Reading mode: Light Dark

Hot Mirror Filters & Blue/Purple Fringe

Go to last post
J

JimC

Forum Enthusiast
Messages
406
Reaction score
0
Has anyone actually tried a Tiffen "Hot Mirror Filter", designed to filter I/R?

After viewing pictures from virtually all Digital Cameras, and seeing a "blue/purple" fringing in thin tree limbs against a bright sky (in overexposed areas of a photo), I found quite a few posts discussing the effect of a Hot Mirror Filter's use on various forums.

Supposedly, part of the problem is that CCD's are very sensitive to Infrared, and when CCD "Blooming" occurs, invisible infrared light is contributing to problem. The Hot Mirror Filter is supposed to filter virtually all infrared light, while allowing all visible light to pass, minimizing the blue/purple fringe (CCD Blooming, Chromatic Distortion, CCD Binning, or whatever you want to call this phenomenon ) seen in digital camera pictures, with high contrast, usually overexposed areas.

The most popular description of this filter is this:

"Tiffen 49mm hot mirror Lens Filter - Hazy, bluish casts in images are often caused by the charged coupled device of the camera, which is highly sensitive to light in the infrared range. This lens filter limits this bluish cast and allows for the image to have normal color."

While I've found lots of sites with descriptions of the filter, and plenty of discussion about the use of this filter (everyone says that it's supposed to help in the forums and articles that I've read)....I've found nobody that claims to have actually used one of these filters yet (everybody is quoting everybody else, but nobody I've found has any actual experience with this filter)....

Comments? I want to know if it REALLY helps, before buying one.
 
See the following, under IR Contamination...

http://www.cliffshade.com/dpfwiw/ir.htm

Larry
Jim Cockfield said:
Has anyone actually tried a Tiffen "Hot Mirror Filter", designed to
filter I/R?

After viewing pictures from virtually all Digital Cameras, and seeing a
"blue/purple" fringing in thin tree limbs against a bright sky (in
overexposed areas of a photo), I found quite a few posts discussing the
effect of a Hot Mirror Filter's use on various forums.

Supposedly, part of the problem is that CCD's are very sensitive to
Infrared, and when CCD "Blooming" occurs, invisible infrared light is
contributing to problem. The Hot Mirror Filter is supposed to filter
virtually all infrared light, while allowing all visible light to pass,
minimizing the blue/purple fringe (CCD Blooming, Chromatic Distortion,
CCD Binning, or whatever you want to call this phenomenon ) seen in
digital camera pictures, with high contrast, usually overexposed areas.

The most popular description of this filter is this:

"Tiffen 49mm hot mirror Lens Filter - Hazy, bluish casts in images are
often caused by the charged coupled device of the camera, which is
highly sensitive to light in the infrared range. This lens filter limits
this bluish cast and allows for the image to have normal color."

While I've found lots of sites with descriptions of the filter, and
plenty of discussion about the use of this filter (everyone says that
it's supposed to help in the forums and articles that I've read)....I've
found nobody that claims to have actually used one of these filters yet
(everybody is quoting everybody else, but nobody I've found has any
actual experience with this filter)....

Comments? I want to know if it REALLY helps, before buying one.
Click to expand...
 
Thanks for the link to the article... It sound like the one poster didn't see any difference, but I'm not convinced that the problem is not partially Infrared being picked up by the CCD...I have found some articles discussing this phenomenen in CCD's for Astronomy. Of course, some of the problem is simply CCD Blooming (too much light overloading pixels)...I found some CCD design papers discussing this, and the solution (to drain off excess current from pixels affected) is quite expensive...I found a recent press release on a Camera using a CCD designed to stop the problem -- it listed for over $24k. Another problem seems to be "chromatic distortion" where I guess the lens plays a big part, trying to focus the light on such a small CCD.

I've even observed pictures from pro cameras (like the Kodak DC620) with a blue/purple fringing problem in high contrast areas too.

I believe that my problem is a combination of things....The pics that have the problem are shown here:

http://albums.photopoint.com/j/ViewPhoto?u=896141&a=6599474&p=22109108

However, the problem seems just as bad from most of the other cameras that I've looked at similiar pics from....If there is a solution that may work, then I'll probably give it a try, and take some pictures under controlled conditions, to compare with and without the I/R Filter. I was hoping to find someone with a positive experience with the filters, before I decide to go ahead and spend the money for one.

Thanks again for the link.
Larry Sabo said:
http://www.cliffshade.com/dpfwiw/ir.htm

Larry
Jim Cockfield said:
Has anyone actually tried a Tiffen "Hot Mirror Filter", designed to
filter I/R?

After viewing pictures from virtually all Digital Cameras, and seeing a
"blue/purple" fringing in thin tree limbs against a bright sky (in
overexposed areas of a photo), I found quite a few posts discussing the
effect of a Hot Mirror Filter's use on various forums.

Supposedly, part of the problem is that CCD's are very sensitive to
Infrared, and when CCD "Blooming" occurs, invisible infrared light is
contributing to problem. The Hot Mirror Filter is supposed to filter
virtually all infrared light, while allowing all visible light to pass,
minimizing the blue/purple fringe (CCD Blooming, Chromatic Distortion,
CCD Binning, or whatever you want to call this phenomenon ) seen in
digital camera pictures, with high contrast, usually overexposed areas.

The most popular description of this filter is this:

"Tiffen 49mm hot mirror Lens Filter - Hazy, bluish casts in images are
often caused by the charged coupled device of the camera, which is
highly sensitive to light in the infrared range. This lens filter limits
this bluish cast and allows for the image to have normal color."

While I've found lots of sites with descriptions of the filter, and
plenty of discussion about the use of this filter (everyone says that
it's supposed to help in the forums and articles that I've read)....I've
found nobody that claims to have actually used one of these filters yet
(everybody is quoting everybody else, but nobody I've found has any
actual experience with this filter)....

