• UPDATE



    The forum will be moving to a new domain in the near future (canonrumorsforum.com). I have turned off "read-only", but I will only leave the two forum nodes you see active for the time being.

    I don't know at this time how quickly the change will happen, but that will move at a good pace I am sure.

    ------------------------------------------------------------

Mirror box bokeh clipping haunts my dreams

ahsanford said:
So if you have a large aperture prime, a few questions come to mind:

1) Which Canon lenses demonstrate this phenomenon? Does the 135L as well? Does the Sigma Art 85 or 135 do this? Otus glass, perhaps?

2) How does one get round/soft/large bokeh balls from a large aperture lens? Just stop it down and forego the size/softness of the large aperture to eliminate the clipping? (Wouldn't that defeat the point of buying the fast lens?)

- A

1) As this is strictly geometric, with the same f number always having the same light cone angle from pixel to exit aperture, the placement of rear element makes no difference since it can't be larger than the rectangular hole of the mirror box. The mirror box intersects the light cones for some parts of the image starting at about f/1.8 at the extreme top for the 5DmkIV mirror box. The 5DmkII has a deeper cut out at the bottom but clips slightly more so is probably smaller. Haven't measured.

2) if you want perfectly round bokeh balls, you need to buy slower lenses than f/1.8 (with humongous front elements to avoid cat's eye bokeh), stop down to at least f/2 or use only the centre part of the frame.
 
Upvote 0
hne said:
2) if you want perfectly round bokeh balls, you need to buy slower lenses than f/1.8 (with humongous front elements to avoid cat's eye bokeh), stop down to at least f/2 or use only the centre part of the frame.

I'm not convinced that's true. Here's an 85mm f/1.4 lens on a FF sensored SLR, in this case a Nikon D810 -- I don't see any clipping.

https://www.youtube.com/watch?v=E4RXQNAeCeo

So either Nikon has a much larger mirror box (for the same size sensor?), or there might be a lens design component to this.

I tried to find some background lighting bokeh balls for other fast 85s (including Dustin Abbott's site) and struck out. If someone has 3rd party lens 85 1.4 samples on a Canon FF rig, please share.

- A
 
Upvote 0
ahsanford said:
hne said:
2) if you want perfectly round bokeh balls, you need to buy slower lenses than f/1.8 (with humongous front elements to avoid cat's eye bokeh), stop down to at least f/2 or use only the centre part of the frame.

I'm not convinced that's true. Here's an 85mm f/1.4 lens on a FF sensored SLR, in this case a Nikon D810 -- I don't see any clipping.

https://www.youtube.com/watch?v=E4RXQNAeCeo

So either Nikon has a much larger mirror box (for the same size sensor?), or there might be a lens design component to this.

I tried to find some background lighting bokeh balls for other fast 85s (including Dustin Abbott's site) and struck out. If someone has 3rd party lens 85 1.4 samples on a Canon FF rig, please share.

- A

Any of these any good to you?

https://www.flickr.com/photos/wefwef/5198757019/in/dateposted/
https://www.flickr.com/photos/wefwef/5182085737/in/dateposted/
https://www.flickr.com/photos/wefwef/5150349093/in/dateposted/
https://www.flickr.com/photos/wefwef/5205019328/in/dateposted/
https://www.flickr.com/photos/wefwef/5218131465/in/dateposted/
 
Upvote 0
jd7 said:
ahsanford said:
hne said:
2) if you want perfectly round bokeh balls, you need to buy slower lenses than f/1.8 (with humongous front elements to avoid cat's eye bokeh), stop down to at least f/2 or use only the centre part of the frame.

I'm not convinced that's true. Here's an 85mm f/1.4 lens on a FF sensored SLR, in this case a Nikon D810 -- I don't see any clipping.

https://www.youtube.com/watch?v=E4RXQNAeCeo

So either Nikon has a much larger mirror box (for the same size sensor?), or there might be a lens design component to this.

