Help understanding this lens aberration, please ...

[dmr]

I've had the QL17 GIII for about 8 months now, and it's really my main low-light camera. I've noticed this on some shots, but it's never really shown up this badly until this shot I did recently.

I was doing some urban night scenes last week, and quite intentionally picking scenes that were only lit by relatively dim street lights, stoop lights, and such. I know this is unusual and somewhat challenging. I was shooting wide open or close to wide open.

In the shot below, you'll see that the street lights to the right have a distinct "football" shape to them, with the short axis of the "football" appearing to point to the center of the image.

Is this coma?

I've always thought of coma as a blob or "comet" shape blur with the long axis pointing to the center of the image.

Also, for those who are more familiar with the GIII, is this typical of that lens? I usually end up shooting stopped down a couple stops from wide open. This one was totally wide open, hand-held at 1/30. The right image below is a blow-up of the area I'm questioning.

Thanks in advance, gang. 🙂

 

[Sonnar Brian]

It may be astigmatism.

Look at the second one, with the "football shapes".

 

[jlw]

It looks like astigmatism to me, too. This aberration is characterized by radial lines being brought into a different plane of focus than concentric lines, as shown in Brian's second diagram.

It makes your highlights look like footballs because the lights are well-focused in one direction (where they look nice and compact) and out of focus in a perpendicular direction (causing them to look blurred and spread-out.)

Astigmatism is a manufacturing defect of the lens-grinding process, in which the lens' surfaces don't come out perfectly spherical, so it has more refractive power in one direction than the other. That means there's nothing you can do about it by, e.g., having your camera adjusted.

Incidentally, your eye can get exactly the same vision flaw, and your optician corrects it by prescribing a lens with a "spherocylindric" curvature (sphere plus cylinder) so that the lens applies more correction on one orientation than the other. I suppose it would be possible in theory to do the same thing with a camera lens, but it's probably not worth the bother -- either you learn to live with it or do your low-light photography with a camera having a better-corrected lens.

 

[dmr]

It looks like astigmatism to me, too. This aberration is characterized by radial lines being brought into a different plane of focus than concentric lines, as shown in Brian's second diagram.

It makes your highlights look like footballs because the lights are well-focused in one direction (where they look nice and compact) and out of focus in a perpendicular direction (causing them to look blurred and spread-out.)

Thanks, guys, I think I understand now. I wasn't really sure.

As you can probably see from the photo, I focused on the brick of the near building, and realized that the depth of field would be small. It looks like the effect gets much worse as the distance increases, as the farthest lights are very much elongated.

I guess I've found the limit of that lens, huh? 🙂

I would still be interested to hear any opinions as to whether this degree of astigmatism is typical of the GIII.

 

[clintock]

I see the same shapes in the lights when I use my Jupiter 3 at night..
google : sagittal coma flare

In this thread- http://rangefinderforum.com/forums/showthread.php?t=34064
I was thinking it was weird out-of-focus behavior, but it's the in-focus part that is making the weird abberations.

As to the GIII being normally like that, maybe- since the lens is likely a sonnar copy, as the J3 is. Maybe it takes more than six elements to make such a fast lens not exibit coma flare?

 

[Xmas]

The GIII is a six element four group which seems to indicate a double Gauss design rather then a triplet. A 40mm 1.7 in six elements is a 'racy' design in the 70-80s but I'd not have expected these results - Canon were quite conservative.

I'll have to try mine, in a similar situation.

Noel

edit
P.S. astigmatism is a feature of the lens curvature being spherical if they were aspheric the effect would be reduced for the same element/group complexity

 

[Xmas]

DMR

My 1.7 seem worse than yours, I now need to try a better test...

Noel

 

[Finder]

It is not astimatism. It is vignetting. The aperture is being vignetted and causing the out-of-focus image to take on its shape. This is why it seems to "rotate" around the optic axis. Think of it as vignetted Bokeh. If you stop down it will go away.

Why not astigmatism? That would also be affecting all the image. The brick work on the right has no sign of astigamism.