Will we ever get a Pro-level alternative to Canon + Nikon?

[Kyle Nagel]

This Foveon sensor debate is an old and passionate one from both sides of the fence, I would say conventional sensors are "true" in regards to their effective megapixels, this can be seen by viewing the recorded un-interpolated pixel dimensions, you will see Fuji's dimensions are not 14 mp worth of dimensions without interpolation, theirs is closer to 6 or 7, I believe the camera setting that produces close to 14 mp of resolution states "interpolated" in their specs.

As far as a traditional sensor the way they work is sometimes miss-understood. here is a quote from the Luminous Landscape that actually explains it rather well:

When a digital camera makes an exposure the imaging chip (whether it's CCD or CMOS) records the amount of light that has hit each pixel, or photo site. This is recorded as a voltage level. The camera's analog to digital circuitry now changes this analog voltage signal into a digital representation. Depending on the camera's circuitry either 12 or 14 bits of data are recorded. Incidentally, if the camera records 12 bits of data then each pixel can handle 4,096 brightness levels (2^12), and if 14 bit then it can record 16,384 different brightness levels (2^14). (To my knowledge no current imaging chip records a true 16 bits worth of data)... ...imaging chips use what is called a Bayer Matrix or colour Filter Array in order to record colour. The way this is done is to place red, blue and green filters over each pixel. Half of the pixels are filtered green and the remainder are either red or blue. Through a very complex algorithm the values recorded by each pixel are compared with its neighbors, and full colour information is derived

I have played with images from the Foveon sensors and while the colors seem really nice (actually very nice) I find artifacts appear much sooner when enlarged than with a traditional CCD or CMOS sensor.

I'm still not sure how a 4.7 mp sensor can be a 14 mp sensor, it seems like a stretch to me, Just one mans opinion.

Kyle

 

[John Sheehy]

If I were Nikon or Canon I'd be very nervous about the advent of the Foveon chip and their association with Sigma. The potential for this thing is ridiculous, if the press reports are true. Already 14 mp's on a chip smaller than APS-C.

The Sigma SD14 is a 4.7MP camera. The 14MP part is a marketing lie.

Imagine a full-frame or even 1.3 crop sensor. And I've looked at the SD14, and it seems to be a very well-designed camera. Mirror lock-up on a knob at the top of the camera, go figure. Integrated sensor dust protector. And a 14mp p&s, to boot. ("Why?" might be a reasonable question, but still, people will flock to it)

They didn't flock to the SD10 when it was a 3.43MP camera marketed as 10.3MP, in the days of 6MP APS-sized sensors.

 

[Ray West]

I think it is the same sort of marketing hype as in measuring/promoting cpu performance in MHz. Folk think that higher MHz means faster performance, so higher MPixels means better images. The Fovean sensor can be considered as having fewer pixels, but each pixel site is red, green and blue, so to equate the resulting image resolution to the more usual Bayer layout, which, more or less has a seperate site for each colour, then its fair enough to multiply the foveon value by three. I think it may be better to divide the Bayer resolutions by 4, since there are usually twice as many green sites as red and blue.

And, in the same way as hard discs are sold in 'unformatted sizes', I expect the camera makers are not quoting the usable pixels. It reminds me of many years ago, I bought a five transistor radio. It indeed had five transistors, but only three were actually in the circuit, the other two had all legs soldered to ground.

Best wishes,

Ray

 

[John Sheehy]

The Fovean sensor can be considered as having fewer pixels, but each pixel site is red, green and blue, so to equate the resulting image resolution to the more usual Bayer layout, which, more or less has a seperate site for each colour, then its fair enough to multiply the foveon value by three. I think it may be better to divide the Bayer resolutions by 4, since there are usually twice as many green sites as red and blue.

With the same number of 2D spatial witnesses, the Foveon only has higher resolution of color against color, not of light against dark. It is very poor at discriminating green from blue, but does very well resolving red from blue, and red from green. Most real-world scenes do not have what it takes for the Foveon to out-resolve color-wise, and much of its apparent extra detail is just aliasing artifacts.

The best way, IMO, is to number pixels by 2D spatial witnesses, and then also describe what these pixels can resolve. Bending the number of pixels is ridiculous, IMO.

 

[Kyle Nagel]

OK somebody check me if I'm wrong here, but as I understand it here is what is happening with the Foveon vs Traditional sensors. We will assume for arguments sake that the Foveon is the new "so-called" 14mp sensor, and the traditional sensor has approximately 14 million actual pixels (we know it isn't exact, but very close). In the case of the standard sensor data is captured from 14 million pixels as voltages, for the sake of this analogy let's say "Black" and "White" information. At this point three different color filters are applied, again I'll quote Mr. Riechman's statement:

The way this is done is to place red, blue and green filters over each pixel. Half of the pixels are filtered green and the remainder are either red or blue. Through a very complex algorithm the values recorded by each pixel are compared with its neighbors, and full colour information is derived

So even though some pixels are filtered green, and others red and blue, there is still 14 million pixels of image information (read: detail), this three color filtering is then used to determine the actual color information for each pixel. When you are done there have been no pixels added or taken away (interpolated), so there are 14 million pixels worth of actual image data, it is only the color information that was created through this "filtering" process.

