I FOUND the little SQUARES

[O'Malley, Donald]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Gang,

Apologies for prolonging this thread (and I do agree with both parties, depending on whose email I read last).  I thus thought I might relate the finding of little squares in my MRC600 confocal images.  Because my analytical abilities were largely limited to what the BioRad confocal software (SOM) could do at that time (1994), I could not easily display the individual pixel values: all the images we published of nuclear calcium transients and neuronal activity in zebrafish (from 1994 to 1996) had a smooth aesthetically pleasing appearance (more accurate or less accurate depends I think on the semantic position one takes).   I had wanted to see the individual pixels by zooming up (enabled by SOM481B), but they were all smoothed out to my annoyance.  Somehow I discovered that if one slightly tilted the images, then that knocked out the SOM smoothing algorithm and I could now see the individual pixel values.  This enabled me to see what was happening on a pixel to pixel basis, in both space and time, in my line scans of calcium dynamics.  For anyone wishing to see how these two displays of the data vary in a side-by-side, real world sense, you might take a look at Fig. 3B and Fig. 3C of my Methods article on Calcium Imaging at the Limits of Resolution, O'Malley, et al., 2003, at:
http://www.digital-maze.com/id3.html
[you can actually see the slight tilt of the image in Fig. 3B, which I should have adjusted to perfectly register it to the space and time axes, but this was the best I could do at that time.].

I get the point that the sharp boundaries we see between adjacent pixels are physically impossible (in this case, for sure), but seeing the actual recorded data, presented as recorded, was useful at least to me.  Of course, the colors in the images are a figment of our artistic imaginations!

Enjoyed all the discussion,
Don

Don O'Malley
Assoc. Professor
Dept. Biology
[hidden email]<mailto:[hidden email]>
617-373-2284

[Andreas Bruckbauer]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****


Hi,
with some software you see pixels as squares (imageJ) and with others not (Imaris). I rather like to see them as this is a good warning that the pixel-resolution limit is reached. I had students interpreting interesting triangular patterns created by the software by linking  pixel points, they just did not notice that they zoomed in too much.  This was with a scanning probe microscope. Square pixel also make some good background image for an overlay with higher resolution single molecule tracking data, they give a better idea of the noise in the image. But i think for displaying a most likely version of the object under the microscope i think we we should forget about square pixels. Some flexibility in the image analysis software would be good.

best wishes

Andreas



 

 

 

-----Original Message-----
From: O'Malley, Donald <[hidden email]>
To: CONFOCALMICROSCOPY <[hidden email]>
Sent: Tue, 17 Apr 2012 16:12
Subject: I FOUND the little SQUARES


*****





To join, leave or search the confocal microscopy listserv, go to:





http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy





*****











Hi Gang,











Apologies for prolonging this thread (and I do agree with both parties,





depending on whose email I read last).  I thus thought I might relate the





finding of little squares in my MRC600 confocal images.  Because my analytical





abilities were largely limited to what the BioRad confocal software (SOM) could





do at that time (1994), I could not easily display the individual pixel values:





all the images we published of nuclear calcium transients and neuronal activity





in zebrafish (from 1994 to 1996) had a smooth aesthetically pleasing appearance





(more accurate or less accurate depends I think on the semantic position one





takes).   I had wanted to see the individual pixels by zooming up (enabled by





SOM481B), but they were all smoothed out to my annoyance.  Somehow I discovered





that if one slightly tilted the images, then that knocked out the SOM smoothing





algorithm and I could now see the individual pixel values.  This enabled me to





see what was happening on a pixel to pixel basis, in both space and time, in my





line scans of calcium dynamics.  For anyone wishing to see how these two





displays of the data vary in a side-by-side, real world sense, you might take a





look at Fig. 3B and Fig. 3C of my Methods article on Calcium Imaging at the





Limits of Resolution, O'Malley, et al., 2003, at:





http://www.digital-maze.com/id3.html





[you can actually see the slight tilt of the image in Fig. 3B, which I should





have adjusted to perfectly register it to the space and time axes, but this was





the best I could do at that time.].











I get the point that the sharp boundaries we see between adjacent pixels are





physically impossible (in this case, for sure), but seeing the actual recorded





data, presented as recorded, was useful at least to me.  Of course, the colors





in the images are a figment of our artistic imaginations!











Enjoyed all the discussion,





Don











Don O'Malley





Assoc. Professor





Dept. Biology





[hidden email]<mailto:[hidden email]>





617-373-2284






 
 

[Michael Schell]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

ImageJ offers a choice.  Go to Edit--options--appearance and you can  
toggle "interpolate zoomed images."  I like to see interpolated images  
when working (they DO look better), but this option does not change  
the underlying data set.

