510 meta: scan speed, saturation & stripes

[Maria Jimena Ortega]

Search the CONFOCAL archive at http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Hello,

 

I am having a situation with a new 510 Meta from Zeiss and would like to receive some feedback from other users…

 

First thing I notice is that the intensity level varies a lot with scan speed (from 3.2 usec/pixel to .64 usec/pixel).  So, parameters choosed at a speed doesn`t work in the other…  it really  annoying!

 

The other situation is that when I am working at high speeds (0.8 – 1.64 usec/pix), average 4 line, if my image has saturating conditions ( I =255, 8 bit) I see some regular “stripes”.  

Zeiss tells there is a security system (to protect the PMT), but I do not understand why I see it in some speeds and not others, or why I don`t get them in my Pascal…  any clue?

The situation is that I need to saturate some parts of my sample in order to see the parts I am interested in!!  What should I do?

 

The also tell me not to make average and slow down speed… but I don`t get the same results… Can someone help me in this scan speed vs average discussion?

 

 

Any clue would be gratefully received!

 

Maria

[Teng-Leong Chew]

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Hi Maria,

The increase in fluorescence intensity with slower scan speed
is to be expected. The slower the scan speed, the higher the
pixel dwell time of the laser, you will always get higher
intensity with that no matter what confocal you use.

I am not sure about the stripes. Are you doing serpentine scan
by any chance, meaning scanning the sample back and forth and
not always from left to right? Serpentine scan tends to create
scan lines. I am not sure if that's what you are describing.

Leong


Teng-Leong Chew, Ph.D.
Director,
Cell Imaging Facility
Nikon Imaging Center, Chicago
Northwestern University Feinberg School of Medicine
303 E. Chicago Avenue
Chicago, IL 60611



Maria Jimena Ortega wrote:

>  ...
> First thing I notice is that the intensity level varies a lot with scan
> speed (from 3.2 usec/pixel to .64 usec/pixel).  So, parameters choosed
> at a speed doesn`t work in the other…  it really  annoying!
>
> The other situation is that when I am working at high speeds (0.8 – 1.64
> usec/pix), average 4 line, if my image has saturating conditions ( I
> =255, 8 bit) I see some regular “stripes”.  
>
> Zeiss tells there is a security system (to protect the PMT), but I do
> not understand why I see it in some speeds and not others, or why I
> don`t get them in my Pascal…  any clue?
>
> The situation is that I need to saturate some parts of my sample in
> order to see the parts I am interested in!!  What should I do?
>
> The also tell me not to make average and slow down speed… but I don`t
> get the same results… Can someone help me in this scan speed vs average
> discussion?

[Julio Vazquez]

Search the CONFOCAL archive at http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal -

Hi Maria, 

My experience with the 510 is that changing the scan speed should not affect the intensity levels (contrary to what the other lister posted; this may be instrument dependent-see paragraph below). In addition, slowing down the scan speed should be equivalent to increasing the averaging (for example, one scan at 3.2 microseconds per pixel should give you an image almost identical to 1.6 microsecond per pixel with an average of two). The intensity again should not change, and the effect on noise reduction should be almost identical.  Therefore, either something is wrong with your confocal, or Zeiss changed the way scan speed settings are handled...

Contrary to a CCD where signals are truly integrated (you keep adding photons until the well is full or saturated), PMTs read currents over relatively short time intervals (shorter than the pixel dwell time).  When you increase the pixel time, you could either increase the read interval time, or keep the read interval fixed and either sum the different reads, or average the different reads). If you sum the different reads, you could use the sum as the final output, or normalize it to the  read time to provide similar output independently of the pixel dwell time.  I could never figure out how Zeiss handles the scan speed issue, but the fact is that it is somewhat equivalent to averaging, at least in practice (for our 510 META). Our old Leica SP, on the other hand, would give higher intensities at slower speeds.  There was a thread on this topic sometime last year...

In any event, could you specify whether you see this effect on the META detector or on the regular detector(s)? Are you doing fluorescence, reflected light, or transmitted light imaging?  Do you see these effects on one specific channel or on any channel? Do you see this effect with any laser or a specific laser? What are the "stripes" you mention (are they dark, bright)? Do you see a difference if you average in line or frame mode? In other words, I am trying to see if this issue is related to the lasers, the scanner, or the PMTs.  

Even though this does not change the fact that you may have a problem with your instrument, I don't understand why you can't follow Zeiss' advice and slow down the scan speed, as long as you adjust the image parameters for your new conditions...   Also, are your bright and dark regions intermingled, or could you restrict your imaging to the low intensity regions you are interested in (by zooming in or by drawing a region of interest)? 

