Position Available (Updated) Light Microscopy Specialist/Technician, Whitehead Institute, Cambridge, MA

[Rodney Byrd]

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

Whitehead Institute for Biomedical Research (Cambridge, MA) is a leading,
nonprofit research and educational institution that has defined the cutting
edge of biomedical science, creating a legacy of research excellence and
academic eminence since 1982. Wholly independent in its governance, finances
and research programs, Whitehead shares a teaching affiliation with
Massachusetts Institute of Technology (MIT), offering the intellectual,
collegial and scientific benefits of a leading research university.
Whitehead’s Faculty Members and Fellows run pioneering programs in cancer
research, immunology, developmental biology, stem cell research,
regenerative medicine, genetics and genomics—programs with a record of
success. Director David C. Page likens the Institute to an artists’ colony.
 “What we do here at Whitehead is to attract the best possible intellectual
capital and empower maximally creative—really wildly creative—individuals to
realize their dreams within these walls”.

Overall responsibilities for Technician and Specialist levels
• Support and train users on the light microscopes in the Whitehead
Institute’s, Keck Imaging Facility (including widefield, TIR-FM, spinning
disc, and confocal microscopes) for imaging of both fixed and live samples.
• Assist in the general maintenance of Keck microscopy facility including
updating the user database, maintaining and cleaning the scientific
instruments, and ensuring the maintenance of necessary reagents.
• Other tasks as required or assigned.

Incumbent will train users on how to properly use the microscopes in the
facility (both initial training and more advanced sessions), working with
users to discuss the way to use microscopy to maximize their research goals.
 The incumbent may also provide users with services that complement their
work.  The incumbent must be able to recognize and report problems and/or
issues that arise with the microscopes, and must be able to work with other
facility personnel and microscope service engineers to correct such
problems.  In addition to working with users, the incumbent is responsible
for keeping the microscopes clean, for every day lab cleaning, and for
preparing and maintaining stock solutions for facility use.

Additional responsibilities for Specialist level
• Consult with users on experiment design, materials and methods, and
instrument optimization.
• Assist in data acquisition and analysis, troubleshooting problems with
microscopes or samples.
• Organize/conduct periodic courses explaining core principles of light
microscopy and specific approaches/applications.
• Work with the other members of the Keck facility to develop future plans
for the acquisition, implementation, and use of additional light microscopes.

Qualifications
• Strong knowledge of advanced light microscopy techniques including TIR-FM,
confocal microscopy, and image processing software.
• A Bachelor’s Degree in biological and/or physical sciences, or a related
discipline.
• For Technician level, a minimum of one year’s experience working in a
biomedical research laboratory preferred with understanding of basic lab
techniques and principles of cell biology.
• Strong computer and record keeping skills for maintaining an up to date
facility database and web site.
• Strong written and verbal communications, interpersonal and customer
service skills.
• Strong troubleshooting and problem solving skills.  
• The ability to work independently under deadlines with minimal guidance,
and to work as a team member.

Additional qualifications for Specialist level
• 5 -7 years of related experience.
• A working knowledge of design, theory and functioning of instrumentation
and how instrumentation can support research goals as put forth by users.
• Drive and ability to stay current on, and apply advances in the field of
light microscopy that can enhance research locally, and support faculty,
research support staff and students, by assisting and training them to carry
out advanced microscopy techniques in support of their studies.

[Nina Allen]

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

Mailed you a package today.
Part is breakable so open carefully.
Did Eileen get the socks?

Nina Allen
Professor Emerita
Department of Plant Biology
North Carolina State University

