Reading mode: Light Dark

707 Purple Fringe examples....

Go to last post
Ron,

If you have a picture to post, you can send it to me ([email protected]) and I will put it on my FotoTime.com account.

Paul
Ron C said:
II believe that there are a number of sources of the dreaded
"PURPLE FRINGE".

In some camera systems the optics are to blame for what is normally
termed CA chromatic aberration. I feel that the Sony 505v/707 have
little CA because of the super quality Zeiss lenses.

But, we still see the PF in many images. My theory is that a
combination of what is being shot and the internal sofwate causes
the PF. If you will look at most of the posted PF problems, the
green channel is DEAD or nearly so -- thus PURPLE. How would the
green get suppressed so much? Some have said blooming causes the
PF but again why is the green channel dead? Phil's blackboard used
to test PF is exhibited along side the bright white "L".

What happens in the dsp while the image is being finished for
output to the Memory Stick?

Three or more algorithms are passed by the image.
The image is interpolated from the Bayer mask (GRGB) form to a
pixel of each color at each input pixel location.

The image is white/color balanced using some statistical scheme
based on user setup.

The image is sharpened, based on user setup (-2,-1,0,1,2), using
something like Unsharp Mask (very poor name for it).

I believe that a combination of bright areas along side of rather
weak areas show the PF as the red and blue channels are accentuated
and green it killed. I have a simple synthetic test to show a poor
mans white balance will do this but nowhere to upload.

Enough of my rambling.

CHEERS,
RON C
Ron Parr said:
Ron C said:
I have a theory about the "Purple Fringe" and sharpening and white
balancing.
Click to expand...
What's the basis for this theory? It seems to contradict a lot of
the body of knowledge that has been developed about this phenomenon.

Ron Parr
Click to expand...
Click to expand...
 
Ron C said:
Ron,
I believe that a combination of bright areas along side of rather
weak areas show the PF as the red and blue channels are accentuated
and green it killed. I have a simple synthetic test to show a poor
mans white balance will do this but nowhere to upload.
Click to expand...
White balance is applied to the entire image. If the green is knocked down so much that it is almost completely eliminated in the brightest parts of the image, then it would totally kill it in the rest of the image.

Ron Parr
 
Well now that more pictures are being posted.... It seems to be that there is a lot of blooming/ca. I still can't wait to get mine and try it out, but should we be worried?

Take a look at 105,106,107, 100, 102.

http://www.pbase.com/snyders2000/sony_f707_2
Howdy said:
Brian Smith said:
Does the purple fringing seem normal. I noticed it right off as
did my wife. Would you guys considers this "normal". The purple
seems very noticeable.

What do you guys think? Is this acceptable?

-Brian Smith
Click to expand...
Brian,
First off, thanks for the shots, much appreciated.

Normal, I guess, is a relative question. Normal to what? I agree
the purple is very visible. But, as others have noted, the
pictures not only have an 'out-of-focus' quality, but you picked
the perfect subject to exhibit purple fringing. I assume your
asking whether the 707 is exhibiting any more fringing than another
camera would given this same composition. I think, however, given
the 'extreme' example/composition you discovered, the only way to
answer that question is to go back with a 707 and some other
digital cameras, take lots of test photos with each camera,
discover the best settings (ie, those that best balance a reduction
of the purple fringing while still producing a great shot), and
then compare the best results produced by each camera.

I am intrigued. Once my 707 arrives, I'ld like to get my hands on
a G2 and a 995 for an afternoon, and see how the cameras compare in
real-world 'extreme purple fringing' settings.
Click to expand...
 
I took a quick peek at a couple. Hopefully I can look tonight when I have more time. Pic#107, EV +.7, F2.3, shutter 1/50th, background like it was, seems like trouble all the way. Seems like this as extreme to me. My S70 would give some CA too.

Randy
 
Hi Randy-

Those are actually my pictures and I am glad you pointed that out. Because of some discussion about CA, I intentionally tried to produce shots that would exaggerate this condition so we all would have more extreme example to analyze. Apparently, I may have convinced some that the problem is worse than it really is. :-)

Jeff
MN Randy said:
I took a quick peek at a couple. Hopefully I can look tonight when
I have more time. Pic#107, EV +.7, F2.3, shutter 1/50th,
background like it was, seems like trouble all the way. Seems like
this as extreme to me. My S70 would give some CA too.

Randy
Click to expand...
 
Jeff,

Thanks for the update. I meant no harm by my post. I missed the thread that stated you were going to exaggerate this condition. Whew, you freaked me on that one.

