Bad glass is... bad glass.

[Mike Buckley]

Yes but you live in DC where they've repealed the laws of physics

They've only repealed the laws of common sense.

[AlwaysOnAuto]

I agree with Mike, it's a double rainbow. I can see the start of it on the left and the finish on the right on my monitor.

[NorthernFocus]

Hmm. Not my recollection, so I looked it up. What I found was that rainbows occur at a deviation of 40-42 degrees from the angle of the sunlight. ...

The visible prism effect appears at that angle. In other words, if you set up a shot to take in the full arc of a rainbow centered in the frame, the sun will be directly behind you. If you center one end or the other of a rainbow in the frame at sunrise/sunset, then the sun would be coming over your shoulder at an angle of 40-42 degrees.

Bottom line is that in this shot, the full arc of the prism effect is centered on the center of the frame while the sun was at nearly a right angle. Not atmospheric. At the risk of further confusing the issue, here's a diagram that explains it.

[NorthernFocus]

They've only repealed the laws of common sense.

Unfortunately that is not unique to DC.

[dem]

It cannot be a rainbow. To catch both sides of a rainbow (never mind a double rainbow), you need a lens with the angle of view more than 84 degrees. That is anything wider than 24 mm on a full frame. We've got 55 mm on a crop sensor.

Edit: cross posting, Dan have already explained it with a diagram. Indeed, apart from the lens being telephoto, the sun is not behind the camera but on the left hand side. So definitely not a rainbow

[tao2]

Hi Dan,

It looks tae me like an artifact caused by extraneous light on the sensor edge. Didye use a lens hood? In particular conditions, ye see banding like this on some digital cameras. Extreme contrast, excess exposure comp are also causes.

Didye get carried away lifting shadows? That can be as bad as flare and veiling (which is what this problem resembles).

If it was the glass, ye'd have noticed it before now and it would be noticeable tae a greater/lesser degree, on many more shots. Ah think it's PP.

A rainbow? Naaah...

[tao2]

It cannot be a rainbow. To catch both sides of a rainbow (never mind a double rainbow), you need a lens with the angle of view more than 84 degrees. That is anything wider than 24 mm on a full frame. We've got 55 mm on a crop sensor.

Edit: cross posting, Dan have already explained it with a diagram. Indeed, apart from the lens being telephoto, the sun is not behind the camera but on the left hand side. So definitely not a rainbow

To catch both sides of a rainbow (never mind a double rainbow),

you need a lens with the angle of view more than 84 degrees

. That is anything wider than 24 mm on a full frame. We've got 55 mm on a crop sensor.

Nah, ye don't...



Canon Ixus L-1 APS, 26MM/F2.8 (FF equiv - 34mm) But yer correct dem...not a rainbow...

[NorthernFocus]

...It looks tae me like an artifact caused by extraneous light on the sensor edge. Didye use a lens hood? In particular conditions, ye see banding like this on some digital cameras. Extreme contrast, excess exposure comp are also causes.

I did have the hood on but it is indeed very high contrast.

Didye get carried away lifting shadows? That can be as bad as flare and veiling (which is what this problem resembles).

If it was the glass, ye'd have noticed it before now and it would be noticeable tae a greater/lesser degree, on many more shots. Ah think it's PP. ...

No shadow lifting. The opposite as a matter of fact. PP doesn't create something like this. It may enhance it, make it more visible. But the underlying issue has to be there to be enhanced. I've not shot more than a couple of dozen frames with this lens since I got it. This was some extreme lighting. But I shoot this kind of stuff a good bit and have never seen this happen.

Nah, ye don't...



Canon Ixus L-1 APS, 26MM/F2.8 (FF equiv - 34mm) ...

The 84 degree angle only occurs with the sun right on the horizon when the full semi-circle of the refraction cone shows up. The higher the sun is above the horizon, the less of the circle show up and the narrower the angle is. In this shot only a fraction of the circle of the rainbow is visible so the angle was clearly much less.

[george013]

I did have the hood on but it is indeed very high contrast.


No shadow lifting. The opposite as a matter of fact. PP doesn't create something like this. It may enhance it, make it more visible. But the underlying issue has to be there to be enhanced. I've not shot more than a couple of dozen frames with this lens since I got it. This was some extreme lighting. But I shoot this kind of stuff a good bit and have never seen this happen.


The 84 degree angle only occurs with the sun right on the horizon when the full semi-circle of the refraction cone shows up. The higher the sun is above the horizon, the less of the circle show up and the narrower the angle is. In this shot only a fraction of the circle of the rainbow is visible so the angle was clearly much less.

It took me a while to discover something like a rainbow in the picture. If what I see is what you mean, that rainbow is passing the 180 degree corner. Is that possible?

George

[Mike Buckley]

that rainbow is passing the 180 degree corner. Is that possible?

I hadn't noticed that until you mentioned it. I've never seen a rainbow do that so maybe it's really not a rainbow as I thought.

[DanK]

I just 'Sony a6000 rainbow artifact' without the quotes in Duckduckgo. The fifth hit was this: http://www.dpreview.com/forums/thread/3905765. Note that the lens is different, which suggests that if this is an artifact, the body is the issue. I didn't search further.

[NorthernFocus]

.... If what I see is what you mean, that rainbow is passing the 180 degree corner. Is that possible?

It is possible if you are above ground level. Which I was when I took the photo. Our house is about 1500ft up the mountainside. But as already discussed, you can only see the full circle of a rainbow if the sun is directly behind you.

I just 'Sony a6000 rainbow artifact' without the quotes in Duckduckgo. The fifth hit was this: http://www.dpreview.com/forums/thread/3905765. Note that the lens is different, which suggests that if this is an artifact, the body is the issue. I didn't search further.

Thanks for that link, Dan. It looks like the same phenomenon discussed in the thread. And with different glass. Well that's even worse news if it is something inherent to the body. There was a good bit of discussion regarding the vignette correction. I believe they meant in-camera. It definitely warrants further investigation. Well this will keep me busy for a while.

[travelfinn]

Dan
Check www.atoptics.com. You might find your answer there.
Ilmari

[george013]

It is possible if you are above ground level. Which I was when I took the photo. Our house is about 1500ft up the mountainside. But as already discussed, you can only see the full circle of a rainbow if the sun is directly behind you.


Thanks for that link, Dan. It looks like the same phenomenon discussed in the thread. And with different glass. Well that's even worse news if it is something inherent to the body. There was a good bit of discussion regarding the vignette correction. I believe they meant in-camera. It definitely warrants further investigation. Well this will keep me busy for a while.

If it's in the equipment you have it must be possible to reproduce it.

I know of the Nikon 50mm 1.8 that it produces a spot in the centre of the image. Normally not visible, but in a light background it is.

George

[travelfinn]

Sorry wrote false address www.atoptics.co.uk is the correct one

[NorthernFocus]

Sorry wrote false address www.atoptics.co.uk is the correct one

I was wondering what lasers had to do with the subject Thanks for the corrected link. Much more relevant though in this case I think we've established that atmospheric effects aren't the culprit.