Comments? I want to know if it REALLY helps, before buying one.
Click to expand...
Click to expand...
 
Jim - The heat that bothers the CCD is generated within the camera. The camera is running a little computer and you don't use current without generating heat. Any IR coming in the lens is at most a tiny fraction of what's there.

I suggest that if you want to use a digital camera it would be sensible to use it for what it does well? Unless tree branches are somehow essential to your work, it's easy to avoid the kind of shots that are known to be difficult and troublesome.

Digital cameras can't do everything! Personally I wish they had better light gathering ability since I like to do available light, but there you are. I do very much enjoy my 990 and try to humor it.
Coly
Jim Cockfield said:
I've even observed pictures from pro cameras (like the Kodak DC620) with
a blue/purple fringing problem in high contrast areas too.

I believe that my problem is a combination of things....The pics that
have the problem are shown here:

http://albums.photopoint.com/j/ViewPhoto?u=896141&a=6599474&p=22109108

However, the problem seems just as bad from most of the other cameras
that I've looked at similiar pics from....If there is a solution that may
work, then I'll probably give it a try, and take some pictures under
controlled conditions, to compare with and without the I/R Filter. I
was hoping to find someone with a positive experience with the filters,
before I decide to go ahead and spend the money for one.

Thanks again for the link.
Larry Sabo said:
http://www.cliffshade.com/dpfwiw/ir.htm

Larry
Jim Cockfield said:
Has anyone actually tried a Tiffen "Hot Mirror Filter", designed to
filter I/R?

After viewing pictures from virtually all Digital Cameras, and seeing a
"blue/purple" fringing in thin tree limbs against a bright sky (in
overexposed areas of a photo), I found quite a few posts discussing the
effect of a Hot Mirror Filter's use on various forums.

Supposedly, part of the problem is that CCD's are very sensitive to
Infrared, and when CCD "Blooming" occurs, invisible infrared light is
contributing to problem. The Hot Mirror Filter is supposed to filter
virtually all infrared light, while allowing all visible light to pass,
minimizing the blue/purple fringe (CCD Blooming, Chromatic Distortion,
CCD Binning, or whatever you want to call this phenomenon ) seen in
digital camera pictures, with high contrast, usually overexposed areas.

The most popular description of this filter is this:

"Tiffen 49mm hot mirror Lens Filter - Hazy, bluish casts in images are
often caused by the charged coupled device of the camera, which is
highly sensitive to light in the infrared range. This lens filter limits
this bluish cast and allows for the image to have normal color."

While I've found lots of sites with descriptions of the filter, and
plenty of discussion about the use of this filter (everyone says that
it's supposed to help in the forums and articles that I've read)....I've
found nobody that claims to have actually used one of these filters yet
(everybody is quoting everybody else, but nobody I've found has any
actual experience with this filter)....

Comments? I want to know if it REALLY helps, before buying one.
Click to expand...
Click to expand...
Click to expand...
 
My understanding is that Blue/Purple Fringe, aka chromatic abberation, is caused by RGB bending at different angle through the lens and hitting the CCD at different location. The effect is worse at wider angles. Try a blue sky and tree shot with lens set at narrow and wide and see for yourself if the effect is worse at wide angle. There is a tool that can help with this in post-processing.

Check out panorama tools and this web page:
http://www.fh-furtwangen.de/~dersch/pscan/pscan.html

Gordon
 
Been trying for the last fiew days to recreate the purple fringing effect with little success. Most of the time it is just not there even on branches that are only a few pixels wide. I had noticed that on most peoples samples that the fringing seems to be consistent in terms of what side of the object it appears on. This made me think that the problem was camera motion during shutter release. But most of the sample pictures were with trees in the upper part of the picture only so that was inconclusive.

After taking many shots into the centre of groupings of branches I discovered that the effect is spherical and radiating from the centre. In a shot where it happens it does not occur near the centre and increases towards the edges. Also it only happens in shots where the sky is overexposed. This problem has happened in some of my old 35mm shots though it is much harder to distinguish.

I suspect that this is not a problem of digital cameras but is a problem of optics and photography in general. We now have the ability to zoom in and analyse each individual pixel with ease. Maybe we are learning things about photography we didn't know were there before.

Can anyone else verify if this purpling is spherical in nature? If it is I don't think it is the CCD's fault.
GC said:
My understanding is that Blue/Purple Fringe, aka chromatic abberation, is
caused by RGB bending at different angle through the lens and hitting the
CCD at different location. The effect is worse at wider angles. Try a
blue sky and tree shot with lens set at narrow and wide and see for
yourself if the effect is worse at wide angle. There is a tool that can
help with this in post-processing.

Check out panorama tools and this web page:
http://www.fh-furtwangen.de/~dersch/pscan/pscan.html

Gordon
Click to expand...
 
I've read several reasons for it....One of them being Chromatic Distortion (caused by lens accuracy - with the problem being more common towards the edges at full wide angle), and another cause being "CCD Blooming", where the Pixels are overloaded, causing the electric charge to spread to nearby pixels. I read a recent press release about a new camera using a CCD designed to drain excessive energy from affected CCD's to minimize this affect... Unfortunately the Camera is selling for $24k (not within my budget for a digital camera!).....and finally, the description for "Hot Mirror Filters" blames Infrared Sensitivity of CCD's for the problem...