Independently of the averaging/scan speed issue, there is also the stripe issue. Do you also observe the "stripes" when you have grossly saturated pixels?  

Did you try to put a normal fluorescent sample and try to get an image?  Would you see the stripes then too? This also would give clues as to whether you have a problem with the instrument, or whether, as Zeiss seems to think, this is a result of oversaturating your PMTs 

Another thought is that we sometimes observe strange things for specific combinations of scan speed and image size. We also have had cases were specific configurations get corrupted and start giving very strange results. In that situation, we delete the configuration, restart the software, and create 

a new configuration from scratch.

As a side note, I personally tend to avoid the fastest scan speeds (faster than 1.6 microsecond) because at this speed most images are quite noisy, and I also don't like to hear the hight pitch sound of the vibrating mirrors. As with a car, you can probably run it at the highest possible speed, but it may not be very good to do that for too long...

--

Julio Vazquez

Fred Hutchinson Cancer Research Center

Seattle, WA 98109-1024

http://www.fhcrc.org/

 

==

On Mar 11, 2008, at 11:33 AM, Maria Jimena Ortega wrote:

Search the CONFOCAL archive at http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Hello,

 

I am having a situation with a new 510 Meta from Zeiss and would like to receive some feedback from other users…

 

First thing I notice is that the intensity level varies a lot with scan speed (from 3.2 usec/pixel to .64 usec/pixel).  So, parameters choosed at a speed doesn`t work in the other…  it really  annoying!

 

The other situation is that when I am working at high speeds (0.8 – 1.64 usec/pix), average 4 line, if my image has saturating conditions ( I =255, 8 bit) I see some regular “stripes”.  

Zeiss tells there is a security system (to protect the PMT), but I do not understand why I see it in some speeds and not others, or why I don`t get them in my Pascal…  any clue?

The situation is that I need to saturate some parts of my sample in order to see the parts I am interested in!!  What should I do?

 

The also tell me not to make average and slow down speed… but I don`t get the same results… Can someone help me in this scan speed vs average discussion?

 

 

Any clue would be gratefully received!

 

Maria

[Jacqueline Ross]

Search the CONFOCAL archive at http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Hi Maria,

 

What happens if you switch to 12bit imaging instead of 8bit? This would give you a greater dynamic range so may help with the saturation problem.

 

Kind regards,

 

Jacqui

 

Jacqueline Ross

Biomedical Imaging Microscopist
Biomedical Imaging Research Unit 
School of Medical Sciences 
Faculty of Medical & Health Sciences
The University of Auckland
Private Bag 92019
Auckland, NEW ZEALAND

Tel: [phone] Ext 87438
Fax: [phone]

<A title=http://www.health.auckland.ac.nz/biru/ href="blocked::http://www.health.auckland.ac.nz/biru/">http://www.health.auckland.ac.nz/biru/

 

---

From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Maria Jimena Ortega
Sent: Wednesday, 12 March 2008 7:34 a.m.
To: [hidden email]
Subject: 510 meta: scan speed, saturation & stripes

Search the CONFOCAL archive at http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Hello,

 

I am having a situation with a new 510 Meta from Zeiss and would like to receive some feedback from other users…

 

First thing I notice is that the intensity level varies a lot with scan speed (from 3.2 usec/pixel to .64 usec/pixel).  So, parameters choosed at a speed doesn`t work in the other…  it really  annoying!

 

The other situation is that when I am working at high speeds (0.8 – 1.64 usec/pix), average 4 line, if my image has saturating conditions ( I =255, 8 bit) I see some regular “stripes”.  

Zeiss tells there is a security system (to protect the PMT), but I do not understand why I see it in some speeds and not others, or why I don`t get them in my Pascal…  any clue?

The situation is that I need to saturate some parts of my sample in order to see the parts I am interested in!!  What should I do?

 

The also tell me not to make average and slow down speed… but I don`t get the same results… Can someone help me in this scan speed vs average discussion?

 

 

Any clue would be gratefully received!

 

Maria

[Dale Callaham]

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Double check the manual to see that this really gives you more dynamic
range and not just finer resolution of the same signal range.

