Doesn't look like bokeh . . . is it a reflection?

<p>Look at the blobs in the upper left of this image. The edges are sharp, so I don't think they represent out-of-focus highlights. Reflections between the lens and the UV filter perhaps? This image was taken with a 100 mm f/2 lens, wide open. There were a bunch of people in the background, in a banquet room with can-lights in the ceiling. As I recall, there was one of these lights fairly close to me, roughly in the direction of the sharp blobs.</p>

<p>Having trouble posting the actual image. Let's see if I can do it in this response.</p><div></div>

<p>They are out of focus highlights or light bulbs, not flare. </p>

<p>That's what I thought at first, but if that's so, why are the edges sharp? Compare them to the blobs lower down -- those definitely have fuzzy edges, and are what I am accustomed to seeing.</p>

That's the way highly-corrected lenses often render specular highlights.

<p>Indeed, this is exactly why people sometimes choose one lens over another - because, if it matters to them, they may find that one lens handles this sort of thing better than another. Hard outlines on OoF rendering is a fact of life with certain lenses used at certain apertures, and it's particularly evident on high contrast things (like points of light, etc). It doesn't matter to most people, but if it does, there are other lenses to choose from that don't do that, or which do it differently.<br /><br />Some of these artifacts are made much more obvious when you down-sample an image for small prints or web viewing, and then apply sharpening to the whole image. If you're selective in your post-production sharpening, you can - ahem - take the edge off a bit.</p>

<p>http://toothwalker.org/optics/bokeh.html<br>

http://toothwalker.org/optics/vignetting.html</p>

<p>I believe the common term is "cat's eye" bokeh.</p>

 

<p>Really? It looks like lens flare to me, as if you've got some bright lights out of shot, top left (which kind of tallies with the lighting of the subject), bouncing between glass surfaces. The same would apply to the illumination shining off the subject's breast, unless that's an unusual anatomical feature. The bokeh on the remaining out-of-focus light sources looks normal. That said, I'm used to faking this kind of stuff with computer graphics; my knowledge of it in real life is more limited. :-)</p>

<p>Nope, that's <em>definitely</em> out-of-focus stuff, being rendered harshly by the lens. This is completely consistent with many tests I've done using different lenses. Nikon's 50/1.8, for example, is <em>really</em> good at doing that trick. And yes, that harshness can look very different in a corner of the frame while looking not bad at all in the center. Different lens recipes handle it in completely different ways.</p>

As far as I recall, there was a room full of people behind her, mostly seated at tables. Any specular highlights would

have been reflections of things, because there were no light sources (such as Christmas lights) in the field of view. As

I mentioned in my original posting, I think there was a fairly bright, point-like source above and to the left of the

subject, which comports with the very directional lighting on her. I don't think the light source was all that close to the

lens axis, and, of course, it was a 100 mm lens, so it had a relatively narrow field of view. But it appears that there was

definitely something different about the source of those blobs than of the fuzzy ones below.

 

I'll check out the links to the cat-eye bokeh.

<p>The difference between the rendering of the objects at the top of the frame and the ones at the middle is the difference between how close they are to the edge of the lens's image circle. More extreme things are happening in that more distant corner, optically. The blob at the lower left is also starting to exhibit that same behavior, but it's lower contrast, so you're not seeing it as clearly.</p>

<p>Out of focus image disks from pin-point light sources have sharp edges. Big light sources deliver fuzzy edge OOF blobs. I'd conjecture that the light sources in the upper left corner were small-bulb halogens or LED's.</p>

Andrew: I have no clue as to the source of the unusual anatomical feature you noticed! She appears to have a strap of

some sort over her arm; perhaps we're seeing a reflection from a buckle.

 

Matt: support for your theory is found in the fact that I was using a full-frame camera, which would mean that the

corners were in fact near the edge of the image circle

<p>Also: if you look at the orientation of those football-shaped artifacts, you'll see that the axis running through each of them - if you connect the dots - defines a large circle. Not at all the same behavior you'd expect from typical reflection-type artifacts.</p><div></div>

<p>http://www.bobatkins.com/photography/technical/bokeh.html</p>

<p> </p>

<p>Matt: That's true! So this does appear to be a bokeh sort of thing rather than a reflection, which then leaves the question of why some of the footballs are sharp and others are fuzzy. It can't be just distance from the axis, as there is at least one fuzzy one in the upper left of the image, just inside the white circle you drew, and farther out from the center of the image than some of the sharp footballs. Would that be because, as Maris suggests above, the sharp footballs come from point sources? </p>

<p>Rob: Very helpful link. It looks like my mystery areas must indeed come from point sources. The lens, at f/2 was wide open. Atkins has an example that looks almost exactly what I have in my image.</p>

<p>Sorry guys, I'm still not buying it. The <i>shape</i> of the blobs is a function of it being near the edge of the frame and the light cones being truncated, and this applies both to the soft blobs (like the grey, top left) and to the hard-edged blobs which appear to have awful bokeh (little light in the centre, lots at the edges) which I still claim are flare. That both appear in the same place in the frame suggests it's not merely that the lens behaves very differently in different places. It's true that a large light source will partly smooth out nervous bokeh, but for this to be the issue, the smooth grey blob would have to be larger but closer to the lens than the sharp yellow blobs (technical term), and also the same shape as the lens's cat's eye bokeh, which would be a remarkable coincidence.<br />