One review of the Canon Pro-70, refers to the problem as "CCD Binning" - see this page:

http://www.dpreview.com/reviews/canonpro70/default.asp?page=7

I also see the blue/purple fringing in the leaves of the tree pictures on the Pro-70 Gallery page here (if you count the pics from left to right, you can see a good example of the problem in pictures 11, 13, 14, 17 - the pics with tree limbs against a bright sky):

http://www.dpreview.com/gallery/pro70_samples/

I tend to go with the "CCD Blooming" theory (pixels being overloaded in high contrast areas due to overexposure, causing the electrical charge to spread to nearby pixels). However, I'm not convinced that infrared is not contributing to this "overload" condition. I'll probably go ahead and order one of the Hot Mirror Filters, and perform some test shots (with and without the filter) to see if it helps.
David A S said:
After taking many shots into the centre of groupings of branches I
discovered that the effect is spherical and radiating from the centre.
In a shot where it happens it does not occur near the centre and
increases towards the edges. Also it only happens in shots where the sky
is overexposed. This problem has happened in some of my old 35mm shots
though it is much harder to distinguish.

I suspect that this is not a problem of digital cameras but is a problem
of optics and photography in general. We now have the ability to zoom in
and analyse each individual pixel with ease. Maybe we are learning
things about photography we didn't know were there before.

Can anyone else verify if this purpling is spherical in nature? If it is
I don't think it is the CCD's fault.
GC said:
My understanding is that Blue/Purple Fringe, aka chromatic abberation, is
caused by RGB bending at different angle through the lens and hitting the
CCD at different location. The effect is worse at wider angles. Try a
blue sky and tree shot with lens set at narrow and wide and see for
yourself if the effect is worse at wide angle. There is a tool that can
help with this in post-processing.

Check out panorama tools and this web page:
http://www.fh-furtwangen.de/~dersch/pscan/pscan.html

Gordon
Click to expand...
Click to expand...
 
So far CCD Binning, or Blooming has not been a factor in any sample I have seen. If it were then the purpling would happen all around an object and not just to one side of it. The overloading of adjacent pixels would not have side favoritism: it would happen all the way around dark objects. Once again I am seeing the purpling occurring on the center frame side of dark objects. I am sure CCD blooming occurs but have not seen a clear example of it.

Also there is another method for the excitation of adjacent pixels. When light strikes a semi opaque surface such as a piece of paper, a camera film, or a CCD there is a tendency for the light to scatter. This is most obvious if you shine a laser pointer at a piece of paper. Even though the laser light is all traveling in parallel it scatters within the substance and illuminates the surrounding area. I think this effect or CCD blooming may account for some of the narrower purpling but not the larger ones in the samples. Again, because of the one side favoritism of the more obvious form of purpling the scattering of light within the CCD could not be the cause.

Again, I am pointed back towards the lens.
Jim Cockfield said:
One review of the Canon Pro-70, refers to the problem as "CCD Binning" -
see this page:

http://www.dpreview.com/reviews/canonpro70/default.asp?page=7

I also see the blue/purple fringing in the leaves of the tree pictures on
the Pro-70 Gallery page here (if you count the pics from left to right,
you can see a good example of the problem in pictures 11, 13, 14, 17 -
the pics with tree limbs against a bright sky):

http://www.dpreview.com/gallery/pro70_samples/

I tend to go with the "CCD Blooming" theory (pixels being overloaded in
high contrast areas due to overexposure, causing the electrical charge to
spread to nearby pixels). However, I'm not convinced that infrared is
not contributing to this "overload" condition. I'll probably go ahead
and order one of the Hot Mirror Filters, and perform some test shots
(with and without the filter) to see if it helps.
David A S said:
After taking many shots into the centre of groupings of branches I
discovered that the effect is spherical and radiating from the centre.
In a shot where it happens it does not occur near the centre and
increases towards the edges. Also it only happens in shots where the sky
is overexposed. This problem has happened in some of my old 35mm shots
though it is much harder to distinguish.

I suspect that this is not a problem of digital cameras but is a problem
of optics and photography in general. We now have the ability to zoom in
and analyse each individual pixel with ease. Maybe we are learning
things about photography we didn't know were there before.

Can anyone else verify if this purpling is spherical in nature? If it is
I don't think it is the CCD's fault.
GC said:
My understanding is that Blue/Purple Fringe, aka chromatic abberation, is
caused by RGB bending at different angle through the lens and hitting the
CCD at different location. The effect is worse at wider angles. Try a
blue sky and tree shot with lens set at narrow and wide and see for
yourself if the effect is worse at wide angle. There is a tool that can
help with this in post-processing.

Check out panorama tools and this web page:
http://www.fh-furtwangen.de/~dersch/pscan/pscan.html

Gordon
Click to expand...
Click to expand...
Click to expand...
 
I believe the chromatic abberation (CA) effect is dominated by the lens construction. Check out this review about experience with an IR cut filter.

http://www.cliffshade.com/dpfwiw/ir.htm#contamination

There was another test I read, couldn't find the source. The argument it present is that IR appears white on CCD (take a picture of the remote and see for yourself), and it should appear uniformly across the scene. If it's IR, why would the effect be red and blue and appears worse at the edges.

CA explain the error as prism effect of different wavelengths striking the lens and defracted at different angles. The error is worse at the edge of the lens, where the entry angle is at the greatest, thus the correlation between the error and the location of the pixels.The only good way is to search for the best lens you can find that has minimum CA on all focal lenght. Since you are stuck with the fixed lens on most digital cam, the next best thing is to fix it in postprocessing. If you buy a ill-matched add-on lens, CA can be much worse. See some sample photos here, also check out the fix that can be done:
http://www.zing.com/album/?id=4294407587

CA is fact of life. I think it's safe to say all digital cam lens has some CA errors. The amount could vary base on focal lenght and lighting condition. Maybe it's more productive to learn how to deal with it. But by all mean, carry on your experiment and let us know if the IR cut filter helps. I imagin it would help on low light photos more than trees with blue sky shots.

