i am talking SHARP

[munro harrap]

If your gear list is genuine, you already HAVE sharp.

 

[Bird pics]

I'm chiming in with Mediaarchivist. The Sigma 70 is sharper than my 1.8G primes and very close to the Sigma 18-38 1.8 which is my next indulgence. B&H has those right now for $450.00.

I bought it not only because every review had nothing bad to say except auto focus was a little slow but no slower than average for a macro and at 105 equiv. on a 7100 its the perfect focal length. I think it gives a better fov than the 85 1.8 on a DX which I find a little long.

Steve

 

[ultimitsu]

Ok, talk to me people

I want the sharpest prime lens/camera combination

Lets park ISO noise, colour, blah blah blah, just want sharp

With Leica M240 in your gear list I am not sure if you are taking the pisss.

Assuming you are not, you need to first understand that when you use lay term sharp, it can be construed in two different wants - better pixel level sharpness, or more overall resolution. The two concept are in fact at odds because for the same given lens and format, the larger the pixels, the better pixel level sharpness but less overall resolution; conversely the smaller the pixels the worse pixel level sharpness but more overall resolution.

IF you downsize the higher resolution image to the lower resolution, you end up with even better pixel level sharpness.

To put this in a real world example, Nikon 50/1.8D wide open has mediocre sharpness. If you shot it with a D700, pixel level sharpness will be better than D800, but if you downsize D800 image to 12mp, the D800 image will have better pixel level sharpness.

So essentially, you should always go with the largest format and highest pixel count, if you want more sharpness. I say go with 645Z.

 

[Reilly Diefenbach]

The problem with the Sigma Arts is that although they are sharp wide open and everywhere else shooting a flat test chart, they have wretched background bokeh under a lot of conditions, which defeats the whole purpose of shooting wide open and spending that much money.

I read about this, and certainly there are examples of poor bokeh to be found online (even on this very site). I'm a bit of a bokeh snob myself, and I have not found the 35mm Art lens to be a slouch in this department at all. All lenses have poor bokeh with certain backgrounds, the ∑35 is no different but not more so. It handles harsh uneven lighting and foreground bokeh rather well IMHO, even after extreme tone mapping (which usually destroys the bokeh):

This is the worst bokeh I have been able to coax out of it (so far):



I have no doubt that clinical sharpness was one of the design goals of this lens, perhaps at the expense of bokeh (is that ever a primary design goal?).

Yes. The new Nikon 58 1.4 was specifically designed for people shooting with narrow DOF and excellent bokeh. It is far superior for that purpose, but at a steep price :^)

I think they hit the ball out of the park on this lens. The bokeh is not "the best ever for any lens" but I think it gets unfairly knocked for no good reason.

Both pics above prove the point, unfortunately. I certainly wouldn't mind the Sigma for landscape at f5.6, but for wide open work, I wouldn't be happy at all.

 

[XLB]

More MP are likely going to give you the best results, but I would reference something other than DXO. Many issues have been taken with their results, at least when comparing one camera to another. PZ or lenstip might be better options. For example, DXO scores the Panasonic 20mm F1.7 on a GX7 with a 6 for P-mPix, yet I have some extremely sharp samples showing there is no way that's accurate.

This was not downsampled, and wasn't overly processed for sharpening, as you can see with the bokeh showing no artifacting. This is one of the sharpest lenses in the center I have owned, on any system. There is no way a 6 is accurate here.

The same lens gets 10 P-mPix on a GH2. This seems a bit strange.

 

[LTZ470]

More MP are likely going to give you the best results, but I would reference something other than DXO. Many issues have been taken with their results, at least when comparing one camera to another. PZ or lenstip might be better options. For example, DXO scores the Panasonic 20mm F1.7 on a GX7 with a 6 for P-mPix, yet I have some extremely sharp samples showing there is no way that's accurate.

This was not downsampled, and wasn't overly processed for sharpening, as you can see with the bokeh showing no artifacting. This is one of the sharpest lenses in the center I have owned, on any system. There is no way a 6 is accurate here.

The same lens gets 10 P-mPix on a GH2. This seems a bit strange.

Sharp lens...slow AF is the killer...

--
--Really there is a God...and He loves you..
FlickR Photostream:
www.flickr.com/photos/46756347@N08/

 

[John Sheehy]

indeed the resolution on larger sensors plays different. I remember how sharp the 5D and D700 were and the A7S also has a richness of image that is difficult to describe

lastly the lower resolution is more tolerant of lens as well

That doesn't make any sense.

With the exception of the A7s, which does special high-ISO read noise stunts, the cameras you mention all have inferior IQ to higher-MP cameras which have replaced them. A 50.6MP 5Dsr capture is better in every way than a 12.6MP 5D capture.

