This is amazingly difficult to believe, and I would look for other causes than the 1Ds sharpness being deficient. Unless it's particular to that copy of the camera body ... I wonder if an AA filter can get "fogged" like an old lens? Doubtful.
Sure you can, if you're trying to make good photographs and can only use one of them at a time.You can't logically compare a 50mm 1.8 with 50mm equivalency on a
zoom, that's comparing apples and oranges
The 1.7x crop has a .7x "smaller" circle falling on the sensor than the full frame. Now are you talking about circle of definition or circle of confusion? LOL. I think I know what you are getting at, but you can't avoid the fact that that portion of the image circle coming from the lens with distortions is outside the range of the sensor.
Ostensibly by mounting the 50mm canon on the Sigma it "should" also out-perform the 15-30 though it's difficult to say exactly because I haven't modified a Canon 50mm 1.8 to try it, but judging from the PhotoDo tests, I would expect that it would. I've tested a 105mm macro Sigma on each in past years with similar results. The acutance is simply better on the SD10 at pixel level than on my 1D'S (which is a pretty average good copy of the camera).
A lot of people like the 1.7 crop.
There are only missing a few fast ones.
less
10mp equiv in a prosumer dslr at a competitive price. For the people with SA lenses its a home run as long as it has improved features.
Just because bayer was first does not mean it is best. Always a better mouse trap. People seem to forget some of the other benefits.
Bayer does the job okay, and camera body and lens quality comes into it too, but even with the problems I still prefer foveon images.
--
You are buying a system that has its pluses and minuses and you should consider those aspects in a hierarchy that fits your needs.
Not necessarily. It all depends on how you define market acceptance.
Limits are approaching, lightwaves are only so large and it takes 3 times more pixel area to produce one pixel on a bayer which means a foveon style chip has a lot more room to grow.
Labratory curiosity governed by the obvious knowledge that bayer will come to an end soon.
I agree the 3.4 against a 10.2 is too much, but if the chip were worked on to produce say 8-10mp, it could rival almost anything bayer.
I said "History is bunk". Obviously, reincarnation is fact.Reincarnation is bunk!
Lithwaves are 0.4 to 0.7 microns. DSLR pixel pitches are 5-10 microns. That leaves room for pixel counts to grow by a factor of around a hundred.
Except that it doesn't, for two reasons. First, Foveon sensors are considerably noisier than Bayer sensors for a given pixel size. This has to do with math and physics (I won't get into details, but it has to do with spectral responses and the large off diagonal matrix coefficients needed to convert Foveon sensor output to "human eye" red, green, and blue). Added electrical noise also tends to come out in the Bayer sensor's favor. For example, Canon's approach of having a crude buffer amplifier at every pixel to reduce CMOS switching noise simply can't be applied to the Foveon sensor.
Well, except for the minor detail of that being neither "knowledge" or "obvious".
If it were 8-10mp, it would produce 24-30 million samples of data. While that would make for a very nice image (if the noise could be controlled), it would also mean you'd need more processing power than any current Bayer camera has to preprocess that data in a reasonable amount of time and get it stored on a card, and more PC power to get it off the card, archive, and process it to printable form.
Yeah, you're always going on about that.Reincarnation is bunk!
Just because lightwaves are .4 to .7 microns doesn't mean you can make the pixel pitch that small, as a matter of fact, not even close.....
Like I said, the technology just needs to be improved on.
Like I said, computing power is cheap, so is storage. RAW workflow nowadays is just as simple as Jpeg. In fact, I don't even shoot Jpeg anymore.....Second, there is diffusion between the lower layers of Foveon
sensors (that is why the raw processing software requires red
sharpening). This increases as pixel pitch decreases.
Well, except for the minor detail of that being neither "knowledge"
or "obvious".
If it were 8-10mp, it would produce 24-30 million samples of data.
While that would make for a very nice image (if the noise could be
controlled), it would also mean you'd need more processing power
than any current Bayer camera has to preprocess that data in a
reasonable amount of time and get it stored on a card, and more PC
power to get it off the card, archive, and process it to printable
form.
That's the Foveon Achilles heal, to match a Bayer sensor, it always
needs more computing resources, both in the camera and in the PC.
I don't think I'm going out on a limb when I say that you shirley must be over-excercising your imagination.Yeah, you're always going on about that.Reincarnation is bunk!
Yes. It seems that these past life memories are typically much more glamorous than those of ordinary folks. (Though the movie, "Defending Your Life", had a pretty good gag going on that issue.)
Again this isn't something I can disagree with, and speculation about the future of your system, while more relivant in this case, should float its way downward in the hierarchy you describe; certainly below "real" factors.
But we're not in the film days anymore; the camera and the sensor are married. If there were Canon and Nikon bodies with Foveon chips, this wouldn't have come up.
That's the issue we loose sight of too easily ... but if "man is an enterprise" the question you just asked is "the bottom line."
I see things a bit differently because of the ability to load Tech Pan 8 when you want it and Provia 100 when you want that ... and yet this is a valid point you make. You obviously don't shoot birds at dusk, so it would be silly on your point to spend any time worrying about stabilized 400 mm lenses. The same is of course true for the body and sensor: it's a good thing they make more than one so we can each choose what comes closest to meeting our full set of needs.
I'm really baffled that Foveon don't make sensors for Nikon. Canon seems to take pride in their in-house CMOS fab, but Nikon would be well suited to offer a lower-res alternative to the Sony chip ... and that would get the sensor more wide recognition ( which translates to software support and whatnot ), but also get Foveon more R&D funding.
Sigma amortizes its lens development costs over many different lens mounts. I am not aware of any "dedicated" Sigma lens made just for their own DSLR's, as they offer the same lenses in Nikon and Canon mount, as well as others. So any agument regarding lenses is really not valid.
If Sigma sells enough cameras to be happy, then it is all good. If they don't, then they will stop making them. It really is of no consequence trying to debate who has "taken" over the front seat. Remember quality alone does not sell. It is marketing that sells.
A "bit late" for what? Sigma is a small camera maker, with modest goals. They don't expect to compete in volume with Canon and Nikon. Come on, with ONE camera model, how could they?
Bayer does the job OK. Foevon might be better, if it had the same aparrent resolution as 16mp or 22mp Bayer. I need more resolution than Foevon currently offers, even with the slight improvement of the new unreleased camera from Sigma. I also need more robust camera body, and more lens choices than just Sigma. But I am not your average "entry level" consumer.
Of course, close. And even smaller. Half wave dipoles are wonderful resonating structures. I wonder if anyone has ever speculated about photon resonators as a color separation mechanism.
You're missing the point. Those same improvements can also be made to Bayer sensors. There's limits imposed by physics that say the Foveon will remain noisier.
Not inside a camera, it isn't. That's why only $3000+ monsters like Canon 1D II, 1Ds II, and Nikon D2X currently have enough processing power to hit 60 million pixels/sec. $1500 gets you 40-50 million pixels/sec (Canon 30D, Nikon D200). $1000 gets you 30 million pixels/sec. At that speed, your 8-10mp Foveon (which "consumes as much processing power as a 24-30mp Bayer) would just hit 1 frame/second. Canon, Nikon, Pentax, Sony, etc. have established that 3 frames/sec is "entry level" and even folks like Thomas Mottle (shot Sigmas for panorama work) found the 1.8 frames/sec of SD10 to be restrictive).
Computing power and storage are cheap, if you're a pro, or willing to devote substantial "hobby" resources to photography. But many people are more casual. And many pros produce tremendous amounts of data.
All well and good, but if the memory buffer is large enough - who cares? And memory is cheap.
