Jump to content

Controlling Noise - lens or body who really is responsible?


gunjankv

Recommended Posts

<p>Controlling Noise is the smart work done by the intelligence in camera body? Or is it because of the quality of glass which does not let noise signal to enter<br>

There is something called Signal to Noise ratio.. which makes perfect sense and I guess it is directly related to quality of image w.r.t noise. This ratio would be in favor (less noise) if glass is good. Body would not play any role here.<br>

Does this u/s make sense? Are there any other factor worth considering w.r.t Noise?<br>

I understand that high ISO means higher noise.</p>

Link to comment
Share on other sites

<p>FWIW, All of the above:<br>

- Lens distortion are noise<br>

- Lens or body glare are noise<br>

- Shutter and mirror vibration which resulted in motion blur are noise<br>

- Sensor related noise are noise (dark current, anti-alias, amp noise and etc.)<br>

- JPEG conversion software noise (bayer/chroma luma interpalation)<br>

Everything can contribution to noise. </p>

 

Link to comment
Share on other sites

<p>Bob and the rest are absolutely on the point.<br>

Matthew is particularly right in that noise will show up at a 100% view for someone looking at the image with their nose touching the screen, so to speak, for images where it would likely not be visible at all in any print you're likely to make of it.</p>

<p>Noise may be slightly less felicitous than "grain" was back in film days, but it is often better to accept a little noise when the alternative is no picture at all.</p>

 

Link to comment
Share on other sites

<p>Use a lower ISO setting, and expose your shots in a way that does not require you to increase the brightness of the image in your computer, and you will get less noise. The camera has a lot to do with it too. A Nikon D700 will capture images with much less noise than a Nikon D7000, because the D700 has a full-frame 12 megapixel sensor, and the D7000 has a smaller sensor with more (smaller) photo-sites. This means that the D700 will not be as susceptible to photonic noise (related to Shot noise - see Photon noise at Wikipedia for an explanation).<br>

-<br>

You may want to read this stuff:<br>

http://www.luminous-landscape.com/essays/sensor-design.shtml<br>

http://en.wikipedia.org/wiki/Image_noise#In_digital_cameras<br>

http://en.wikipedia.org/wiki/Photon_noise<br>

http://www.imageskill.com/articles/HTML/camera_noise.html<br>

-<br>

Don't listen to people who tell you the lens is responsible for noise. It's the aperture that is responsible for the amount of light and the depth of field (all else being equal). Yes, some lenses have the capability of opening the aperture more, but at a cost (in sharpness and depth of field). This is not to say that the smaller the aperture the sharper the photo, but usually, if you set your aperture to about the middle of the lens range, and you set your ISO to 100 or 200, then you will get the clearest pictures, so long as the camera and subject are not moving. It is all compromises, of course. You will understand as you experiment more. The most important thing to remember is that no matter what you do, you will never get a camera and lens that is perfect for all situations, no matter how much you spend. You can generally expect newer digital SLR camera bodies to allow you to capture photos with less noise than older cameras. You can expect to spend more money for better cameras. You can expect to spend more money for better lenses. Of course, depending on your needs, you may be able to make decisions which can save you money, such as a decision to buy fixed focal length lenses with f1.8 or f2 apertures, rather than zoom lenses with f2.8 apertures. Of course, for a given quality, zooms are often less expensive than a comparable set of fixed focal length lenses, because one lens can pretty much do the job of two or three lenses.<br>

-<br>

I digress. One more thing to keep in mind is that technology is changing. Full frame sensors do not necessarily win in the war against noise. My friend's Canon 7 D, which has a 1.6 crop factor sensor (about half the size of a full-frame sensor) produces less noise than my Canon 5 D. This is because it has newer technology. The sensor in my 5 D is an older design, which means it probably has smaller photo-sites (the little solar cells that make up a sensor's light-catching area) than the sensor in the 7 D, even though the 7 D sensors is smaller and has more photo-sites. Weird, huh? That's technological development for you!<br>

-<br>

The new Sigma SD1 is a special camera with a special sensor that also seems to be a very "low noise" camera. Check it out, if you have a budget of thousands, rather than hundreds.<br>