Perhaps you are one of those people who view images at 100% pixel view on your monitor, and think that you are looking at "image quality"?

 

[ultimitsu]

More MP are likely going to give you the best results, but I would reference something other than DXO. Many issues have been taken with their results, at least when comparing one camera to another. PZ or lenstip might be better options. For example, DXO scores the Panasonic 20mm F1.7 on a GX7 with a 6 for P-mPix, yet I have some extremely sharp samples showing there is no way that's accurate.

You ought to learn how to read DXO better. 6pmp is for lens wide open, your image is F2.8 considerably sharper in the centre region, as shown in DXO's graph.

 

[Neil-H]

More MP are likely going to give you the best results, but I would reference something other than DXO. Many issues have been taken with their results, at least when comparing one camera to another. PZ or lenstip might be better options. For example, DXO scores the Panasonic 20mm F1.7 on a GX7 with a 6 for P-mPix, yet I have some extremely sharp samples showing there is no way that's accurate.

You ought to learn how to read DXO better. 6pmp is for lens wide open, your image is F2.8 considerably sharper in the centre region, as shown in DXO's graph.

The difference between F1.7 and F2.8 wasn't massive, and, that still doesn't explain why the same lens scores vastly different on two sensors that are of similar generation, both MFT size, and both 16mp. Just bc they are in different bodies?? I call BS. You can defend their odd results if you want, but IMO DXO is doing something wrong.

EDIT: This is what I get for taking people at their word. Perhaps you should learn how to read, period. Here is what DXO says for their P-mPix score:

SHARPNESS-

The DxOMark resolution score shows sharpness performance of a lens-camera combination averaged over its entire focal length and aperture ranges.

The resolution score is computed as follows:
For each focal length and each f-number, we first compute sharpness and then weight it throughout the field, tolerating less sharpness in the corners than in the center. This gives one number for each focal and aperture combination.
Then, for each focal length, we select the maximal value of sharpness over the range of available apertures. We average this value over the whole range of focal length to obtain the DxOMark resolution score that we report (in P-MPix).

Note that for a wide-range zoom, there are huge differences between the resolutions for different focal lengths.

Sharpness is expressed in PMpix and is typically between 50% and 100% of the sensor pixel count, Differences below 1 P-MPix are usually not noticeable.

 

[ultimitsu]

The difference between F1.7 and F2.8 wasn't massive, and, that still doesn't explain why the same lens scores vastly different on two sensors that are of similar generation, both MFT size, and both 16mp. Just bc they are in different bodies?? I call BS. You can defend their odd results if you want, but IMO DXO is doing something wrong.

DXO does not cherry pick best results. they report whatever they get. Who knows why G7X's result is different? maybe it was an aliegnemnt issue between that particular 20mm and that G7X. One way or another, it would be odd to discount their entire library of test results just because you dont agree with one.

 

[Neil-H]

The difference between F1.7 and F2.8 wasn't massive, and, that still doesn't explain why the same lens scores vastly different on two sensors that are of similar generation, both MFT size, and both 16mp. Just bc they are in different bodies?? I call BS. You can defend their odd results if you want, but IMO DXO is doing something wrong.

DXO does not cherry pick best results. they report whatever they get. Who knows why G7X's result is different? maybe it was an aliegnemnt issue between that particular 20mm and that G7X. One way or another, it would be odd to discount their entire library of test results just because you dont agree with one.

But it's not just one, there are many examples of the same lens performing very different just bc it's on a different camera, a different camera that has the same size/rez sensor. I mean you obviously haven't spent much time actually reading over DXO's website or you would not have falsely stated their testing perameters, so perhaps you should look over the test results too. You will find that the GX7 isn't the only "alignment" issue.

Im not saying they do it on purpose, but nobody can tell us why near identical testing rigs have such great differences in results, especially when we see real world samples that completely disagree. On top of this, why anybody would judge "sharpness" by a smoothed number that is an average of all apertures is beyond me, unless you look at individual charts you have no idea if it's sharp wide open or stopped down.

Back to my original post, if you are comparing one camera to another, people are better off using alternate lens charts such as PZ or lenstip or even SLRgear.

 

[MediaArchivist]

Yes. The new Nikon 58 1.4 was specifically designed for people shooting with narrow DOF and excellent bokeh. It is far superior for that purpose, but at a steep price :^)

Perhaps my best overall lens for bokeh is the older ∑50/1.4 EX (not the Art version). It gets knocked for "soft corners" but I have found it is really just focal plane curvature. I think the Nikon 58 is similar in that respect... for a lot more money