Sent from my iPhone

On Feb 2, 2011, at 1:42 PM, Rodney Byrd <[hidden email]> wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Whitehead Institute for Biomedical Research (Cambridge, MA) is a  
> leading,
> nonprofit research and educational institution that has defined the  
> cutting
> edge of biomedical science, creating a legacy of research excellence  
> and
> academic eminence since 1982. Wholly independent in its governance,  
> finances
> and research programs, Whitehead shares a teaching affiliation with
> Massachusetts Institute of Technology (MIT), offering the  
> intellectual,
> collegial and scientific benefits of a leading research university.
> Whitehead’s Faculty Members and Fellows run pioneering programs in c
> ancer
> research, immunology, developmental biology, stem cell research,
> regenerative medicine, genetics and genomics—programs with a record  
> of
> success. Director David C. Page likens the Institute to an artists’  
> colony.
> “What we do here at Whitehead is to attract the best possible intell
> ectual
> capital and empower maximally creative—really wildly creative—
> individuals to
> realize their dreams within these walls”.
>
> Overall responsibilities for Technician and Specialist levels
> •    Support and train users on the light microscopes in the Whitehe
> ad
> Institute’s, Keck Imaging Facility (including widefield, TIR-FM, spi
> nning
> disc, and confocal microscopes) for imaging of both fixed and live  
> samples.
> •    Assist in the general maintenance of Keck microscopy facility i
> ncluding
> updating the user database, maintaining and cleaning the scientific
> instruments, and ensuring the maintenance of necessary reagents.
> •    Other tasks as required or assigned.
>
> Incumbent will train users on how to properly use the microscopes in  
> the
> facility (both initial training and more advanced sessions), working  
> with
> users to discuss the way to use microscopy to maximize their  
> research goals.
> The incumbent may also provide users with services that complement  
> their
> work.  The incumbent must be able to recognize and report problems  
> and/or
> issues that arise with the microscopes, and must be able to work  
> with other
> facility personnel and microscope service engineers to correct such
> problems.  In addition to working with users, the incumbent is  
> responsible
> for keeping the microscopes clean, for every day lab cleaning, and for
> preparing and maintaining stock solutions for facility use.
>
> Additional responsibilities for Specialist level
> •    Consult with users on experiment design, materials and methods,
>  and
> instrument optimization.
> •    Assist in data acquisition and analysis, troubleshooting proble
> ms with
> microscopes or samples.
> •    Organize/conduct periodic courses explaining core principles of
>  light
> microscopy and specific approaches/applications.
> •    Work with the other members of the Keck facility to develop fut
> ure plans
> for the acquisition, implementation, and use of additional light  
> microscopes.
>
> Qualifications
> •    Strong knowledge of advanced light microscopy techniques includ
> ing TIR-FM,
> confocal microscopy, and image processing software.
> •    A Bachelor’s Degree in biological and/or physical sciences,  
> or a related
> discipline.
> •    For Technician level, a minimum of one year’s experience  
> working in a
> biomedical research laboratory preferred with understanding of basic  
> lab
> techniques and principles of cell biology.
> •    Strong computer and record keeping skills for maintaining an up
>  to date
> facility database and web site.
> •    Strong written and verbal communications, interpersonal and cus
> tomer
> service skills.
> •    Strong troubleshooting and problem solving skills.
> •    The ability to work independently under deadlines with minimal  
> guidance,
> and to work as a team member.
>
> Additional qualifications for Specialist level
> •     5 -7 years of related experience.
> •    A working knowledge of design, theory and functioning of instru
> mentation
> and how instrumentation can support research goals as put forth by  
> users.
> •    Drive and ability to stay current on, and apply advances in the
>  field of
> light microscopy that can enhance research locally, and support  
> faculty,
> research support staff and students, by assisting and training them  
> to carry
> out advanced microscopy techniques in support of their studies.

[Jerry (Gerald) Sedgewick]

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

I would like to make a table of manufacturer's image formats that
separate scaled 16-bit images from those image formats in which 12-bits
of tonal data (0-4095 tones) are placed in a 16-bit (0-65,535) "package."

As an example, images saved as 16-bit TIFF files in SPOT software are
scaled from 12-bits to 16-bits, so these contain the full range of tones
(at least in earlier versions of the software).  Olympus .oib and
exported TIFF files, on the other hand, contain 12-bits of tonal range
within a 16-bit image, leaving 4 "empty" bits.

In programs that auto-scale for display purposes, like Image J, the
12-bit inside 16-bit looks great on the screen; but, in programs like
Photoshop in which images are NOT auto-scaled for display purposes, the
images are completely black, because only 1/16th of the 16-bit tonal
range is displayed.

I don't have access to all image formats that retain the 12-bit original
tonal range within a 16-bit image: I know of only Olympus and the DVC
company.

I would greatly appreciate it if participants on this list could let me
know of other manufacturer's image formats that retain 12-bit tonal
range inside a 16-bit image.  Or maybe a good source for downloading
Zeiss, Leica, Nikon, Andor, Hamamatsu, etc., sample images. This table
will be posted on the web for those who are interested.

Thanks!

Jerry

--
IMAGING: Image Integrity, Quantitation, Digital Imaging Instruction

Jerry Sedgewick
Sedgewick Initiatives
965 Cromwell Avenue
Saint Paul, MN  55114
651-788-2261

[hidden email]
[hidden email]

http://www.imagingandanalysis.com
http://www.quickphotoshop.com

Author of: "Scientific Imaging with Photoshop: Methods, Measurement, and
Output"

[leoncio vergara]

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

Metamorph also maintains the 12 bit data leaving the 4 extra bits empty.

The software in the Zeiss LSM 510 gives you the option of doing that (12 bit Tiffs) or expanding the scale to fill up the 16 bit range.

I don't think there is interpolation or "data creation" involved, just a simple linear scaling, right? x 16?

I think the motivation is precisely to deal with programs like photoshop that do not scale the displays and confuse the users... I have dealt with that in our imaging core for too long, some users would not hesitate to throw away information by converting all their 12 bit images to 8 bits just so they display right in photoshop... dumb

Leoncio A. Vergara MD
Assistant Director
Center for Biomedical Engineering
Assistant Professor
Microbiology and Immunology
University of Texas Medical Branch
409-750-2153 (cell)


-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Jerry (Gerald) Sedgewick
Sent: Thursday, February 03, 2011 11:12 AM
To: [hidden email]
Subject: Re: 12- to 16-bit images

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

I would like to make a table of manufacturer's image formats that
separate scaled 16-bit images from those image formats in which 12-bits
of tonal data (0-4095 tones) are placed in a 16-bit (0-65,535) "package."