Gordon
----------------------
Jim Cockfield said:
One review of the Canon Pro-70, refers to the problem as "CCD Binning" -
see this page:

http://www.dpreview.com/reviews/canonpro70/default.asp?page=7

I also see the blue/purple fringing in the leaves of the tree pictures on
the Pro-70 Gallery page here (if you count the pics from left to right,
you can see a good example of the problem in pictures 11, 13, 14, 17 -
the pics with tree limbs against a bright sky):

http://www.dpreview.com/gallery/pro70_samples/

I tend to go with the "CCD Blooming" theory (pixels being overloaded in
high contrast areas due to overexposure, causing the electrical charge to
spread to nearby pixels). However, I'm not convinced that infrared is
not contributing to this "overload" condition. I'll probably go ahead
and order one of the Hot Mirror Filters, and perform some test shots
(with and without the filter) to see if it helps.
David A S said:
After taking many shots into the centre of groupings of branches I
discovered that the effect is spherical and radiating from the centre.
In a shot where it happens it does not occur near the centre and
increases towards the edges. Also it only happens in shots where the sky
is overexposed. This problem has happened in some of my old 35mm shots
though it is much harder to distinguish.

I suspect that this is not a problem of digital cameras but is a problem
of optics and photography in general. We now have the ability to zoom in
and analyse each individual pixel with ease. Maybe we are learning
things about photography we didn't know were there before.

Can anyone else verify if this purpling is spherical in nature? If it is
I don't think it is the CCD's fault.
GC said:
My understanding is that Blue/Purple Fringe, aka chromatic abberation, is
caused by RGB bending at different angle through the lens and hitting the
CCD at different location. The effect is worse at wider angles. Try a
blue sky and tree shot with lens set at narrow and wide and see for
yourself if the effect is worse at wide angle. There is a tool that can
help with this in post-processing.

Check out panorama tools and this web page:
http://www.fh-furtwangen.de/~dersch/pscan/pscan.html

Gordon
Click to expand...
Click to expand...
Click to expand...
 
Thanks for all of the responses...

I've been searching the forums, and occasionally found users saying that they were going to order one... I've sent emails to these users, asking about their experience with it, but haven't gotten any replies yet. Unfortunately, most of the posts were old, and some of the Email addresses are no longer valid...Maybe some of the users with good email addresses will respond.

Despite my "digging through the forums" to try and find comparison photos (with and without Hot Mirror Filters), I've been unable to locate any. Interestingly, I understand that Nikon included an I/R Block Filter in the 990 (which prevents you from using a 990 for Infrared Photography)... Since the purple fringing affect has gotten a lot of space concerning the Nikon 950, I'm wondering if Nikon introduced I/R Blocking in the 990, because they know that it helps... From the photo comparisons that I've made of 950 versus 990 shots of tree limbs against a bright sky, the 990 does seem to handle these shots better than the 950.

In any event, I'm going to order a Hot Mirror Filter anyway tomorrow for my new digital camera. When I get the filter, I'll take test shots of tree tops against a bright sky, with and without the filter, and upload the results to a photo site, so that all can see what (if any) improvement is made. Probably none..but if that's the case, then I'll have an extra lens protector to use, in case I scratch or damage my UV Filter.
David A S said:
Also there is another method for the excitation of adjacent pixels. When
light strikes a semi opaque surface such as a piece of paper, a camera
film, or a CCD there is a tendency for the light to scatter. This is
most obvious if you shine a laser pointer at a piece of paper. Even
though the laser light is all traveling in parallel it scatters within
the substance and illuminates the surrounding area. I think this effect
or CCD blooming may account for some of the narrower purpling but not the
larger ones in the samples. Again, because of the one side favoritism of
the more obvious form of purpling the scattering of light within the CCD
could not be the cause.

Again, I am pointed back towards the lens.
Jim Cockfield said:
One review of the Canon Pro-70, refers to the problem as "CCD Binning" -
see this page:

http://www.dpreview.com/reviews/canonpro70/default.asp?page=7

I also see the blue/purple fringing in the leaves of the tree pictures on
the Pro-70 Gallery page here (if you count the pics from left to right,
you can see a good example of the problem in pictures 11, 13, 14, 17 -
the pics with tree limbs against a bright sky):

http://www.dpreview.com/gallery/pro70_samples/

I tend to go with the "CCD Blooming" theory (pixels being overloaded in
high contrast areas due to overexposure, causing the electrical charge to
spread to nearby pixels). However, I'm not convinced that infrared is
not contributing to this "overload" condition. I'll probably go ahead
and order one of the Hot Mirror Filters, and perform some test shots
(with and without the filter) to see if it helps.
David A S said:
After taking many shots into the centre of groupings of branches I
discovered that the effect is spherical and radiating from the centre.
In a shot where it happens it does not occur near the centre and
increases towards the edges. Also it only happens in shots where the sky
is overexposed. This problem has happened in some of my old 35mm shots
though it is much harder to distinguish.

I suspect that this is not a problem of digital cameras but is a problem
of optics and photography in general. We now have the ability to zoom in
and analyse each individual pixel with ease. Maybe we are learning
things about photography we didn't know were there before.

Can anyone else verify if this purpling is spherical in nature? If it is
I don't think it is the CCD's fault.
GC said:
My understanding is that Blue/Purple Fringe, aka chromatic abberation, is
caused by RGB bending at different angle through the lens and hitting the
CCD at different location. The effect is worse at wider angles. Try a
blue sky and tree shot with lens set at narrow and wide and see for
yourself if the effect is worse at wide angle. There is a tool that can
help with this in post-processing.

Check out panorama tools and this web page:
http://www.fh-furtwangen.de/~dersch/pscan/pscan.html

Gordon
Click to expand...
Click to expand...
Click to expand...
Click to expand...
 
