[Simon]

For the main part of the assignment, you’ll be loading a configuration file called config.scm every time you start Festival. This loads a better-sounding voice.

You can add those commands to the end of config.scm and they will be run as you start Festival.

Until then, you could create a file called startup.scm, put those two commands in that file, then start Festival like this:

$ festival startup.scm

[Simon]

Try changing the method that Festival uses to play audio:

festival> (Parameter.set 'Audio_Method 'Audio_Command)                      
festival> (Parameter.set 'Audio_Command "play -t raw -r 16000 -b 16 -c 1 -e signed-integer $FILE")
festival> (SayText "hello world")

which tells Festival to save the audio to a file, then use play (i.e., sox) to play that file. So, if playing audio on the command line works, this should work.

[Simon]

I’ll start you off with this and you try to complete it – I’m going to use square brackets to index everything to emphasise that we are dealing with discrete (= digital) signals only, and to match the notation in the sp-m1-4-discrete-fourier-transform notebook.

x[n] is the waveform. n is indexing time. To plot x[n], we place n on the horizontal axis and label it “time (samples)” then plot the value of x[n] at each value of n. We have drawn the waveform.

X[k] is the Discrete Fourier Transform of x[n]. k is indexing BLANK. To plot X[k], we place BLANK on the horizontal axis and label it “BLANK” then plot the value of X[k] at each value of k. We have drawn the BLANK.

Can you fill in some of the BLANKs?

[Simon]

You can use the terminal feature on Edina, then follow these instructions as if you were updating on your own machine, but using the terminal on Edina (in a browser).

There are instructions on using the terminal in Edina in the sp-m0-how-to-start notebook (look for the heading “Alternative: Using the Noteable terminal”).

There are more tips on using Edina Noteable in their user guide.

[Simon]

First, try rebooting the Virtual Machine.

Some other possible ways to diagnose the problem:

1. can you reach websites (e.g., http://www.ed.ac.uk or google.com) using Firefox in the VM?

2. are you in University accommodation (in which case you don’t need the VPN – you are already inside the University network)?

3. use a browser on your personal computer, and also in the VM, and type “what is my ip” into the Google search box

[Simon]

The $ indicates the shell prompt – the command to run is the part after it.

[Simon]

Looks like you are trying to install the VPN inside the Virtual Machine. That’s not the way to do it. Install the VPN on your personal computer.

To use the VPN with the Virtual Machine (e.g., for copying the files needed for the first assignment), connect to the VPN on your personal computer, then start the Virtual Machine. The Virtual Machine uses your personal computers network connection, so will also be using the VPN.

[Simon]

Thanks for testing. Those errors should not matter. Did you see rsync copying a lot of files the first time you ran it?

The second time you run rsync it won’t copy anything that is already copied, so it runs really fast.

Please try it one more time. You will always get the “Could not chdir…” error – ignore that. The other error “…failed to open…” should now be solved.

[Simon]

That’s a philosophical question, to which I have no answer. But I do have an engineer’s answer:

We have a toolbox. Our tools are techniques like short-term analysis, the Fourier transform, IIR filters, etc.

In any toolbox, there are usually several tools that could do the job. You could hit a nail with a spanner, but you’d be a bad engineer and the nail probably would not go in straight!

Being a good engineer usually means selecting the best tool for the job at hand, and having the skill to use it correctly. