Number of aperture blades - important or not?

<p>I understand that the number of aperture blades supposedly dictates the shape of the out of focus elements, and Nikon mention this in the instruction books as 'natural looking blur', however on my 105mm f2.5, the aperture shape is very angular, but the blur shapes are not harsh. The 50mm f1.8 in contrast has very clear hexagons when there are highlights in the background, yet the aperture is not so different to that in the 105mm lens...<br />So, on to my question. Concerning Nikon 35mm f2.8 lenses, would the older version with 9 blades be a better proposition than the newer versions with 6 blades for environmental portraits where there is very often bright areas in the background?</p>

<p>Your thoughts are welcomed. Ian</p>

<p>Well for point light sources it's desirable to have as round an aperture as possible. Large format lenses usually have many blades and produce an almost round aperture. Many rangefinder lenses are like this, too.</p>

<p>Nikon for some stupid bloody reason seems to take the pinheaded view that saving money on production is more important than their customers' images. Me, I'm glad of it in some way: the discerning photographers will always have nicer OOF point sources and others will keep scratching their heads as to why.</p>

<p>I don't think that the number of blades affects the quality of the OOF background as a whole, though. Usually the lens design determines that. I could be wrong.</p>

<p>Don`t know if I`d say the aperture blades "dictates", as it`s part of the phenomenon. I suppose the level of spherical correction dictates in most cases the brightness and shape of this highlight rings and where this "doughtnuts" are placed... background or foreground. Most Nikkors are corrected, thought, but imagine a portrait taken with an undercorrected lens; the background highlights will be smooth whatever the aperture shape is. It could explain why some of my Schneider lenses has almost squared apertures with no special bokeh issues and that differences between your lenses.</p>

<p>Well Karim, 9 blades certainly sounds better than 6 on paper - and cutting the amount of blades <em>sounds</em> like a cost cutting excersise.</p>

<p>I know that choice of background is important, and it is down to the photographer to take care, but I had an otherwise wonderful 'spur of the moment' candid of my parents with backlit trees behind them ruined by very intusive aperture shapes (50mm f1.8) and I am wanting to ensure that for future photos I can take portrais in somewhat demanding conditions without the photo being too badly spoiled by jagged hexagons. The new 50mm AF-S is loads better and I am glad I upgraded (9 blades).</p>

<p>EDIT:</p>

<p>Jose - I have just ordered today an adaptor to use my Schneider lenses on my Nikons - those Schniders are just so smooth and forgiving (but the cameras they fit are so fragile).</p>

<p>Odd numbers create sunstars with more "points". 6 blades: 6 sunstars. 7 blades: 14 sunstars.</p>

<p>Just one more thing.</p>

<p>Well, it depends on your personal taste, really. Like what has been mentioned before, it can affect the sunstars, but generally speaking, the more blades, the more round the point of light when you stop down a bit. If you see some exotic lens, the blades are super curved like a scimitar, it's amazing.</p>

<p>But less blades is not always bad. One of my 50mm lenses is even only five-bladed, and can't be happier with it and it still produces great bokeh wide open or stopped down. Better than the usual Nikon 50mm I think. Here are the samples:<br /> <a href="http://img44.imageshack.us/i/hallucinogen1.jpg/"><img src="http://img44.imageshack.us/img44/6504/hallucinogen1.jpg" border="0" alt="Image Hosted by ImageShack.us" /> </a></p>

<p><a href="http://img242.imageshack.us/i/hallucinogen2.jpg/"><img src="http://img242.imageshack.us/img242/8118/hallucinogen2.jpg" border="0" alt="Image Hosted by ImageShack.us" /> </a></p>

<p>What lens is that Albert? Out of interest, how many sunstars would a 9 blade lens have?</p>

<p>Ian</p>

<p>My 50/1.8AFD definitely is not my favourite bokeh-wise, it could be one of the worst bokeh performers in the Nikon line... don`t know if its extreme sharpness has to do something here.<br>