I've speculated in the past that fringing effects could be the result of the processing that is done on the ccd outputs. Each pixel measures only one color and a sophisticated interpolation is used to get a color at each pixel. Cameras also do various filtering and sharpening operations. It is quite conceivable that this processing particularly the sharpening could result in fringing. If this is related to sharpening, then one would expect that fringing might be worse with manufacturers that have sharper pictures. I'm not that familiar with this aspect across the various makes but there are likely others who are. If it is ccd related, then the effects should correlate with ccd manufacturer. Leon
Jim Cockfield said:
Has anyone actually tried a Tiffen "Hot Mirror Filter", designed to
filter I/R?

After viewing pictures from virtually all Digital Cameras, and seeing a
"blue/purple" fringing in thin tree limbs against a bright sky (in
overexposed areas of a photo), I found quite a few posts discussing the
effect of a Hot Mirror Filter's use on various forums.

Supposedly, part of the problem is that CCD's are very sensitive to
Infrared, and when CCD "Blooming" occurs, invisible infrared light is
contributing to problem. The Hot Mirror Filter is supposed to filter
virtually all infrared light, while allowing all visible light to pass,
minimizing the blue/purple fringe (CCD Blooming, Chromatic Distortion,
CCD Binning, or whatever you want to call this phenomenon ) seen in
digital camera pictures, with high contrast, usually overexposed areas.

The most popular description of this filter is this:

"Tiffen 49mm hot mirror Lens Filter - Hazy, bluish casts in images are
often caused by the charged coupled device of the camera, which is
highly sensitive to light in the infrared range. This lens filter limits
this bluish cast and allows for the image to have normal color."

While I've found lots of sites with descriptions of the filter, and
plenty of discussion about the use of this filter (everyone says that
it's supposed to help in the forums and articles that I've read)....I've
found nobody that claims to have actually used one of these filters yet
(everybody is quoting everybody else, but nobody I've found has any
actual experience with this filter)....

Comments? I want to know if it REALLY helps, before buying one.
Click to expand...
 
Jim:

If you check around on this forum you will find some nice IR pictures taken with the 990,If Nikon put a filter in the camera it's not very affective. I think if you do a search for a 89B Wratten filter you will come up with the posts.
Don.
Jim Cockfield said:
I've been searching the forums, and occasionally found users saying that
they were going to order one... I've sent emails to these users, asking
about their experience with it, but haven't gotten any replies yet.
Unfortunately, most of the posts were old, and some of the Email
addresses are no longer valid...Maybe some of the users with good email
addresses will respond.

Despite my "digging through the forums" to try and find comparison photos
(with and without Hot Mirror Filters), I've been unable to locate any.
Interestingly, I understand that Nikon included an I/R Block Filter in
the 990 (which prevents you from using a 990 for Infrared Photography)...
Since the purple fringing affect has gotten a lot of space concerning the
Nikon 950, I'm wondering if Nikon introduced I/R Blocking in the 990,
because they know that it helps... From the photo comparisons that I've
made of 950 versus 990 shots of tree limbs against a bright sky, the 990
does seem to handle these shots better than the 950.

In any event, I'm going to order a Hot Mirror Filter anyway tomorrow for
my new digital camera. When I get the filter, I'll take test shots of
tree tops against a bright sky, with and without the filter, and upload
the results to a photo site, so that all can see what (if any)
improvement is made. Probably none..but if that's the case, then I'll
have an extra lens protector to use, in case I scratch or damage my UV
Filter.
David A S said:
Also there is another method for the excitation of adjacent pixels. When
light strikes a semi opaque surface such as a piece of paper, a camera
film, or a CCD there is a tendency for the light to scatter. This is
most obvious if you shine a laser pointer at a piece of paper. Even
though the laser light is all traveling in parallel it scatters within
the substance and illuminates the surrounding area. I think this effect
or CCD blooming may account for some of the narrower purpling but not the
larger ones in the samples. Again, because of the one side favoritism of
the more obvious form of purpling the scattering of light within the CCD
could not be the cause.

Again, I am pointed back towards the lens.
Jim Cockfield said:
One review of the Canon Pro-70, refers to the problem as "CCD Binning" -
see this page:

http://www.dpreview.com/reviews/canonpro70/default.asp?page=7

I also see the blue/purple fringing in the leaves of the tree pictures on
the Pro-70 Gallery page here (if you count the pics from left to right,
you can see a good example of the problem in pictures 11, 13, 14, 17 -
the pics with tree limbs against a bright sky):

http://www.dpreview.com/gallery/pro70_samples/

I tend to go with the "CCD Blooming" theory (pixels being overloaded in
high contrast areas due to overexposure, causing the electrical charge to
spread to nearby pixels). However, I'm not convinced that infrared is
not contributing to this "overload" condition. I'll probably go ahead
and order one of the Hot Mirror Filters, and perform some test shots
(with and without the filter) to see if it helps.
David A S said:
After taking many shots into the centre of groupings of branches I
discovered that the effect is spherical and radiating from the centre.
In a shot where it happens it does not occur near the centre and
increases towards the edges. Also it only happens in shots where the sky
is overexposed. This problem has happened in some of my old 35mm shots
though it is much harder to distinguish.

I suspect that this is not a problem of digital cameras but is a problem
of optics and photography in general. We now have the ability to zoom in
and analyse each individual pixel with ease. Maybe we are learning
things about photography we didn't know were there before.

Can anyone else verify if this purpling is spherical in nature? If it is
I don't think it is the CCD's fault.
GC said:
My understanding is that Blue/Purple Fringe, aka chromatic abberation, is
caused by RGB bending at different angle through the lens and hitting the
CCD at different location. The effect is worse at wider angles. Try a
blue sky and tree shot with lens set at narrow and wide and see for
yourself if the effect is worse at wide angle. There is a tool that can
help with this in post-processing.

