Originally posted by blackcloudbrew Interesting thread. I have a number of these lenses, some of which I've added only recently and not enough time on them to speak intelligently about them, but really I haven't be concentrating on bokeh for my 645 and 6x7 lenses. Sounds like I may have to though.
In looking at the PF ratings for a lot of the modern 645 DFA and FA lenses, most have 9+ numbers for bokeh. A few have rounded blades, notably the DFA 55, DFA 90, DFA 25 (good luck finding one), and the DA 28-45. One would think that rounded blades should produce a more pleasing bokeh. That has been my experience in my 35mm lenses between the DFA 100, and the DFA 100 WR lenses, (not rounded, rounded). The user ratings on the 645 and 6x7 lenses don't really show that though, as the ratings are all very high and very close together for rounded blades or not rounded blades. The exception being the DFA 28-45, as also noted above by SylvianB and PF members rating it 1 full point below most of the other modern lenses. Curiously, the best bokeh lens rated by PF members is the A 75 LS lens at 10, while the FA 75, which I believe has the same optical formula, only gets a 9. However, that may be attributed to the small number of reviewers for the 75 LS lens versus the other 75's. The point being here that I'm not really sure that user ratings can be counted on to be as sensing a difference in ratings between say 9.5 or 9.6, I can't. Not going to criticize the PF member ratings, because I've tossed out my share of them and it's hard to be objective or even have the same value of a number between raters.
So, I'm interested in the comments above because the consensus is not what I would have expected, based on what I've said above. Seems like I need to start shooting a lot of flowers and really observe and judge the bokeh of the lenses I have for myself. Until then, I'm going to follow this discussion and absorb the wisdom of those who post here.
My experience is that ratings are not reliable here. One reason is that faster and longer lenses make more blur, and more blur is often confused with the quality of that blur. Another is that bokeh, while real, is subjective.
I have done some detailed bokeh comparisons of a range of non-Pentax lenses, looking for patterns related to lens design (finding not too much), and looking for better ways to describe it. I came up with three categories of bokeh:
1. Bright-edge bokeh, where out-of-focus point highlights present as a disk with a bright edge. From my perspective, this is the worst bokeh, resulting in false details and swirls, but some old lenses that produce are revered, so what do I know?
2. Neutral-edge bokeh, where out-of-focus point highlights present as a smooth disk with a sharp, but utterly un-emphasized edge.
3. Faded-edge bokeh, where out-of-focus point highlights present as a smooth blob with no distinct edge. The old Sonnar design fits in this category, and is the reason I describe it as a wide, smooth brush.
And then there are other artefacts, including double lines, where the edges of out-of-focus details sum up to create false edges that make the out-of-focus areas noisy and distracting. I've seen that effect quite a lot on lenses praised for bokeh just because they are fast. The 200/4 image I posted above reveals a touch of this problem, which is why my praise of its bokeh is muted.
And some lenses seem to make the out-of-focus details look bumpy or clumpy, instead of creamy and smooth.
One of the myths I debunked was that the number of aperture blades is the primary determinant of bokeh. This is just false--many use fast lenses wide open when they want the bokeh effect, and no matter how many blades are in the aperture, it's always round when wide open. Sometimes, bright-edge bokeh is particularly distracting when few blades made sharp points, but in that case it's the bright edge that is the problem, and the aperture just makes it worse. With true faded-edge bokeh, nobody notices the number of blades, or can even tell how many there are. The worst lens in my test has a dozen or more blades and makes a very round aperture at all settings.
Here's a link to the test I conducted:
Kiev Cameras
If I tested lenses for the 645z in the 150-200 range, I'd do something similar to what I did before. And that was at least a dozen years ago, maybe more.
Candidate lenses:
645 FA 150/2.8
645 FA 200/4
67 165/2.8
CZJ Sonnar 180/2.8 (with P6-Pentax 645 adapter)
Maybe: 645 A 80-160, at 160.
And, for shorter lenses:
645 FA 45-85 at 85
67 105/2.4
645 A 120 Macro (don't own the DFA 90, unfortunately)
645 A 75LS
67--I have actually forgotten what lenses I own in the 75-90 range, but I have at least the 75-Shift and one other. I'll have to look.
Maybe: 645 A 80-160, perhaps at 100.
I don't think I still have that bottle of Lindeman's, but I do still have the kachina dolls.
Rick "but somewhat down the project list" Denney