This is similar to taking a 120mm medium format image with Black and White film and then using red, green, and blue filtering to create a color image from a single frame, I know this is not a perfect analogy since with film you can't have all three filters on the camera at the same time, but you get the idea.

With the Foveon sensor it is like taking three separate images, one on red film, one on blue, and one on green, then matching up the grains containing the same image information of each color, only in this case you are shooting 35mm film instead of 120mm. So you are using 3 identical image grains to create ONE color grain, but keep in mind these three different colored grains(pixels) all contain the same piece of image information, so in the end the actual image information is higher on the 120 film, and where you may possibly get better color with the 35mm process described above (in theory), the medium format image still has more image information.

Kyle

 

[Asher Kelman]

Kyle,

The question I have with layered pixels in the Foveon solution is what is the charge capacity of the sensels so created. That will effect the noise and dynamic range.

Asher

 

[John Sheehy]

Kyle,

The question I have with layered pixels in the Foveon solution is what is the charge capacity of the sensels so created. That will effect the noise and dynamic range.

Asher

I don't know what the actual charge is, but the individual channels in the Sigma are actually quite clean. I've inspected the deep shadows of ISO 800 shots, and each channel is fairly clean, and the sum of the three channels, as a greyscale, is also quite clean. It's when RGB color is extrapolated from the Foveon data that most of the noise starts to become visible, being mainly chromatic in nature. I would seriously consider getting a Foveon-based camera for B&W (quite ironic, I think), if it had a mild AA filter.

 

[Asher Kelman]

I'm very interested in the Sigma DP-1 which should be a wonderful camera to throw in one's bag for travel!

Asher

 

[Kyle Nagel]

I would seriously consider getting a Foveon-based camera for B&W (quite ironic, I think), if it had a mild AA filter.

Hadn't thought of that, it's an interesting idea though.

Kyle

 

[Asher Kelman]

I really wouldn't worry too much about Moire as it is mostly avoidable or can be corrected.

I suspect that the new camera will be great. But why the dealy in release?

asher

 

[Kyle Nagel]

Earlier in this thread I mentioned that my next lens purchase would be the Pentax 31mm FA Limited. Well it arrive last night. I will be shooting like mad this weekend to put it though it's paces with the K10D attached. I'm hoping it turns out to be a killer combination as an everyday walk around. I'll let you all know what I find out.

Kyle

 

[Rob Riley]

with the digicam market topping out this year (predicted)
the focus is on the still expanding dSLR market
hence new market entrants have been appearing left right and centre

the new entrants are set to compete on the market entry and middle ground
they appear to be using an upscale set of features against the major players
who have started to compete on price, cite the canon xti with reduced featureset (spot metering etc)

as we all know, pro photographers (as opposed to their equipment) make the best of the kit they have
could we conceive that there will be new pro entrants as a result
I think yes, but have no idea how the numbers might pan out

to me the question is, not about the equipment but about the photographers

 

[Benjamin Kanarek]

I am pleased that you received your lens. I will be getting mine sent to me in about three weeks. You'll have to tell me how you like it...It has been called one of the best lenses in the world...

 

[Kyle Nagel]

I absolutely love it! It lives on my K10D about 85% of the time, the focal length is perfect and it is razor sharp with beautiful contrast, I'd give up just about any other lens before I gave up this one.

Kyle

 

[Benjamin Kanarek]

Alright I'll Take Them

I absolutely love it! It lives on my K10D about 85% of the time, the focal length is perfect and it is razor sharp with beautiful contrast, I'd give up just about any other lens before I gave up this one.

Kyle

I think you are so right Kyle. That should be the only lens you should keep. Why be encumbered with all that other material stuff. You are obviously much more evolved than I am on the ladder of spiritually evolution and as I am just a mere plebeian, pedestrian speck of insignificant meat scavenging the face of this earth of ours, I can't help but feel this deep urge to assist you in your quest for "The Ultimate Perfection". Therefor as an act of great empathy and compassion, I hereby offer you my humble service and will take all of those other nasty lenses off your hands. It is the least I can do for you...

Your humble servant...b....

 

[Kyle Nagel]

That should be the only lens you should keep. Why be encumbered with all that other material stuff. You are obviously much more evolved than I am on the ladder of spiritually evolution and as I am just a mere plebeian, pedestrian speck of insignificant meat scavenging the face of this earth of ours, I can't help but feel this deep urge to assist you in your quest for "The Ultimate Perfection". Therefor as an act of great empathy and compassion, I hereby offer you my humble service and will take all of those other nasty lenses off your hands. It is the least I can do for you...

Funny you mention it, I am going to be selling my Sigma EX DG 105mm f2.8 1:1 Macro sometime in the near future. It is also a very nice lens.

Kyle