I tried to find some background lighting bokeh balls for other fast 85s (including Dustin Abbott's site) and struck out. If someone has 3rd party lens 85 1.4 samples on a Canon FF rig, please share.

- A

Any of these any good to you?

https://www.flickr.com/photos/wefwef/5198757019/in/dateposted/
https://www.flickr.com/photos/wefwef/5182085737/in/dateposted/
https://www.flickr.com/photos/wefwef/5150349093/in/dateposted/
https://www.flickr.com/photos/wefwef/5205019328/in/dateposted/
https://www.flickr.com/photos/wefwef/5218131465/in/dateposted/

They all show clipped OOF balls.
 
Upvote 0
ahsanford said:
hne said:
2) if you want perfectly round bokeh balls, you need to buy slower lenses than f/1.8 (with humongous front elements to avoid cat's eye bokeh), stop down to at least f/2 or use only the centre part of the frame.

I'm not convinced that's true. Here's an 85mm f/1.4 lens on a FF sensored SLR, in this case a Nikon D810 -- I don't see any clipping.

https://www.youtube.com/watch?v=E4RXQNAeCeo

So either Nikon has a much larger mirror box (for the same size sensor?), or there might be a lens design component to this.

I tried to find some background lighting bokeh balls for other fast 85s (including Dustin Abbott's site) and struck out. If someone has 3rd party lens 85 1.4 samples on a Canon FF rig, please share.

- A

hen you record a video in 16:9 aspect ratio, you cut off 5% top and bottom which likely hides rather efficiently also on your Canon DSLR.
The Nikon F mount has a 2.5mm longer flange focal distance and as such ought to be more susceptible to this. It could very well be the case that Nikon is using a wider opening on its mirror boxes than Canon does.

That said, the phenomenon is there on the...
D700: https://www.flickr.com/photos/geoff_sills/11939255546/
D750: https://www.flickr.com/photos/stephanrudolph/31296328703/
D800: https://www.flickr.com/photos/hishnessmahal/21399122753/
D800E: https://www.flickr.com/photos/charliemcmahon/26360753869/
D810: https://www.flickr.com/photos/51290195@N07/15574134554/
 
Upvote 0
hne said:
2) if you want perfectly round bokeh balls, you need to buy slower lenses than f/1.8 (with humongous front elements to avoid cat's eye bokeh), stop down to at least f/2 or use only the centre part of the frame.

My understanding is that the Sigma Art 85mm f/1.4 has perfectly round bokeh balls wide open, with no mirror box clipping. If that is true, is it due to the large front element (86mm)?
 
Upvote 0
bholliman said:
My understanding is that the Sigma Art 85mm f/1.4 has perfectly round bokeh balls wide open, with no mirror box clipping. If that is true, is it due to the large front element (86mm)?

Your understanding is incorrect. For example, see the set of Flickr images posted by jd7, which are with the Sigma 85A. A couple of images from that set are highlighted below.
 

Attachments

  • Clipping1.png
    Clipping1.png
    207.9 KB · Views: 145
  • Clipping2.png
    Clipping2.png
    219.1 KB · Views: 154
Upvote 0
bholliman said:
hne said:
2) if you want perfectly round bokeh balls, you need to buy slower lenses than f/1.8 (with humongous front elements to avoid cat's eye bokeh), stop down to at least f/2 or use only the centre part of the frame.

My understanding is that the Sigma Art 85mm f/1.4 has perfectly round bokeh balls wide open, with no mirror box clipping. If that is true, is it due to the large front element (86mm)?

Then your understanding is incorrect: https://www.lensrentals.com/blog/wp-content/gallery/nicks-wedding-photography-sigma-85mm-f1-4-examples/Sigma-85mm-Art-Series-Example-Photo-3.jpg
 
Upvote 0
ahsanford said:
Now that this blasted forum I love has told me what mirror box bokeh clipping is, I can't *not* see it in portraits.

https://petapixel.com/2017/12/01/shooting-portraits-christmas-lights-ordinary-bedroom/

IT'S EVERYWHERE! (85 f/1.2L II I believe for these shots)

- A

LMAO! Seriously. I kind of wish I can unread the thread that mentioned the bokeh ball clipping concept.
 
