Talk to me about this bokeh...

[Derrel]

We want to keep in mind there is a lot of information on the depth of field issue. There is much less information available on the Internet regarding background blur. There are just a few photo illustration examples that I know of. Imagine that we have a 50 mm lens, an 85 mm lens, a 200 mm lens, and a 300 mm lens. If we shoot at F4 with each of the lenses and move the camera farther and farther and farther away from the subject, to maintain a full body person in the picture, what you'll find is that the longer the focal length at F4, the more blurred the background. Now, if you do the mathematics and use 1/4 of the length value for each lens, it will tell you the approximate aperture width in millimeters at F/4. Yes, the exposures will be the same, because the relative value is f/4, but the larger the physical width of the aperture-- the blurrier the background. This is why for several decades, the 300 mm lens has been used to blow out backgrounds, because even at F4 it has a very blurry background rendering when used on people, at normal people photographing distances

 

[Derrel]

At f/4 with a 50 mm lens, the aperture width is roughly 12.5 mm. With a 200 mm lens set at f/4, the actual aperture width is 50 mm. With a 300 mm lens set to f/4, the actual aperture width is 75mm. The wider the actual, physical,measured aperture width, the less chance of diffraction, and the blurrier the background rendering. Basic physical science. Not talked about much.

 

[mrca]

Derrel is absolutely right on lens length having an effect on background blur. How close you are to the subject and getting the bg further away does as well. However the QUALITY of the oof is a function of the lens and two of perhaps the absolute best are the Nikon 105 and 135 2.0 dc's. Older lenses, the 135 continues to be made unchanged since 1994. Guess why, it's amazing. It is a 7 element lens, that produces not only perhaps the best bokeh out there, but also superb micro contrast, which is critical for b&w since it is all about contrast. Don't listen to the gear heads and test chart people who worry about CA, it is correctible with a click in Lightroom. I am usually shooting at at least 3.2 in the real world to get a safe subject dof and it is gone there and you can adjust the bokeh ring to keep gorgeous bokeh. Like many low element lenses, you just have to watch high contrast subj to bg. Also, what is the quality of bokeh, is it creamy, cats eye, octoginal? Google 135 2.0 dc and look at the photos. The "feel" of the bokeh is soft and gentle, not harsh and disturbing. Compare it to the bokeh of a 70-200 and the difference is huge. For a 100 range, I use the zeiss 100 2.0 makro planar and the 85 1.4. Stunning bokeh and also, when closed down, subj bg separation with incredible micro contrast often referred to as zeiss pop. The 100 dusts the twice as expensive Nikon 14 element 105 1.4 at half the price. The 100 is incredibly sharp on my d850, so sharp it will make your eyes bleed. Here's a 100 shot, stopped down and hand held.

 

[zulu42]

For @Derrel and others:
Recently I saw a post where a lens was described as having a rapid fall off from the in focus areas to the OOF areas.

What physical characteristic of a lens creates a more or less rapid fall off from in focus to out of focus areas? Is that component affected by focal length?

Apologies to the OP for inserting my own question in your thread.

 

[vintagesnaps]

When I first saw the thumbnails I thought it looked like someone's trying to beam the kid up to the mother ship. Seriously, I see a whitish/lighter brighter streak vertically above and below the subject thru the image - when I click on it to enlarge it I don't see it.

The other two are so edited/processed... how in the world could you get a butterfly to pose right in front of someone like that?? and keep a child there waiting til a butterfly happens a long? (Or hold a cat long enough? lol) The blue butterfly? fake, and the pink foreground? yeah, boofed up cotton candy sounds about right... And why is the girl glowing?? where is that light coming from??

I can see the edge around the blue sweater of the other girl. And again, the light around her makes no sense. I agree about the backgrounds and foregrounds not going together, at least it's hard to tell. I don't know what was edited together or how or why people like it.

At best, it's a photo illustration.

 

[Derrel]

For @Derrel and others:
Recently I saw a post where a lens was described as having a rapid fall off from the in focus areas to the OOF areas.

What physical characteristic of a lens creates a more or less rapid fall off from in focus to out of focus areas? Is that component affected by focal length?

Apologies to the OP for inserting my own question in your thread.

I am actually the person who,just a few days ago on TPF, mentioned the rapid focus fall off from the focused area to the defocused area; I was referring to the 90mm f/2 Fujinon lens for APS-C digital, which is the equivalent of a 135mm focal length on full frame. The 135mm f/2 Canon and Nikon lenses are well-known for this rapid defocusing. Longer focal length lenses do this, yes.

