 07:52 PM
I think you misunderstood me. What I meant is that I have found iPhoto to do a poor job with RAW files from my 300D. I thought you might be getting the same problem I have seen where the files seem overly noisy coming out of iPhoto, but they really are not that noisy.

Basically this is what I saw and have vowed never to use ISO 400 or above unless I really really needed to.

mkrishnan
Feb 11, 2006, 09:22 PM
RAW is typically unprocessed, or less processed, than JPG. Noise isn't much of an issue at ISO400 with a Canon dSLR. I don't believe RAW has any more or less noise than Fine quality JPG, but it does have more versatility as far as post-processing.

Hmmm, what I meant to say is that JPGs can have color noise associated with compression, that looks like sensor noise. Fine quality should be pretty good, though. And another thing that can cause the appearance of these sorts of problems is the quality of glass.... Pictures from my Canon kit lens are noticeably noisier than pictures taken under similar conditions with my 50mm f/1.4.

form
Feb 11, 2006, 10:40 PM
Hmmm, what I meant to say is that JPGs can have color noise associated with compression, that looks like sensor noise. Fine quality should be pretty good, though. And another thing that can cause the appearance of these sorts of problems is the quality of glass.... Pictures from my Canon kit lens are noticeably noisier than pictures taken under similar conditions with my 50mm f/1.4.

Canon JPG compression is exceptional, and you'd be hard pressed to notice any such extra noise in them. Same goes for higher end Nikon cameras.

Noise CAUSED by a lens...That might be worth consideration...er, if one actually had anything at all to do with the other. Please tell me, if you put more expensive tires on your car, can you also see better through the windshield?

Realistically, the claimed increase in noise with the kit lens can be attributed to the extra post-processing or in-camera processes, such as boosting saturation, contrast, sharpening, etc., which might be required with a cheaper lens that doesn't have the same capacity for capturing light, detail, contrast and color as the more expensive and highly regarded 50mm f/1.4. However, the actual lens itself can't possibly have anything at all to do with noise, in the purest sense; it simply picks up whatever's out there, and nothing more.

And just in case you would suggest such a thing, the AF/Zooming/Metering circuitry within the lens has nothing to do with your claimed extra noise, either.

Adobe Camera Raw is very popular for RAW importing, and according to what I've read it's one of the very best plugins/apps available for that purpose.

mkrishnan
Feb 11, 2006, 10:55 PM
Noise CAUSED by a lens...That might be worth consideration...er, if one actually had anything at all to do with the other. Please tell me, if you put more expensive tires on your car, can you also see better through the windshield?

Heh...thanks for the lesson in engineering. ;)

jared_kipe
Feb 11, 2006, 11:38 PM
Heh...thanks for the lesson in engineering. ;)
That would be very weird. You sure you're just not confusing real noise with noise that comes up when you sharpen an inferior shot? I have never heard of noise being caused by the lens, and do not believe it exists. From a physics standpoint, that is.

mkrishnan
Feb 12, 2006, 10:15 AM
That would be very weird. You sure you're just not confusing real noise with noise that comes up when you sharpen an inferior shot? I have never heard of noise being caused by the lens, and do not believe it exists. From a physics standpoint, that is.

I meant noise caused by things like chromatic aberration. I might have a loser concept of noise than you, though. I definitely don't mean noise in the sense of cosmic rays hitting the sensor and over-exposing it. Of course that has little to do with optics. Basically only ISO and shutter time.

But there was little / none of this in the original picture, to begin with.

jared_kipe
Feb 12, 2006, 10:23 AM
I meant noise caused by things like chromatic aberration. I might have a loser concept of noise than you, though. I definitely don't mean noise in the sense of cosmic rays hitting the sensor and over-exposing it. Of course that has little to do with optics. Basically only ISO and shutter time.

But there was little / none of this in the original picture, to begin with.
Ahh, I see. Just to put your fears to rest ;), cosmic rays or any relatively high energy photons, are not actually a problem. Reason being glass has such an awesome low pass cut off right at the UV range.

mkrishnan
Feb 12, 2006, 11:04 AM
Ahh, I see. Just to put your fears to rest ;), cosmic rays or any relatively high energy photons, are not actually a problem. Reason being glass has such an awesome low pass cut off right at the UV range.

