Processing & reconstruction
Automated phase correction
Determine zero- and first-order phase automatically, or set it by hand with live feedback.
Phasing is quick when a spectrum is clean and genuinely difficult when it is crowded, and a first-order error that looks acceptable across the middle of a spectrum can invert baseline behaviour at the edges, where your weakest peaks are.
We offer both routes. A convolutional network proposes P0 and P1 across the whole spectrum, and the interactive viewer lets you take over and adjust them against live contours. Whichever you use, the applied values are stored with the spectrum.
What it needs
- Fourier-transformed spectrum
What you get
- Phased spectrum with the applied P0/P1 recorded
What does the work
DEEP Picker
Open source, redistributed under its licenceGPL-3.0Convolutional phase estimation
Related in Processing & reconstruction
- Spectral processingTake raw time-domain data from the spectrometer to a phased, baseline-corrected, referenced spectrum in the browser.
- Non-uniform sampling reconstructionReconstruct a full spectrum from a sparsely sampled indirect dimension, recovering resolution without the acquisition time.
- Deep-learning spectral enhancementNeural reconstruction that decouples, sharpens and denoises spectra that classical processing cannot rescue.