Which one is a sharper image?

[brian_m.1]

<p>I have a set of pictures taken by two classic cameras, one an RF and the other an SLR. It is hard to duplicate the exact conditions but the season, lighting and time of the day were almost the same but years apart. The lenses are of course different. I am pretty sure I had the focus right on all. To me B looks sharper than A but could it be the sharp sun glint and somewhat closer focusing leaves that impression? C and D are a little harder to tell, perhaps because of different perspective.<br>

<strong>A</strong><br>

<img src="https://dl.dropbox.com/u/415093/07950026.jpg" alt="" /></p>

<p> <strong>B</strong><br>

<img src="https://dl.dropbox.com/u/415093/74420011_011.jpg" alt="" /></p>

<p>Here is another set:</p>

<p><strong>C </strong><br>

<img src="https://dl.dropbox.com/u/415093/74420002_002.jpg" alt="" /><br>

<strong> D</strong><br>

<img src="https://dl.dropbox.com/u/415093/07940008.jpg" alt="" /></p>

[Robin Smith]

A short depth of field always makes the sharp part look sharper. That's one of the reasons to pick a wider aperture.

[Mike Gammill]

<p>The point of focus in the first two photos looks different. Hard to really judge sharpness here. I would try with the subject perpendicular to the film plane rather than at an angle so depth of field won't be an issue. I do applaud your efforts to provide useful information those who have the same question. </p>

[john_robison3]

<p>It is hard to tell from a small online picture much about sharpness. In fact the term 'sharpness' can have several components. Are we talking about resolution, as in a lens ability to image very fine detail? Or the overall impression produced by both contrast and fine detail. How were these samples put up here for evaluation? Are these scans direct from the original negative or slide, or from prints? Was all the detail available from the original negative captured in reproduction? In the old days I always wanted to examine the original negative or slide when questions arose comparing one lens with another. Of course, with a negative even that does not provide much information about the lens color rendering.</p>

[brian_m.1]

<p>John, your points are well taken. We can probably come up with an objective metric to characterize sharpness. I am looking at it purely subjectively. The pictures were taken on Fuji 200 film and developed and scanned at Costco. In case people are wondering, A and D were taken with the Canonet 28 with a 40mm f2.8 lens and B and C were taken with the Yashica FX-103 with a 55mm f1.7 lens. Overall, I think Yashica comes out sharper but as Robin mentioned may be its longer focal length is a factor. I recall it can focus closer than Canonet as well.</p>

[lex_jenkins]

<p>The objective metrics are well established: resolution and contrast for the minimally accurate subjective assessments; <a href="http://www.luminous-landscape.com/tutorials/understanding-series/understanding-mtf.shtml">modulation transfer function (MTF)</a> for objective testing.</p>

<p>"Sharpness" is a subjective term that's almost as imprecise as "IQ". At least with sharpness we're usually referring to the combined results of resolution, contrast, color saturation (affected by flare resistance), and various types of distortion, usually as these apply to the finished product - the photograph.</p>

<p>The matter is confounded by the medium used to display the photograph: the monitor, which varies with calibration; the camera LCD, for which there is no standard or calibration; inkjet or other non-light-sensitive prints from varied sources; optical enlargements from negatives; scans of negatives, which rarely do justice to or equal the potential of optical enlargements with b&w negatives.</p>

<p>With "IQ", it can mean whatever the user wants it to mean, including how many megapickles are stored in the sensor.</p>

[lex_jenkins]

<p>The Canonet 28 and GIII QL17 lenses are fairly vulnerable to flare from low light in the frame, typical of dawn and dusk conditions. The front element is very close to the front edge of the barrel. These lenses are coated and resist flare pretty well, but may suffer from veiling flare in some conditions - this can cause loss of contrast and saturation, which may be perceived as "unsharpness".</p>

<p>Most SLR normal lenses like the 55mm f/1.7 Yashica have a bit of barrel extension beyond the front element, which helps act as a short lens shade. In fact, some lenses like the 50mm f/2 AI Nikkor have such deeply recessed front elements an optional hood is almost unnecessary.</p>

<p>That said, I can't see any significant differences in the four photos you posted, and any real differences are masked by the differing compositions. All appear very similar in apparent sharpness - focus, resolution, contrast and color saturation. Not enough difference to worry about.</p>

[JDMvW]

<p>I think that there are just too many uncontrolled variables here, not excluding the effects of reducing images to fit on the www, to make any meaningful judgment about the lenses and cameras themselves.<br>

Like Lex, I don't see any reason not to "worry about" it.</p>

[brian_m.1]

