Igor: It is probably the best when fitting is concerned, but you have to write the fit function by your own as a limited number of system built fitting equations are given.

peakfit - jandel scientific software - http://www.sigmaaldrich.com/catalog/product/sigma/z365815?lang=en®ion=RO

PeakFit , i think, is the best software for comprehensive peak fitting. Disadvantages of this program - you can't do batch or packet's analysis of several similar spectra. You should do it one by one.

you can also use The unscramble (CAMO).

i suggest you to use igor.....the best for handling spectroscopic data.....

Thanks to all,

Yup i have heard about Peakfit but ii mixed it with Systat program rather its a separate program.

Thanks in initial stage, i find Peakfit is doing better job as compared to Origin .

Abhigyan and Shahnawaz thanks for the suggestion, i will surely give a try to IGOR as well.

The ROOT data analysis system (root.cern.ch) has a TSpectrum class with an extensive peak find/fit capabilities (among other things). This would be a good choice, if you are familiar with programming in C/C++. Documentation on TSpectrum: http://root.cern.ch/root/htmldoc/TSpectrum.html (you can find also many tutorials on the website). Example of the peak finder program: http://root.cern.ch/root/html/tutorials/spectrum/peaks.C.html

BTW, ROOT is free and open source.

oops ! i am not much practiced on C/C++ so i guess Andrej it will rather take too much time to learn it first to use above mentioned software application

Hi Ashish,

I agree that origin is sufficient for peak fitting but its rather more time consuming .

What exactly is giving you problems in fitting the peaks? In my experience, fitting Gaussian profiles to arbitrary line shapes is problematic when the line profile is offset from zero. That is to say, a Gaussian profile requires that the value of your data be exactly zero far from the center of the peak, which often times is not the case. Either adjust your data accordingly or add an additional DC offset to the Gaussian model. It should not matter what program you choose to use.

Hello A. Apponi,

Previously i was having problems in fitting peaks which were very less intense and merged with highly intense peaks so i was not even able to precisely locate them in UV-Visible spectra. But now i am using peakfit (by sigmaplot) and its user interface allows me to satisfactorily fit those peaks .

Taking second derivatives of peaks provides consistent results. You have to select the parameters, and the same parameters should be used for comparing the different spectra. You may have to use two different methods if there are both narrow and broad bands in one spectrum. This avoids variable judgement when using peak fit programs. This resolves peaks and gives reproducible peak positions as well as peak "heights" in second derivative, the higher concentration gives more negative peaks) for quantitative analysis.

I have used PeakEasy and we also use ROOT. Peak Easy has a relatively good fit, but at this time only gives a simplified statistical error of 1 std dev and is not derived from the fit co-variance.

Can any one suggest me how to do power law spectral fitting to multiple data sets??