Michael

On Apr 18, 2012, at 6:16 PM, Andreas Bruckbauer wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
>
> Hi,
> with some software you see pixels as squares (imageJ) and with  
> others not (Imaris). I rather like to see them as this is a good  
> warning that the pixel-resolution limit is reached. I had students  
> interpreting interesting triangular patterns created by the software  
> by linking  pixel points, they just did not notice that they zoomed  
> in too much.  This was with a scanning probe microscope. Square  
> pixel also make some good background image for an overlay with  
> higher resolution single molecule tracking data, they give a better  
> idea of the noise in the image. But i think for displaying a most  
> likely version of the object under the microscope i think we we  
> should forget about square pixels. Some flexibility in the image  
> analysis software would be good.
>
> best wishes
>
> Andreas
>
>
>
>
>
>
>
>
>
> -----Original Message-----
> From: O'Malley, Donald <[hidden email]>
> To: CONFOCALMICROSCOPY <[hidden email]>
> Sent: Tue, 17 Apr 2012 16:12
> Subject: I FOUND the little SQUARES
>
>
> *****
>
>
>
>
>
> To join, leave or search the confocal microscopy listserv, go to:
>
>
>
>
>
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>
>
>
>
>
> *****
>
>
>
>
>
>
>
>
>
>
>
> Hi Gang,
>
>
>
>
>
>
>
>
>
>
>
> Apologies for prolonging this thread (and I do agree with both  
> parties,
>
>
>
>
>
> depending on whose email I read last).  I thus thought I might  
> relate the
>
>
>
>
>
> finding of little squares in my MRC600 confocal images.  Because my  
> analytical
>
>
>
>
>
> abilities were largely limited to what the BioRad confocal software  
> (SOM) could
>
>
>
>
>
> do at that time (1994), I could not easily display the individual  
> pixel values:
>
>
>
>
>
> all the images we published of nuclear calcium transients and  
> neuronal activity
>
>
>
>
>
> in zebrafish (from 1994 to 1996) had a smooth aesthetically pleasing  
> appearance
>
>
>
>
>
> (more accurate or less accurate depends I think on the semantic  
> position one
>
>
>
>
>
> takes).   I had wanted to see the individual pixels by zooming up  
> (enabled by
>
>
>
>
>
> SOM481B), but they were all smoothed out to my annoyance.  Somehow I  
> discovered
>
>
>
>
>
> that if one slightly tilted the images, then that knocked out the  
> SOM smoothing
>
>
>
>
>
> algorithm and I could now see the individual pixel values.  This  
> enabled me to
>
>
>
>
>
> see what was happening on a pixel to pixel basis, in both space and  
> time, in my
>
>
>
>
>
> line scans of calcium dynamics.  For anyone wishing to see how these  
> two
>
>
>
>
>
> displays of the data vary in a side-by-side, real world sense, you  
> might take a
>
>
>
>
>
> look at Fig. 3B and Fig. 3C of my Methods article on Calcium Imaging  
> at the
>
>
>
>
>
> Limits of Resolution, O'Malley, et al., 2003, at:
>
>
>
>
>
> http://www.digital-maze.com/id3.html
>
>
>
>
>
> [you can actually see the slight tilt of the image in Fig. 3B, which  
> I should
>
>
>
>
>
> have adjusted to perfectly register it to the space and time axes,  
> but this was
>
>
>
>
>
> the best I could do at that time.].
>
>
>
>
>
>
>
>
>
>
>
> I get the point that the sharp boundaries we see between adjacent  
> pixels are
>
>
>
>
>
> physically impossible (in this case, for sure), but seeing the  
> actual recorded
>
>
>
>
>
> data, presented as recorded, was useful at least to me.  Of course,  
> the colors
>
>
>
>
>
> in the images are a figment of our artistic imaginations!
>
>
>
>
>
>
>
>
>
>
>
> Enjoyed all the discussion,
>
>
>
>
>
> Don
>
>
>
>
>
>
>
>
>
>
>
> Don O'Malley
>
>
>
>
>
> Assoc. Professor
>
>
>
>
>
> Dept. Biology
>
>
>
>
>
> [hidden email]<mailto:[hidden email]>
>
>
>
>
>
> 617-373-2284
>
>
>
>
>
>
>
>

Michael J. Schell
Dept. Pharmacology
USUHS
4301 Jones Bridge Rd.
Bethesda, MD  20814
Tel: [phone]
[hidden email]
http://www.usuhs.mil/pha/mschell.html
*please note new email address*

[Arvonn Tully]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

COMMERCIAL RESPONSE..
Dear All,

Just to add a few bits to Andrea's statement, -- Imaris defaults to an
"Interpolated" view of the data, but this can be switched to a "square
pixel" view by changing a preference.
From Imaris, Preferences-->Display-->Interpolate (check on, or off)
This is purely a visualization, and the actual data is not changed by this
Display setting.