One does not get rapid defocusing from something like 28mm,35mm,or 50mm lens at normal distances of 5m to 30m. There is a reason that 135mm, 180mm, 200mm, and 300mm prime lenses have been used for people photos and fashion photos for a long time. These focal lengths offer the photographer the ability to easily use selective focus, or shallow depth of field, or background blowout.

 

[vintagesnaps]

If you use a wide open aperture (or close to that) you may see some falloff; it depends on where the camera was focused how much of that is in front of the subject and how much is in the background.

But the photo with the pink smooshy stuff along the bottom edge makes no sense; the falloff woudn't look like that because it would be a little more gradual. The subject isn't that far away from the pink mushy foreground. If a subject was further distance from the camera and an object was sticking up in the foreground close to the camera, you could get a similar look (edot - depending on vantage point and framing). But not that soft, that was just processed to look like fluff.

 

[JBPhotog]

And to add to the mix, all lenses have the same DOF at the same aperture, IF and only IF the subject is the same size at each focal length, whether a 24mm or a 300 mm.

All of your examples look as if they were shot that way. Sure some post was done for colour grading but in essence they were shot with the intended compression and background bokeh.

 

[Derrel]

HERE you go: one of the BEST on-line tutorials explaining how the actual,physical width of the lens aperture affects the degree of "background blur". Not the depth of field, but the degree of background blur. This article is one of a series written by Bob Atkins, a man who understands the science behind photography. His articles are not full of bullshit, like so many web-based things these days.

This is from 2006...before all the instant expert types started flooding the web with BS.

Bokeh and Background Blur - Bob Atkins Photography

Take careful note: the SAME, exact size of camera box is shown in all three photos! The SAME subject size was achieved by adjusting the camera-to-subject distance, using a 50mm lens, and 85mm lens, and a 135mm lens.

The aperture widths were 17.8mm in width on the 50mm lens; 30.4 mm on the 85mm lens; 48.2mm on the 135mm lens.

LET THE PHOTOS PROVE TO YOU that the degree of background blurring is greater with a longer lens, with a physically wider aperture diameter, even if all three lenses are shot at the same f/value.

And take note that the depth of field is considered "the same". But as I said earlier today, depth of field and background blur are NOT the same things!

"The following images show the effect quite clearly. All three were shot to produce the same magnification of the camera box and so have essentially the same depth of field (region of "acceptably sharp" focus), but the image shot with the larger physical aperture (longer focal length) lens shows the greatest degree of background blur."

 

[zulu42]

Thanks Derrel.

Cleared up my mistaken assumption that one controls the amount of blur in the background primarily by controlling DOF.

 

[TreeofLifeStairs]

It never occurred to me that the aperture was wider at the same f/stop in longer lenses. I just checked my 16-35 which has a constant 2.8 aperture, and sure enough, as I increase the focal length the aperture opens up more. It makes complete sense though since with a wider FoV there's more light entering, but as the FoV narrows less light is collected and therefore the aperture needs to open more to compensate.

Wow, mind blown.

 

[JBPhotog]

This is precisely why, given the same DOF required to get critical elements of the subject sharp at the same magnification, a longer lens produces softer backgrounds. And people wonder why I shoot portraits with my 300 mm f2.8?

 

[JBPhotog]

It never occurred to me that the aperture was wider at the same f/stop in longer lenses. I just checked my 16-35 which has a constant 2.8 aperture, and sure enough, as I increase the focal length the aperture opens up more. It makes complete sense though since with a wider FoV there's more light entering, but as the FoV narrows less light is collected and therefore the aperture needs to open more to compensate.

Wow, mind blown.

If it didn’t, your 16-35 would likely be an f2.8-f4 lens

 

[Braineack]

Even at shorter lengths, the lens design can come into play.

I just took two quick shots at 35mm @ 2.8 using two different lenses:

The $1,700 Nikon 35mm 1.4g smashes the Tamron $1,200 24-70 2.8 VC in terms of bokeh.


So while your 85mm 1.8g is good in the regard, the 85mm 1.4g is gooder.

 

[Derrel]

In direct, side-by-side, same-time comparison against the 35mm f/1.4 Nikkor AF-S G-series, the Tamron 24-70 Di VC has unpleasant bokeh. However in the Nikkor G-series 85mm side-by-side tests I have seen, the f/1.8 model and the f/1.4 model are pretty close in terms of the type of bokeh and the amount of blur possible. In fact, I really do not think the 1.4 model has much advantage whatsoever.