As an example, images saved as 16-bit TIFF files in SPOT software are
scaled from 12-bits to 16-bits, so these contain the full range of tones
(at least in earlier versions of the software).  Olympus .oib and
exported TIFF files, on the other hand, contain 12-bits of tonal range
within a 16-bit image, leaving 4 "empty" bits.

In programs that auto-scale for display purposes, like Image J, the
12-bit inside 16-bit looks great on the screen; but, in programs like
Photoshop in which images are NOT auto-scaled for display purposes, the
images are completely black, because only 1/16th of the 16-bit tonal
range is displayed.

I don't have access to all image formats that retain the 12-bit original
tonal range within a 16-bit image: I know of only Olympus and the DVC
company.

I would greatly appreciate it if participants on this list could let me
know of other manufacturer's image formats that retain 12-bit tonal
range inside a 16-bit image.  Or maybe a good source for downloading
Zeiss, Leica, Nikon, Andor, Hamamatsu, etc., sample images. This table
will be posted on the web for those who are interested.

Thanks!

Jerry

--
IMAGING: Image Integrity, Quantitation, Digital Imaging Instruction

Jerry Sedgewick
Sedgewick Initiatives
965 Cromwell Avenue
Saint Paul, MN  55114
651-788-2261

[hidden email]
[hidden email]

http://www.imagingandanalysis.com
http://www.quickphotoshop.com

Author of: "Scientific Imaging with Photoshop: Methods, Measurement, and
Output"

[Jerry (Gerald) Sedgewick]

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

The increase from 12- to 16-bits is simple linear scaling.  This is no
different than what is done by manufacturers who use a 10-bit chip and
then who scale to 12-bits for the image.  Often the CCD/CMOS chips used
produces a bit range less than that of what is used in a final image,
such as a chip with a 14-bit range that is then scaled to a 16-bit image.

If the image is intended for measuring the optical density or intensity,
then a change to the bit depth (whether or not it has been scaled
without you knowing it by the manufacturer) would be inappropriate.

If, however, like the vast majority of images created, the intent is to
create a representative image like what was seen under the microscope, a
simple linear scaling is perfectly appropriate.  It could arguably be
considered not just appropriate, but a necessary change, because the
greater range of tones then minimizes rounding errors when a tonal
correction is made.

For those microscopists who make necessary corrections to images, such
as color corrections for common hue shifts introduced by the camera for
brightfield images, corrections for color fringing (to correct unnatural
colors at edges due to light scattering), noise filtering, and
brightening/contrast corrections to fill the dynamic range (when
appropriate), the additional headroom minimizes rounding errors and the
amount of "destruction" to the image.  Some of these corrections cannot
be made in scientific software, and so Photoshop is often used.

I hope this offers a better explanation.

Jerry

On 2/3/2011 2:20 PM, Vergara, Leoncio A. wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Metamorph also maintains the 12 bit data leaving the 4 extra bits empty.
>
> The software in the Zeiss LSM 510 gives you the option of doing that (12 bit Tiffs) or expanding the scale to fill up the 16 bit range.
>
> I don't think there is interpolation or "data creation" involved, just a simple linear scaling, right? x 16?
>
> I think the motivation is precisely to deal with programs like photoshop that do not scale the displays and confuse the users... I have dealt with that in our imaging core for too long, some users would not hesitate to throw away information by converting all their 12 bit images to 8 bits just so they display right in photoshop... dumb
>
> Leoncio A. Vergara MD
> Assistant Director
> Center for Biomedical Engineering
> Assistant Professor
> Microbiology and Immunology
> University of Texas Medical Branch
> 409-750-2153 (cell)
>
>
> -----Original Message-----
> From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Jerry (Gerald) Sedgewick
> Sent: Thursday, February 03, 2011 11:12 AM
> To: [hidden email]
> Subject: Re: 12- to 16-bit images
>
> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> I would like to make a table of manufacturer's image formats that
> separate scaled 16-bit images from those image formats in which 12-bits
> of tonal data (0-4095 tones) are placed in a 16-bit (0-65,535) "package."
>
> As an example, images saved as 16-bit TIFF files in SPOT software are
> scaled from 12-bits to 16-bits, so these contain the full range of tones
> (at least in earlier versions of the software).  Olympus .oib and
> exported TIFF files, on the other hand, contain 12-bits of tonal range
> within a 16-bit image, leaving 4 "empty" bits.
>
> In programs that auto-scale for display purposes, like Image J, the
> 12-bit inside 16-bit looks great on the screen; but, in programs like
> Photoshop in which images are NOT auto-scaled for display purposes, the
> images are completely black, because only 1/16th of the 16-bit tonal
> range is displayed.
>
> I don't have access to all image formats that retain the 12-bit original
> tonal range within a 16-bit image: I know of only Olympus and the DVC
> company.
>
> I would greatly appreciate it if participants on this list could let me
> know of other manufacturer's image formats that retain 12-bit tonal
> range inside a 16-bit image.  Or maybe a good source for downloading
> Zeiss, Leica, Nikon, Andor, Hamamatsu, etc., sample images. This table
> will be posted on the web for those who are interested.
>
> Thanks!
>
> Jerry
>

--
IMAGING: Image Integrity, Quantitation, Digital Imaging Instruction

Jerry Sedgewick
Sedgewick Initiatives
965 Cromwell Avenue
Saint Paul, MN  55114
651-788-2261

[hidden email]
[hidden email]

http://www.imagingandanalysis.com
http://www.quickphotoshop.com

Author of: "Scientific Imaging with Photoshop: Methods, Measurement, and
Output"

[James Pawley]

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

Hi all,

I would like to hear a little about where this data is coming from.

