spatialaudio / digital-signal-processing-lecture Goto Github PK

In the paragraph "Auto-correlation function" I assume f(x[k1],y[k2]) has to be f(x[k1],x[k2]) in the first sentence.

In the second code example maybe a further explanation for the estimate of the auto-correlation function is useful, i needed very long to figure it out. maybe something like:

acf = np.zeros((L, L))
for n in range(L):
    for m in range(L):
        acf[n, m] = 1/N * np.sum(x[:, n]*x[:, m], axis=0)
        # x[0, n]*x[0, m] is the product of the amplitudes at two different time-steps in the same sample function x0
        # np.sum sums up all the products of the above time-step-combination of every sample function (x0 to x63)

Introducing the quantization step als denoted in equation (3) into the variance of the errorfunction (2) does not lead to the expression in (4) for me. Is there an additional approximation used? Another explanation or calculation step would be helpful.

Using binder seems to be broken - or do I miss something?

This is the binder error: The legacy buildpack was removed in January 2020. Please see https://repo2docker.readthedocs.io/en/latest/configuration/index.html for alternative ways for configuring your repository

not quite sure about it: at the end of the paragraph of the properties of the ACF doesn't the variance and the squared linear mean have to be multiplied with the Diraq?

In the second exercise i think the period time is 0,01 s and not 0,1 s.

in the first code box the second comment should be #plot PDF

In the example about Amplifier Noise it would be easier to use simply
mean_np = np.mean(noise) variance_np = np.var(noise)
or to make a comment that mean, sigma = stats.norm.fit(noise) is actually doing only that in case of a normal distribution but can be used for other distributions as well. I spent quite some time to understand what stat.norm.fit does and I am still not very sure about it.

in the solution of the example "Power Spectral Density of a Speech Signal" i would suggest that the "vocal folds" are meant and not the "vowel folds" - at least i couldn't figure out what that would be.

I suppose that the term "Quadratic Mean" is not the right term for the second raw moment. If I am right then you mean the Mean Square. In this case this term has to be changed in the reference card as well.

For a clear differentiation you may add a note, that the square root of a mean square is known as the root mean square (RMS or rms) that is also known as 'quadratic mean'.

Hey folks,

I noticed you tried out treebeard the other day but decided it wasn't for you. Also that you are the developers of nbsphinx!

I'd love to hear any misconceptions/frustrations you had about treebeard, as I am currently prototyping a lower-level pytest plugin for parallel testing + building jupyterbooks.

Any feedback is really valuable, thanks.

Im Übergang von Gleichung (1) zu Gleichung (2) fehlt der Hinweis, dass Ergodizität vorausgesetzt wird.