CMOS is less noisy because it consumes less power?

I've been doing some searching trying to figure out why DSLR CMOS sensors are less noisy at higher ISO's when compared to CCD sensors in compact cameras.

I found this in a thread on this forum:
"Still, there are some differences. By their nature CMOS use less power and, due to that, are less likely to have "cross talk" between photosites (which causes image "noise" at higher ISOs in particular)."

So they are less noisy because there is less crosstalk. I also read that the size of the sensor has something to do with it, and that pixels that are packed on a small sensor are more noisy as well, is that also due to 'crosstalk'?

I'm trying to figure out exactly why CMOS is less noisy than CCD, I appreciate your help.

Really, you have to look at two different things: the difference between the two sensor types in the same class of cameras, and the difference in how the large-sensor DSLR cameras perform noise-wise compared to the small-sensor compact camera performance, and what causes this difference.

Canon uses CMOS in its DSLRs, CCD in its compact. Nikon has, until the upcoming D3, used CCD. Up 'till now, it seems like Canon has had the edge in noise, and this is possibly partly due to the sensor type, but not completely.

The big gap in noise is evident between sensors that use small "wells" to collect light. This, rather than sensor type, is why compact cameras will always suffer from noise at above ISO 100 and why cameras like Canon's 5D and 1D series will outperform -- each pixel site collects more light information, and more good light data proportionately, because the wells that collect the light are larger than other DSLRs and several times larger than compact cameras.

So, it seems like, historically, the CMOS has an edge, but the difference with the compact camera is much more than the CCD model.

Beyond that, I can't address all the technical differences between CMOS and CCD. The broader internet will probably give you some in-depth discussions that you probably won't get on a user forum!

Ok if that theory is correct then the CMOS sensor used in cell phone camera should produce less noise.

I believe that Canon achieve the result w/ CMOS is a combination of energy efficiency of CMOS and their ability to manufcture their CMOS with intergrated AD circuit to handle it right of each photosite. Where as CCD convertions are done externally. I also believe that CMOS are live only moments before the shutter open which allows it be cooler then live CCD use in P&S cams since they don't have physical shutters.

THIS might be of interest.

lostdoggy wrote in post #4268316
Ok if that theory is correct then the CMOS sensor used in cell phone camera should produce less noise.

I believe that Canon achieve the result w/ CMOS is a combination of energy efficiency of CMOS and their ability to manufcture their CMOS with intergrated AD circuit to handle it right of each photosite. Where as CCD convertions are done externally. I also believe that CMOS are live only moments before the shutter open which allows it be cooler then live CCD use in P&S cams since they don't have physical shutters.

That may be, and may explain why Canon has the rep for lowest noise -- and why Nikon's vaunted D3 now has CMOS!

Still, the biggest factor in noise is the fact that there is innate noise in every sensor that is directly related to the size of the well and the number of photons collected. New technologies, as I understand, are working to make light collection more efficient, and the 1Ds Mk III might be using these new technologies, so we may be seeing this whole field of analysis being ratcheded up a notch.

Like I said, though, I haven't done a lot of study on CMOS vs CCD, just a general take-away that CMOS is more expensive but does a better job at less noise.

I beieve CMOS is less expensive then CCD for equiv size.

lostdoggy wrote in post #4268421
I beieve CMOS is less expensive then CCD for equiv size.

Is it? I just don't have all the facts. If that were true, why would compacts use CCD?

CMOS sensors has been around a long tim often used in inexpensive cameras and video cameras. Canon was able t over come CMOS's short coming by intergrating the AD and Amp directly on the CMOS. I don't remeber the exact technology behind it but this much I do remember.

lostdoggy wrote in post #4268449
CMOS sensors has been around a long tim often used in inexpensive cameras and video cameras. Canon was able t over come CMOS's short coming by intergrating the AD and Amp directly on the CMOS. I don't remeber the exact technology behind it but this much I do remember.

Well, Canon has, and continues to, put work in the bottom line of image quality, and the high performance at high ISO has been a big part of that. It'll be interesting to see how Nikon's D3 stacks up.

Yes, from what I have read, CMOS is less expensive to produce than CCD.

Okay, so if the "light wells" are smaller and they collect less light, how does that contribute to noise? Is it because the sensor doesn't collect accurate data about the light hitting it and spits out a faulty signal?

Also, I read that CCD and CMOS work by transforming photons to electrons, like a solar panel. So if you crank up the ISO, it means you are also cranking the power being fed to the sensor right? How does that make it more sensitive when it is collecting light to produce electricity in the first place?

I'm sorry if this is too technical, I'll be doing more searching on Google, I just thought someone here might know. If I find out why, I'll post it here for those that are interested.

pete.mod wrote in post #4271397
Yes, from what I have read, CMOS is less expensive to produce than CCD.

Okay, so if the "light wells" are smaller and they collect less light, how does that contribute to noise? Is it because the sensor doesn't collect accurate data about the light hitting it and spits out a faulty signal?

It's a physics/physical property that shows the amount of noise is inversely proportionate to the amount of light collected by a receptor. There is a good discussion of this, with examples, at this site.

When you turn up the ISO you are, in essence, amplifying the light collected at your given exposure to "look like" you got more light. It's not "real" sensitivity in the sense that film is more sensitive. It's electronic amplification. Noise becomes increasingly noticeable at higher ISOs because you are amping less and less "good" signal to produce an acceptable equivalent exposure: the less "good" light that the pixel sites collect the more noise comes up proportionately. You see the same thing if you have an image at any ISO with deep shadows and try to amp up the shadows in post-processing.

Check out the site I pointed you to. There is a lot of testing, analysing, discussing and referencing going on there!

Tonylong,
That basically explains what I was trying to figure out. I'll check out that website you mentioned as well.
Thanks again.

Just so you guys no...Nikon uses CMOS sensors in their D2 series also...CMOS isn't a first for nikon DSLR.

lostdoggy wrote in post #4268316
Ok if that theory is correct then the CMOS sensor used in cell phone camera should produce less noise.

lol, that' doesnt really make sense. the sensor inside a cell phone camera is about 1/10th the size, and the noise there comes from basically a built in digital zoom/crop. take a 1024x768 pic with your cell phone, then a 640x480 cell phone. the 1024res is just an enlarged 640 capture..that accounts for the additional noise and blurriness on cell phone pics...phones dont exactly have a DigicIII processor in them, lol

::: connor

Bukka wrote in post #4272891
Just so you guys no...Nikon uses CMOS sensors in their D2 series also...CMOS isn't a first for nikon DSLR.

canon has been using CMOS in video for about 2 years now. maybe DSLR will catch up.

::: connor