# Amplitude of derivatives approximated by continuous wavelet transform

I do have a question regarding the meaning of the amplitude of an approximated derivative derived from a continuous wavelet transform. From what is known, applying the CWT (Haar wavelet) to a signal results in an approximation of the first derivative. Applying it several times result in higher order derivatives.

Since the amplitude of the resulting derivative is altered by the scale factor, the real amplitude of the derivative is not known.

My goal is to calculate the highest value of a noisy signal's first derivative. However, the noise is hurting a lot. I found the Savitzky Golay algorithm that should work out for me.

However, I was wondering if there is any way to get the real amplitude of a derivative derived from a CWT.

Ideas are welcome? Thanks