Raman spectroscopy: data analysis: convolution

[liss]

hey guys,
i hope you can help.

my task is to analyse data of raman spectroscopy. therefor i have to deconvolute it. that means the data must have been convoluted somewhere.

is it true that the raw data which i receive is convoluted already? or is it common to convolute the data "active"?
i guess its a stupid question but i am not quite sure.

thanks in regard

 

[ardie]

its a very good question if you haven't come across convolution method before. i had the same question a few months ago on my medical imaging course. the idea is that the data you take from any detector, is imperfect and as a result, the raw data you get is a convolution of the Fourier transform of your detector and the data. so information about the Fourier transform of your detector will allow you to do the inverse transformation, and then the raw data is just multiplied by the inverse Fourier transform to give your final data. i am unforunately not very familiar with the raman spectrometer detectors, but I am sure they have basic experiments run on them to determine the information about their detectors

 

[liss]

thank you so much for your detailed answer, i got it now.

as far as i can remember we used the gaussian curve to deconvolute the data based on a hint of my prof. and it worked out.
only when i prepared my talk i have wondered where the convolution takes place.
so your answer suffices absolutly :)

 

[ardie]

woopy :P