Dale

Jacqui Ross wrote:

> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
> Hi Maria,
>  
> What happens if you switch to 12bit imaging instead of 8bit? This would
> give you a greater dynamic range so may help with the saturation problem.
>  
> Kind regards,
>  
> Jacqui
>  
>
> *Jacqueline Ross *
>
> Biomedical Imaging Microscopist
> Biomedical Imaging Research Unit
> School of Medical Sciences
> Faculty of Medical & Health Sciences
> The University of Auckland
> Private Bag 92019
> Auckland, NEW ZEALAND
>
> Tel: [phone] Ext 87438
> Fax: [phone]
>
> http://www.health.auckland.ac.nz/biru/ 
> <blocked::http://www.health.auckland.ac.nz/biru/>
>
>  
>
> ------------------------------------------------------------------------
> *From:* Confocal Microscopy List [mailto:[hidden email]]
> *On Behalf Of *Maria Jimena Ortega
> *Sent:* Wednesday, 12 March 2008 7:34 a.m.
> *To:* [hidden email]
> *Subject:* 510 meta: scan speed, saturation & stripes
>
> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>
> Hello,
>
>  
>
> I am having a situation with a new 510 Meta from Zeiss and would like to
> receive some feedback from other users…
>
>  
>
> First thing I notice is that the intensity level varies a lot with scan
> speed (from 3.2 usec/pixel to .64 usec/pixel).  So, parameters choosed
> at a speed doesn`t work in the other…  it really  annoying!
>
>  
>
> The other situation is that when I am working at high speeds (0.8 – 1.64
> usec/pix), average 4 line, if my image has saturating conditions ( I
> =255, 8 bit) I see some regular “stripes”.  
>
> Zeiss tells there is a security system (to protect the PMT), but I do
> not understand why I see it in some speeds and not others, or why I
> don`t get them in my Pascal…  any clue?
>
> The situation is that I need to saturate some parts of my sample in
> order to see the parts I am interested in!!  What should I do?
>
>  
>
> The also tell me not to make average and slow down speed… but I don`t
> get the same results… Can someone help me in this scan speed vs average
> discussion?
>
>  
>
>  
>
> Any clue would be gratefully received!
>
>  
>
> Maria
>

[S. Pagakis (IIBEAA)]

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Since it is the same PMT used either in 8bit ot 12bit imaging, it is  
unlikely it will give you greater dynamic range.

S. Pagakis

On 12 Mar 2008, at 04:41, Dale Callaham wrote:

> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>
> Double check the manual to see that this really gives you more dynamic  
> range and not just finer resolution of the same signal range.
>
> Dale
>
> Jacqui Ross wrote:
>> Search the CONFOCAL archive at  
>> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>> Hi Maria,
>>  What happens if you switch to 12bit imaging instead of 8bit? This  
>> would give you a greater dynamic range so may help with the  
>> saturation problem.
>>  Kind regards,
>>  Jacqui
>>  *Jacqueline Ross *
>> Biomedical Imaging Microscopist
>> Biomedical Imaging Research Unit School of Medical Sciences Faculty  
>> of Medical & Health Sciences
>> The University of Auckland
>> Private Bag 92019
>> Auckland, NEW ZEALAND
>> Tel: [phone] Ext 87438
>> Fax: [phone]
>> http://www.health.auckland.ac.nz/biru/ 
>> <blocked::http://www.health.auckland.ac.nz/biru/>
>>  
>> ----------------------------------------------------------------------
>> --
>> *From:* Confocal Microscopy List  
>> [mailto:[hidden email]] *On Behalf Of *Maria Jimena  
>> Ortega
>> *Sent:* Wednesday, 12 March 2008 7:34 a.m.
>> *To:* [hidden email]
>> *Subject:* 510 meta: scan speed, saturation & stripes
>> Search the CONFOCAL archive at  
>> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>> Hello,
>>  I am having a situation with a new 510 Meta from Zeiss and would  
>> like to receive some feedback from other users…
>>  First thing I notice is that the intensity level varies a lot with  
>> scan speed (from 3.2 usec/pixel to .64 usec/pixel).  So, parameters  
>> choosed at a speed doesn`t work in the other…  it really  annoying!
>>  The other situation is that when I am working at high speeds (0.8 –  
>> 1.64 usec/pix), average 4 line, if my image has saturating conditions  
>> ( I =255, 8 bit) I see some regular “stripes”.  Zeiss tells there is  
>> a security system (to protect the PMT), but I do not understand why I  
>> see it in some speeds and not others, or why I don`t get them in my  
>> Pascal…  any clue?
>> The situation is that I need to saturate some parts of my sample in  
>> order to see the parts I am interested in!!  What should I do?
>>  The also tell me not to make average and slow down speed… but I  
>> don`t get the same results… Can someone help me in this scan speed vs  
>> average discussion?
>>   Any clue would be gratefully received!
>>  Maria
>>