Actually, it's true that glass stops near UV... when you get to high energies, though, the photon passes through the glass, through the sensor, through the back of the camera, and through you. But it can leave a little energy in its wake and generate a blip on the image. If you take an extended image with the shutter closed, or with the cap on, if that's not possible, you'll see some bright pixels...there's lower level electrical noise, but the bright pixels are cosmic rays. The higher you turn up the ISO, the lower the excitation threshold is, and you get more of these bright pixels in a shorter period of time. That's the way I understand it, at least. When I was working in lasers...this kind of thing occasionally became a real nuisance. :)

But this is waaaay OT. We all agree that there really isn't noise, per se, in this photo, and there are other issues which have to do with optimizing the light into the camera, that would make a picture like this better. :)

jared_kipe
Feb 12, 2006, 12:27 PM
Actually, it's true that glass stops near UV... when you get to high energies, though, the photon passes through the glass, through the sensor, through the back of the camera, and through you. But it can leave a little energy in its wake and generate a blip on the image. If you take an extended image with the shutter closed, or with the cap on, if that's not possible, you'll see some bright pixels...there's lower level electrical noise, but the bright pixels are cosmic rays. The higher you turn up the ISO, the lower the excitation threshold is, and you get more of these bright pixels in a shorter period of time. That's the way I understand it, at least. When I was working in lasers...this kind of thing occasionally became a real nuisance. :)

But this is waaaay OT. We all agree that there really isn't noise, per se, in this photo, and there are other issues which have to do with optimizing the light into the camera, that would make a picture like this better. :)
This is not true. The hot pixels you see are pixels in the camera, with the lens cap on, are pixels that are always on, or much much noisier than the rest of the pixels.

Borrowing from my nuclear physics lab lectures... The quantum efficiency curve for glass cuts photons with wavelength under 350nm, which just so happens to be where UV spectrum begins. This is why we use Photomultiplier tubes in nuclear research that have more exotic casings than plain glass. http://silicon.phys.washington.edu/lubatti/Phy433/Lectures/Spring_05/lecture_3-05.pdf All materials absorb and pass through a band of photonic energies. It is an important feature of glass that it should pass through the visible specturm, and at the same time that puts a great limit on the other spectrum it can pass (just so happens it is very close to the visible spectrum).

As for the "cosmic rays".
I think we can rule out neutrinos, since they don't interact with mater very much at all.
Muons are one of the only particles that can make it down from the atmosphere. They will penetrate most matter, but will eventually decay, or get caught and then decay into electrons. But I can assure you, the odds that they would be stopped in your camera are low, though one does reach the surface at about 1 per cm per min. ( I think from memory)
Much of the so called "cosmic rays" are actually just elements from super nova's and obviously singular atoms or groupings of atoms are never gong to get inside through your camera.

SpAtZ
Feb 12, 2006, 12:53 PM
I am using a 70-300 DO. Like I said before I had a large crop and edited the picture alot, so thats probalay explains that.

Could the DO cause more noise?

jared_kipe
Feb 12, 2006, 01:00 PM
I am using a 70-300 DO. Like I said before I had a large crop and edited the picture alot, so thats probalay explains that.

Could the DO cause more noise?
nah, out of curiosity, why do you have that lens?

SpAtZ
Feb 12, 2006, 04:26 PM
nah, out of curiosity, why do you have that lens?

I heard some good reviews and I got a very very good price on it because of the Canon 1-2-3 rebate. It is also very light compared to those L lenses, which people say are similar quality.

jared_kipe
Feb 12, 2006, 05:41 PM
I heard some good reviews and I got a very very good price on it because of the Canon 1-2-3 rebate. It is also very light compared to those L lenses, which people say are similar quality.
How do you like it?
I've heard mixed reviews about it. The problem with it isn't its diffraction, but rather the choice of a fresnel lens; which seem to give some mixed results.

Just that given the price, I'd probably buy the 70-300mm USM IS, which costs about 1/2 the price and turns out similar features and performance.

ibilly
Feb 12, 2006, 10:02 PM
I am using a 70-300 DO. Like I said before I had a large crop and edited the picture alot, so thats probalay explains that.

Could the DO cause more noise?

I've seen some pics where it produces funkiness in thinks like high contrast bokeh and fine oof detail, like a brick wall

Then again, that image was very nice for a highly procesed 400 shot (from any consumer or semi-consumer digital camera)

SpAtZ
Feb 12, 2006, 10:56 PM
I've seen some pics where it produces funkiness in thinks like high contrast bokeh and fine oof detail, like a brick wall

Then again, that image was very nice for a highly procesed 400 shot (from any consumer or semi-consumer digital camera)

Yes, when shooting at the sun. You get used to it. other than that it is a very good lens. The IS is very fast.

Chip NoVaMac
Feb 12, 2006, 11:05 PM
Sorry, but what noise? Try this with 400 ISO film and tell us which you prefer. :)

If you want to try and see if you can smooth the image out more, try something like Noise Ninja.

ChrisA
Feb 12, 2006, 11:33 PM
My 300d has continued to give me noise issues even when shooting jpegs. The shot that I have attached was at 400 ISO. Is this normal or a well known problem? Is there any way to resolve the issue?]

I don't see a lot of noise. The image is a a bit dark and a little soft. Even the hightlight in the bird's eye is grey. Did you change the levels in an editor? Trying the liten the image will add quatiitazaton noise

If you want to measure noise shot a grey card in very even lightting and plot a histogram. Try grossly under exposing the grey card and then us the editor to liten the image t 18% grey. A grey card it the best subject if you are looking for nise.