Actually, I've got a pretty decent handle on the physics of light, it appears you have little handle on what improving fill factor means in reality (Larger effective sensor sites at a given resolution and sensor size). And yes, bigger IS better. That's exactly why the new sensors have improved performance over the old ones, they have a larger percentage of the sensor area capturing light because fill factor has been improved and there's less wasted space on the sensor. So you're getting the 'bigger' part with the same sensor size, due to improved manufacturing capabilities.
And yes, these improvements apply to the FF cameras as well, thus the D3 with its performance advantage over the 5D at the same resolution and sensor size. Improved sensor production tech is the sort of thing that benefits most, if not all, sensor producers, and in this case us, as the improvements affect the efficiency of the sensors gathering light. Improving microlens designs also improves the light-gathering capability of the sensors, and microlenses are a fairly new addition to mainstream sensors, so there's likely some improvements still to be made there. Pentax's more film like image is entirely a product of the camera's processing, this isn't going to change when they come out with a camera based on some variation of the new Sony sensor (or somebody else's with the process improvements that allow a better fill factor). As to Oly, they're continuously one step behind in noise performance and I don't see that changing. They also don't have all that huge an advantage when it comes to size, only the E-410 is truly smaller than other current bodies and it achieves this at the cost of ergonomics. As to the D300, I think you're putting significantly too much into the idea that its performance comes primarily from improved processing. yes it has a new, better, faster processor. And yes, it most likely has improved noise reduction and this is most likely benefiting the sample images. But witness the 1DmIII, which also gains some noise performance via its new DiGiC III processors, and still turns in resolution numbers that are high for its MP count across the board, improved NR does not necessarily degrade image quality (Otherwise it wouldn't be 'improved'). There's a lot more to these new cameras than just better NR software. You're grasping at straws here to try and find reasons why the new generation sensors are not noticable improvements over the previous, 2+ year old, sensors that populate most of the current DSLR market outside of Oly/Panasonic (Who have finally matched the older APS-C 10MP sensors for noise performance). High ISO noise performance is an area where most of the market has improved while adding MP. Since there are hard, physical, limits on how much light will hit a sensor site of theoretical max size for the resolution and sensor size, this improvement indicates that we haven't hit the actual physical limits of sensor design. We probably have hit the current max for current sensor production technology, but that's not to say that the production technology won't improve. -Adam P. J. Alling wrote: > You really don't understand much about the physics of light. You can't > get something for nothing. But in a way that's besides the point. > Something else you don't get about bigger is always better weather > you're talking about film or some other sensor technology, bigger will > be better. No matter what your starting point! Pentax has been selling > their cameras on a better more film like and more natural representation > in the captured image. What will they sell on now. We're smaller and > cheaper. Didn't they try that before? Hasn't Olympus already staked that > territory out with their new line of cameras, and don't they already > have a built in advantage there? You seem to be impervious. A large > amount of the Nikon D300's improved performance is due to improved anti > noise algorithms. running on a faster more efficient processor. You > don't get that improvement for nothing, my guess is that it costs you in > real detail being replaced with some kind of phony detail If you don't > care about how real the image is it won't matter very much to you as > long as it looks good. > > Adam Maas wrote: >> P. J. Alling wrote: >> >>> The Nikon D300 of course first look on a magazine rack. I read the >>> article on paper, in a book store, you remember paper don't you? >>> >> Yeah, I remember paper. I read a few paper mags. I've yet to see one >> that does equipment reviews even a tenth as good as the major >> sites(DPReview et al)(heck, DPreviews D300 preview is far more extensive >> and in-depth than any mag's final review will be). And determining this >> from a 'first look' is bloody laughable, especially when it's at odds >> with what the manufacturer is stating about the new sensor (improved >> fill factor, more efficient microlenses) and the images posted online >> (which are for the most part in-camera JPEG's and thus inherently >> overprocessed with regards to noise reduction, but still show good >> resolution). Review hardware isn't going to be in peoples hands for a >> couple weeks (D300 ships in november), we'll see then if reality matches >> the samples we have now. It has in the case of the 40D and A700, the >> former matching its FF brethren(5D) for high ISO performance according >> to reputable sources (like LL) >> >> >> >>> Look simple math, assuming the sensor sites have no spacing between them >>> will tell you that the capture sites are too small already, to be as >>> effective as they have to be. >>> >> And there's your problem. You're assuming something that simply isn't >> the case, in fact the major improvement in sensor design from the crop >> of '05 to the new ones is that the spacing between sensor sites has >> decreased in favour of larger actual sensor sites and smaller gaps. >> Therefore the new sensors can capture more light per site at the same >> resolution and sensor size. Instant potential improvement in noise >> performance. >> >> > I'm not going to cite my sources, I'd have >> >>> to consult books I've currently got in storage. Rod Studdart did a >>> masterful job of explaining this to the list a few years ago anyway. >>> It's should be in the archives. It doubt I explain it as clearly. >>> Telling me that they've re-designed the sensors to use more space for >>> the individual sites only makes it worse. I've already been assuming >>> they took all the space available. >>> >> The space available has changed. Previous sensor manufacturing tech >> required larger gaps between sensor sites, (fill factor). They've gotten >> closer to the theoretical max. Once again, reality proves theory wrong, >> because theory makes certain assumtions (like your assuming that fill >> factor was theoretical maximum, rather than the actual lower number, >> then mapping current performance on to an idealized theoretical model). >> >> More than likely, fill factor will incrementally improve in the future, >> leading to more light-efficient sensors, microlens tech will also likely >> improve. I doubt the current models are the best we're going to see,a nd >> the new crop is already shaping up as a real improvement over the last >> (Already quite good) generation. >> >> -Adam >> >> > > -- PDML Pentax-Discuss Mail List PDML@pdml.net http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
