 Spectral data coming straight from the lab can have many artifacts. It can often be noisy, we can have baseline variation, scattering and so on. In short, there are many problems we need to solve before the analysis. This is why we need preprocessing. We will use deliver data from the datasets widget. Open datasets, type liver in the filter, select the data and press send data. First, let us observe the data in the spectra widget. Here, we see a subset of the spectra, colored by the predefined label. The data looks nice, but let's see how it can be better prepared for the analysis. Add the preprocess spectra widget to the canvas and connect it to file. Now, all the data is sent for preprocessing. In the top right corner, you see the original data. In the lower right corner, you see the effect of the preprocessing. You might wonder why we see so few curves here. This is for efficiency. You can increase the number of the displayed curves in the bottom left corner. But let us keep 3 for now. Let us add a preprocessor from the top left menu. Say baseline correction. You immediately see the effect on the right. Now the spectra in the bottom are better aligned. We will add a couple more preprocessors to create a simple preprocessing pipeline. Say normalize spectra. We will use area normalization with the closest value. We can drag the red line to select the X value at which all the spectra will be normalized. Look at the effect in the bottom plot. To observe the effects of just one preprocessing step, I can click on the triangle next to the preprocessor and see how the plot changes. Finally, we will use the Savitsky-Golei filter to calculate second derivatives. This is a commonly used procedure to accurately find peak positions or remove low-order data variation, such as baselines. The result looks completely different from the input data. You can also reorder preprocessors to explore and tune their effect on the final result. Be careful, as the order can be quite important and can affect the final result of the analysis. To see the final output, you can click the final preview button. This shows the original data at the top and the final data with all the preprocessing steps applied at the bottom. Of course, there are multiple options in each preprocessor and we encourage you to explore their effect as well as try other methods. You can also read help for more detailed information and examples. At the end, don't forget to press the commit button to apply the preprocessing workflow to all the spectra. Now, we can observe our preprocessed data in the spectra widget and use it for further analysis. Today we have learned how to apply the preprocessors to the data and how to visually explore their effect in the preprocessed spectra widget. Now, you are ready for the analysis. In the next video, we will show you how to do the PCA transformation for spectroscopy.