Since the Pentax Q was announced, I have been thinking about getting one for testing lens performance of my EF mount lenses. I started out with the TS-E 24 II, and the results show that it has plenty of resolution for much higher pixel count than say the Nikon D800. The test method and initial findings are described here:
The pixel density of the Q corresponds to a 360 MP 36x24 mm sensor.
When I today tested my Distagon 2.8/21 ZE, my mind was totally blown away by the sharpness on the tiny sensor. Center sharpness peaks wide open, and the 100% crops are soft but with a very stable and healthy impression. This is almost unbelievable. I would expect f/4 to be slightly better than f/2.8. Obviously only the center is tested.
On a side note, the lens has been in Oberkcochen for repair after a barrel damage, and was calibrated for MTF.
Wide Distagons are the sharpest Zeiss lenses in the image centre and at small image heights. They typically show around 80% contrast at 40 lpmm, that's routine, for f2.8-f4. The soon to arrive 15mm does this at f2.8, and improves to 85% at f5.6. That's the best one with the latest technology and design expertise.
Your ZE 21/2.8 is best in the centre at f2.8, and drops contrast at every level down at f5.6. Very well corrected also, an understatement.
Do you notice how the image retains integrity - colour relationships especially - at smaller apertures though? Only the dimensionality suffers a little.
Sensors can get much more dense before in-focus material resolves poorly.
Yeah, I think so. Incredibly soft all around.
But I guess it's good to see what the Q looks like.
Now I don't have to worry about ever wanting one.
Well, until they're available for $50 or something. Then it
might be fun to try. I think I'd still be asking myself "why" tho.
Bifurcator wrote:
Yeah, I think so. Incredibly soft all around.
But I guess it's good to see what the Q looks like.
Now I don't have to worry about ever wanting one.
Well, until they're available for $50 or something. Then it
might be fun to try. I think I'd still be asking myself "why" tho.
You are misunderstaning. It is not the camera that is impressive. It is the lens. Think about what you are looking at. It is like a 410% view from the 5D II
philip_pj wrote:
Wide Distagons are the sharpest Zeiss lenses in the image centre and at small image heights. They typically show around 80% contrast at 40 lpmm, that's routine, for f2.8-f4. The soon to arrive 15mm does this at f2.8, and improves to 85% at f5.6. That's the best one with the latest technology and design expertise.
Your ZE 21/2.8 is best in the centre at f2.8, and drops contrast at every level down at f5.6. Very well corrected also, an understatement.
Do you notice how the image retains integrity - colour relationships especially - at smaller apertures though? Only the dimensionality suffers a little.
Sensors can get much more dense before in-focus material resolves poorly. ...Show more →
Thank you for your observations and comments. Your bottom line says it all.
trenchmonkey wrote:
I've seen sharper SOOC 200% crops with a kit lens. (
Nope. Even the sharpest lens on a Leica M9 interpolated to 200% will be softer and with less detail than the 2.8 sample. The rest is diffraction, and that relates to DoF and image size. There is no way to get around those physics with any other lens or sensor size for visible wavelength light.
You guys are killing me. I tested a lens using a pixel density that was supposed to outresolve the lens by a huge margin. I find that it barely did. And you comlain that diffraction limited images are soft
That's pretty cool actually. It reminds me about when I used a 4/3 camera to check the center performance of legacy 50mm lenses. It's a pity we can only see the very center of the image circle though.
These small sensors (to an even greater degree than micro 4/3 and 4/3) are pretty demanding!
Jonas B wrote:
That's pretty cool actually. It reminds me about when I used a 4/3 camera to check the center performance of legacy 50mm lenses. It's a pity we can see see the very center of the image circle though.
These small sensors (to an even greater degree than micro 4/3 and 4/3) are pretty demanding!
Yeah, smaller sensors have always been our peeking device. They only tell half the truth as you say, as sharpness across the frame may be important.
deadwolfbones wrote:
I know it's not really relevant to the purpose of your test, but could we see the full shot at f/2.8?
Sure. Sorry about the blown highlights.
Standard resize and sharpening for web script. This is still only about 6 % of the area of the image circle, not far away from the equivalent of a 100% crop from the 5DII
alundeb wrote:
You are misunderstaning. It is not the camera that is impressive. It is the lens. Think about what you are looking at. It is like a 410% view from the 5D II
Hmm, OK, I get it. So do some critical focusing with the 5dII in liveview and lets see a 400% crop next to the Q image.
I'm betting the 5dII at 400% will look better.
But out of curiosity why is the Q sensor supposed to out-resolve the lens? What data is that based on?