Am I right in assuming that it is related to a photon image of some
sort. Probably not confocal, where we virtually never get as many as
4,096 photons/pixel.

But even if it is the output from some widefield LM on a CCD or
sCMOS, and the saturation level really does correspond to 65k photons
(or at least 65k photoelectrons), I can't get all that worried about
whether it is 65,000 or 65,016, when the Poisson Noise on such a
signal is +/- 256 photoelectrons. Even if you manage somehow to get
your staining to be so specific that there is almost no background
(say 100:1), then the darkest pixel would be 650 but the Poison Noise
will still be +/-25.6 photoelectrons.

Usually in fluorescence microscopy, we are working at the other end
of the scale: maybe 2,000 PE/pixel peak, 20 PE in the darkest area.

In this case, I don't see a lot of reason to worry about the
interpolation levels, as long as you have the camera-gain set so that
one count in the memory equals about half the read noise of the
camera in e/pixel.

Cheers,

Jim Pawley

***************************************************************************
Prof. James B. Pawley,                          Ph.
608-238-3953                        
21. N. Prospect Ave. Madison, WI 53726 USA
[hidden email]
3D Microscopy of Living Cells Course, June 11-23, 2011, UBC, Vancouver Canada
Info: http://www.3dcourse.ubc.ca/            Applications due by March 15, 2011
               "If it ain't diffraction, it must be statistics." Anon.

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Metamorph also maintains the 12 bit data leaving the 4 extra bits empty.
>
>The software in the Zeiss LSM 510 gives you the option of doing that
>(12 bit Tiffs) or expanding the scale to fill up the 16 bit range.
>
>I don't think there is interpolation or "data creation" involved,
>just a simple linear scaling, right? x 16?
>
>I think the motivation is precisely to deal with programs like
>photoshop that do not scale the displays and confuse the users... I
>have dealt with that in our imaging core for too long, some users
>would not hesitate to throw away information by converting all their
>12 bit images to 8 bits just so they display right in photoshop...
>dumb
>
>Leoncio A. Vergara MD
>Assistant Director
>Center for Biomedical Engineering
>Assistant Professor
>Microbiology and Immunology
>University of Texas Medical Branch
>409-750-2153 (cell)
>
>
>-----Original Message-----
>From: Confocal Microscopy List
>[mailto:[hidden email]] On Behalf Of Jerry
>(Gerald) Sedgewick
>Sent: Thursday, February 03, 2011 11:12 AM
>To: [hidden email]
>Subject: Re: 12- to 16-bit images
>
>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>I would like to make a table of manufacturer's image formats that
>separate scaled 16-bit images from those image formats in which 12-bits
>of tonal data (0-4095 tones) are placed in a 16-bit (0-65,535) "package."
>
>As an example, images saved as 16-bit TIFF files in SPOT software are
>scaled from 12-bits to 16-bits, so these contain the full range of tones
>(at least in earlier versions of the software).  Olympus .oib and
>exported TIFF files, on the other hand, contain 12-bits of tonal range
>within a 16-bit image, leaving 4 "empty" bits.
>
>In programs that auto-scale for display purposes, like Image J, the
>12-bit inside 16-bit looks great on the screen; but, in programs like
>Photoshop in which images are NOT auto-scaled for display purposes, the
>images are completely black, because only 1/16th of the 16-bit tonal
>range is displayed.
>
>I don't have access to all image formats that retain the 12-bit original
>tonal range within a 16-bit image: I know of only Olympus and the DVC
>company.
>
>I would greatly appreciate it if participants on this list could let me
>know of other manufacturer's image formats that retain 12-bit tonal
>range inside a 16-bit image.  Or maybe a good source for downloading
>Zeiss, Leica, Nikon, Andor, Hamamatsu, etc., sample images. This table
>will be posted on the web for those who are interested.
>
>Thanks!
>
>Jerry
>
>--
>IMAGING: Image Integrity, Quantitation, Digital Imaging Instruction
>
>Jerry Sedgewick
>Sedgewick Initiatives
>965 Cromwell Avenue
>Saint Paul, MN  55114
>651-788-2261
>
>[hidden email]
>[hidden email]
>
>http://www.imagingandanalysis.com
>http://www.quickphotoshop.com
>
>Author of: "Scientific Imaging with Photoshop: Methods, Measurement, and
>Output"

--
***************************************************************************
Prof. James B. Pawley,                          Ph.
608-238-3953                        
21. N. Prospect Ave. Madison, WI 53726 USA
[hidden email]
3D Microscopy of Living Cells Course, June 11-23, 2011, UBC, Vancouver Canada
Info: http://www.3dcourse.ubc.ca/            Applications due by March 15, 2011
               "If it ain't diffraction, it must be statistics." Anon.