<br />

In addition to that, the hard-edged blobs are effectively overlaid on the rest of the image, which is not behaviour you get from bokeh. And I still claim that the blobs on the subject's arm can only be internal reflections.<br />

<br />

Of course, it's hard to prove it without going back and setting up the same scene with a bigger lens hood. :-)</p>

<p>Andrew: Interesting you should mention a lens hood. I wasn't using one, so I was definitely setting myself up for internal reflections. That's what made me think of the can light as a possible source of the sharp blobs.</p>

<p>Incidentally, here, roughly, is what the rest of the room looked like. I was standing near the front, taking pictures of anything that looked interesting (this was a holiday party I was attending, and I just had the camera for my amusement). I had only the one lens with me, a 100 mm prime. This photo is more or less in the same direction as the other one, but somewhat to the right, which would be away from the ceiling light. The dark shapes in the foreground are heads, standing about where the woman in the puzzle picture is standing, though obviously not quite, as they aren't being lit so directly from the upper left.</p>

<p>Note the harsh directional lighting on the people, and the arcs of light you can see on the back wall. This is from can lights mounted in the ceiling, similar to the one that was near me. I was using only available light, which is why I was shooting at f/2, and why I had the prime lens.</p><div></div>

<p>Andrew: look carefully at her hair. The "football" nearest her head shows <em>behind </em>the strands of hair that intersect it. This doesn't mean we aren't also seeing reflections (as on the arm), but those footballs are OoF background stuff. I've seen it a thousand times. Do <em>not</em> make me get out two different lenses and starting taking picture of our Christmas tree! Because I will, and then Les Berkley will show up and give me trouble.</p>

<p>Matt: I agree that a lens's bokeh can look like this. I just don't think, from evidence elsewhere in the picture, that <i>this</i> lens's bokeh looks like this. I agree that the overlap with the hair looks odd, but - at least with this image resolution - hair is a very nebulous subject, and it's not enough to persuade me. It's also weird that the two types of blobbly things in the background are roughly the same size (for either possible explanation). I think we're going to have to agree to differ: I'm prepared to believe I could be wrong, I just don't think I am! Like I said, I doubt we'll be able to prove it without recreating the conditions. If you can get the same lens to show both types of bokeh at the same time in roughly the same place in the image, I'll be persuaded, but also impressed (and deeply confused about the physics). If it <i>is</i> bokeh, I'm also pretty unimpressed by any lens with that kind of bokeh characteristic - except possibly a defocus control lens misadjusted.<br />

<br />

Barry: out of interest, what was the lens? A review site may have a comment on what the bokeh is supposed to look like.<br />

<br />

This may not be the first time I've said something authoritative and completely embarrassed myself on this forum. :-)</p>

<p>Now you've done it, Andrew. We're lucky that it happens to be that time of year when we have a large pagan solstice celebration object in the living room, what with Festivus coming and all!<br /><br />So I deliberately got out my most obnoxious lens (when it comes to OoF rendering wackiness), a Nikkor 50/1.8 AF-D. I give you foreground, and background, complete with hard and soft blur artifacts from things at roughly the same distance (lights, and ornaments, on the same tree). A complex background with various shapes/sizes of objects can definitely produce a mix of bokehblob flavors.</p><div></div>

<blockquote>

<p>" it happens to be that time of year when we have a large pagan solstice celebration object in the living room, what with Festivus coming and all!"</p>

</blockquote>

<p><br /><br />Crikey, what Wordsmithery! On Novemenber 30, I simply get told to: "put the lights on the Christmas Tree" . . .<br>

That second photo seems to have a lot of Lens Flare - did you have a filter on the lens? . . .<br>

(Merry Christmas)</p>

 

<p>Come on, William, Festivus: it's the holiday for the rest of us! Actually, we do call it the Christmas tree, all pagan roots aside. <br /><br />No, no filter. That's just lots of reflective bits happening near a window. Not really much flare, to my eyes, but those 50/1.8's, used at f/2 (as in this case) are always a little iffy. Hey, it's a 25 year old lens to which I've been horrible, all its life. It's really quite amazing that it still works at all. But these days, I just break it out to show stuff like this. I'm happier with the 50/1.4 HSM when I actually mean business.</p>

<p>Sorry Matt.<br />Perhaps I might not have sent my message correctly. Perhaps being used to my straight and pedantic answers confused.<br />I was making a joke.<br />I have been reading this thread from the beginning, and was quite intrigued with your particular Bokeh experience with that particular Nikkor lens.<br />As the conversation between you and Andrew developed and you "threatened" to use the lens to demonstrate your point, I could not help myself but to suggest that you had Lens Flare.<br />I see no Lens Flare whatsoever<br />The comments about your Wordsmithery were a compliment, also. <br />I think Aussie humour might be misunderstood sometimes - when we appear to be criticising we often are in awe and complimentary.<br />I don’t do these “ : ) ", but on this occasion I should have used a smiley, perhaps.<br />The “Merry Christmas” was simple and with no hidden meaning and totally sincere - and should be interpreted as "Happy Holidays" if that better suits. It is a relaxed custom down here to use "Merry Christmas" generically and across all cultures and beliefs.<br />WW</p>

<p> </p>