<p>Good to know Lex, JDM. I love both cameras. I did a walk around with the Canonet this past weekend and it acquitted itself nicely. Wonder if anyone noticed. A and B were taken 3 years apart almost to the day. Notice a piece of casting is missing in A!</p>

[lex_jenkins]

<p>I had the Canonet 28 and GIII QL17 at one time. I always wished the GIII QL17 had the 28's knurled focus ring rather than that lever. The lever never did suit me for quick focusing. The 28 is very underrated - the build quality is identical to the GIII QL17, with only a few differences in features.</p>

[brian_m.1]

<p>Some of you might have seen this picture. Took it with my EOS 650 and the 50mm/1.8 lens</p>

<p><img src="https://dl.dropbox.com/u/415093/98480004.jpg" alt="" /></p>

<p>Now (almost) the same scene with the Canonet weeks later. Have to say it is just as sharp, if not more. Funny thing, the red car is still there.</p>

<p><img src="https://dl.dropbox.com/u/415093/07950007.jpg" alt="" /></p>

[rick_drawbridge]

<p>Ansel Adams: "There is nothing more useless than a sharp photograph of a fuzzy concept".<br>

H.Cartier Bresson: "Sharpness is a bourgeois concept."</p>

<p>They all look pretty "sharp" to me, <strong>Brian</strong>...<br>

</p>

[rick_drawbridge]

<p><strong>Brian</strong> , on reading my above post I think it might be construed as criticism of this thread; it was not intended to be such. It's just that I get a little saddened by the repetitious emphasis on the "quality" of the lenses we use, be it the resolving power, "bokeh", colour rendition, corner sharpness, distortion...et al. For most of us, the lenses we use are perfectly adequate for the work we do; I would be at a loss to describe the difference in results between a Yashinon, Canon or Rokkor of similar specifications and, so far as the images we post on this Forum, lenses of much simpler construction supply much the same results. In specialised fields such as architecture or science the finer points of a lens performance does become more critical, but I just feel we should be concentrating on the enjoyment we derive from using these old cameras, and not a pre-occupation with comparing the micro-fine points of their performance.</p>

[User_502260]

<p>I don't agree that MTF testing is more objective than other measures. From what I remember the origin of MTF is that it was used to weed out obvious lemons when lenses were mass produced. It is no more objective than lp/mm or % contrast. Years ago Modern Photography used lp/mm (later in combination with contrast) in its testing while Popular Photoghraphy used % contrast. Some lenses like the famous 50/2 Dual Range Summicron have very high resolution but low contrast. Other lenses like the 50/1.8 Zuikos for the OM cameras have lower resolution but higher contrast. It is the final result you are looking for and there is usually more than one way to design a lens of a particular focal length and type to get there. Digital scanning and printing allow us to make many useful adjustments which were either more difficult or not possible with film before. The result is that some subtle diferences between different lenses and different films can be less obvious. </p>

[Robin Smith]

<p>The advantage of the MTF is it simultaneously shows resolution across the whole field rather than just at particular points (center, edge etc). Also it does not rely on subjective measures of the eye to tell whether two lines are resolved or not as it shows the function as a whole.</p>

[Robin Smith]

<p>Sharpness is a subjective term, but the MTF test itself is objective.</p>

[cosmin_saveanu]

<p>The sharpest lenses will give fuzzy pictures if focusing is wrong or the camera is not perfectly steady. Both situations happen all the time with me, but the camera quality can sometime make it happen less often. I agree with Rick, in most situations the performance of the lens is more than adequate for what we do. However, testing lenses is part of the pleasure of using a new - classic old - camera. It is also useful to know what look the pictures will have with a given camera - for heavy nice vignetting I love the Olympus XA, for reliability, the Nikkormat with its f/2 Nikkor is perfect. For distortion and excessive contrast, nothing beats a Yashica T4 Super compact. And when image quality really matters an old TLR, like the Ricohmatic 225 is perfect. Testing cameras and lenses is part of knowing the material and its limitations, so that you know what are your expectations.</p>

[brian_m.1]

<p>>>Sharpness is a subjective term, but the MTF test itself is objective.<<<br>

We need subjective tests, i.e. what people see with their eyes. Nobody computes MTF in their head then decide if they like the picture. This happens in other fields too. JPEG and MP3 standards were developed using subjective tests based on the so called Mean Opinion Score.</p>

[Troll]

<p>If you can't tell which image is "sharper," it doesn't make any difference. Quit obsessing and just go make some pictures.</p>

[Troll]

<p>If you can't tell which image is "sharper," it doesn't make any difference. Quit obsessing and just go make some pictures.</p>