Pixel density - can the playing field be leveled???

Started Jun 6, 2009 | Discussions thread
Daniel Browning Senior Member • Posts: 1,058
[6/6] Myth busted: small pixels bad, 4 legs good - part 6

[Part 6 out of 6.]

On the other hand we have noise, dynamic range, and QE. For these, G. Agranov's paper showed no difference from 5.6 microns to 1.7 microns in typical situations, which agrees with my own reading of available sensor data and image analysis. Now data on 1.4 micron sensors is available and they too appear to have performance that scaled in proportion with their size. I chalk this up to the ingenuity of sensor designers. I think we're getting close, though, so perhaps 1.2 microns will finally be the point at which diminishing returns kick in for typical (photon shot noise limited) situations.

One important consideration is performance in low light across all types of cameras. There is a certain sensor design that reduces read noise to very low levels, but it only happens to occur in some sensors with big pixels (4.7+ microns), analog gain, and is accompanied by high read noise at low gain (compared to small pixels). Not all big pixels have this design, and not all big pixels with analog gain have this characteristic either. Yet for every decrease in pixel size, this design continues to have effect. Since pixel size in large sensors is limited by the processing power, the pixel size is much larger than it could be. This limitation is moved yearly thanks to Moore's law, but we can't know where it will finally end until it does. So until there is any evidence of that happening, it would be premature to worry that the benefit will disappear if sensor designers continue shrinking pixels.

Another consideration is angle of response. Pixels of all sizes tend to have lower response from oblique angles, such as an ultra wide angle f/1.4 lens, but smaller pixels have even more difficulty because it's hard to scaling the depth (z dimension) in proportion with the area (x and y, width and height).

There is a strong correlation between all the performance metrics and sensor size; such that larger sensors (with a proportionately larger lens and thinner DOF) have much improved image quality in low light.

Another interesting consideration is that, for one, Canon's DSLR sensitivity (QE) improved only 5% in the last 1 year, 30% in the last 3 years, and 115% in the last 6 years. [10D: 0.042 photo-electrons per 12-bit ADU per square micron. 50D: 0.100 e- adu/um^2.] The dynamic range improvement on Canon was a bit less a single stop in 4.5 years. [Area-referred base ISO read noise 20D: 1.81, 40D: 1.15, 50D: 0.97. Area referred full well stayed about the same.]

Going back to the main point again: performance is correlated with the level of detail. DPR editors have previously stated:

Ultimately the main dispute is between whether you measure
noise at the 1:1 level or at an arbitrary image size. dpreview
has always worked at the pixel level because it doesn't make
any assumptions for what you're going to do with the images.

I think that is an incorrect description of the dispute. On one side the level of detail is ignored, and noise is measured at an arbitrary size image size (1:1 AKA 100% crop); the other advocates any method that accounts for the level of detail, such as comparing noise at the same image size (or fixed spatial frequency).

Imagine if the same were said about a car reviews:

Ultimately the main dispute is whether you measure noise at the
100% speed level or at an arbitrary speed. dpreview has always
worked at 100% speed because it doesn't make any assumptions
for what you're going to do with the car.

100% speed on a sports car might be 200 MPH, while a commuter sedan may top out at 100 MPH. It would be wrong to say that the sports car is noisier than the sedan, because you haven't even driven them at the same speed. At 100 MPH, the sports car might be quieter than the sedan.

Analogies (especially cars) can go awry quickly, but after all the attempts at explaining this concept, anything is worth a shot.

Overall, DPR is a great site and highly informative; but there are some important flaws, and the DPR war on pixel density is one of them. I'm disappointed to see that Bob Newman has been banned from DPR. I hope I wont be the next one up on the chopping block.

Four legs good, two legs better.

Kind regards,

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow