Sigma 16mm/f1.4 Querie?

[BAM]

I bought the Sigma 16mm/f1.4 DC DN/C for low light photography. So far I’m very impressed with it but when there are street lights in shot there is a globe or halo, for want of a better description, of light around the street lights. Circled in the images below. I find this a bit distracting.

Is this a by-product of the lens and the coatings on them and something I have to live with or is there a way of eliminating or reducing this phenomena?.

All hand held with a Sony a6300 at f1.4

 

1/160sec-f1.4-ISO800-ExpComp -1.7EV

1/15sec-f1.4-ISO500-ExpComp -1EV

1/15sec-f1.4-ISO100-ExpComp -1EV

1/15sec-f1.4-ISO2000-ExpComp -1.3EV

http://www.lowlightfotography.co.uk/25-london-30nov17/4594133244

Brian

 

[Chrissie]

Are these cropped or uncropped images?

 

My guess is, these are cropped images. I'm also guessing, that these halos are lense flares. In which case all flares would have to "point" at the uncropped image's center.

I've marked the visible image's center as the intersection of the green lines.

 

If my theory holds any water, the red "pointing" lines of the lense flares would all have to pass through the uncropped image's center, which they don't.

You will notice a distinct deviation from a circular shape of the halos, which gets more pronounced the farther away the light source is from the (uncropped) image's center.

 

If you take the largest diameter of any halo, then the "pointing" is done at right angles from the center of this largest diameter. I hope the annotated images can clarify this:

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

 

There are probably experts around who can tell you if there's anything you can do about lense flares.

[BAM]

Lens corrected in Lightroom 6 and then cropped.

 

Brian

[Chrissie]

You can try my theory on the uncropped images. Please do report back.

Thanks.

[Guest Jaf-Photo]

Halos and flares are par for the course when shooting street lights. The unusual thing is the defined circle. Flares usually extend from the light source towards the opposite edge of the frame, which is not the case here.

 

Is there any difference if you use the mechanical shutter vs the electronic?

 

I mean, it could be a result of the lens construction, for instance that it suppresses the flares, except close to the light source. Perhaps if you try finding similar photos taken with the same lens?

[BAM]

Chrissie

 

Many thanks for your time and explanation and the photo below is uncropped/un-straightened. You nearly got the centre point right  .

 

 

Rather than lens flare, I'm now more inclined to go with it being spherical or coma aberration which is common at a very wide aperture and apparently cured by stopping down to about f2, but as I'm hand holding, I was hoping to get away with f1.4 all the time. Just have to remember the mantra now, 'Street Light's Stop Down'. 

 

Brian

[BAM]

Is there any difference if you use the mechanical shutter vs the electronic?

Street night shooting I like to use silent shutter but I did turn it off to see if there was any difference but there wasn't. One thing I didn't try was stopping down the aperture which might be a cure to it.

 

I'm going round Lumiere London (link below) on Thursday so I'll see how I get on then.

 

 https://www.visitlondon.com/lumiere#2qW9KPfRbEA4QP2q.97

 

Brian

[Chrissie]

Rather than lens flare, I'm now more inclined to go with it being spherical or coma aberration which is common at a very wide aperture and apparently cured by stopping down to about f2, but as I'm hand holding, I was hoping to get away with f1.4 all the time. Just have to remember the mantra now, 'Street Light's Stop Down'. 

 

 

Brian, as a matter of courtesy I would have liked to agree with you the same way you did with me.  , but alas, I can't do that.:

 

This is not , in my view, a chromatic aberration, as those would show an object as having red and blue color fringes. Which your images don't show at all. The halos are all of the same and uniform hue, the same as the actual light sources.

 

And I also can't agree on spherical aberrations either, as these would not be confined to your highlights but would manifest themselves as a progressive loss of sharpness towards the edges of your images.

[Guest Jaf-Photo]

It's unusual. Coma and spherical aberrations often radiate and dissipate smootly. The look more like defined bokeh balls. I've never had this look with any lens, although I don't do much night-time photography.

 

I found a couple of normal-looking shots of street lights in this review:

https://lensvid.com/gear/sigma-16mm-f1-4-dc-dn-contemporary-lens-review/

[Chrissie]

@BAM : were you using any filters?  I found this:

 

 

 

Lens filters, as with lens elements, need to have a good anti-reflective coating in order to reduce flare. Inexpensive UV, polarizing, and neutral density filters can all increase flare by introducing additional surfaces which light can reflect from.

[Username]

You can easily see the shape of the exit pupil in the 

the images that have similar lights from near-center 

to near edge. Stop down about 1.5 stops. 

[BAM]

Chrissie - I don't use filters for night shooting.

 

Username - I am stopping down now when I need to.

 

But

 

I went around Lumiere London last week, twice. This time I had my Sigma 16mm attached to the front of my A7rII, in crop mode, as I wanted access to the 5-axis stabilisation so I could go hand holding at lower ISO and shutter speeds. No matter how hard I tried, even deliberately getting street lights into shot, I could not repeat the 'halo' phenomenon on the A7rII at any aperture or shutter speed setting. 

 

(Images straight out of the camera and just reduced in size for posting)

Taken with the a6300 @f1.4

Taken with the A7rII @f1.4

 

Before any one asks, anti-alias filter was turned off on the A7rII.

 

Lumiere London 2018 - 

Web Site - http://www.lowlightfotography.co.uk/27-lumiere-jan18/4594189834?preview=Y;use_flash=1

Facebook -https://www.facebook.com/lowlightfotography/posts/247647885773299

 

Brian

[Username]

The area you circled in red does indeed show the halo effect.  

   

I still believe it's a lens effect, but different cameras especially 

of different formats, will use different image processing which 

can either hide or reinforce various image flaws.  

    

If you look thru a lens at a blank wall and tilt the lens slowly 

off axis til the aperture changes shape and then goes black, 

you will see the exit pupil changing shape exactly as the halo 

changes shape as your subway lamps get further off-axis in

the scene with Santa Claus. 

   

IMPORTANT thing: The blank wall demo shows the shifting 

shape same as the subway lamps, not only at max wide open

aperture but also 1, 2, or even 3 stops down. It's only at rather  

smallish apertures that the shape shifting is negligible, or not 

at all visible. When you watch the off-axis changes to the exit

pupil, and compare them to the halos on the row of subway 

lamps, you should experience a true "ah ha" moment :-)    

    

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

    

I don't know whether this will further your understanding, but 

technically speaking, all the wide-ish apertures that show any 

shape change when you turn the lens off-axis are apertures at 

which your lens has some degree of corner shading ... even if 

extremely slight at certain apertures. After you stop it down to  

smaller apertures where the exit pupil does not change shape 

off-axis, you have truly eliminate all vignetting. Corner shading 

is caused by the way the round on-axis exit pupil delivers more 

light than the gibous shape of the off-axis exit pupil. To optically 

reduce the problem requires using larger, thus heavier and also 

more costly, front elements ... BECAUSE, the gibous shape is 

effect of the limited size of the front elements "shaving off" part 

of the round shape of the aperture. Thaz why it happens more

as you get further off-axis and encounter more and more of the 

physical edge of the front elements infringing on the aperture. 

   

Since the lens in question is speed-priority design, and a wide 

angle, providing front elements large enuf to nearly eliminate

this off-axis infringment would result in a heavy, monster-sized,   

monster-priced lens doomed to instant marketing failure. IOW,  

it never gets built in the first place. 

[Guest Jaf-Photo]

At least the halo looks normal in the A7rII photo, i.e. diffused.

[Chrissie]

I still believe it's a lens effect ...

 

Same lens, different bodies/sensors. One body/sensor does show the incriminated behaviour, the other doesn't.

 

Although I fought very hard to prove this to be a lens effect, the above makes it very hard to maintain that claim.