Check out panorama tools and this web page:
http://www.fh-furtwangen.de/~dersch/pscan/pscan.html

Gordon
Click to expand...
Click to expand...
Click to expand...
Click to expand...
Click to expand...
 
Jim:

Sorry for the wrong info in my last post, do a search for 990 infrared and you will find some infrared shots done with the 990 and a wratten 89B filter.
Don.
Jim Cockfield said:
Has anyone actually tried a Tiffen "Hot Mirror Filter", designed to
filter I/R?

After viewing pictures from virtually all Digital Cameras, and seeing a
"blue/purple" fringing in thin tree limbs against a bright sky (in
overexposed areas of a photo), I found quite a few posts discussing the
effect of a Hot Mirror Filter's use on various forums.

Supposedly, part of the problem is that CCD's are very sensitive to
Infrared, and when CCD "Blooming" occurs, invisible infrared light is
contributing to problem. The Hot Mirror Filter is supposed to filter
virtually all infrared light, while allowing all visible light to pass,
minimizing the blue/purple fringe (CCD Blooming, Chromatic Distortion,
CCD Binning, or whatever you want to call this phenomenon ) seen in
digital camera pictures, with high contrast, usually overexposed areas.

The most popular description of this filter is this:

"Tiffen 49mm hot mirror Lens Filter - Hazy, bluish casts in images are
often caused by the charged coupled device of the camera, which is
highly sensitive to light in the infrared range. This lens filter limits
this bluish cast and allows for the image to have normal color."

While I've found lots of sites with descriptions of the filter, and
plenty of discussion about the use of this filter (everyone says that
it's supposed to help in the forums and articles that I've read)....I've
found nobody that claims to have actually used one of these filters yet
(everybody is quoting everybody else, but nobody I've found has any
actual experience with this filter)....

Comments? I want to know if it REALLY helps, before buying one.
Click to expand...
 
I was basing that assumption on some of the posts I found yesterday, while searching for posts on "hot mirror filters"....Apparently, it's not "totally" infrared insensitive, but shots taken with the 87c and a Nikon 950, did not work with the 990 at the same exposures.

According to Bryan Biggers, since the Nikon 950 benefited from the use of Hot Mirror Filters, Nikon included one in the 990 (whether you want it or not):

http://www.dpreview.com/forums/read.asp?forum=1007&message=236402&query=990+hot+mirror+biggers

http://www.steves-digicams.com/forum/Forums/Thread.cfm?CFApp=2&Thread_ID=1219&mc=4#Message6404

http://www.dpreview.com/forums/read.asp?forum=1007&message=232767&query=infrared+990
Don Schultz said:
Jim Cockfield said:
I've been searching the forums, and occasionally found users saying that
they were going to order one... I've sent emails to these users, asking
about their experience with it, but haven't gotten any replies yet.
Unfortunately, most of the posts were old, and some of the Email
addresses are no longer valid...Maybe some of the users with good email
addresses will respond.

Despite my "digging through the forums" to try and find comparison photos
(with and without Hot Mirror Filters), I've been unable to locate any.
Interestingly, I understand that Nikon included an I/R Block Filter in
the 990 (which prevents you from using a 990 for Infrared Photography)...
Since the purple fringing affect has gotten a lot of space concerning the
Nikon 950, I'm wondering if Nikon introduced I/R Blocking in the 990,
because they know that it helps... From the photo comparisons that I've
made of 950 versus 990 shots of tree limbs against a bright sky, the 990
does seem to handle these shots better than the 950.

In any event, I'm going to order a Hot Mirror Filter anyway tomorrow for
my new digital camera. When I get the filter, I'll take test shots of
tree tops against a bright sky, with and without the filter, and upload
the results to a photo site, so that all can see what (if any)
improvement is made. Probably none..but if that's the case, then I'll
have an extra lens protector to use, in case I scratch or damage my UV
Filter.
David A S said:
Also there is another method for the excitation of adjacent pixels. When
light strikes a semi opaque surface such as a piece of paper, a camera
film, or a CCD there is a tendency for the light to scatter. This is
most obvious if you shine a laser pointer at a piece of paper. Even
though the laser light is all traveling in parallel it scatters within
the substance and illuminates the surrounding area. I think this effect
or CCD blooming may account for some of the narrower purpling but not the
larger ones in the samples. Again, because of the one side favoritism of
the more obvious form of purpling the scattering of light within the CCD
could not be the cause.

Again, I am pointed back towards the lens.
Jim Cockfield said:
One review of the Canon Pro-70, refers to the problem as "CCD Binning" -
see this page:

http://www.dpreview.com/reviews/canonpro70/default.asp?page=7

I also see the blue/purple fringing in the leaves of the tree pictures on
the Pro-70 Gallery page here (if you count the pics from left to right,
you can see a good example of the problem in pictures 11, 13, 14, 17 -
the pics with tree limbs against a bright sky):

http://www.dpreview.com/gallery/pro70_samples/

I tend to go with the "CCD Blooming" theory (pixels being overloaded in
high contrast areas due to overexposure, causing the electrical charge to
spread to nearby pixels). However, I'm not convinced that infrared is
not contributing to this "overload" condition. I'll probably go ahead
and order one of the Hot Mirror Filters, and perform some test shots
(with and without the filter) to see if it helps.
David A S said:
After taking many shots into the centre of groupings of branches I
discovered that the effect is spherical and radiating from the centre.
In a shot where it happens it does not occur near the centre and
increases towards the edges. Also it only happens in shots where the sky
is overexposed. This problem has happened in some of my old 35mm shots
though it is much harder to distinguish.