Thanks again,
David
Jeff S. said:
Those are actually my pictures and I am glad you pointed that out.
Because of some discussion about CA, I intentionally tried to
produce shots that would exaggerate this condition so we all would
have more extreme example to analyze. Apparently, I may have
convinced some that the problem is worse than it really is. :-)

Jeff
MN Randy said:
I took a quick peek at a couple. Hopefully I can look tonight when
I have more time. Pic#107, EV +.7, F2.3, shutter 1/50th,
background like it was, seems like trouble all the way. Seems like
this as extreme to me. My S70 would give some CA too.

Randy
Click to expand...
Click to expand...
 
Hi Jeff,

Perhaps a caption that read "Shot to accentuate CA" might have helped. It's too bad people are jumping to these conclusions without first reading the entire thread, but as some are "seeking" out imperfections, I can understand the reaction somewhat.

Oh, and thanks again for posting all the originals. They are wonderful shots. The golf course picture is excellent. Looks so lucious and serene. I'd rather be there than at work now. And I don't even play golf!

Regards,
Pat
Jeff S. said:
Those are actually my pictures and I am glad you pointed that out.
Because of some discussion about CA, I intentionally tried to
produce shots that would exaggerate this condition so we all would
have more extreme example to analyze. Apparently, I may have
convinced some that the problem is worse than it really is. :-)

Jeff
MN Randy said:
I took a quick peek at a couple. Hopefully I can look tonight when
I have more time. Pic#107, EV +.7, F2.3, shutter 1/50th,
background like it was, seems like trouble all the way. Seems like
this as extreme to me. My S70 would give some CA too.

Randy
Click to expand...
Click to expand...
 
I wonder if the purple fringe problem is more apparent
in digital cameras than with film (e.g. 35mm film) because
the CCD sensor is smaller in size than film. Is this the case?

What is the best design solution? Let's give Sony engineers some
recommendations now that they are probably on to designing
the F707V (or the F909)... :-) (People should be perfectionists
and be rather arrogant in discussion forums like this, shouldn't they?)

And, if the explanation of this chromatic aberration problem
in the glossary here is correct, then it seems to me that there
should be no purple fringe in the center of images, through
the center of the lens. I this the case? I'll have to experiment.
 
Ron,

If you look at the Phil's CA target, you will see that the green is missing in the area around the white L. Because the target is mostly black, you can't see what happened in that area.

If you look at the 505v leaf problem, you'll see that the "Fringe" is zero in green, and the greens in other areas are reduced by some amount as they are very dark.

I don't know what algorithma are used in the dsp but they can't be very powerful.

I sent Brian some test examples to post but haven't heard from him.

The test used a very simple white balance method than equalizes the channels but maintaining the overall level of the images.
Hopefully, I hear or see the examples here soon.

RON C
Ron Parr said:
Ron C said:
Ron,
I believe that a combination of bright areas along side of rather
weak areas show the PF as the red and blue channels are accentuated
and green it killed. I have a simple synthetic test to show a poor
mans white balance will do this but nowhere to upload.
Click to expand...
White balance is applied to the entire image. If the green is
knocked down so much that it is almost completely eliminated in the
brightest parts of the image, then it would totally kill it in the
rest of the image.

Ron Parr
Click to expand...
 
That should have been Paul not Brian,
RON C
Ron C said:
If you look at the Phil's CA target, you will see that the green is
missing in the area around the white L. Because the target is
mostly black, you can't see what happened in that area.

If you look at the 505v leaf problem, you'll see that the "Fringe"
is zero in green, and the greens in other areas are reduced by some
amount as they are very dark.

I don't know what algorithma are used in the dsp but they can't be
very powerful.

I sent Brian some test examples to post but haven't heard from him.

The test used a very simple white balance method than equalizes the
channels but maintaining the overall level of the images.
Hopefully, I hear or see the examples here soon.

RON C
Ron Parr said:
Ron C said:
Ron,
I believe that a combination of bright areas along side of rather
weak areas show the PF as the red and blue channels are accentuated
and green it killed. I have a simple synthetic test to show a poor
mans white balance will do this but nowhere to upload.
Click to expand...
White balance is applied to the entire image. If the green is
knocked down so much that it is almost completely eliminated in the
brightest parts of the image, then it would totally kill it in the
rest of the image.

Ron Parr
Click to expand...
Click to expand...
 
Hi David-

No problem, I wasn't offended. :-)

I'm glad I could clear it up a bit.