Upvote 0
neuroanatomist said:
bholliman said:
My understanding is that the Sigma Art 85mm f/1.4 has perfectly round bokeh balls wide open, with no mirror box clipping. If that is true, is it due to the large front element (86mm)?

Your understanding is incorrect. For example, see the set of Flickr images posted by jd7, which are with the Sigma 85A. A couple of images from that set are highlighted below.

I think all of those images were actually taken with the older Sigma 85 1.4 EX rather than the Art ... although I do not expect there would be much difference in terms of bokeh clipping.

bholliman - perhaps the comments you have seen about round bokeh balls are about the ability of the lens to produce "perfectly" round bokeh balls if located near the centre of the frame? I would expect all of the higher end 85s (at least) to be able to do that. I believe the clipping occurs as you get closer to the edge of the frame. See the images Neuro posted in this thread http://www.canonrumors.com/forum/index.php?topic=33939.0
 
Upvote 0
jd7 said:
bholliman - perhaps the comments you have seen about round bokeh balls are about the ability of the lens to produce "perfectly" round bokeh balls if located near the centre of the frame? I would expect all of the higher end 85s (at least) to be able to do that. I believe the clipping occurs as you get closer to the edge of the frame. See the images Neuro posted in this thread http://www.canonrumors.com/forum/index.php?topic=33939.0

I guess both 85L lenses aren't 'higher end' then:

http://www.canonrumors.com/forum/index.php?topic=33869.msg696463#msg696463

Lovely shot, btw, but the clipping is clearly present. And as best I can tell, it's not a corner or central issue, but somewhere in between. That shot shows cat eye bokeh up top (fine by me) and some lovely central circles, but the sides in particular have the bokeh clipped pretty hard. The bokeh sadly becomes a distraction when this happens. No knock on the photographer in this case -- it's a great shot that the lens couldn't fully realize, I guess.

- A
 
Upvote 0
neuroanatomist said:
bholliman said:
My understanding is that the Sigma Art 85mm f/1.4 has perfectly round bokeh balls wide open, with no mirror box clipping. If that is true, is it due to the large front element (86mm)?

Your understanding is incorrect. For example, see the set of Flickr images posted by jd7, which are with the Sigma 85A. A couple of images from that set are highlighted below.

Thanks for setting the record straight. That's what I get for speed reading the 85mm f/1.4 IS threads.
 
Upvote 0
ahsanford said:
jd7 said:
bholliman - perhaps the comments you have seen about round bokeh balls are about the ability of the lens to produce "perfectly" round bokeh balls if located near the centre of the frame? I would expect all of the higher end 85s (at least) to be able to do that. I believe the clipping occurs as you get closer to the edge of the frame. See the images Neuro posted in this thread http://www.canonrumors.com/forum/index.php?topic=33939.0

I guess both 85L lenses aren't 'higher end' then:

http://www.canonrumors.com/forum/index.php?topic=33869.msg696463#msg696463

Lovely shot, btw, but the clipping is clearly present. And as best I can tell, it's not a corner or central issue, but somewhere in between. That shot shows cat eye bokeh up top (fine by me) and some lovely central circles, but the sides in particular have the bokeh clipped pretty hard. The bokeh sadly becomes a distraction when this happens. No knock on the photographer in this case -- it's a great shot that the lens couldn't fully realize, I guess.

- A

I take your point, although it does seem the bokeh circles which are very close to the centre are round ... it's just that you don't have to get very far at all from the centre before you see significant clipping. It seems to me that compared to the Canon 85Ls, the older Sigma 85 EX (and the newer Art?) is showing clipping in a smaller area of the frame, primarily near the top ... or am I imagining it? Anyway, I don't mean to suggest the Canon 85Ls are not high end, regardless of bokeh clipping!
 
Upvote 0