I suspect that this is not a problem of digital cameras but is a problem
of optics and photography in general. We now have the ability to zoom in
and analyse each individual pixel with ease. Maybe we are learning
things about photography we didn't know were there before.

Can anyone else verify if this purpling is spherical in nature? If it is
I don't think it is the CCD's fault.
GC said:
My understanding is that Blue/Purple Fringe, aka chromatic abberation, is
caused by RGB bending at different angle through the lens and hitting the
CCD at different location. The effect is worse at wider angles. Try a
blue sky and tree shot with lens set at narrow and wide and see for
yourself if the effect is worse at wide angle. There is a tool that can
help with this in post-processing.

Check out panorama tools and this web page:
http://www.fh-furtwangen.de/~dersch/pscan/pscan.html

Gordon
Click to expand...
Click to expand...
Click to expand...
Click to expand...
Click to expand...
Click to expand...
 
It could be that the 89b has a cutoff frequency closer to visible light, and the I/R Block in the 990 is a little higher frequency (allowing a "small window" for infrared photos with the 89b, since it appears to work, and the 87c doesn't appear to work with a 990 (according to posts from users on the Steve's Digicam Forum.)
Jim Cockfield said:
According to Bryan Biggers, since the Nikon 950 benefited from the use of
Hot Mirror Filters, Nikon included one in the 990 (whether you want it or
not):

http://www.dpreview.com/forums/read.asp?forum=1007&message=236402&query=990+hot+mirror+biggers

http://www.steves-digicams.com/forum/Forums/Thread.cfm?CFApp=2&Thread_ID=1219&mc=4#Message6404

http://www.dpreview.com/forums/read.asp?forum=1007&message=232767&query=infrared+990
Don Schultz said:
Jim Cockfield said:
I've been searching the forums, and occasionally found users saying that
they were going to order one... I've sent emails to these users, asking
about their experience with it, but haven't gotten any replies yet.
Unfortunately, most of the posts were old, and some of the Email
addresses are no longer valid...Maybe some of the users with good email
addresses will respond.

Despite my "digging through the forums" to try and find comparison photos
(with and without Hot Mirror Filters), I've been unable to locate any.
Interestingly, I understand that Nikon included an I/R Block Filter in
the 990 (which prevents you from using a 990 for Infrared Photography)...
Since the purple fringing affect has gotten a lot of space concerning the
Nikon 950, I'm wondering if Nikon introduced I/R Blocking in the 990,
because they know that it helps... From the photo comparisons that I've
made of 950 versus 990 shots of tree limbs against a bright sky, the 990
does seem to handle these shots better than the 950.

In any event, I'm going to order a Hot Mirror Filter anyway tomorrow for
my new digital camera. When I get the filter, I'll take test shots of
tree tops against a bright sky, with and without the filter, and upload
the results to a photo site, so that all can see what (if any)
improvement is made. Probably none..but if that's the case, then I'll
have an extra lens protector to use, in case I scratch or damage my UV
Filter.
David A S said:
Also there is another method for the excitation of adjacent pixels. When
light strikes a semi opaque surface such as a piece of paper, a camera
film, or a CCD there is a tendency for the light to scatter. This is
most obvious if you shine a laser pointer at a piece of paper. Even
though the laser light is all traveling in parallel it scatters within
the substance and illuminates the surrounding area. I think this effect
or CCD blooming may account for some of the narrower purpling but not the
larger ones in the samples. Again, because of the one side favoritism of
the more obvious form of purpling the scattering of light within the CCD
could not be the cause.

Again, I am pointed back towards the lens.
Jim Cockfield said:
One review of the Canon Pro-70, refers to the problem as "CCD Binning" -
see this page:

http://www.dpreview.com/reviews/canonpro70/default.asp?page=7

I also see the blue/purple fringing in the leaves of the tree pictures on
the Pro-70 Gallery page here (if you count the pics from left to right,
you can see a good example of the problem in pictures 11, 13, 14, 17 -
the pics with tree limbs against a bright sky):

http://www.dpreview.com/gallery/pro70_samples/

I tend to go with the "CCD Blooming" theory (pixels being overloaded in
high contrast areas due to overexposure, causing the electrical charge to
spread to nearby pixels). However, I'm not convinced that infrared is
not contributing to this "overload" condition. I'll probably go ahead
and order one of the Hot Mirror Filters, and perform some test shots
(with and without the filter) to see if it helps.
David A S said:
After taking many shots into the centre of groupings of branches I
discovered that the effect is spherical and radiating from the centre.
In a shot where it happens it does not occur near the centre and
increases towards the edges. Also it only happens in shots where the sky
is overexposed. This problem has happened in some of my old 35mm shots
though it is much harder to distinguish.

I suspect that this is not a problem of digital cameras but is a problem
of optics and photography in general. We now have the ability to zoom in
and analyse each individual pixel with ease. Maybe we are learning
things about photography we didn't know were there before.

Can anyone else verify if this purpling is spherical in nature? If it is
I don't think it is the CCD's fault.
GC said:
My understanding is that Blue/Purple Fringe, aka chromatic abberation, is
caused by RGB bending at different angle through the lens and hitting the
CCD at different location. The effect is worse at wider angles. Try a
blue sky and tree shot with lens set at narrow and wide and see for
yourself if the effect is worse at wide angle. There is a tool that can
help with this in post-processing.

Check out panorama tools and this web page:
http://www.fh-furtwangen.de/~dersch/pscan/pscan.html

Gordon
Click to expand...
Click to expand...
Click to expand...
Click to expand...
Click to expand...
Click to expand...
Click to expand...
 
Jim,

Thanks for for digging into this issue. I don't believe IR has been ruled out by the sample shots we have seen so far. IR could exibit one side favouritism as it would refract differently from the visible light due to the different wavelength.

