Baseline Correction

Baseline Corrections can flatten the baseline on the Fourier Transformed data in both dimensions of 2D experiments. Typically they are more useful in F2 than F1, but they can be applied in both dimensions..


The Baseline Correction menu is located under the Processing tab then Baseline Correction.

Is Baseline Correction necessary?

To determine the need for baseline correction, increase the vertical scale until you can see the noise of the spectrum. Then analyze the noise to determine what is causing it.

1) If the noise is spread across the spectrum randomly, there is no need for baseline correction. This is random noise and can be changed with more scans in acquisition, or different 2D Apodization in processing.

2) If there is noise on both sides of peaks, and the color is the same so red or blue on both sides of the resonance, then this is potentially a baseline correction problem in the dimension of the noise- if the noise goes up and down the screen, this is the y-axis normally F1, if it is across, this is the x-axis and normally F2.

3) If there is noise on both sides of peaks, and the color is different- so red noise on one side, blue on the other, this is a problem with phasing, see:
2D Phase Correction

Normally, correct the phasing in both dimensions before the baseline correction.

4) If there is alternating blue and red noise on the same side of the resonance, this is either t1 noise if it is going across the F1 dimension or truncation of the FID potentially in either dimension; the truncation should be fixed with 2D Apodization . Truncation noise will decrease further away from the peak, while t1 noise will be approximately the same across the entire dimension.


Applying Baseline Correction

1) To accomplish 2D Baseline Correction, first choose the dimension to correct. This is under Processing Menu, then Set Processing Dimension. First choose F2.

2) Next, open the Baseline Correction tab.

3) As in 1D Baseline Correction Whittaker-Smoother is generally the best option and the default values are likely fine.

4) To best see the effects of the Baseline Correction, increase ther vertical scale to analyze the noise as described above.

5) Repeat the steps above for F1, by choosing F1 under Set Processing Dimension if necessary.