*********************************
Stamatis Pagakis Ph.D.
Biological Imaging Unit
Biomedical Research Foundation, Academy of Athens
Soranou Efessiou 4, Athens 115 27 - Greece
M: [phone] W: [phone]
FAX: [phone]             [hidden email]

[Michael Weber-4]

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Dale, you are right, 12 bit doesn't increase the dynamic range, it just
gives you higher intensity resolution. However, this might be a good way
to go in this case, since it could help to distinguish different intensity
levels. In most cases it's not necessary to saturate regions of the scan
field, it's just about displaying the low intensities (look-up-tables
etc.).

Maria, could you provide a sample image? Do the stripes only appear in the
lines with saturation ("comet-tails")? Or in a regular way over the whole
image?

Regarding the scan speed vs. averaging, I have the impression that
averaging is the better way to go. With a constant light source as sample,
reducing the scan speed results in a rather linear decrease of noise,
whereas averaging gives exponential decrease. Based on these measurements,
up to 8 times averaging is better than decreasing the speed to the similar
period.

Michael

> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>
> Double check the manual to see that this really gives you more dynamic
> range and not just finer resolution of the same signal range.
>
> Dale
>
> Jacqui Ross wrote:
>> Search the CONFOCAL archive at
>> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>> Hi Maria,
>>
>> What happens if you switch to 12bit imaging instead of 8bit? This would
>> give you a greater dynamic range so may help with the saturation
>> problem.
>>
>> Kind regards,
>>
>> Jacqui
>>
>>
>> *Jacqueline Ross *
>>
>> Biomedical Imaging Microscopist
>> Biomedical Imaging Research Unit
>> School of Medical Sciences
>> Faculty of Medical & Health Sciences
>> The University of Auckland
>> Private Bag 92019
>> Auckland, NEW ZEALAND
>>
>> Tel: [phone] Ext 87438
>> Fax: [phone]
>>
>> http://www.health.auckland.ac.nz/biru/
>> <blocked::http://www.health.auckland.ac.nz/biru/>
>>
>>
>>
>> ------------------------------------------------------------------------
>> *From:* Confocal Microscopy List [mailto:[hidden email]]
>> *On Behalf Of *Maria Jimena Ortega
>> *Sent:* Wednesday, 12 March 2008 7:34 a.m.
>> *To:* [hidden email]
>> *Subject:* 510 meta: scan speed, saturation & stripes
>>
>> Search the CONFOCAL archive at
>> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>>
>> Hello,
>>
>>
>>
>> I am having a situation with a new 510 Meta from Zeiss and would like to
>> receive some feedback from other users…
>>
>>
>>
>> First thing I notice is that the intensity level varies a lot with scan
>> speed (from 3.2 usec/pixel to .64 usec/pixel).  So, parameters choosed
>> at a speed doesn`t work in the other…  it really  annoying!
>>
>>
>>
>> The other situation is that when I am working at high speeds (0.8 – 1.64
>> usec/pix), average 4 line, if my image has saturating conditions ( I
>> =255, 8 bit) I see some regular “stripes”.
>>
>> Zeiss tells there is a security system (to protect the PMT), but I do
>> not understand why I see it in some speeds and not others, or why I
>> don`t get them in my Pascal…  any clue?
>>
>> The situation is that I need to saturate some parts of my sample in
>> order to see the parts I am interested in!!  What should I do?
>>
>>
>>
>> The also tell me not to make average and slow down speed… but I don`t
>> get the same results… Can someone help me in this scan speed vs average
>> discussion?
>>
>>
>>
>>
>>
>> Any clue would be gratefully received!
>>
>>
>>
>> Maria

[Aryeh Weiss]

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Michael Weber wrote:

>
> Regarding the scan speed vs. averaging, I have the impression that
> averaging is the better way to go. With a constant light source as sample,
> reducing the scan speed results in a rather linear decrease of noise,
> whereas averaging gives exponential decrease. Based on these measurements,
> up to 8 times averaging is better than decreasing the speed to the similar
> period.
>

This surprises me. How can averaging give an exponential decrease? I would
expect the S/N to increase as the square root of the number of averages.

In addition, if one considers the sum of N scans, it will include the sum of N
readouts (ie, N x readout noise). On the other hand, increasing dwell time x N
will increase the signal by a factor of N, but will only subject the signal to a
single readout. So I dont understand the above observation.

