Time Measurement in CoreLog

[Peter Haub]

Hi,

we are still looking for an improvement of our systems performance.

We logged some scans and found that in the time stamps the smallest
difference is 15.6 msec.

It seems as if this time resolutions is based on the std::clock function
on Windows in our case, which is typically 16 msec.

(The
         boost::posix_time::ptime bt =
boost::posix_time::microsec_clock::local_time();
uses the  std::clock  .  )

The question is if we can increase the time stamp resolution by some
configuration?
Is the fairly large resolution effecting other functions in
MicroManager, e.g. time delays.

Regards,
Peter



------------------------------------------------------------------------------
Get your Android app more play: Bring it to the BlackBerry PlayBook
in minutes. BlackBerry App World™ now supports Android™ Apps
for the BlackBerry® PlayBook™. Discover just how easy and simple
it is! http://p.sf.net/sfu/android-dev2dev

_______________________________________________
micro-manager-general mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/micro-manager-general

[Peter Haub]

The std::clock function  seems not be the reason for the 15.6 msec timing interval.

I run a test on a Win7 32bit machine in which the debug log shows time differences of 1 msec.

DBG(2011-11-01T20:46:40.638616, 4148, t:00254760:) > "(. org.micromanager.mm/mmc popNextImageMD #<Metadata mmcorej.Metadata@b0f2b2>)"
DBG(2011-11-01T20:46:40.640616, 4148, t:00254760:) > "  --> [B@10dbef1"
DBG(2011-11-01T20:46:40.640616, 4148, t:00254760:) > "(. org.micromanager.mm/mmc getSystemStateCache)"
DBG(2011-11-01T20:46:40.641616, 4148, t:00254760:) > "  --> mmcorej.Configuration@172e08e"
DBG(2011-11-01T20:46:40.644616, 4148, t:00254760:) > "(. org.micromanager.mm/mmc isBufferOverflowed)"
DBG(2011-11-01T20:46:40.645616, 4148, t:00254760:) > "  --> false"

We do some more investigations to speed up the communication.

Peter

On 31.10.2011 12:35, Peter Haub wrote:

Hi,

we are still looking for an improvement of our systems performance.

We logged some scans and found that in the time stamps the smallest
difference is 15.6 msec.

It seems as if this time resolutions is based on the std::clock function
on Windows in our case, which is typically 16 msec.

(The
        boost::posix_time::ptime bt =
boost::posix_time::microsec_clock::local_time();
uses the  std::clock  .  )

The question is if we can increase the time stamp resolution by some
configuration?
Is the fairly large resolution effecting other functions in
MicroManager, e.g. time delays.

Regards,
Peter

------------------------------------------------------------------------------
Get your Android app more play: Bring it to the BlackBerry PlayBook 
in minutes. BlackBerry App World™ now supports Android™ Apps 
for the BlackBerry® PlayBook™. Discover just how easy and simple 
it is! http://p.sf.net/sfu/android-dev2dev

_______________________________________________
micro-manager-general mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/micro-manager-general

------------------------------------------------------------------------------
RSA&#174; Conference 2012
Save $700 by Nov 18
Register now&#33;
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
micro-manager-general mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/micro-manager-general