Jeff
r00t said:
Thanks again,
David
Jeff S. said:
Those are actually my pictures and I am glad you pointed that out.
Because of some discussion about CA, I intentionally tried to
produce shots that would exaggerate this condition so we all would
have more extreme example to analyze. Apparently, I may have
convinced some that the problem is worse than it really is. :-)

Jeff
MN Randy said:
I took a quick peek at a couple. Hopefully I can look tonight when
I have more time. Pic#107, EV +.7, F2.3, shutter 1/50th,
background like it was, seems like trouble all the way. Seems like
this as extreme to me. My S70 would give some CA too.

Randy
Click to expand...
Click to expand...
Click to expand...
 
Hi Pat--
Pat said:
Perhaps a caption that read "Shot to accentuate CA" might have
helped. It's too bad people are jumping to these conclusions
without first reading the entire thread, but as some are "seeking"
out imperfections, I can understand the reaction somewhat.
Click to expand...
That is a great idea. I will make the update now. :-)
Pat said:
Oh, and thanks again for posting all the originals. They are
wonderful shots. The golf course picture is excellent. Looks so
lucious and serene. I'd rather be there than at work now. And I
don't even play golf!
Click to expand...
Thanks. :-)

Jeff
 
Ron,

Sorry for the delay. I wasn't expecting the web page. I don't think I can upload a website page to Fototime.com. But I also have a little website. My webspace is limited and I plan to change hosts at some point in the future. But I uploaded your WB examples to my current website anyway just to get it out there. So here is the link:
http://www.niez.org/misc/rcwbex/wb.html

Thanks for putting this together!

Paul
Ron C said:
Ron C said:
If you look at the Phil's CA target, you will see that the green is
missing in the area around the white L. Because the target is
mostly black, you can't see what happened in that area.

If you look at the 505v leaf problem, you'll see that the "Fringe"
is zero in green, and the greens in other areas are reduced by some
amount as they are very dark.

I don't know what algorithma are used in the dsp but they can't be
very powerful.

I sent Brian some test examples to post but haven't heard from him.

The test used a very simple white balance method than equalizes the
channels but maintaining the overall level of the images.
Hopefully, I hear or see the examples here soon.

RON C
Ron Parr said:
Ron C said:
Ron,
I believe that a combination of bright areas along side of rather
weak areas show the PF as the red and blue channels are accentuated
and green it killed. I have a simple synthetic test to show a poor
mans white balance will do this but nowhere to upload.
Click to expand...
White balance is applied to the entire image. If the green is
knocked down so much that it is almost completely eliminated in the
brightest parts of the image, then it would totally kill it in the
rest of the image.

Ron Parr
Click to expand...
Click to expand...
Click to expand...
 
Paul,

Thanks for putting it up. I don't have a clear answer to CA, PF, WB etc. I do have a lot of questions though.

I wonder if the early film types went through this stuff.

CA will always be around in the optics but the other stuff I thank are artifacts.
RON C
IROC'N said:
Sorry for the delay. I wasn't expecting the web page. I don't
think I can upload a website page to Fototime.com. But I also have
a little website. My webspace is limited and I plan to change
hosts at some point in the future. But I uploaded your WB examples
to my current website anyway just to get it out there. So here is
the link:
http://www.niez.org/misc/rcwbex/wb.html

Thanks for putting this together!

Paul
Ron C said:
Ron C said:
If you look at the Phil's CA target, you will see that the green is
missing in the area around the white L. Because the target is
mostly black, you can't see what happened in that area.

If you look at the 505v leaf problem, you'll see that the "Fringe"
is zero in green, and the greens in other areas are reduced by some
amount as they are very dark.

I don't know what algorithma are used in the dsp but they can't be
very powerful.

I sent Brian some test examples to post but haven't heard from him.

The test used a very simple white balance method than equalizes the
channels but maintaining the overall level of the images.
Hopefully, I hear or see the examples here soon.

RON C
Ron Parr said:
Ron C said:
Ron,
I believe that a combination of bright areas along side of rather
weak areas show the PF as the red and blue channels are accentuated
and green it killed. I have a simple synthetic test to show a poor
mans white balance will do this but nowhere to upload.
Click to expand...
White balance is applied to the entire image. If the green is
knocked down so much that it is almost completely eliminated in the
brightest parts of the image, then it would totally kill it in the
rest of the image.

Ron Parr
Click to expand...
Click to expand...
Click to expand...
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