strange grid pattern

On 2012-05-22 09:34:41 -0700, sobriquet said:


Hi.

Does anyone have an explanation for the partial grid pattern visible in
the background of this picture (consisting of horizontal and vertical
lines)?

http://i.imgur.com/EUJBA.jpg

I am not sure that describing what you are seeing as a "partial grid
pattern" is quite correct.
Confirm that the areas I have marked are what you are talking about, so
we are on the same page.
http://db.tt/MDPLIqNp

For now I can only guess that what you are seeing is some type of
aliasing artifact. I am sure that others will have a much better idea,
along with a possible fix.


--
Regards,

Savageduck

On 5/22/2012 2:21 PM, sobriquet wrote:
On Tuesday, May 22, 2012 8:03:27 PM UTC+2, Savageduck wrote:
On 2012-05-22 09:34:41 -0700, said:


Hi.

Does anyone have an explanation for the partial grid pattern visible in
the background of this picture (consisting of horizontal and vertical
lines)?

http://i.imgur.com/EUJBA.jpg

I am not sure that describing what you are seeing as a "partial grid
pattern" is quite correct.

Well ok, no regular grid but rather an implied grid of irregularly spaced horizontal and vertical line segments.

Confirm that the areas I have marked are what you are talking about, so
we are on the same page.
http://db.tt/MDPLIqNp

Yes, these lines.


For now I can only guess that what you are seeing is some type of
aliasing artifact. I am sure that others will have a much better idea,
along with a possible fix.


--
Regards,

Savageduck

Interesting, try as I will I can only see one pattern that looks like a
few small squares in out of focus top left area. It's surely some sort
of aliasing. Is it visible at different resolutions?

--
Jim Silverton (Potomac, MD)

Extraneous "not" in Reply To.

On 5/22/2012 2:47 PM, James Silverton wrote:
On 5/22/2012 2:21 PM, sobriquet wrote:
On Tuesday, May 22, 2012 8:03:27 PM UTC+2, Savageduck wrote:
On 2012-05-22 09:34:41 -0700, said:


Hi.

Does anyone have an explanation for the partial grid pattern visible in
the background of this picture (consisting of horizontal and vertical
lines)?

http://i.imgur.com/EUJBA.jpg

I am not sure that describing what you are seeing as a "partial grid
pattern" is quite correct.

Well ok, no regular grid but rather an implied grid of irregularly
spaced horizontal and vertical line segments.

Confirm that the areas I have marked are what you are talking about, so
we are on the same page.
http://db.tt/MDPLIqNp

Yes, these lines.


For now I can only guess that what you are seeing is some type of
aliasing artifact. I am sure that others will have a much better idea,
along with a possible fix.


--
Regards,

Savageduck

Interesting, try as I will I can only see one pattern that looks like a
few small squares in out of focus top left area. It's surely some sort
of aliasing. Is it visible at different resolutions?

Can I add that opening with IrfanView does very little but Windows Photo
Viewer and Paint show an overall square grid most apparent in Paint but
switching to Full Screen almost removes the pattern.

--
Jim Silverton (Potomac, MD)

Extraneous "not" in Reply To.

On 22/05/2012 19:47, James Silverton wrote:
On 5/22/2012 2:21 PM, sobriquet wrote:
On Tuesday, May 22, 2012 8:03:27 PM UTC+2, Savageduck wrote:
On 2012-05-22 09:34:41 -0700, said:


Hi.

Does anyone have an explanation for the partial grid pattern visible in
the background of this picture (consisting of horizontal and vertical
lines)?

http://i.imgur.com/EUJBA.jpg

I am not sure that describing what you are seeing as a "partial grid
pattern" is quite correct.

Well ok, no regular grid but rather an implied grid of irregularly
spaced horizontal and vertical line segments.

Confirm that the areas I have marked are what you are talking about, so
we are on the same page.
http://db.tt/MDPLIqNp

Yes, these lines.


For now I can only guess that what you are seeing is some type of
aliasing artifact. I am sure that others will have a much better idea,
along with a possible fix.