[Tao Tong]

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

I believe the 12-bit or 16-bit business refers to the quantification levels
of the A/D conversion, 12-bit pixel dynamic range is 0-4095, 4096 levels of
the converted signal level, not necessarily 4096 photos / pixel. It would be
very strange to measure pixels by photos, unless a very well calibrated
photon counting scheme.

16-bit refers to the pixels dynamic range, 0-65535, 65536 levels of the
detected signal. Most likely it will not refers to 65k photoelectrons.

Unless we are doing photon counting, pixels are usually not measured by
photons.

Tao

On Thu, Feb 3, 2011 at 7:36 PM, James Pawley <[hidden email]> wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Hi all,
>
> I would like to hear a little about where this data is coming from.
>
> Am I right in assuming that it is related to a photon image of some sort.
> Probably not confocal, where we virtually never get as many as 4,096
> photons/pixel.
>
> But even if it is the output from some widefield LM on a CCD or sCMOS, and
> the saturation level really does correspond to 65k photons (or at least 65k
> photoelectrons), I can't get all that worried about whether it is 65,000 or
> 65,016, when the Poisson Noise on such a signal is +/- 256 photoelectrons.
> Even if you manage somehow to get your staining to be so specific that there
> is almost no background (say 100:1), then the darkest pixel would be 650 but
> the Poison Noise will still be +/-25.6 photoelectrons.
>
> Usually in fluorescence microscopy, we are working at the other end of the
> scale: maybe 2,000 PE/pixel peak, 20 PE in the darkest area.
>
> In this case, I don't see a lot of reason to worry about the interpolation
> levels, as long as you have the camera-gain set so that one count in the
> memory equals about half the read noise of the camera in e/pixel.
>
> Cheers,
>
> Jim Pawley
>
> ***************************************************************************
> Prof. James B. Pawley,                                      Ph.
> 608-238-3953                               21. N. Prospect Ave. Madison, WI
> 53726 USA [hidden email]
> 3D Microscopy of Living Cells Course, June 11-23, 2011, UBC, Vancouver
> Canada
> Info: http://www.3dcourse.ubc.ca/           Applications due by March 15,
> 2011
>               "If it ain't diffraction, it must be statistics." Anon.
>
>
>  *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> Metamorph also maintains the 12 bit data leaving the 4 extra bits empty.
>>
>> The software in the Zeiss LSM 510 gives you the option of doing that (12
>> bit Tiffs) or expanding the scale to fill up the 16 bit range.
>>
>> I don't think there is interpolation or "data creation" involved, just a
>> simple linear scaling, right? x 16?
>>
>> I think the motivation is precisely to deal with programs like photoshop
>> that do not scale the displays and confuse the users... I have dealt with
>> that in our imaging core for too long, some users would not hesitate to
>> throw away information by converting all their 12 bit images to 8 bits just
>> so they display right in photoshop... dumb
>>
>> Leoncio A. Vergara MD
>> Assistant Director
>> Center for Biomedical Engineering
>> Assistant Professor
>> Microbiology and Immunology
>> University of Texas Medical Branch
>> 409-750-2153 (cell)
>>
>>
>> -----Original Message-----
>> From: Confocal Microscopy List [mailto:[hidden email]]
>> On Behalf Of Jerry (Gerald) Sedgewick
>> Sent: Thursday, February 03, 2011 11:12 AM
>> To: [hidden email]
>> Subject: Re: 12- to 16-bit images
>>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> I would like to make a table of manufacturer's image formats that
>> separate scaled 16-bit images from those image formats in which 12-bits
>> of tonal data (0-4095 tones) are placed in a 16-bit (0-65,535) "package."
>>
>> As an example, images saved as 16-bit TIFF files in SPOT software are
>> scaled from 12-bits to 16-bits, so these contain the full range of tones
>> (at least in earlier versions of the software).  Olympus .oib and
>> exported TIFF files, on the other hand, contain 12-bits of tonal range
>> within a 16-bit image, leaving 4 "empty" bits.
>>
>> In programs that auto-scale for display purposes, like Image J, the
>> 12-bit inside 16-bit looks great on the screen; but, in programs like
>> Photoshop in which images are NOT auto-scaled for display purposes, the
>> images are completely black, because only 1/16th of the 16-bit tonal
>> range is displayed.
>>
>> I don't have access to all image formats that retain the 12-bit original
>> tonal range within a 16-bit image: I know of only Olympus and the DVC
>> company.
>>
>> I would greatly appreciate it if participants on this list could let me
>> know of other manufacturer's image formats that retain 12-bit tonal
>> range inside a 16-bit image.  Or maybe a good source for downloading
>> Zeiss, Leica, Nikon, Andor, Hamamatsu, etc., sample images. This table
>> will be posted on the web for those who are interested.
>>
>> Thanks!
>>
>> Jerry
>>
>> --
>> IMAGING: Image Integrity, Quantitation, Digital Imaging Instruction
>>
>> Jerry Sedgewick
>> Sedgewick Initiatives
>> 965 Cromwell Avenue
>> Saint Paul, MN  55114
>> 651-788-2261
>>
>> [hidden email]
>> [hidden email]
>>
>> http://www.imagingandanalysis.com
>> http://www.quickphotoshop.com
>>
>> Author of: "Scientific Imaging with Photoshop: Methods, Measurement, and
>> Output"
>>
>
>
> --
> ***************************************************************************
> Prof. James B. Pawley,                                      Ph.
> 608-238-3953                               21. N. Prospect Ave. Madison, WI
> 53726 USA [hidden email]
> 3D Microscopy of Living Cells Course, June 11-23, 2011, UBC, Vancouver
> Canada
> Info: http://www.3dcourse.ubc.ca/           Applications due by March 15,
> 2011
>               "If it ain't diffraction, it must be statistics." Anon.
>

