The importance of non-equivalence

Hey, it's a ball, and it illustrates bokeh. Seriously though...

Apparently you had some trouble with the link I had at the end of an earlier reply to you. Click here.

Apparently you don't try to photograph anything that moves very fast. Also, if the subject is closer and moving relatively fast the AF is challenged more. Your feeble attempts to impress me don't show me that the lens you are using can keep up when even a slow target gets within twenty feet of where you are, and even if it did (which I strongly doubt) I wouldn't be happy with the bokeh.

which part of the bokeh dont you like ? are you not happy with your own bokeh ? do like the ball 😁

I already spelled out the issues I had with the bokeh on the bird photo.

I'm happy with the bokeh I get from my Nikon 70-200mm f/2.8G VR and Tamron 150-600mm f/5-6.3 VC; I'm a bit less impressed by the bokeh from my Nikon 24-70mm f/2.8G, but I accept that over the better bokeh from the Tamron 24-70mm f/2.8 because I want my photos to be more consistently in-focus and bokeh is less of an issue at shorter focal lengths at f/4 and smaller apertures (narrow DOF also contributes to too much of the subject being out-of-focus).

As to your invalid comparison, NR is not bokeh. In spite of that, here is a 100% enlarged crop post processed to be shown as such:

Another reason your attempted comparison is invalid is that the backgrounds are different (distances and details) and therefore not comparable.

One last point here. The lens you are using isn't even available for F-mount, which makes it a non-starter for me and illustrates the complete waste of time this avenue of discussion has been.

Nice to see there are things we can agree about.

At some point I'm going to spend my time actually comparing the DOF of the two shorter focal length lenses in the OP more precisely to evaluate the character of those two lenses (and I intend to throw in the soon to arrive Nikon AF-S DX NIKKOR 18-140mm f/3.5-5.6G VR I have ordered which cost me about $250, and is actually relevant to what I intend to do). The plan will be:

My big Tamron zoom lens at 150mm & f/8
My about to acquired DX NIKKOR at 140mm & f/8 (a very negligible difference in the pupil diameter)
My DX NIKKOR in the OP at 85mm & f/5.6
My NIKKOR 24-70mm at 70mm & f/4.5

Again, this doesn't represent how I intend to use the lenses, and the aperture diameters are about as close as I can get them and are not going to be identical. Ironically, the testing might be entirely moot as I may give up on using either of the DX lenses entirely if their AF motors can't keep up with the dogs, and I will know that before I do any equivalence/bokeh testing.

i never use NR on any of my a7iv files, the a7iv files are the cleanest ive seen from any camera, its why its the best camera on the market shooting extreme macro at 10x stacking 400 images. even the files from the a6300 are not bad but not to the same standard as the 74.

Let me rephrase. Noise and bokeh are not the same thing.

I don't have the luxury of trying to capture fast moving dogs at 1/320s. ISO 560 on a DX sensor is going to be a little noisy if viewed closely, and any cropping will magnify that. Add to that end of the day lighting and that's one of the reasons I am switching my Tamron 150-600mm over to my D850, and why I'm looking at a Nikon 18-140mm on my D500 to fill the gap I will have when the dogs get closer than 20 feet from me (shooting at 70mm or 85mm means there is going to be a lot of cropping of the DX frame).

I havnt shot any real fast moving subjects yet over the years most of my work is of dancers in my studio where i still action with flash, so this ambient light shooting is a real challenge for me at the moment. i did capture my first BIF image yesterday, wow that was tedious ,dont know if i will be able to do it for extended periods as my hands/arms suffer from over-use from being a carpenter over the years, its the reason ive bought the lightest 300mm zoom i could . but the a7iv defiantly lives up to its claims as being a baby A1 it can track a radical moving butterfly tenaciously, truly amazing.

I have also found that there can be a real difference in which side of the image plane that blur is found.
Some lens have good both front and back while another back might be its only pleasing bokeh
What has surprised me is that 1 or 2 of my lens I felt would have poor bokeh was not too bad take the sigma 150-600 sport.
It can really depend on the objects and their repeating patterns twigs and grass can pose a problem when combined with front and back bokeh.

Other lens maybe better for bokeh when stopped down, one of my favorites is the old 105 2.5 nikor.
When shooting flowers and stopped down to ƒ16 I really prefer it over more of my modern lens.

