Quote:
Originally Posted by Brian Petersen
Barry,
Great info. Is there anything that we can pull from the current cameras out in regards to the higher res/lower sensitivity issue. I mean, how does the XL2 compare to the XL H1 in that regard.
|
Well, there's the theoretical, and there's the actual. With the XL2/XLH1, the theoreticals would say that you're looking at likely having substantially reduced dynamic range on the XLH1, and reduced sensitivity, in exchange for having so many more pixels and such higher resolution. However, that would demand that all other things be equal, and they rarely are. If the XLH1's CCDs are using some sort of Super-HAD technology or something, perhaps they can perform even better than we would otherwise expect.
Quote:
Or the Sony FX1 to the PD-150 (don't know if there are more appropriate cameras to compare).
|
The FX1 is about two stops less sensitive than the PD150.
Quote:
Don't we have some cameras out that are the same chip size, but one is SD and the other HD. Is there substantial loss in light sensitivity? Noise?
|
Well, yes, but there's very few apples-to-apples comparisons. JVC's HD1, for example, is a 1/3" single-chip, and it exhibits very very low sensitivity, low dynamic range, and excessive chroma noise. But the FX1 is a 1/3" 3-chip camera and it has comparatively better sensitivity, better dynamic range, and very very low noise. So you can't just do a direct comparison and be able to draw any reliable conclusions just off of the theory behind it.
In practice, the FX1 has lower sensitivity and lower observable latitude than a PD170 does. However, the FX1 is much cleaner in noise performance. Noise and sensitivity go together hand in hand -- the more sensitive, the more noise, so as the engineers tweak for noise performance they also have to give up sensitivity.
In general, no blanket statements can be made as to predicting what actual performance will be, unless all other things are kept equal (which they never are). So you could say that a 1/3" megapixel CCD will be noisier with lower sensitivity and lower dynamic range than the exact same CCD if that CCD was produced in a 350,000-pixel version. But that argument only holds up in the lab because in the real world, that case doesn't exist. You have to judge the final output of each on its own merits, because there are so many interrelated ways that the engineers can tweak performance that a quick glance at a spec sheet just doesn't let you accurately predict what the actual performance is. I think it's safe to say we were all pretty much blown away by the low-noise performance of the FX1 and didn't expect that at all, given the chip's specs. But the lower sensitivity is a side effect of that.
And technology marches on. The PD170's sensors were first released on the market around six years ago, in the first VX2000. The FX1's have been on the market only a year. Lots can change in that amount of time!
So, basically -- we just have to wait and see.