Savageduck

Interesting, try as I will I can only see one pattern that looks like a
few small squares in out of focus top left area. It's surely some sort
of aliasing. Is it visible at different resolutions?

I don't think it is aliasing. The source image is completely out of
focus there with no sharp features to alias. The reason this quirk
stands out is because there is no genuine high frequency component to
hide it. The in focus insect makes it impossible to follow the lines
although every one is present on the other side at a weaker level than
on the side where it is most obvious.

The features are sharp single pixel lines with adjacent ripples and not
regularly spaced in terms of JPEG index or any other obvious sensor
metric. My best guess is that some kind of anti-shake mechanism has
moved the sensor during the exposure or there is some kind of subtle
electronic interference occurring.

For those wondering what all the fuss is about easily visible lines
occur horizontally at (14, 604) (541,305) delta 300 pixels
(183,111) (448, 816) (556, 272) (815, 778) delta 265, 112, 256
taking top left as (0,0) and coordinates as (x,y)

256 is the only one of these spacings that might be a memory related
issue. The lines are accurately parallel to the sensor array and can be
picked up with difficulty on any out of focus zone.

It seems to me like the lines have a different exposure on either side
so my instinct is some kind of anti-blur sensor movement or active IS
lens affecting exposure. That or you were being illuminated by fairly
high power pulsed radar at the time of taking the picture.

--
Regards,
Martin Brown

On Tue, 22 May 2012 11:03:27 -0700, Savageduck
wrote:
: On 2012-05-22 09:34:41 -0700, sobriquet said:
:
:
: Hi.
:
: Does anyone have an explanation for the partial grid pattern visible in
: the background of this picture (consisting of horizontal and vertical
: lines)?
:
: http://i.imgur.com/EUJBA.jpg
:
: I am not sure that describing what you are seeing as a "partial grid
: pattern" is quite correct.
: Confirm that the areas I have marked are what you are talking about, so
: we are on the same page.
: http://db.tt/MDPLIqNp
:
: For now I can only guess that what you are seeing is some type of
: aliasing artifact. I am sure that others will have a much better idea,
: along with a possible fix.

My guess is a sloppy compression algorithm that results in one or more rows or
columns of pixels being jettisoned. In effect, it's roundoff error.

Bob

On 23/05/2012 5:08 PM, Martin Brown wrote:
On 22/05/2012 19:47, James Silverton wrote:
On 5/22/2012 2:21 PM, sobriquet wrote:
On Tuesday, May 22, 2012 8:03:27 PM UTC+2, Savageduck wrote:
On 2012-05-22 09:34:41 -0700, said:


Hi.

Does anyone have an explanation for the partial grid pattern
visible in
the background of this picture (consisting of horizontal and vertical
lines)?

http://i.imgur.com/EUJBA.jpg

I am not sure that describing what you are seeing as a "partial grid


I have had this sort of thing its just the image has been corrupted and
the original is more than likely fine.

Mine have come from a slow transfer from media to media or multiple
transfers.

On 24/05/2012 05:05, Rob wrote:
On 23/05/2012 5:08 PM, Martin Brown wrote:
On 22/05/2012 19:47, James Silverton wrote:
On 5/22/2012 2:21 PM, sobriquet wrote:
On Tuesday, May 22, 2012 8:03:27 PM UTC+2, Savageduck wrote:
On 2012-05-22 09:34:41 -0700, said:


Hi.

Does anyone have an explanation for the partial grid pattern
visible in
the background of this picture (consisting of horizontal and vertical
lines)?

http://i.imgur.com/EUJBA.jpg

I am not sure that describing what you are seeing as a "partial grid


I have had this sort of thing its just the image has been corrupted and
the original is more than likely fine.

Mine have come from a slow transfer from media to media or multiple
transfers.

No. A corrupted JPEG stream looks entirely different and *must* change
state over a JPEG 8x8 block boundary. These don't.

--
Regards,
Martin Brown