Camera Labs

I have noticed that often, the Pixel Density of a camera(sensor)
is said to be 35mp/sq. cm or 10mp/sq.cm and so on.
That seems to me that there are various sizes for the pixels
or light gathering units, as well as the sensor size itself,
particular to that camera.
How is this supposed to give one an idea of the quality
of these pixels and corresponding images created?
Is it a known fact that the more MP /sq cm is a "good" thing
or is it a matter of the quality of those pixels within that sq.cm which counts? If the latter, how does one tell
if a 10MP/sq.cm has a "better" image quality in it's rendition than
one with a higher count?
Does it really matter in the total scheme of things anyway?
Otto

_________________
Noise to one is music to another...

For a good tutorial on how sensor size and pixel size affects photos, read http://www.cambridgeincolour.com/tutori ... r-size.htm

A smaller sensor with a higher pixel density will be much more prone to noise. There's a reason why the Nikon D700 full frame camera has "only" 12.1 mega pixels.

_________________
PhilipGoh.com

As a general guide, the bigger the pixels, the better the light gathering capability of the individual pixels and the dynamic range. There are other factors such as the gap between photosites (smaller gaps mean bigger photosites for a given sensor area), micro lens design (focus more light onto the photosites themselves and reduce the amount of "wasted" light), sensor cooling and the analog to digital converters in the sensor or on the processor. These aren't so easily quantified and are usually not mentioned for this reason but they are important as well.

More MP/cm^2 can be a good thing if your priority is image size and resolution over noise and dynamic range but some people are willing to sacrifice some resolution for better noise control and a larger DR and for these, a lower MP/cm^2 is better.

All other things being equal, the sensor with larger photosites will always be better but whether it's important in the grand scheme of things can only be determined by you and what you expect from your camera.

It's easier to let Gordon figure out theses things!

_________________
Graham
Flickr

Don't believe what you read. Try it yourself. In average daylight, take a photo of a scene with trees in the background, in focus, about 50-75 yards away. Then look at the results at the pixel level and the image level. i.e. with the screen viewing at 1:1, then with the whole image sized to be viewable at once. Irfanview, the best photo viewer (and free), makes this a one-key operation.

As I have said here, pixel density is basically a trade-off between good light effective resolving power and low-light effective(!) resolving power. It depends on your personal needs/preferences which suits you best.

Ben
_________________
When in doubt..... Press the shutter.

This is a misleading statement. What suits me is a camera exactly like the LX3 (note: exactly) but with 15 mp. Now, I have plenty of money, so where can I buy it?

Ah! That's the whole point. There IS no camera exactly like the LX3 with 15 Mp. Because if it would have the same pixel density then the sensor would be larger and the lens would have to be bigger etc.

If you would want an LX3 with the same sensor size and lens size but with higher pixel density it would behave differently. Yes you would gain good light resolving power but at the expense of low light resolving power. You can't have your cake and eat it. It's either/or. The LX3 was designed to fill the needs of people 'whose needs it fills'. It's as blatantly simple as that. If it doesn't fill your needs that's bad luck. Let's hope a camera manufacturer will make one soon that does. Because that's one of the points of the LX3 too. To diversify the spectrum of cameras on offer. Let's hope the trend will continue and more such more 'specialised' cameras will be made to suit the needs of their respective target audiences.

Ben
_________________
When in doubt..... Press the shutter.

Not exactly. The point, which Panasonic broadcast loudly, was to lessen pixel density, which in turn would significantly improve quality. And the result of their efforts, the Lx3, is a failure.

Lowering pixel count lowers the signal-to-noise ratio, which makes the quality worse. Those persons who expect quality to improve due to less density will be disappointed, because the decreased density will not compensate equally for the decrease in pixel count. The only way to get better quality is to get a much larger sensor (dslr-size etc.), or increase the pixel count.

This thread isn't about the pixel count, it's about the pixel density. The pixel density is what differentiates DSLR sized sensors and compact camera sensors. Lowering the pixel count for a given sensor area has the effect of reducing pixel density anyway so we're just going around in circles here.

