For Masshunter quantitation software at what point does the signal to noise go from a real value to infinity?

We are validating a method and are having an issue with our masshunter data.  Some of the signal to noise values come back at infinity which is not a value we can use for our validation.  Is there a specific number it targets that if it exceeds will throw infinity back?  Thank you very much.

  • This is an excellent question. I will find out and get back to this forum with an answer.

  • Hi David,

    After speaking with R&D - the signal to noise ratio will report "infinity" if the noise value is calculated as 'zero'. You should be able to confirm this by exposing the noise column in Quant in addition to the signal to noise column. Be sure to expose many digits (perhaps 16) of the noise column (Tools -> Number Formats) to make sure there's no rounding going on. Any non-zero noise value should not evaluate to infinity.


    Hope this helps.



    Matt Leyden

  • Hi @all,

    I've got the same problem. I ran a SIM method with several ion groups. When defining the noise ranges, I made sure to only include ranges that are inside their own ion group. There is definitely some noise, as I can see when opening the file(s) in the Enhanced Data Analysis. Despite this, two compounds seem to have no noise at all (at least, for MassHunter Quant), therefore giving a S/N of infinity. How can I fix this?

  • Hello  ,

    Are the noise ranges you specified for a target within the target's data extraction window? This is defined in the Retention Time Setup portion of your method. MassHunter will allow you to specify a noise range that is not included in a target's extraction window. It will find no noise since signal for that time range will not exist for the target and it will calculate infinite signal to noise.

    Here is an example where I show some of the columns used for noise calculation. For a compound with a +/- 1 delta RT window at 7.6 minutes I specified a noise range of 5.0-5.8 minutes. No error is given but the noise is calculated to be 0, so S/N is infinity.

    If I move the noise range to 6.8-7.4 minutes, then noise can be found and S/N is calculated as expected.

    If this isn't what you are observing, try adding the Noise and Noise Region columns to your batch table to get a better idea of what is happening and let us know what you find. 

  • Dear howard_sanford,

    thank you for your answer. You found the solution! I had defined the noise regions within the correct SIM group, but not inside the extraction window. Now I know what to do. Thank you very much!

Was this helpful?