--
http://tongtao.com

[Tao Tong]

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

Very sorry, I typed photos instead of photons, :-(

Also, need to point out photons and photoelectrons are not to be confused,
neither is commonly used in referring to pixel value.

Tao

On Thu, Feb 3, 2011 at 10:40 PM, Tao Tong <[hidden email]> wrote:

> I believe the 12-bit or 16-bit business refers to the quantification levels
> of the A/D conversion, 12-bit pixel dynamic range is 0-4095, 4096 levels of
> the converted signal level, not necessarily 4096 photos / pixel. It would be
> very strange to measure pixels by photos, unless a very well calibrated
> photon counting scheme.
>
> 16-bit refers to the pixels dynamic range, 0-65535, 65536 levels of the
> detected signal. Most likely it will not refers to 65k photoelectrons.
>
> Unless we are doing photon counting, pixels are usually not measured by
> photons.
>
> Tao
>
>
> On Thu, Feb 3, 2011 at 7:36 PM, James Pawley <[hidden email]> wrote:
>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> Hi all,
>>
>> I would like to hear a little about where this data is coming from.
>>
>> Am I right in assuming that it is related to a photon image of some sort.
>> Probably not confocal, where we virtually never get as many as 4,096
>> photons/pixel.
>>
>> But even if it is the output from some widefield LM on a CCD or sCMOS, and
>> the saturation level really does correspond to 65k photons (or at least 65k
>> photoelectrons), I can't get all that worried about whether it is 65,000 or
>> 65,016, when the Poisson Noise on such a signal is +/- 256 photoelectrons.
>> Even if you manage somehow to get your staining to be so specific that there
>> is almost no background (say 100:1), then the darkest pixel would be 650 but
>> the Poison Noise will still be +/-25.6 photoelectrons.
>>
>> Usually in fluorescence microscopy, we are working at the other end of the
>> scale: maybe 2,000 PE/pixel peak, 20 PE in the darkest area.
>>
>> In this case, I don't see a lot of reason to worry about the interpolation
>> levels, as long as you have the camera-gain set so that one count in the
>> memory equals about half the read noise of the camera in e/pixel.
>>
>> Cheers,
>>
>> Jim Pawley
>>
>>
>> ***************************************************************************
>> Prof. James B. Pawley,                                      Ph.
>> 608-238-3953                               21. N. Prospect Ave. Madison, WI
>> 53726 USA [hidden email]
>> 3D Microscopy of Living Cells Course, June 11-23, 2011, UBC, Vancouver
>> Canada
>> Info: http://www.3dcourse.ubc.ca/           Applications due by March 15,
>> 2011
>>               "If it ain't diffraction, it must be statistics." Anon.
>>
>>
>>  *****
>>> To join, leave or search the confocal microscopy listserv, go to:
>>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>> *****
>>>
>>> Metamorph also maintains the 12 bit data leaving the 4 extra bits empty.
>>>
>>> The software in the Zeiss LSM 510 gives you the option of doing that (12
>>> bit Tiffs) or expanding the scale to fill up the 16 bit range.
>>>
>>> I don't think there is interpolation or "data creation" involved, just a
>>> simple linear scaling, right? x 16?
>>>
>>> I think the motivation is precisely to deal with programs like photoshop
>>> that do not scale the displays and confuse the users... I have dealt with
>>> that in our imaging core for too long, some users would not hesitate to
>>> throw away information by converting all their 12 bit images to 8 bits just
>>> so they display right in photoshop... dumb
>>>
>>> Leoncio A. Vergara MD
>>> Assistant Director
>>> Center for Biomedical Engineering
>>> Assistant Professor
>>> Microbiology and Immunology
>>> University of Texas Medical Branch
>>> 409-750-2153 (cell)
>>>
>>>
>>> -----Original Message-----
>>> From: Confocal Microscopy List [mailto:[hidden email]]
>>> On Behalf Of Jerry (Gerald) Sedgewick
>>> Sent: Thursday, February 03, 2011 11:12 AM
>>> To: [hidden email]
>>> Subject: Re: 12- to 16-bit images
>>>
>>> *****
>>> To join, leave or search the confocal microscopy listserv, go to:
>>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>> *****
>>>
>>> I would like to make a table of manufacturer's image formats that
>>> separate scaled 16-bit images from those image formats in which 12-bits
>>> of tonal data (0-4095 tones) are placed in a 16-bit (0-65,535) "package."
>>>
>>> As an example, images saved as 16-bit TIFF files in SPOT software are
>>> scaled from 12-bits to 16-bits, so these contain the full range of tones
>>> (at least in earlier versions of the software).  Olympus .oib and
>>> exported TIFF files, on the other hand, contain 12-bits of tonal range
>>> within a 16-bit image, leaving 4 "empty" bits.
>>>
>>> In programs that auto-scale for display purposes, like Image J, the
>>> 12-bit inside 16-bit looks great on the screen; but, in programs like
>>> Photoshop in which images are NOT auto-scaled for display purposes, the
>>> images are completely black, because only 1/16th of the 16-bit tonal
>>> range is displayed.
>>>
>>> I don't have access to all image formats that retain the 12-bit original
>>> tonal range within a 16-bit image: I know of only Olympus and the DVC
>>> company.
>>>
>>> I would greatly appreciate it if participants on this list could let me
>>> know of other manufacturer's image formats that retain 12-bit tonal
>>> range inside a 16-bit image.  Or maybe a good source for downloading
>>> Zeiss, Leica, Nikon, Andor, Hamamatsu, etc., sample images. This table
>>> will be posted on the web for those who are interested.
>>>
>>> Thanks!
>>>
>>> Jerry
>>>
>>> --
>>> IMAGING: Image Integrity, Quantitation, Digital Imaging Instruction
>>>
>>> Jerry Sedgewick
>>> Sedgewick Initiatives
>>> 965 Cromwell Avenue
>>> Saint Paul, MN  55114
>>> 651-788-2261
>>>
>>> [hidden email]
>>> [hidden email]
>>>
>>> http://www.imagingandanalysis.com
>>> http://www.quickphotoshop.com
>>>
>>> Author of: "Scientific Imaging with Photoshop: Methods, Measurement, and
>>> Output"
>>>
>>
>>
>> --
>>
>> ***************************************************************************
>> Prof. James B. Pawley,                                      Ph.
>> 608-238-3953                               21. N. Prospect Ave. Madison, WI
>> 53726 USA [hidden email]
>> 3D Microscopy of Living Cells Course, June 11-23, 2011, UBC, Vancouver
>> Canada
>> Info: http://www.3dcourse.ubc.ca/           Applications due by March 15,
>> 2011
>>               "If it ain't diffraction, it must be statistics." Anon.
>>
>
>
>
> --
> http://tongtao.com
>

