Andor ixon fast acquisition (isolated crop mode)

[Holden Seamus John]

Hi

I am very interested in using Micromanager for small field of view (128x128 pix) very fast (~500Hz) data acquisition on an EMCCD camera.

I am currently investigating Andors relatively new "Isolated crop mode" for their Ixon+ and Ixon3 cameras. This allows high frame rates by restricting readout to only a small region at the bottom left of the CCD. I have observed significant differences between the max frame rate possible with the Andor Solis software and Micromanager 1.4.7 32bit Windows, which seem to be related to the ReadoutTime/Andor-ActualInterval calculated by the driver.

Using identical settings for Andor Solis and Micromanger (vertical shift speed, readout mode etc),with an isolated crop window of 128x128, the minimum frame rate is 2.5ms (although I have found 3ms to work most reliably). Micromanger gives me an apparent minimum readout time of 7ms, and "Andor-ActualInterval" of 7.85ms with "internal trigger". However with these settings the "Andor Elapsed time" after recording a short stack is completely haywire, giving different intervals between frames ranging from 1-10ms. 

Similarly, with crop mode 64x64, I can achieve 1ms frame rate on Solis, with 2ms working reliably. MM gives apparent minimum readout time of 4ms, and "Andor-ActualInterval" of 4.3ms with "internal trigger". Again the haywire "Andor Elapsed time" is observed at this frame rate. If I choose "Software trigger" (which means what?) the minimum readout time  drops to 3ms, but the "Andor-ActualInterval" remains at 4.3ms.

It seems like there is some sort of bottleneck in the MM Andor driver which prevents the same frame rates being achieved as with the Solis software.

Any advice much appreciated. Also, if any has had any success in performing these sorts of measurement with micromanger (perhaps with a dedicated small FOV cam), input is much appreciated.

I also have 2 small FOV 128x128 cams (Andor and Photometrics) on demo at the moment, so I will report back on whether I have any more success with these. 

Regards

Seamus

-- 
Dr Seamus Holden
Post-doctoral Fellow
Bâtiment des Sciences Physiques
École Polytechnique Fédérale de Lausanne
CH-1015 Lausanne, Switzerland

Phone: [phone]

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2dcopy1
_______________________________________________
micro-manager-general mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/micro-manager-general

[P.Almada]

Hi,

I'd like to add that this isn't a Micro manager problem. I see the same difference with an iXon2 EMCCD under Andor iQ 2 versus Andor SOLIS. So far their software team is stumped on what's the problem.

Best,

Pedro Almada

On 10 October 2011 17:34, Seamus Holden <[hidden email]> wrote:

Hi

I am very interested in using Micromanager for small field of view (128x128 pix) very fast (~500Hz) data acquisition on an EMCCD camera.

I am currently investigating Andors relatively new "Isolated crop mode" for their Ixon+ and Ixon3 cameras. This allows high frame rates by restricting readout to only a small region at the bottom left of the CCD. I have observed significant differences between the max frame rate possible with the Andor Solis software and Micromanager 1.4.7 32bit Windows, which seem to be related to the ReadoutTime/Andor-ActualInterval calculated by the driver.

Using identical settings for Andor Solis and Micromanger (vertical shift speed, readout mode etc),with an isolated crop window of 128x128, the minimum frame rate is 2.5ms (although I have found 3ms to work most reliably). Micromanger gives me an apparent minimum readout time of 7ms, and "Andor-ActualInterval" of 7.85ms with "internal trigger". However with these settings the "Andor Elapsed time" after recording a short stack is completely haywire, giving different intervals between frames ranging from 1-10ms. 

Similarly, with crop mode 64x64, I can achieve 1ms frame rate on Solis, with 2ms working reliably. MM gives apparent minimum readout time of 4ms, and "Andor-ActualInterval" of 4.3ms with "internal trigger". Again the haywire "Andor Elapsed time" is observed at this frame rate. If I choose "Software trigger" (which means what?) the minimum readout time  drops to 3ms, but the "Andor-ActualInterval" remains at 4.3ms.

