Pulling amount from two quantitation steps

So, for whatever reason, you can't mix A and B into a single calibration standard, but can you not use a single processing method that looks for both?  I haven't had an issue where I need separate calibration standards for multiple components, but wouldn't Empower work just fine to calibrate for both components and in A, it just ignores B since it is missing and vice versa in creating their respective calibration curves/response factors?  As long as you can use a single processing method that looks for both components, I don't think this would be a problem.  That would then eliminate your issue of having multiple results on the sample, simplify your ratio calculation since it is all in one result, simplify the processing instructions for routine analysis, and eliminate the likely confusion/error that is bound to happen with analysts where you'd have multiple results calculating a custom field result.

I could otherwise draft up what I'd see a custom field looking like between injections, but I don't think it would be necessary, assuming I'm correct that you would be able to simply process using a processing method that looks for both components.

I didn't mean to imply you had to change the method, but maybe I just wasn't very clear on how it would work with a single PM which, unless your QC method has been written to that crazy level of detail as to specify processing method parameters such that you must use two for this analysis instead of one, really requires very little change on your part and you probably do 90% of what you'd need to do already.  Simply don't enter a standard amount for component B in the Std A column in the Amounts table in your sample set.  Empower will allow the processing method to integrate peak B in the Std A injection, but won't add it to the calibration without that amount value, so, when it eventually hits the Std B injection where there is an amount, it would then initiate the calibration curve for that component (and vice versa with no A in Std B's row in the Amounts table).  Like I said, you probably do exactly this already since you make separate Std injections for A and B, so it ultimately just means using a single processing method and simplifying your sample set processing instructions rather than effectively duplicating them between components A and B.  You could easily "test" my theory for yourself by simply analyzing the data in the review window by using a single processing method with both components and clicking through both standards and then viewing the calibration curve and sample results all without saving anything in your QC environment.

But, back to your original question:

If you ultimately want to continue using multiple processing methods and getting two results on each sample, then consider the following:

So for a ratio of A to B...I assume something like A/B*100 to convert to a % or something?  If you want to share any other details, then the following could be tweaked accordingly as I'm still making a few assumptions with what you have provided:

Peak field, Real data type, Calculated, Search Order=Result Set Only (I'd also set the Sample Type to Unknowns only to improve performance as you don't need this on standards as well as Peak type to Founds or Founds and Groups for the same reason).

SAME..(A[Amount])/SAME..(B[Amount])*100 where "A" and "B" are your peak names spelled exactly like in your processing method(s).

Effectively, using the "A[Amount]" approach will provide the calculated amount for component A in the formula.  Using that within the SAME..() nomenclature allows Empower to do an intersample calculation on a peak component value between injections with identical sample labels.  You could go further to specify SAME.SAME.SAME() if needed should you ever inject a sample multiple times from the same sample set row or if you use multiple channels/wavelengths for a single sample (first is sample label, second is injection, third is channel).

Assumptions:

-Since you say it is a QC method, A and B are always present in samples.  If not, then you may want to consider adding some conditional elements to the calculation to generate an alternate response...like "100%A" if B is "missing."  This would require a boolean test of some sort like "GT(A[Amount], 0)*....." in order to complete it.  The language in Empower is just not similar to something like Excel which most people are going to be more familiar with, but it is still something that can be done just like an If/Then statement.

-You must assign unique labels to each sample within the sample set to identify.

-You'll likely need to add a "Summarize Custom Fields" line to your Sample set "Function" column or else you'll get inconsistencies.

I haven't "tested" this, but based ...