Perhaps there might be some other reason why the Panasonic sensors are below par when it comes to noise? I don't think there is any sensor of Matsushita design that is known for it's noise control, not in their DSLRs, not in their compacts. The LX3 may not be the best performing compact at high ISOs but it is better than Panasonic's previous pixel-cramming attempts so what makes you think it is failure?

_________________
Graham
Flickr

Lowering the pixel count while retaining the same sensor size (i.e. lowering the pixel density) increases the SNR for each pixel. A larger pixel collects more photons, thus providing a stronger signal. This is why full frame DSLRs with a larger sensor and correspondingly lower pixel density always provide better colour reproduction and low light performance.

Cramming that many pixels onto a compact sized sensor drastically reduces the SNR of each pixel. That would explain why you resize your photo from 15 MP to 5 MP to get acceptable results.

_________________
PhilipGoh.com

What's the point of having a super high resolution camera when half the pixels produce output you cannot use?

_________________
PhilipGoh.com

From the wide "variations upon the theme"
in the replies and answers, it seems to me that
Pixel Density in itself is not entirely what image quality is dependent on.
Correct?
There are other factors such as the lens, the image processing
engine (NR) and the combined results of all of the above
in relation to resolving power. Also how big the intended print
is going to be.
So my initial question is still quite "unresolved"...which is to say,
WHY bother to publish a camera's Pixel Density as a number
when this number does not necessarily reflect
the Image quality of your pictures...(at least it might only partially
tell you something)?
What I am trying to infer is that, given the complexity of the equation,
is it not slightly misleading from the marketing point of view,
to infer that a high Pixel Density goes to make a "better" sensor and therefore better camera?
So a consumer is led to believe 35MP/sq.cm is of course
BETTER than a sensor which is rated at 5MP/sq cm...joining the MP race.
This might be another Red Herring which helps to confound the buyer
in the interest of generating sales/perceived disireability of a product.
I can see I might never fully understand this.
For the moment, it seems that;

Less Density= better noise control=less resolving power=not so good for BIG prints.
More Density=potentially more noisy=more resolving power=better for BIG prints.

BUT the ball game changes the moment the sensor itself is made
say 50% larger as in the case of the DP1 and S2.
Otto

_________________
Noise to one is music to another...

Yes Otto, that is exactly right.

And we've been conveniently ignoring the fact that the sensor is only half the story. It doesn't matter if you have the greatest sensor in the world if it's paired with rubbish optics. There's a reason pro-grade lenses cost a fortune.

To illustrate this point, head on over to the sample comparison page at Imaging Resource. Compare the images shot by the 10 MP Sony A200 vs 14 MP Sony A350. Look at all the sample images, particularly the still life shots at the bottom of the list which are designed to compare resolving power of the cameras. Tell me that you can say for a fact that the 14 MP has more resolving power. If you didn't compare them side by side, you'd have trouble knowing which one was which.

Now compare the INB photos. Notice that at ISO 1600, the A200 produces less noise than the A350. Look at the dark areas of the shot, e.g. under the leaves, the black scale in her hand. See the speckles that exist on the A350 but not on the A200? Notice that yet again, the resolving power of both cameras are the same?

Obviously, the situation will change if you put a better lens on the cameras, say the Sony SAL16105. This lens has much greater resolution and I believe it outresolves the sensors of both cameras. Only with such a good lens will you will see a difference in the resolving powers of both cameras since the sensor is brought to its limit. You can look at another site (*cough* dee *cough* pee revue), to see the resolving power of these lenses and you'll see that the resolution of the SAL1870 (which is the lens used by both these cameras) comes no where near the resolution of the sensor.

Where does this leave us?
Is the sensor important?
Yes it is.

Does a bigger MP sensor automatically mean better resolving power?
No it does not. Judge the images with your own eyes.

Does a bigger MP sensor produce more noise?
Yes it does at high ISO speeds. Again look at the test images for yourself.

What else influences the IQ of your photos?
The lens! Use a crappy lens, take crappy photos.

As a corollary of the lens issue, upgrading from an old to new camera usually gives you an improvement in quality. Nevertheless, unless you can control all other factors like the guys at Imaging Resource did, it is impossible to conclude that a higher MP count resulted in a better looking image. It is much more likely that the image quality improvement came from a better lens that was present in the newer higher MP camera.

QED.

_________________
PhilipGoh.com