northern lights - weird thing - DPChallenge Forums

Tamron SP AF 28-75mm f/2.8 XR Di for Canon

11/09/2004 09:38:42 AM · #26

i am curious which filters you are talking about?
part of the sensor itself?

it couldn't be an effect of the northern lights themselves?
that crossed my mind. i hear there is a ton of energy involved in them.

strangeghost

Canon EOS-7D

11/09/2004 11:35:53 AM · #27

The sensor in your camera has layers of filters sandwiched on top of the actual chip itself, which is how digital cameras build up a color image. I know there's lots of good info out there on the web that will show this schematically, but I don't have time to search for representative sites right now. What we're theorizing is that these layers are interfering with each other, but it's not clear. Just for fun, you might check out Canon's website and send your pic to them and see what their engineers think.

Originally posted by soup:

jimmythefish

Canon EOS-10D

11/09/2004 11:46:15 AM · #28

I too tink it's probably the low-pass filter and/or some other layer in the sensors' structure being sucked back onto itself as an effect of the cold, and showing up only due to the small aperture.

11/09/2004 11:53:06 AM · #29

yeah - i have read about the layering of the sensor - i just wasn't sure if that's what was meant here. the UV filter has been mentioned a few times, and i am certain it isn't that. ( unless it's the UV filter itself - but that seems to have been ruled out. )

so is this something to be concerned about as far as damgeing the sensor? is the sensor already dameged? i mean i live in a pretty cold area - have used it in the cold lots of times. until i got the 50mm prime i never shot anything smaller than f:11 normally. this is the only time i have seen this happen.

i appreciate all the feedback. i will see if can get canon to look at it.

jimmythefish

Canon EOS-10D

Canon EF 70-200mm f/2.8L IS USM

11/09/2004 11:56:36 AM · #30

try a shot of the wall, inside, and then autolevels it.

Gil P

Canon EOS-1Ds Mark II

11/09/2004 12:03:24 PM · #31

This looks like a Jpeg compression issue to me.

Canon RF 24-70mm f/2.8 L IS

11/09/2004 12:07:54 PM · #32

here is a 6" exposure of a blue wall @ f:22
is see some dust - the pentagons - most likely are a result of a window to the right of the frame? ran autolevels.

it's nothing of the sort. the rings are there right off the camera with no edits. unless it's the cameras compression - they were shot in JPEG fine mode.

Originally posted by GilP:

This looks like a Jpeg compression issue to me.

Message edited by author 2004-11-09 12:08:20.

kirbic

Canon EOS R5

11/09/2004 01:12:02 PM · #33

Try that same idea with the camera thoroughtly equilibrated to cold temps, and see if the effect shows up.

Gordon

RIM BlackBerry Curve

Canon EF 85mm f/1.8 USM

11/09/2004 01:44:41 PM · #34

Originally posted by Gil P:

This looks like a Jpeg compression issue to me.

That would be extremely unlikely. JPEG works on the idea of block regions, so you tend to get square artefacts, not rings.

It strikes me as much more likely to be some sort of constructive/ destructive interference pattern, which would certainly jive with the comments about Newton's rings. From what I understand these rings form due to defraction between two lenses (generically, bits of glass) across a small air gap. There might well be enough of a gap between the filter and front element of the lens to cause this, with the wavelength of the light you are recording in the northern lights, along with the extended effects of the long exposure. The effect is going to be wavelength dependant, so that dark, greeny, essentially monochromatic light might be required for it to happen.

It probably doesn't have to be a double layered filter. Though the Canon 50 f1.8 front lens is quite recessed from the filter elements.

It is also a cheap, plastic element lens, which might have an impact compared to other, better constructed and coated lenses (it is a great lens for its price, but this could be a factor) In fact, it could be between elements/ groups within the lens.

Message edited by author 2004-11-09 13:55:36.

Gauti

Canon EOS-7D

11/09/2004 01:50:44 PM · #35

Soup can you, just for fun, post a 100% non-compressed crop of the artefact(wow, don't I sound scientific!)

Message edited by author 2004-11-09 13:50:59.

11/09/2004 03:03:28 PM · #36

640x640
100% crop
saved at 100%

about 225K

i'll wait till after dark to try the wall shot again ( so i can do the longer exposure ) - but i am leary of damaging the thing now...

Message edited by author 2004-11-09 15:05:40.

jimmythefish

Canon EOS-10D

11/09/2004 03:19:11 PM · #37

I'm pretty sure the 50 1.8 is entirely glass.

Originally posted by Gordon:

It is also a cheap, plastic element lens, which might have an impact compared to other, better constructed and coated lenses (it is a great lens for its price, but this could be a factor) In fact, it could be between elements/ groups within the lens.

Message edited by author 2004-11-09 15:19:34.