--
http://tongtao.com

[George McNamara]

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

Hi Jerry,

Zeiss Axiovision (4.7.2 through 4.8.2) take the camera data and report
the number in Axiovision, and export it to TIFF, if the user chooses
16-bit format (there is a checkbox in Export if the user wants 8-bit).
The LSM510 (aim software) let the user acquire in 8-bit or 12-bit mode
(I almost always select 12-bit). LSM710 ZEN lets the user choose 8-, 12-
or 16-bit acquisition mode.

Most back-illuminated CCD's offer 16-bit output. Many EMCCD's are
back-illuminated, when operating in EM mode typically outputting 16-bit
mode, eben though as Jim Pawley is likely to point out, they might have
collected 0, 1, 2, ... 10?, ... 16? photons in a likely fast exposure
time. Maybe the digitizer should just spit out the square root of tyhe
raw value.

Back to Axiovision (and MetaMorph, which has always given users the
value from the camera) and cameras. My core now has a Hamamatsu ORCA-II
ERG with both fast 12-bit mode (slightly slower readout than an ORCA-ER)
and a slower 14-bit mode. Same number of photons coming in are reported
as different numbers of intensity level, with 14-bit mode clearly having
less readout noise. The camera has offset of about 200 in fast mode and
about 580 in slow mode. If slow mode was just doing some silly x16, that
value would be close to 1600. Why run slow? Better data. Does it really
matter? Hard to say. I have been very underwhelmed with the Zeiss
Apotome optical sectioning device with the ORCA-ER. Basically, so little
dynamic range left after the Apotome math, that practically a binary
image. Now with 14-bit data to do math on, Apotome's output impresses me
a little bit (but I'd still rather use our LSM710 or MP/SP5).


