Multipoint Background Count Times clarification

[BenjaminWade]

Hi John
I was under the assumption (which now turns out to be totally wrong I think), that when acquiring MPB, the Hi/Lo count times entered were divided by the number of background points acquired. For example, if I am counting 3 points to the Lo side, and I enter in 30s in the "Lo Off Peak Time" box, it would be 10s per background point.

However I have only just noticed in the count times box (and again...it may have been there for a long time) in red font it reads "MultiPoint background count times are based on the corresponding Hi and Lo Off Peak Times divided by two"

Does that mean then in the aforementioned example that by putting a value of 30s in the Lo box, each of the 3 points on the Lo side would be counting for 15 seconds? And if so is there a reason it is done this way? Maybe it has always been like this and I have embarrassingly never noticed...?

Cheers

[John Donovan]

I was under the assumption (which now turns out to be totally wrong I think), that when acquiring MPB, the Hi/Lo count times entered were divided by the number of background points acquired. For example, if I am counting 3 points to the Lo side, and I enter in 30s in the "Lo Off Peak Time" box, it would be 10s per background point.

However I have only just noticed in the count times box (and again...it may have been there for a long time) in red font it reads "MultiPoint background count times are based on the corresponding Hi and Lo Off Peak Times divided by two"

Does that mean then in the aforementioned example that by putting a value of 30s in the Lo box, each of the 3 points on the Lo side would be counting for 15 seconds? And if so is there a reason it is done this way? Maybe it has always been like this and I have embarrassingly never noticed...?

Hi Ben,
Yes, it's been this way for several years, in fact since the beginning I think, but I must say, that my method of dividing by two is completely arbitrary.  But no one has complained yet!   

I guess my thinking was along the lines of: since we are acquiring 8 off-peaks by default, and iterating to 4 off-peaks by default (rather than two), we should still make the on-peak count time equal to the total off-peak count time (statistically required for measuring a zero concentration), by dividing by 2.

But I am completely open to another line of thinking...
john

[BenjaminWade]

Ok thanks John. It sounds like its just me being stupid again. I swear I read it somewhere...
Anyway I am fine with how it is, now that I know how to enter it properly!

Cheers

[John Donovan]

Ok thanks John. It sounds like its just me being stupid again. I swear I read it somewhere...
Anyway I am fine with how it is, now that I know how to enter it properly!

If you do find it documented incorrectly somewhere, please let me know and I will fix it.
john

[BenjaminWade]

Hi John
Unless I am reading it wrong it is in this pdf halfway down page 3? To myself that reads as if it is dividing by the iterate number but I could be wrong.

http://probesoftware.com/download/manuals/Multipoint%20background%20acquisition.pdf

Cheers

[John Donovan]

Hi John
Unless I am reading it wrong it is in this pdf halfway down page 3? To myself that reads as if it is dividing by the iterate number but I could be wrong.

http://probesoftware.com/download/manuals/Multipoint%20background%20acquisition.pdf

You are correct sir!

I can't remember when this might have changed, but I will correct the document.
john

[BenjaminWade]

Ok great, I thought I was going nuts. I probably still am but thought I read it somewhere. No problems, I will start using the correct way now and let users know.

Cheers

[John Donovan]

Ok great, I thought I was going nuts. I probably still am but thought I read it somewhere. No problems, I will start using the correct way now and let users know.

Hi Ben,
I can't tell if you're just pulling my chain here or not...     

You did read it!  The document was written by Julien Allaz in 2011, but hasn't been updated since then. Until now that is.  Thanks for catching this!

The whole document really ought be edited to reflect all the other changes in the multi-point background method since then.  I'll mention it to Julien.   
john

[John Donovan]

Unless I am reading it wrong it is in this pdf halfway down page 3? To myself that reads as if it is dividing by the iterate number but I could be wrong.

http://probesoftware.com/download/manuals/Multipoint%20background%20acquisition.pdf

I updated the above document and also the PFE distribution.
john

[BenjaminWade]

Great, thanks John.
No its fine, now that I know how it should be entered its all good.

Cheers