In the other hand, the 105/2.5 (2nd version and on), seems to be specially designed for portraiture, with an intentional undercorrection of spherical aberration in order to bring softer OOF background areas.<br>

I suspect (my knowledge here is close to nothing) that the relationship between aberration corrections and contrast is too close , then even undercorrected, this lenses are not as pleasant and soft as some vintage lenses.</p>

<p>I understand that Bokeh is also related to the MTF performance. If the sagittal and meridonial lines are close together the out of focus areas are "better". I am not sure how this relates to number of aperture blades. It is interesting to note that the Nikkor 85f1.4 MTF and the 85f1.8 MTF are very similar (perhaps favoring the f1.8 lens in this regard, at f8), but most people are convinced that the 85f1.4 has better bokeh than the 85f1.8. Both lenses have 9 blades, the f1.4's are rounded. The rounded blades do not seem show up as better in the MTF charts at f8.</p>

<p>As a follow up, the MTF for 105f2.5 (7 blades) at Photodo has closer meridonial and sagittal lines at f2.5 than f8. Perhaps this is how the number of blades makes a difference.</p>

<p>K.R. has some info ...<br>

Look up on the site we dare not mention in ... /tech/sunstars.htm</p>

<p>I've never really noticed, have to start looking.<br>

Jim</p>

<p>As stated by others, the shape of the lens aperture determines the shape of out of focus highlights. And the optical aberrations determine the distribution of intensity within the OOF highlight e.g. even, or bright centre or bright edge.<br>

Nikon seem to be taking Bokeh more seriously. The older 60mm micro lens was appalling in that respect. The latest 60mm micro is good.<br>

Albert: that second image has OOF highlights with bright edges which in my opinon is ugly. That is caused by either under or over correction of spherical aberration (I forget which). A perfectly corrected lens has (I think) OOF highlights with even brightness.<br>

By the way, despite what Rockwell says, pointing a camera directly at the sun while looking through the viewfinder can be a very stupid thing to do.</p>

<p>Ian, the lens is Carl Zeiss Jena.<br /> 9 blades should give you 18-bladed stars.<br /> Leif, I don't really want to stray this topic to bokeh talk but the so-called rings are not that obvious compared to Nikon 50mms. But then again, too many people are too hung up on the shape of the bokeh but really, what you need to consider is the context of the bokeh in relation to the picture itself.</p>

<p>At work here and cameraless.<br>

Do odd number of aperture blades produce 2 X the number of rays while even number produce 1 X ? (as indicated by you know who)<br /> Any mathematician able to explain that to an old geezer? The only thing I can imagine is that in any even number sided polygon, opposite sides are parallel to each other. There my antediluvian college degree gets mired.<br>

Jim</p>

<p>Just out of interest, look at the aperture shape on the lens in the left hand photo. That is 9 blades instead of 6 on the later lenses (after 1960). That should help in diffucult lighting conditions, no?</p>

<p><a href="http://www.photosynthesis.co.nz/nikon/f35b.jpg">http://www.photosynthesis.co.nz/nikon/f35b.jpg</a></p>

<p>

<p>I`m afraid your question can only be answered by someone who knows (=use) both lenses. Therefore, I can`t.</p>

<p>Just some thoughts: Generally speaking, with this lens <em>highlight rings</em> could be more round than another with 6 blades... this is not as important as the lens design by itself. It is a 40+ years old lens, its optical design has been changed several times since then. Another later design could perform better or worst, whatever the blades under the same conditions. Flare could be an issue with this lens.</p>

<p>Anyway, it`s a beautiful lens. Despite of its performance, I think it`s interesting, whatever it is. You should test it to see if you like how it performs. Nine or six blades... without checking the lens character... it only says that you`ll (probably) have more or less round "doughtnuts" on a given optical design, thought.</p>

