50mm f/1.8 ugly brokeh

Didn't even know these had hit the shops yet but the bokeh looks absolutely great... The wider aperture of the f/1.2 wide open still has the edge IMHO...

YES!!! Buttery smooth.........and a definite improvement from the 50 primes.

Ro1

Here are 2 I shot this morning with my Rebel & 50mm/1.8:

http://www.photography​-on-the.net/forum/showthre​ad.php?t=70071

thanks Roanjohn...
I will be looking into one of these.

Take a look at this thread!

Psst ... DO NOT look at the picture from the 135/2.0 L

I warned you.

Best regards,
Andy

Andythaler wrote:
Take a look at this thread!

Psst ... DO NOT look at the picture from the 135/2.0 L

I warned you.

Best regards,
Andy

My gosh Andy..
I remember reading that post, That 135 is indeed freaking awesome..

looking up price on B~n~H.

Andythaler wrote:
Take a look at this thread!

Psst ... DO NOT look at the picture from the 135/2.0 L

I warned you.

Best regards,
Andy

ARGH! Yes, THAT is good bokeh. Notice that there is no double-line effect, no bright edges on out-of-focus highlights, and everything looks like it was painted with a broad, smooth brush.

The others are not on the same planet.

But I do wish he'd conducted the experiment adjusting the f/stop to provide the same apparent blur on the background for each lens. People sometimes confuse more blur with better bokeh, but even though more blur is aesthetically usually better, it's not the same thing as better bokeh. Differences in bokeh should be different even at the same apparent blur.

Rick "who thinks the 100/2.8 Macro came in second in that comparison" Denney

With respect, the shape of the aperture affects the shape of the bokeh, not the quality.

Let's take a fixed highlight that is a point source of light against a dark background. The sky shining through leaves on a distant tree is an example. The perfect lens would render this out of focus highlight as a round disk that is evenly illuminated across its diameter. This is the bokeh provided by most of Canon's best lenses.

Some lens designs that are optimized for contrast, however, create a disk that is brighter on the edges than in the middle. I call this "bright-edge" bokeh. Other lenses create a disk with an edge that fades into the background. That is not a perfect lens for resolution and contrast, but it has good bokeh.

The number of blades only affects the shape of the disk, not the relative brightness of its edge. Yes, a round shape is better, but it's not the most important aspect of bokeh. Those shapes are quite apparent in lenses with bright-edge bokeh, but are not at all apparent in lenses that produced soft-edge bokeh. Thus, a round aperture in a lens with bright-edge bokeh just makes bad bokeh look better, but it's not better bokeh.

I have some lenses for rangefinders and large-format cameras that have perfectly round apertures at all settings that still have terrible bokeh. And even the five-bladed 50/1.8 lens has a perfectly round aperture when used wide open. But the out-of-focus highlights will still have bright edges.

The quality of the bokeh depends in large measure on the character of the background, too. Backgrounds with bright spots against a dark field will reveal bad bokeh in a hurry.

On a medium-format forum, a fellow posted a series of portraits, half of which were made with an excellent Hartblei 150mm/2.8 lens, and half of which were made with a Zeiss Jena Sonnar 180/2.8 lens. The apparent blur was the same for both. He invited us to offer our opinions on which was which, thinking we would not be able to. Several of us who had experience with Sonnars got six out of six, though we all thought the 150 was excellent (and it is a modified Sonnar design). A lens with good bokeh glows in the dark, but it is a subtle glow. It's the reason experienced art directors can identify the lenses on photos they see by their look.

A lens is a broad, smooth brush.

Rick "gestating his own comparison test" Denney

a great deal of it depends on the geometrical characteristics of the diaphragm, but I agree, not all of it. There are certain characteristics of the lens optics that change the appearance of the circles of confusion, thus different OOF.

mr.photoguy wrote:
Not trying to throw stones, but I am much more impressed with my Tamron's bokeh.

HTTP response: NOT FOUND | MIME changed to 'image/gif'

Is the bokeh on the 85 1.8 good also.

The 85mm f/1.8 has the best bokeh without being L glass IHMO.

I think it's nice..

