p.1 #1 · Alternate paths to sharpness from Nikon mount lens
Typical path is get sharpest lens you can and mount it on your Nikon.
Another path some explore is put Nikon mount lens on body equipped with Foveon sensor that has been modified to take Nikon mount lens, like Sigma SD14 aka SD14-F.
Below is one pic from CV 58/1.4 mounted on SD14 body. That is downsized, 100% size is at [url]http://zart.zenfolio.com/img/v6/p800721951.jpg[/url].
This is at F2. What you think of it's sharpness? Worth getting that body for it?
p.1 #3 · Alternate paths to sharpness from Nikon mount lens
cputeq wrote:
Or, you could just use a Nikon body
That's SOOC (D300) JPEG with the 35mm f/1.8 DX, at f/1.8.
"vivid" profile, Normal D-lighting.
I really can't imagine any appreciable gains from using a Foveon camera.
That's really apples to pineapples example you posted. Comparison should be _same_ lens on two different bodies. Foveon does not interpolate like Bayer and does not have anti-aliasing filter like D300 does.
p.1 #5 · Alternate paths to sharpness from Nikon mount lens
Yes, I do feel it does look good for 100% crops, I feel I keep getting crisper results from SD14-F than my D300 _with SAME lens.
Below are two 100% crops that illustrate this.
See in first one the "teeth" on roof shingles that are facing you (not ridges on top of them)? You can _not_ see them in shot taken with same lens and D300.
See horizontal blinds behind window on left and patter on door drapes in second one? You can _not_ see them in shot taken with same lens and D300.
p.1 #7 · Alternate paths to sharpness from Nikon mount lens
That looks very noisy for being ISO 100, but on the other hand, the has a beatiful character! Really like fine grain.
Nov 15, 2009 at 03:29 PM
Andre Labonte Offline • • • • • • Upload & Sell: Off
p.1 #8 · Alternate paths to sharpness from Nikon mount lens
A 5MP Foveon is still only 5MP. The only advantage a Foveon sensor provides is that there is no interpolation of the color for the off-color pixel. Thus, in a bayer array, there is a greater chance for color fringing, but otherwise the sharpness is the same for the same MP count. In black and white, pure pixel count is the only thing that matters all else being the same.
Foveon likes to say that their 5MP sensor is equivalent to a 15MP sensor but that's a marketing gimmick and simply not true.
Makten -- the noise you see is one of the failings of the Foveon sensor due to the challenges of the stacked design. When I worked on the sensor, one of my jobs was to change part of the process that reduces the dark current leakage.
p.1 #9 · Alternate paths to sharpness from Nikon mount lens
Andre Labonte wrote:
A 5MP Foveon is still only 5MP. The only advantage a Foveon sensor provides is that there is no interpolation of the color for the off-color pixel. Thus, in a bayer array, there is a greater chance for color fringing, but otherwise the sharpness is the same for the same MP count. In black and white, pure pixel count is the only thing that matters all else being the same.
Foveon likes to say that their 5MP sensor is equivalent to a 15MP sensor but that's a marketing gimmick and simply not true.
Makten -- the noise you see is one of the failings of the Foveon sensor due to the challenges of the stacked design. When I worked on the sensor, one of my jobs was to change part of the process that reduces the dark current leakage.
...Show more →
Andre, considering you say you worked on Foveon sensor can you then please explain to me how come I see more detail with same lens on Foveon sensor than I do on D300's sensor even though you say D300 should be at least equal if not better?
p.1 #10 · Alternate paths to sharpness from Nikon mount lens
"We see what we expect to see." Richard Gregory
Nov 15, 2009 at 05:09 PM
Andre Labonte Offline • • • • • • Upload & Sell: Off
p.1 #11 · Alternate paths to sharpness from Nikon mount lens
Zoran,
There are many factors to image detail in a digital image. Other factors such as in-camera sharpening, noise reduction, ISO setting, base ISO for the camer, raw vs. jpg, conversion method, etc. Also, if using color images, the bayer sensor will suffer from interpolation defects and possibly color fringing in high-contrast lines. Thus, best way to determin sharpness is with B&W images.
Also, as Roger pointed out:
"We see what we expect to see." Richard Gregory
I wonder what a true double-blind test would result in, especially with cameras at optimal settings.
p.1 #12 · Alternate paths to sharpness from Nikon mount lens
Andre Labonte wrote:
Thus, best way to determin sharpness is with B&W images.
Monochrome images favor Bayer sensors, however we humans do not see in monochrome, do we, we see in color, so why would we care about what produces sharpest result in monochrome, we are supposed to care about what produces sharpest result in color.
Or I am wrong?
P.S. Yes, we do see what we want to see. Goes for Bayer watchers just as same as it does for Foveon ones. Double blind test of _color_ images would be the only way to go. In the meantime I am not the only one that sees it:
Andre Labonte Offline • • • • • • Upload & Sell: Off
p.1 #13 · Alternate paths to sharpness from Nikon mount lens
ZoranC wrote:
Monochrome images favor Bayer sensors, however we humans do not see in monochrome, do we, we see in color, so why would we care about what produces sharpest result in monochrome, we are supposed to care about what produces sharpest result in color.
Or I am wrong?
P.S. Yes, we do see what we want to see. Goes for Bayer watchers just as same as it does for Foveon ones. Double blind test of _color_ images would be the only way to go. In the meantime I am not the only one that sees it:
Monochrome does not favor any sensor over another. I don't know where you get that notion.
As for your point about color print sharpness, then yes, due to color fringing and interpolation, a 5MP Foveon will give you better color sharpness than a 5MP bayer. It is a minor improvent and a gap that is closing as interpolation algorithms impove with time. AND, it is an advantage that will be seen only when you have a sharp-line contrast between two dissimilar colors.
BUT let's be clear that Foveon's claim that its 5MP sensor is like or equivalent to 15MP is NOT true at all, it is still only 5MP. People get resolution and sharpness confused --- they are related to each other but not the same. The Foveon sensor reduces color fringing making color images looks sharper, but resolutions is exactly the same as the equivalent sized bayer. Image detail, expecially if reviewed in B&W, will be the same. That is my point.
Also, to gain that minor improvent in color images with the Foveon, you do so at the expense high ISO and/or long exposure performance. So, the Foveon quickly starts to loose its advantage due to noise as soon as exposure time get long and/or ISO gets much above base ISO. Foveon has made some strides in improving this, but they have a long way to go still to catch up with the bayer.
Anyway ... it's a mute point since Foveon has not been able to get its sensor in any of the camera bodies that really matter ... at least not yet. Rather than piss over minor pixel-peeping details, let's go take some pictures with whatever system suits our fancy and have fun!
Yes, I am aware of them and have looked into it. Only problems are:
1. Just the mod is approximately $500 on top of your D300.
2. Unlike changing the mount on SD14, it is permanent, it is not reversible. You don't like it and tough luck, you are out of approximately $2000 (cost of D300 + cost of mod).
3. You still have Bayer sensor, interpolation is still robbing you from detail.