-<br>

If you want to compare cameras, I suggest checking out Imaging-Resource.com - here's a review of the latest Nikon - the D5100 - http://www.imaging-resource.com/PRODS/D5100/D5100A.HTM<br>

-<br>

Good luck Gunjan!</p>

Link to comment
Share on other sites

<p>Look at it from a systems approach (somewhat).<br>

If the lens were intrinsically responsible for S/N ratio problems, then moving the lens from a body like a Nikon D100 to a Nikon D3 would show no improvement in image noise.<br>

ASA 800 and above on the D100 was horrendous.</p>

Link to comment
Share on other sites

<p>Noise is most of the time a grossly over rated fad of digital photographers ...I never see it unless I'm pushing the process like applying 500%USM twice to a photo. And I used to be very much against grain ... funny that ... maybe it comes with age and maturity. Looking at the photograph's subject matter instead of its technical content.<br>

Just because one has features in one's camera doesn't mean you have to use them ... I shoot at 100ISO.</p>

Link to comment
Share on other sites

<p>Lenses have nothing to do with noise. Distortion is a different phenomenon.</p>

<p>The Signal To Noise ratio that the OP mentioned is the basis of the problem and works in two ways.</p>

<p>Sensors are sensitive enough to pick up electrons generated by the electronic components of the sensor itself. However, for a properly exposed image at a low ISO value, the light hitting the sensor is strong enough to overpower the noise. The sensor records mostly light and very few stray electrons. The Signal (light) to Noise (stray elections) ratio is HIGH: lots of light/very few stray electrons.</p>

<p>If the image is underexposed, the light hitting the sensor is much weaker. The stray electrons now start to show up in the photo because the light was not strong enough to cover them up. The Signal (light) to Noise (stray electrons) ratio is LOWER: less light/the same number of stray electrons.</p>

<p>In order to capture light at higher ISO values (800, 1600, 3200, etc.), the sensor amplifies the light. The problem is that it amplifies the stray electrons at the same time. Therefore, photographs taken at high ISO values exhibit more noise than those taken at lower ISO values. The Signal (light) to Noise (stray electrons) ratio is LOW: amplified low light / amplified stray electrons.</p>

Link to comment
Share on other sites

  • 3 weeks later...

<p>It's not just stray electrons Dan, though this explanation is enough for people to understand the concepts of why noise is higher at high ISO settings.<br>

-<br>

There is something called photonic noise, which is about the distribution of photons that occurs unevenly in nature, just like rain. Ever watch rain landing on something? It doesn't fall evenly, does it? Photons lad on a sensor the same way. When it's pouring, there is no sense that there is an uneven distribution, but when there is very little (rain drops or photons), there is a very uneven distribution. On a lot of small photo-sites (sensor spots) there might be 10 falling on one and 15 falling on another and 5 falling on another. That translates into big differences from one pixel to another. Remember that in Bayer pattern sensor cameras (normal digital cameras), only one color of light is let into each sensor. The number of photons is effectively reduced, because most of the light is filtered out by the micro-lenses on the sensor. On a Foveon sensor, like in the Sigma cameras, there is no "filtering out of photons" by micro-lenses, because the Foveon sensor has no micro-lenses. Still, there is photonic noise anyway, though it is not magnified by micro-lenses filtering out most of the photons. (And that is BEFORE the magnification which takes place in the processes after the initial conversion of photons to electrons.)<br>

-<br>

Gunjan, remember that newer sensors have newer solar panel technology. As solar panels improve in efficiency, because of new technologies, there is a similar type of improvement in the efficiency of sensors in digital cameras. Of course, there is probably not a direct correlation. In fact, the two types of solar panels (yes, basically a sensor is a type of solar panel) may be very different. I'm sure there is SOME leakage of technology from one industry to the other (sensors to high-tech solar panels and vice-versa). Since there is gradual improvement in the photon conversion capabilities of sensors, you will see better and better noise-related image quality, even as the number of megapixels increases. Also remember that the pixels keep getting smaller, as there are more of them, and that means that the noise becomes less and less visible, as it becomes more like a smooth film of "off-color" dots, rather than a bunch of spots on the image.</p>

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...