</p>

<p>I have been looking for this site: <a href="http://www.rickdenney.com/bokeh_test.htm">Bokeh test</a> from Rick Denney. Perhaps you already know it.</p>

<p>Anyway, for those interested, check that different lenses (here he is testing 135mm lenses) perform in different ways; bokeh is interesting but aperture blades are not an issue at all... In my opinion, highlights should be always avoided. Highlight rings are (unless desired for effect as Albert says) almost always awkward to my liking.</p>

<p>He has a "conclusions" chapter, at the bottom of the page "<em>What have we learned?</em>" Point 10 is related to the topic of this thread. I came to the same conclusion.</p>

<p>New Nikon literature states that 9 bladed aperture "...achieves a natural blur for out-of-focus elements." Older literature makes little mention of this feature. I feel the statement is a product of a new Nikon sales department in fierce competition with Canon. These days, even the lens barrel color is a selling point.<br>

I agree flare "elements" will look rounder and the sun will have more rays when the lens uses more aperture blades. Impact on Bokeh is subjective.<br>

I would pick the older lens, for it was made at a time when the Nikon lens was the unquestioned top choice of the professional.</p>

<p> </p>

<p>Shoot wide open, and the aperture blades move out of the light path, so you get perfectly round OOF blur.</p>

<p>Nikon 50/1.4 @ f2.8</p><div></div>

<p>Nikkor 50/1.4 @ f1.4</p><div></div>

<p>If you really want to control bokeh, you will need a DC lens. Nikon 135 DC is a really good one. This was taken of my daughter with an 85 1.4. The background is nice, I think.</p><div></div>

<p><em>"wide open, and the aperture blades move out of the light path, so you get perfectly round OOF blur."</em><br>

Dan, I was thinking the same until I read that the old Nikkor likes <em>comets</em> and <em>rugby</em>... read the "Lens performance" chapter <a href="http://imaging.nikon.com/products/imaging/technology/nikkor/n38_e.htm">here</a> (examples #2 and #3).</p>

<p>Double post. My excuses.</p>

<p>Dan Brown (above) reminds us that DOF affects the quality of bokeh, something we know but often forget. It isn't all about aperture blades. I'm not trying to insult anyone's intelligence here--just reminding us of what we all intuitively know, if we haven't figured it out from experience.</p>

<p>--Lannie</p>

<p>As in point of lights, the characteristics change depending on the distance of those point of lights in relevance to the camera too. So sometimes looking at samples of different bokeh taken from the same distance is a bit misleading because once you decrease the distance, the bokeh might be uglier and the donut is getting more apparent, and vice versa.</p>

<p>By the way, is it only me that thinks the way they describe quality of donut rings as in "distracting" is a bit too strong of a word? I mean unless you are a bokeh freak, if you show an average Joe a portrait of a woman with some "distracting" little donut bokeh at the background, he probably won't even notice it any more than when given the same portrait with the "more pleasing" bokeh donut. (because the background will be blurry and only the portrait is sharp anyway)</p>

<p>Like what I said, if the picture is already good, none of these donut ring or onion ring, etc... will matter, really. Unless you are a pixel peeper bokeh freak. I appreciate good bokeh but saying a picture has a horrible bokeh or very distracting just because of the tiny, thin ring wrapped around the small point of light is a bit over the top I think.</p>

<p>But when it comes to bokeh I think I am more interested in the bokeh as in how it renders the background compared to how it renders to point of light spots. I am still looking for a (rather affordable) lens that can do swirly bokeh, but haven't found one so far.</p>

<p>Back to the original question, to the OP, don't sweat it too much, 6 blades or 9 blades, as long as you like how the lens feel and it is sharp enough and contrasty enough, just get it. Don't let the number of blades be the deciding factor of purchasing a lens.</p>

<p>But if you are specifically after a good bokeh, get a Tamron 90mm f/2.8 DiMacro, that one has excellent bokeh (although it's probably way too long for your taste)</p>