Cheers
arvonn

ps.. there is no bead..

----
Arvonn Tully
Western North America Application Support
Bitplane
[hidden email]


On Wed, Apr 18, 2012 at 3:16 PM, Andreas Bruckbauer <[hidden email]>wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
>
> Hi,
> with some software you see pixels as squares (imageJ) and with others not
> (Imaris). I rather like to see them as this is a good warning that the
> pixel-resolution limit is reached. I had students interpreting interesting
> triangular patterns created by the software by linking  pixel points, they
> just did not notice that they zoomed in too much.  This was with a scanning
> probe microscope. Square pixel also make some good background image for an
> overlay with higher resolution single molecule tracking data, they give a
> better idea of the noise in the image. But i think for displaying a most
> likely version of the object under the microscope i think we we should
> forget about square pixels. Some flexibility in the image analysis software
> would be good.
>
> best wishes
>
> Andreas
>
>
>
>
>
>
>
>
>
> -----Original Message-----
> From: O'Malley, Donald <[hidden email]>
> To: CONFOCALMICROSCOPY <[hidden email]>
> Sent: Tue, 17 Apr 2012 16:12
> Subject: I FOUND the little SQUARES
>
>
> *****
>
>
>
>
>
> To join, leave or search the confocal microscopy listserv, go to:
>
>
>
>
>
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>
>
>
>
>
> *****
>
>
>
>
>
>
>
>
>
>
>
> Hi Gang,
>
>
>
>
>
>
>
>
>
>
>
> Apologies for prolonging this thread (and I do agree with both parties,
>
>
>
>
>
> depending on whose email I read last).  I thus thought I might relate the
>
>
>
>
>
> finding of little squares in my MRC600 confocal images.  Because my
> analytical
>
>
>
>
>
> abilities were largely limited to what the BioRad confocal software (SOM)
> could
>
>
>
>
>
> do at that time (1994), I could not easily display the individual pixel
> values:
>
>
>
>
>
> all the images we published of nuclear calcium transients and neuronal
> activity
>
>
>
>
>
> in zebrafish (from 1994 to 1996) had a smooth aesthetically pleasing
> appearance
>
>
>
>
>
> (more accurate or less accurate depends I think on the semantic position
> one
>
>
>
>
>
> takes).   I had wanted to see the individual pixels by zooming up (enabled
> by
>
>
>
>
>
> SOM481B), but they were all smoothed out to my annoyance.  Somehow I
> discovered
>
>
>
>
>
> that if one slightly tilted the images, then that knocked out the SOM
> smoothing
>
>
>
>
>
> algorithm and I could now see the individual pixel values.  This enabled
> me to
>
>
>
>
>
> see what was happening on a pixel to pixel basis, in both space and time,
> in my
>
>
>
>
>
> line scans of calcium dynamics.  For anyone wishing to see how these two
>
>
>
>
>
> displays of the data vary in a side-by-side, real world sense, you might
> take a
>
>
>
>
>
> look at Fig. 3B and Fig. 3C of my Methods article on Calcium Imaging at the
>
>
>
>
>
> Limits of Resolution, O'Malley, et al., 2003, at:
>
>
>
>
>
> http://www.digital-maze.com/id3.html
>
>
>
>
>
> [you can actually see the slight tilt of the image in Fig. 3B, which I
> should
>
>
>
>
>
> have adjusted to perfectly register it to the space and time axes, but
> this was
>
>
>
>
>
> the best I could do at that time.].
>
>
>
>
>
>
>
>
>
>
>
> I get the point that the sharp boundaries we see between adjacent pixels
> are
>
>
>
>
>
> physically impossible (in this case, for sure), but seeing the actual
> recorded
>
>
>
>
>
> data, presented as recorded, was useful at least to me.  Of course, the
> colors
>
>
>
>
>
> in the images are a figment of our artistic imaginations!
>
>
>
>
>
>
>
>
>
>
>
> Enjoyed all the discussion,
>
>
>
>
>
> Don
>
>
>
>
>
>
>
>
>
>
>
> Don O'Malley
>
>
>
>
>
> Assoc. Professor
>
>
>
>
>
> Dept. Biology
>
>
>
>
>
> [hidden email]<mailto:[hidden email]>
>
>
>
>
>
> 617-373-2284
>
>
>
>
>
>
>
>
>