George

On 2/3/2011 4:46 PM, Jerry (Gerald) Sedgewick wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> The increase from 12- to 16-bits is simple linear scaling.  This is no
> different than what is done by manufacturers who use a 10-bit chip and
> then who scale to 12-bits for the image.  Often the CCD/CMOS chips
> used produces a bit range less than that of what is used in a final
> image, such as a chip with a 14-bit range that is then scaled to a
> 16-bit image.
>
> If the image is intended for measuring the optical density or
> intensity, then a change to the bit depth (whether or not it has been
> scaled without you knowing it by the manufacturer) would be
> inappropriate.
>
> If, however, like the vast majority of images created, the intent is
> to create a representative image like what was seen under the
> microscope, a simple linear scaling is perfectly appropriate.  It
> could arguably be considered not just appropriate, but a necessary
> change, because the greater range of tones then minimizes rounding
> errors when a tonal correction is made.
>
> For those microscopists who make necessary corrections to images, such
> as color corrections for common hue shifts introduced by the camera
> for brightfield images, corrections for color fringing (to correct
> unnatural colors at edges due to light scattering), noise filtering,
> and brightening/contrast corrections to fill the dynamic range (when
> appropriate), the additional headroom minimizes rounding errors and
> the amount of "destruction" to the image.  Some of these corrections
> cannot be made in scientific software, and so Photoshop is often used.
>
> I hope this offers a better explanation.
>
> Jerry
>
> On 2/3/2011 2:20 PM, Vergara, Leoncio A. wrote:
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> Metamorph also maintains the 12 bit data leaving the 4 extra bits empty.
>>
>> The software in the Zeiss LSM 510 gives you the option of doing that
>> (12 bit Tiffs) or expanding the scale to fill up the 16 bit range.
>>
>> I don't think there is interpolation or "data creation" involved,
>> just a simple linear scaling, right? x 16?
>>
>> I think the motivation is precisely to deal with programs like
>> photoshop that do not scale the displays and confuse the users... I
>> have dealt with that in our imaging core for too long, some users
>> would not hesitate to throw away information by converting all their
>> 12 bit images to 8 bits just so they display right in photoshop... dumb
>>
>> Leoncio A. Vergara MD
>> Assistant Director
>> Center for Biomedical Engineering
>> Assistant Professor
>> Microbiology and Immunology
>> University of Texas Medical Branch
>> 409-750-2153 (cell)
>>
>>
>> -----Original Message-----
>> From: Confocal Microscopy List
>> [mailto:[hidden email]] On Behalf Of Jerry (Gerald)
>> Sedgewick
>> Sent: Thursday, February 03, 2011 11:12 AM
>> To: [hidden email]
>> Subject: Re: 12- to 16-bit images
>>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> I would like to make a table of manufacturer's image formats that
>> separate scaled 16-bit images from those image formats in which 12-bits
>> of tonal data (0-4095 tones) are placed in a 16-bit (0-65,535)
>> "package."
>>
>> As an example, images saved as 16-bit TIFF files in SPOT software are
>> scaled from 12-bits to 16-bits, so these contain the full range of tones
>> (at least in earlier versions of the software).  Olympus .oib and
>> exported TIFF files, on the other hand, contain 12-bits of tonal range
>> within a 16-bit image, leaving 4 "empty" bits.
>>
>> In programs that auto-scale for display purposes, like Image J, the
>> 12-bit inside 16-bit looks great on the screen; but, in programs like
>> Photoshop in which images are NOT auto-scaled for display purposes, the
>> images are completely black, because only 1/16th of the 16-bit tonal
>> range is displayed.
>>
>> I don't have access to all image formats that retain the 12-bit original
>> tonal range within a 16-bit image: I know of only Olympus and the DVC
>> company.
>>
>> I would greatly appreciate it if participants on this list could let me
>> know of other manufacturer's image formats that retain 12-bit tonal
>> range inside a 16-bit image.  Or maybe a good source for downloading
>> Zeiss, Leica, Nikon, Andor, Hamamatsu, etc., sample images. This table
>> will be posted on the web for those who are interested.
>>
>> Thanks!
>>
>> Jerry
>>
>
>

--


George McNamara, PhD
Analytical Imaging Core Facility
University of Miami

[Nico Stuurman]

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

On Feb 3, 2011, at 1:46 PM, Jerry (Gerald) Sedgewick wrote:

> The increase from 12- to 16-bits is simple linear scaling.  This is  
> no different than what is done by manufacturers who use a 10-bit  
> chip and then who scale to 12-bits for the image.  Often the CCD/
> CMOS chips used produces a bit range less than that of what is used  
> in a final image, such as a chip with a 14-bit range that is then  
> scaled to a 16-bit image.

It is convenient to organize data in computer memory in bytes (each  
equals 8 bits).  That is why cameras with a 10, 12, or 14 bit AD  
converter will output data as 16 bits (2 bytes).  What you are now  
telling me is that some software will left shift these bits, for  
instance the original number:

00001111 11111111

will now be represented as:

11111111 11110000

All of this is completely valid, as long as everyone agrees on what  
those numbers in computer memory actually mean.  For a well-calibrated  
camera, the photo-electron conversion number (number of photon  
electrons per digital unit) will be known (and really, what everyone  
using a camera for scientific imaging should be concerned about is the  
photon-flux at each pixel).  Representing the digital numbers by left  
shifting them still makes it possible to calculate the photon flux,  
but surely adds extra, unneeded, complexity.  Why does anyone choose  
such a complex representation of the measured data?  Apparently so  
that the image "looks good" when opened in certain picture  
manipulation software!

In my opinion, such a representation of data is bad, since it makes it  
harder to understand what was actually measured.  Instead of trying to  
make things look good by changing data it makes a lot more sense to  
educate users so that they understand how to look at their data.  If  
you really think that Photoshop is a good tool to work with such data,  
I suggest to ask the Photoshop programmers to write tools that make it  
easier to visualize scientific data (I heard they are trying to cater  
to the scientific market lately).  In the mean time, I hope that  
everyone will let their camera and software manufacturers know that  
they want their data as the original numbers and not bit shifted by an  
arbitrary amount!

Best,

Nico



Vale Lab
HHMI / Dept. of Cellular and Molecular Pharmacology
University of California, San Francisco