HTTP response: 404 | MIME changed to 'text/html' | Byte size: ZERO

Okay, I've done a little research (this forum is going to get me fired, heh, heh). The Helios is a classic, highly symmetrical double-gauss lens with six elements in four groups, similar to the early Planar. It has better bokeh that the Canon double-gauss lenses because it has a slight amount of uniformly undercorrected spherical aberration, which is the secret to producing out-of-focus highlights with a soft, fading edge. Double-gauss lenses that are highly optimized for MTF performance (which includes most Japanese lenses) typically are, if anything, overcorrected for spherical aberration, which produces bright-edge and double-line bokeh. This is true for all highly MTF-optimized double-gauss designs, including Planars, Xenotars (which have a particular terrible reputation for bokeh along with their particularly stunning reputation for sharpness), Takumars, Summicrons, and Canon's 50mm lenses. Those that are more symmetrical and are not computer-optimized include the Helios, Volna and Biometar lenses from the old Second World.

The Sonnar is thought to be more highly corrected than the double-gauss designs, but it is not practical for SLR in normal focal lengths. The rear element would not clear the mirror. But the Sonnar is more fully corrected than the Planar double-gauss design, and has fewer air surfaces to cause flare, and it can be made to have a wider aperture (the first f/1.5 lenses were Sonnar designs). It will produce extremely sharp images at the plane of sharp focus while still providing overall uniform undercorrection of spherical aberration, and this is why the Sonnars are famous for their soft rendering of backgrounds. The good news is that Sonnars produced in East Germany are still available in M42 mount and are easily adaptable to the Canon, with the 135/3.5 being perhaps the most useful. Russian Jupiters are also Sonnar designs, but with more recent corrections and also poor quality control, such that they are generally underwhelming, which dilutes the Sonnar experience.

The Tessar was a four-element improvement on the triplets of the day. Tessars were very good at close focus (double-gauss lenses have to be specially corrected for it), but were limited in their speed and field of view. They make good enlarger lenses. But their bokeh is thought to be "clumpy", and I've seen that adjective from too many disparate groups of people to think it too far off the mark. Tessars including, well, Tessars, in addition to Xenars and Industars.

Thus, it would seem that highly optimized double-gauss lenses have the worst bokeh, followed by double-gauss designs are more symmetrical and not computer-optimized. The lenses with the best bokeh seem to be Sonnars, followed by double-gauss designs that are specifically optimized for good bokeh. Uniform undercorrection for spherical aberration seems to be the key. It's important to note that good bokeh results from poor performance--perfect rendering of out-of-focus highlights wold be a uniform disk with a sharp edge. I think that's why the Japanese lenses never really glow in the dark here (with the apparent exception of that 135 f/2, which was probably optimized for bokeh), because their designers are so interested in MTF performance.

Rick "who knows bad bokeh when he sees it" Denney

Rick - you raise some very interesting issues, most of which I had never thought about. I'm going to compare a couple of lenses at 50 mm to see just how their bokeh compares - the 50/1.4, the 24-70L, and the 28-105 f/3.5-4.5. I'm very curious as to how they compare. Obviously, the 50/1.4 will perform well at f/1.4 in comparison to the other two, but I'll try to set the test up to show the quality of the blurred circles - soft-vs-hard edges and such. People sometimes complain that the 50/1.4 is soft wide open (aren't they all?), but that is probably helpful in smoothing the bokeh as well.

I'm also going to google Tessar and Sonnar as I'm not familiar with all the various lens design names/categories. The Gaussian design is about the only one I can identify with some regularity.

I am seriously thinking about acquireing a couple of those lenses and a praktica camera.

Also, extremely sharp lenses actually produce less sharp images on the digital sensors... Has something to do with the minimal resolvable detail and the size of the sensor... And to think I only wrote an exam on that stuff 4 months ago

I would rather have a pleasant bokeh and slight softness, rather than extreme sharpness and distracting ghosting in the background.

Do you have any good links or suggested reading on those designs you describe?

It's nice cause the fence is linear already... it's really hard to screw up that.

But the branch in the upper left corner is all circly and distracting...