Question: why do lights show as stars at f/8?

Discussion in 'Open Gear Talk' started by pniev, Nov 18, 2015.

  1. pniev

    pniev Student for life

    May 13, 2013
    The following photo puzzles me. More specifically, the star effect in the light sources puzzle me. The photo was taken at f8 so with a wider aperture than you would expect with such an effect (beyond f11 to my knowledge). What could be the explanation?

    data: iso100, f8, 20sec, D750, sigma art 35mm lens

    • Winner Winner x 1
  2. Jock Elliott

    Jock Elliott All-Pro

    Jan 3, 2012
    Troy, NY
    Aperture blades?

    I dunno, but I like it!

    Cheers, Jock
    • Useful Useful x 1
  3. pniev

    pniev Student for life

    May 13, 2013
    Yes, the shape of the aperture blades at fast apertures causes this. But f8 is not that fast. I would expect to see these type of stars at f16. Initially I thought the exposure time (perhaps in combination with metering (matrix)) could have caused it but I could not find info on that. I thought about the polarizer that I may have used (not sure) but that shouldn't be a factor either. So I remain being puzzled..
  4. Richard

    Richard Top Veteran

    Feb 1, 2013
    Marlow, UK
    Do the stars have as many points as the aperture has blades?

  5. Lightmancer

    Lightmancer Super Moderator Subscribing Member

    Aug 13, 2011
    Sunny Frimley
    Bill Palmer
    It's diffraction alright. Technically they are "diffraction spikes". f8 is really quite slow (albeit not quite as slow as f11), so I would expect it from a mirror lens and a long exposure with intense points of light; it's less common though not unknown for a refraction lens and tends to be associated with the number of aperture blades. Straight-edged blades give a more pronounced effect.
    • Appreciate Appreciate x 1
  6. Ray Sachs

    Ray Sachs Legend

    Sep 21, 2010
    Not too far from Philly
    you should be able to figure it out...
    I think the physics of sun-stars is roughly as complex as the physics of bokeh. The aperture, the glass itself (optical design, maybe coatings?), the focal length, the distance to the point of focus, the distance beyond the point of focus to the light source, the light source itself, etc all play a role. I know that shooting directly into the sun I usually need to stop down as far as I can get to see a star and it's still usually pretty indistinct. Here are two of the sun taken with the same 25mm lens, same tiny f22 aperture, apparently at roughly the same distance, through somewhat different atmospheric conditions - one is quite diffuse and the other quite sharp.

    22144098818_65e3fbb405_h.jpg Stroud-89-Edit by Ray, on Flickr

    22860061295_71b97818b6_h.jpg Stroud-18-Edit by Ray, on Flickr

    Is the amount of moisture in the air the difference? Beats me...

    But here's one shot at f8 also, with an incredibly distinct little star in a reflection off of the top of this guy's horn - you can't really see it at this size, but if you click through you can view it as big as you want in flickr and it's a really cool little detail in this shot... This is at a much longer focal length (somewhere between 75-150mm) with a much closer subject and a much smaller point of light. For whatever combination of reasons, it worked really well at f8 and I suspect would have at f5.6 also...

    16439411104_9f3df7d6f1_b.jpg NYC - day 5-112-Edit by Ray, on Flickr

    Sometimes you catch them just right, sometimes you don't... I think all of the stars have more points than the lenses have aperture blades though. The top two considerably more. This bottom one maybe not but I think some of the points are obscured by other light coming off the horn... And way more in Peter's shot. I don't know of lenses with more than nine aperture blades...

    Last edited: Nov 18, 2015
    • Appreciate Appreciate x 1
  7. Chris2500dk

    Chris2500dk Top Veteran

    Dec 22, 2011
    Copenhagen, Denmark
    If the lens has an odd number of aperture blades you get twice the number of points (so 14 points for 7 blades), if the lens has an even number you get the same number of points (8 points for 8 blades).
    Something about the blade reflections lining up on an even number of blades so you only see half of them as they're exactly on top of each other, but the details elude me.
    • Like Like x 1
    • Appreciate Appreciate x 1
  8. pniev

    pniev Student for life

    May 13, 2013
    Thank you for your feedback. The Sigma 35mm art lens has 9 aperture blades so that would explain the 18 spikes. But I had never thought it would be visible at f8. But Ray's (btw great) shot of the saxophone player shows otherwise.

    I browsed a bit further based on Bill's remarks and saw this article:

    Although the article refers to telescopes, it also says "normally diffraction spikes aren't noticeable. When viewing objects directly you won't typically notice them. But with long photographic exposures they generally show up."

    Which explains the spikes in my photo but does not explain the spikes in the saxophone...

    Thanks for the clarification and your help!!