Luis Augusto
Well-known member
I've been waiting for manufacturers to go with Foveon-like technology for years, so lately I've been asking myself why didn't they go with it?
My answer is rather simple, Sigma SD14 it's a 4.7 MP camera, regardless of how you see it, that's all, it's a 4.7 MP. However, those 4.7 MP are obtained by 14.1K photodiodes. Sony Alpha A380 has 14.9K photodiodes which produce a 14.2 MP image.
Ok, so both have similar numbers of photodiodes, however, the sigma has a 3D sensor design while the other has a 2D sensor arrange, this automatically means Foveon it's more expensive to make.
The theoretical advantages of Foveon X3 is that it has true color rendition, because it's an analog device (it actually detects the color channels individually), while Bayer just guess them.
On the other hand, is known to have better per-pixel sharpness, this is true, but it's using a lot more photodiodes than a Bayer 4.7 MP camera, so it has to be like that, because it has more light sensors, regardless of the image size.
So, what I'm doing here is comparing cameras with a similar number of light sensors (photodiodes).
First comparative, Resolution:
The differences are negligible, even more since I didn't really process them with much care.
Resolution conclusion? With the same number of photodiodes, it's safe to say both offer the same true detail, even if they differ on spatial (MP) resolution.
Second comparative, High-ISO shots (image order remains the same):
ISO 800:
The A380, at this point, is doing a much better job, but the difference isn't that huge, even tough I don't have any way to confirm this, I would bet that if sensor sizes were the same, noise would be almost identical.
ISO 1600:
At ISO 1600, the Foveon sensor can no longer match the Bayer one, the difference seems to be quite huge, and doesn't seem to be caused by the sensors sizes differences.
ISO 3200:
ISO 3200 on the Sigma it's absolutely unacceptable, the dark parts are completely destroyed by the noise (this is not seen on the crops, but if you want to, go to the Sigma DP2 review, then check the full size high ISO shots to see what I'm talking about). The CCD does a pretty poor job too, but not nearly as bad as the Foveon.
High ISO Conclusions? Foveon can't compete with the CCD of the A380, which isn't stellar, against new CMOS sensors it's going to be at least 2 stops behind, if not more (Sony Alpha A550, Canon 7D, Nikon D5000).
Third comparative, Dynamic Range:
I can't truly compare them, so I will rely on dpreview tests:
Foven X3 (Sigma DP2): 10.5 EV; Marked as usable range.
Sony CCD (Sony A380): 11.9 EV; Marked as usable range.
Seems like Foveon doesn't truly offers better DR, it's even beaten by Sony CCD (that said, Sony DR is probably the highest).
Fourth comparative, Color accuracy:
According to Popphoto, Sigma DP2 (and Foveon X3) color accuracy is 6.56 (rated as excellent), but, they didn't rate, yet, the A380, however, the older and cheaper Sony A200 (the most similar I could find) scores 7.6.
Conclusions:
This may cause problems on the Sigma/Foveon community (as it caused to me, since I truly believed Foveon was the way to go), however, using the same number of photodiodes, the classic bayer sensors offer the same resolution at lowest ISO, much better high-ISO performance, slightly better DR and higher color accuracy.
All of that disadvantages, while still being more expensive to produce.
PS:This is as they stand now, I don't know (and neither do you) if Foveon is going to improve enough to beat Bayer sensors with the same number of photodiodes (like improving high-iso performance for example).
--
Feel free to crush my pics, just tell me why, you will
help me a lot.
I'm working on a project called 'Photography, poetry
and a little bit of philosophy ('Fotografía, poesía y algo
de filosofía' in spanish)'. It will take me a long time to
finish it, since I want it to be good on all three
categories.
My answer is rather simple, Sigma SD14 it's a 4.7 MP camera, regardless of how you see it, that's all, it's a 4.7 MP. However, those 4.7 MP are obtained by 14.1K photodiodes. Sony Alpha A380 has 14.9K photodiodes which produce a 14.2 MP image.
Ok, so both have similar numbers of photodiodes, however, the sigma has a 3D sensor design while the other has a 2D sensor arrange, this automatically means Foveon it's more expensive to make.
The theoretical advantages of Foveon X3 is that it has true color rendition, because it's an analog device (it actually detects the color channels individually), while Bayer just guess them.
On the other hand, is known to have better per-pixel sharpness, this is true, but it's using a lot more photodiodes than a Bayer 4.7 MP camera, so it has to be like that, because it has more light sensors, regardless of the image size.
So, what I'm doing here is comparing cameras with a similar number of light sensors (photodiodes).
First comparative, Resolution:
The differences are negligible, even more since I didn't really process them with much care.
Resolution conclusion? With the same number of photodiodes, it's safe to say both offer the same true detail, even if they differ on spatial (MP) resolution.
Second comparative, High-ISO shots (image order remains the same):
ISO 800:
The A380, at this point, is doing a much better job, but the difference isn't that huge, even tough I don't have any way to confirm this, I would bet that if sensor sizes were the same, noise would be almost identical.
ISO 1600:
At ISO 1600, the Foveon sensor can no longer match the Bayer one, the difference seems to be quite huge, and doesn't seem to be caused by the sensors sizes differences.
ISO 3200:
ISO 3200 on the Sigma it's absolutely unacceptable, the dark parts are completely destroyed by the noise (this is not seen on the crops, but if you want to, go to the Sigma DP2 review, then check the full size high ISO shots to see what I'm talking about). The CCD does a pretty poor job too, but not nearly as bad as the Foveon.
High ISO Conclusions? Foveon can't compete with the CCD of the A380, which isn't stellar, against new CMOS sensors it's going to be at least 2 stops behind, if not more (Sony Alpha A550, Canon 7D, Nikon D5000).
Third comparative, Dynamic Range:
I can't truly compare them, so I will rely on dpreview tests:
Foven X3 (Sigma DP2): 10.5 EV; Marked as usable range.
Sony CCD (Sony A380): 11.9 EV; Marked as usable range.
Seems like Foveon doesn't truly offers better DR, it's even beaten by Sony CCD (that said, Sony DR is probably the highest).
Fourth comparative, Color accuracy:
According to Popphoto, Sigma DP2 (and Foveon X3) color accuracy is 6.56 (rated as excellent), but, they didn't rate, yet, the A380, however, the older and cheaper Sony A200 (the most similar I could find) scores 7.6.
Conclusions:
This may cause problems on the Sigma/Foveon community (as it caused to me, since I truly believed Foveon was the way to go), however, using the same number of photodiodes, the classic bayer sensors offer the same resolution at lowest ISO, much better high-ISO performance, slightly better DR and higher color accuracy.
All of that disadvantages, while still being more expensive to produce.
PS:This is as they stand now, I don't know (and neither do you) if Foveon is going to improve enough to beat Bayer sensors with the same number of photodiodes (like improving high-iso performance for example).
--
Feel free to crush my pics, just tell me why, you will
help me a lot.
I'm working on a project called 'Photography, poetry
and a little bit of philosophy ('Fotografía, poesía y algo
de filosofía' in spanish)'. It will take me a long time to
finish it, since I want it to be good on all three
categories.