It seems like there is some sort of bottleneck in the MM Andor driver which prevents the same frame rates being achieved as with the Solis software.

Any advice much appreciated. Also, if any has had any success in performing these sorts of measurement with micromanger (perhaps with a dedicated small FOV cam), input is much appreciated.

I also have 2 small FOV 128x128 cams (Andor and Photometrics) on demo at the moment, so I will report back on whether I have any more success with these. 

Regards

Seamus

-- 
Dr Seamus Holden
Post-doctoral Fellow
Bâtiment des Sciences Physiques
École Polytechnique Fédérale de Lausanne
CH-1015 Lausanne, Switzerland

Phone: <a href="tel:%2B41-%280%2921-693-0557" value="[phone]" target="_blank">[phone]

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2dcopy1
_______________________________________________
micro-manager-general mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/micro-manager-general

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2d-oct
_______________________________________________
micro-manager-general mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/micro-manager-general

[Holden Seamus John]

Hi

Just wanted to add a conclusion to this mail, in case it comes in handy for others. It is possible to get full frame rates with Andor and MM. The problems I was having seemed to be 2-fold:

1. Incorrect power-saving settings on my PC:
http://valelab.ucsf.edu/~MM/MMwiki/index.php/Andor#Disable_power_savings
and
http://www.austinblanco.com/blog/?p=818
which did seem to be limiting speed and possibly causing problems with timestamps.

2. Slow update of "Andor-ActualInterval" driver parameter adding a bit of confusion to the mix (quite often it does not seem to be updated until after an acquisition)

If you see these problems, check you are not affected by the 1st issue, and be aware of the second. Finally, consider accurately checking the frame rate during acquisition by monitoring the trigger out with an oscilloscope.

Regards
Seamus Holden


On 10/10/11 18:34, Seamus Holden wrote:

Hi

I am very interested in using Micromanager for small field of view (128x128 pix) very fast (~500Hz) data acquisition on an EMCCD camera.

I am currently investigating Andors relatively new "Isolated crop mode" for their Ixon+ and Ixon3 cameras. This allows high frame rates by restricting readout to only a small region at the bottom left of the CCD. I have observed significant differences between the max frame rate possible with the Andor Solis software and Micromanager 1.4.7 32bit Windows, which seem to be related to the ReadoutTime/Andor-ActualInterval calculated by the driver.

Using identical settings for Andor Solis and Micromanger (vertical shift speed, readout mode etc),with an isolated crop window of 128x128, the minimum frame rate is 2.5ms (although I have found 3ms to work most reliably). Micromanger gives me an apparent minimum readout time of 7ms, and "Andor-ActualInterval" of 7.85ms with "internal trigger". However with these settings the "Andor Elapsed time" after recording a short stack is completely haywire, giving different intervals between frames ranging from 1-10ms. 

Similarly, with crop mode 64x64, I can achieve 1ms frame rate on Solis, with 2ms working reliably. MM gives apparent minimum readout time of 4ms, and "Andor-ActualInterval" of 4.3ms with "internal trigger". Again the haywire "Andor Elapsed time" is observed at this frame rate. If I choose "Software trigger" (which means what?) the minimum readout time  drops to 3ms, but the "Andor-ActualInterval" remains at 4.3ms.

It seems like there is some sort of bottleneck in the MM Andor driver which prevents the same frame rates being achieved as with the Solis software.

Any advice much appreciated. Also, if any has had any success in performing these sorts of measurement with micromanger (perhaps with a dedicated small FOV cam), input is much appreciated.

I also have 2 small FOV 128x128 cams (Andor and Photometrics) on demo at the moment, so I will report back on whether I have any more success with these. 

Regards

Seamus

-- 
Dr Seamus Holden
Post-doctoral Fellow
Bâtiment des Sciences Physiques
École Polytechnique Fédérale de Lausanne
CH-1015 Lausanne, Switzerland

Phone: [phone]

------------------------------------------------------------------------------
RSA&reg; Conference 2012
Save &#36;700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
micro-manager-general mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/micro-manager-general