--aryeh
--
Aryeh Weiss
School of Engineering
Bar Ilan University
Ramat Gan 52900 Israel

Ph:  972-3-5317638
FAX: [phone]

[Michael Weber-4]

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Aryeh,

you are right, s/n increases as the square root of the number of averages.
I was using wrong words here - sorry for the confusion.

Michael

> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>
> Michael Weber wrote:
>
>>
>> Regarding the scan speed vs. averaging, I have the impression that
>> averaging is the better way to go. With a constant light source as
>> sample,
>> reducing the scan speed results in a rather linear decrease of noise,
>> whereas averaging gives exponential decrease. Based on these
>> measurements,
>> up to 8 times averaging is better than decreasing the speed to the
>> similar
>> period.
>>
>
> This surprises me. How can averaging give an exponential decrease? I would
> expect the S/N to increase as the square root of the number of averages.
>
> In addition, if one considers the sum of N scans, it will include the sum
> of N
> readouts (ie, N x readout noise). On the other hand, increasing dwell time
> x N
> will increase the signal by a factor of N, but will only subject the
> signal to a
> single readout. So I dont understand the above observation.
>
> --aryeh
> --
> Aryeh Weiss
> School of Engineering
> Bar Ilan University
> Ramat Gan 52900 Israel
>
> Ph:  972-3-5317638
> FAX: [phone]

[Gert van Cappellen]

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

When a Zeiss confocal (LSM510) is setup the system is calibrated in such
a way that it gives the same intensity for all scan speeds. In your case
you could ask Zeiss to do this calibration again. I once asked Zeiss if
it was possible to take this out and this was not recommended (done). I
did some comparisons by making 2 images of the same sample either with
lower scan speed or with a average and if you make a 2D plot of the
pixel intensities for both situations you will see that the average
image gives less noise.

We also make sometimes saturated pictures but never observed the
stripes, so this may be a machine dependent phenomenon.
I always thought that the detector protection that was mentioned in an
earlier mail was only active when you activate a bleaching protocol.

Best regards,
Gert van Cappellen

on 11-3-2008 20:33 Teng-Leong Chew said the following:

> Search the CONFOCAL archive at
> http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal
>
> Hi Maria,
>
> The increase in fluorescence intensity with slower scan speed
> is to be expected. The slower the scan speed, the higher the
> pixel dwell time of the laser, you will always get higher
> intensity with that no matter what confocal you use.
>
> I am not sure about the stripes. Are you doing serpentine scan
> by any chance, meaning scanning the sample back and forth and
> not always from left to right? Serpentine scan tends to create
> scan lines. I am not sure if that's what you are describing.
>
> Leong
>
>
> Teng-Leong Chew, Ph.D.
> Director,
> Cell Imaging Facility
> Nikon Imaging Center, Chicago
> Northwestern University Feinberg School of Medicine
> 303 E. Chicago Avenue
> Chicago, IL 60611
>
>
>
> Maria Jimena Ortega wrote:
>>  ...
>> First thing I notice is that the intensity level varies a lot with
>> scan speed (from 3.2 usec/pixel to .64 usec/pixel).  So, parameters
>> choosed at a speed doesn`t work in the other…  it really  annoying!
>>
>> The other situation is that when I am working at high speeds (0.8 –
>> 1.64 usec/pix), average 4 line, if my image has saturating conditions
>> ( I =255, 8 bit) I see some regular “stripes”.
>> Zeiss tells there is a security system (to protect the PMT), but I do
>> not understand why I see it in some speeds and not others, or why I
>> don`t get them in my Pascal…  any clue?
>>
>> The situation is that I need to saturate some parts of my sample in
>> order to see the parts I am interested in!!  What should I do?
>>
>> The also tell me not to make average and slow down speed… but I don`t
>> get the same results… Can someone help me in this scan speed vs
>> average discussion?

--
wigGert van Cappellen, [hidden email]
!!! NEW TEL [phone]; FAX [phone] !!!
Assistant professor
Dept. of Reproduction and Development; http://www.erasmusmc.nl/rede
Optical Imaging Centre; http://www.erasmusmc.nl/oic/
Erasmus MC, room Ee914, Dr. Molenwaterplein 50, 3015 GE ROTTERDAM, The Netherlands
Delivery adres:
Erasmus MC, Westzeedijk 353, 3015 AA Rotterdam, The Netherlands