black dots - an old problem?

Started Dec 7, 2008 | Discussions
Harry J Regular Member • Posts: 121
black dots - an old problem?

Hi,

just thought this might interest you. When the 1DMarkII came out, it failed my astrophoto test because of black holes next to stars of medium brightness. They were very similar to the black things shown now on forum. These black holes or dots/blotches showed up next to stars of medium brightness. It seemed that the phenomenon occurred when the star image was saturated, but if the stellar image was too large in pixels it overwhelmed the black blotch, so the brightest stars did not show these. The blotches were always on the same side (right) of the stellar image.
The tests were made with a tele and by tracking the stars

1DMark III did not have this problem.

Regards
Harry J

Ken Phillips Forum Pro • Posts: 16,364
Just blacken the sky ...

... ain't no big deal.
KP
--

http://www.ahomls.com/photo.htm
http://www.phillipsphotographer.com
Voted Best of the City 2004 by Cincinnati Magazine
I don't believe in fate, but I do believe in f/8!

'The urge to save humanity is always a false front for the urge to rule it.', H. L. Mencken

David J. Littleboy Veteran Member • Posts: 3,317
Does that mean...

Harry J wrote:

Hi,
just thought this might interest you. When the 1DMarkII came out, it
failed my astrophoto test because of black holes next to stars of
medium brightness.

1DMark III did not have this problem.

Does that mean I should wait for the 5DIII? Since I don't use tracking AF much, I won't mind the AF disaster.

-- hide signature --

David J. Littleboy
Tokyo, Japan

 David J. Littleboy's gear list:David J. Littleboy's gear list
Canon EOS M5 Canon EF-M 18-55mm f/3.5-5.6 IS STM Canon EF-M 11-22mm f/4-5.6 IS STM
Jimmy G Regular Member • Posts: 284
Re: Just blacken the sky ...

Ken Phillips wrote:

... ain't no big deal.
KP

Hi Ken,

Well, speaking as an astrophotgrapher, I would say that this is a pretty big deal.

Let me explain...it is a matter of course when taking long exposures of bright extended objects (that's astro-speak for things like nebulae and galaxies) that foreground stars will often become overexposed in the process. That the over-exposure reaches a point of full-well capacity for the affected pixels the sensor needs to do something with the excess accumulated electrons. In the past, using older CCD sensors as an example, this overflow would cascade down the sensor's (vertical) column causing a long bright bar whose width was that of the overflowing pixels further upstream. This effect is referred to as blooming, and its presence made for unaccceptable astroimages.

Now, over the years a variety of (both hardware and software) solutions and work-arounds were developed to deal with this problem. (I'm condensing here, I know.) Fast-fowarding to today, it is my understanding that modern CMOS APS sensors use a dual-well system to handle this pixel over-saturation. However, it would seem to me that this is not without its faults. It is my guess, due to the uniform distances being encountered between the blown light sources and the corrseponding black spots, that the black spots are nothing more than derived data in the RAW file being caused by an offset read-out error of this overflow information from the chip. (This could be put to the test by creating a sequence of incremental exposures of a light source known to cause this effect. The results would show at which point the overflow occurs causing the resultant black spot effect. It would, of course, be helpful that someone conducting this experiment make centroid to centroid pixel measurements between the source and spot to help quantify what might actually be happening.) If my assumptions as to the nature of the problem are correct, then I suppose a firmware solution that would allow the camera to properly handle the overflow information more appropriately so that it could create a more accurate RAW file could be possible. However, if, in fact, this overflow "charge" is physically affecting photosites down stream then I am less hopeful for a fix.

Regardless of all of this, my point is that whether it is bright column of light or black dot "echoes" of point sources, neither are acceptable in astroimaging. The bottom line is that there is no workable solution to recreate the lost delicate detail of an emission nebulae.

Hopefully, this explains some of our discontent?!

Peace,
Jimmy G

Curator Contributing Member • Posts: 868
Re: Just blacken the sky ...

Jimmy G wrote:

Well, speaking as an astrophotgrapher, I would say that this is a
pretty big deal.

Do you have any examples that show this issue?

Jimmy G Regular Member • Posts: 284
Re: black dots - an old problem?

Harry J wrote:

Hi,
just thought this might interest you. When the 1DMarkII came out, it
failed my astrophoto test because of black holes next to stars of
medium brightness. They were very similar to the black things shown
now on forum. These black holes or dots/blotches showed up next to
stars of medium brightness. It seemed that the phenomenon occurred
when the star image was saturated, but if the stellar image was too
large in pixels it overwhelmed the black blotch, so the brightest
stars did not show these. The blotches were always on the same side
(right) of the stellar image.
The tests were made with a tele and by tracking the stars

1DMark III did not have this problem.

Regards
Harry J

Hi Harry,

It seems you were not alone in your experiences...here's a link to a thread about another photog's experiences with a 1D2N and black spots...

http://www.fredmiranda.com/forum/topic/597080/0

...great photographic examples of the problem.

(Link posted as an FYI for the community.)

Peace,
Jimmy G

Jimmy G Regular Member • Posts: 284
Re: Just blacken the sky ...

Curator wrote:

Jimmy G wrote:

Well, speaking as an astrophotgrapher, I would say that this is a
pretty big deal.

Do you have any examples that show this issue?

Hi Curator,

I stumbled on this fellow's image over at POTN of the stars in Orion's belt...

http://photography-on-the.net/forum/showthread.php?t=609860&page=12

...what's to be particularly noted here is that his exposure is extremely short as far as deep sky imaging goes and the black spots are already in the image. Now, if you're familiar with this grouping of stars you'll know that the Horse Head Nebula is adjacent to the western most star in the belt, and that a longer exposure would be needed to reveal it's presence. As the exposure gets longer the next dimmer stars begin to reach full-well capacity and the black spots begin to increase accordingly. An good example of what this region should look like can be found here (not my image)...

(Remove the space in the URL below to visit the image. Had I not placed the space in the address the image would have appeared in this message, and it's not mine to post as such.)
http://www.cloudynights.com/
ubbthreads/attachments/2780357-B33.jpg

...shot with an astro-modified Rebel XT. A nice capture, no black dots. Now, if it turns out that DPP is helpful in addressing this, as suggested here...

http://forums.dpreview.com/forums/readflat.asp?forum=1032&thread=30247624

...then I am hopeful, as per my original posting, that a firmware (e.g. software) solution might address the issue. It's reassuring to see that a lot of folks are looking into this issue, hopefully Canon is, too!

Peace,
Jimmy G

OP Harry J Regular Member • Posts: 121
Image of the black spots in 1DMkII

Hi all,

it took me a while to dig up the picture from the Ö-files. The image below is a 100% jpg crop of the original raw image taken with a Canon 1DmkII (not III or 5DII) and a Canon 500f4 lens+1.4x, ISO800,300 sec

The image does not have the exif info attached, but from the supernova 2004dj in the galaxy NGC2403 one can easily figure out that the picture was taken during autumn 2004. The supernova is just down and right from the brightest star showing in front of the galaxy. All other stars shown in this picture except for the supernova are in our own galaxy. To be more exact the picture was shot on Oct 13 2004 at 03:09:35UT to be more exact.

The intention of this picture is not to show a beautiful picture, but rather to show that black holes did exist before 5DMkII. No background or other adjustments are made for this very reason. You can easily spot 7 stars with black spots. In brighter stars they are lost in the glare of the star (no such bright stars are shown here). These black spots can be truely annoying in prints.

regards
Harry J

Keyboard shortcuts:
FForum MMy threads