I look forward to hearing about your results with the new filter.
 
P.S.

I just looked up refraction in my physics text to check for certain.

Longer wavelengths (towards infrared) are refracted less than shorter wavelengths. Thus the brighter and presumably hotter sky would have IR waves that were not focussed in as much as visible light. In theory, the result would be that the centre frame side of dark objects would have IR contamination, resulting in the observed one side favouritism of the purple fringe. It remains to be seen if the IR present is enough to cause the purple fringing but at least there is a possible explanation to justify trying the filter.
 
I ordered a Tiffen "Hot Mirror" Filter. It was supposed have shipped 2nd day air yesterday, so it should be here tomorrow... .We'll find out if/how much it helps, as soon I get it (it's pretty easy to reproduce the problem). I'll take photos with and without the filter, and post them in a public album on one of the photo sites.

(probably here - I'll create a new album with the photos labeled as "with hot mirror filter", and "without hot mirror filter"):

http://albums.photopoint.com/j/AlbumList?u=896141
David A S said:
P.S.

I just looked up refraction in my physics text to check for certain.

Longer wavelengths (towards infrared) are refracted less than shorter
wavelengths. Thus the brighter and presumably hotter sky would have IR
waves that were not focussed in as much as visible light. In theory, the
result would be that the centre frame side of dark objects would have IR
contamination, resulting in the observed one side favouritism of the
purple fringe. It remains to be seen if the IR present is enough to
cause the purple fringing but at least there is a possible explanation to
justify trying the filter.
Click to expand...
 
Sorry guys,

Came into this a bit late.. However I've been doing quite a bit of research on this subject, after a recent conversation with a very senior Kodak Engineer this is the conclusion (answer) he gave:

Consumer CCD's have little "pillows" over each pixel called a Microlens, these lenses are designed to improve the sensitivity of the CCD by gathering more light down to the small photodiode. Because these lenses are shaped light can refract / reflect off the edges of the "pillow" and strike adjacent pixels / not enter the lens correctly thus producing the effect we've seen.
Jim Cockfield said:
Has anyone actually tried a Tiffen "Hot Mirror Filter", designed to
filter I/R?

After viewing pictures from virtually all Digital Cameras, and seeing a
"blue/purple" fringing in thin tree limbs against a bright sky (in
overexposed areas of a photo), I found quite a few posts discussing the
effect of a Hot Mirror Filter's use on various forums.

Supposedly, part of the problem is that CCD's are very sensitive to
Infrared, and when CCD "Blooming" occurs, invisible infrared light is
contributing to problem. The Hot Mirror Filter is supposed to filter
virtually all infrared light, while allowing all visible light to pass,
minimizing the blue/purple fringe (CCD Blooming, Chromatic Distortion,
CCD Binning, or whatever you want to call this phenomenon ) seen in
digital camera pictures, with high contrast, usually overexposed areas.

The most popular description of this filter is this:

"Tiffen 49mm hot mirror Lens Filter - Hazy, bluish casts in images are
often caused by the charged coupled device of the camera, which is
highly sensitive to light in the infrared range. This lens filter limits
this bluish cast and allows for the image to have normal color."

While I've found lots of sites with descriptions of the filter, and
plenty of discussion about the use of this filter (everyone says that
it's supposed to help in the forums and articles that I've read)....I've
found nobody that claims to have actually used one of these filters yet
(everybody is quoting everybody else, but nobody I've found has any
actual experience with this filter)....

Comments? I want to know if it REALLY helps, before buying one.
Click to expand...
 
Well...I received the Tiffen Hot Mirror Filter I ordered today, and just finished taking a few quick shots, with and without the filter installed.

At first glance, it does not appear to help at all with the Purple Fringing Problem....I'll take some more pics when I get a chance.

So, it appears that I've just bought an expensive Lens Protector to use if my UV filter gets damaged.

See this link for the photos I just took:

http://albums.photopoint.com/j/AlbumIndex?u=896141&a=6702686

Here's the main album list (with more 3000z pics):

http://albums.photopoint.com/j/AlbumList?u=896141

Phil's explanation is probably correct (see it within this thread):

http://www.dpreview.com/forums/read.asp?forum=1001&message=278022
Jim Cockfield said:
Has anyone actually tried a Tiffen "Hot Mirror Filter", designed to
filter I/R?

After viewing pictures from virtually all Digital Cameras, and seeing a
"blue/purple" fringing in thin tree limbs against a bright sky (in
overexposed areas of a photo), I found quite a few posts discussing the
effect of a Hot Mirror Filter's use on various forums.

Supposedly, part of the problem is that CCD's are very sensitive to
Infrared, and when CCD "Blooming" occurs, invisible infrared light is
contributing to problem. The Hot Mirror Filter is supposed to filter
virtually all infrared light, while allowing all visible light to pass,
minimizing the blue/purple fringe (CCD Blooming, Chromatic Distortion,
CCD Binning, or whatever you want to call this phenomenon ) seen in
digital camera pictures, with high contrast, usually overexposed areas.

The most popular description of this filter is this:

"Tiffen 49mm hot mirror Lens Filter - Hazy, bluish casts in images are
often caused by the charged coupled device of the camera, which is
highly sensitive to light in the infrared range. This lens filter limits
this bluish cast and allows for the image to have normal color."

While I've found lots of sites with descriptions of the filter, and
plenty of discussion about the use of this filter (everyone says that
it's supposed to help in the forums and articles that I've read)....I've
found nobody that claims to have actually used one of these filters yet
(everybody is quoting everybody else, but nobody I've found has any
actual experience with this filter)....

Comments? I want to know if it REALLY helps, before buying one.
Click to expand...
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top