I look at it as if there are two facets to bokeh. The simple one is the unfocused point spread function of a single point of light, which is rendered in various ways by different lenses, but in a predictable manner for any given lens. When the distances and sizes of things in the field allow subject matter to be seen from some parts of the aperture and not from others, such as the twigs and branches you mentioned, this game of peek-a-boo takes the bokeh to a totally new level of distracting visual chaos which is unwelcome in most photography (but of course, it can be exploited artistically when it is part of the intended image).

Large entrance pupils are a double-edged sword; so many desirable photographic qualities benefit from a larger pupil, but the larger pupil, which makes greater subject isolation possible, also brings in a lot of undesirable peek-a-boo effects in the bokeh.

Also, many people are now finding that AF, which always seemed to be helped by having the lowest minimum f-ratio possible with DSLR AF, is now extremely counter-productive for AF in some situations, with image-sensor-based AF. Take a 600/4 and focus it on some bushes 14 meters away, and then try to AF on a a sparrow on top of a fence post 5 meters in front of the background, 5 meters from the lens, and the AF will act like it is already focused and there is nothing for it to do.

HAND.

I don't think you are visualizing the same issue. There is no "bird" for the camera to see; just the color of the bird, because it is so OOF that it is just tinting the frame, and is 95% transparent. DSLRs would better detect that something was there, under the AF point, and carefully and slowly try to find it. Mirror-less cameras don't even know something is there, with any AF point setting you use. The user must use manual focus override, or point away at something else that is relatively flat at the same distance, to get focus close enough to see that there is, in fact, a bird in the FOV. Hence, the double-edged sword of the large pupil (large relative to subject width and distance).

I don't know what's going on with the AF on mirrorless cameras in that regard. It sounds to me like there is no single AF point to fall back on, such as what happens in this comparison I made using my big Tamron zoom and D850:

AF point patterns make no difference in the situations that I was describing. I've cycled through all the AF modes and point patterns and full field, and the AF will not budge. Typically the close object (bird or branch) is a bit more blurred than your hummingbird feeder when AF gets stuck. With DSLRs, with the same lens (100mm pupil), AF will be slow and non-confident, but it at least tries when there is a huge blurred object in front of the background. Even with the ML camera set to seek AF when focus is impossible, it doesn't even try, apparently because it believes that it is already focused, even if it is focused on bushes and the camera is set to focus on animal eyes. There is a very real problem with image-sensor-based AF and very shallow DOF. I'm talking real DOF here (pupil/distance/size-based); not f-ratios.

I don't have one, but Canon's latest 300/2.8 lenses have a 141mm 106mm entrance pupil (at infinity), but they can focus only a couple of meters from the subject, so those lenses could really have an issue, I would think, near minimum focus distance.

HAND.

I'm sorry. 141mm was an error. I was thinking of a 400/2.8's pupil.
Even with a pupil of 106mm on the EF300/2.8L IS II, though, 2 meter minimum focus is very close for a pupil that size, and the angular size of the pupil at minimum focus distance is probably one of the largest in the industry, which is why I predict that it would be one of the worst lenses, at minimum focus, for the particular problem that I was describing, as it is already severe with my 400/4 with a minimum focus of 3 meters.. I assume that most of the lenses with pupils greater than 106mm do not focus close enough to have a larger angle than the EF300/2.8L IS II.

It would be great to have a database of all of these parameters that you could sort by your own criteria, like angular pupil size, especially if the actual close pupil size is used, rather than that at infinity focus, which would make comparisons more practical.

Yes, that's the scenario, although it could apply to some other situations as well.

Links to what?

HAND.

Yes, that could be even worse.in the type of scenario I was describing, unless the pupil size plummets as focus approaches 1.3m.

Yes, that diameter is a base factor of many variables in photography, for example blur circles ...

Lot of talk about blur and bokeh in this thread. Both are ways to describe aberrations caused by being out-of-focus (OOF) although the term bokeh is not usually applied to edges but more normally to OOF pinpoints of light or small well-lit objects. I found this article quite informative:

gegenwind.dk/blog/what-is-bokeh/

This too:

leicasocietyinternational.org/blog/2017/02/the-definition-of-bokeh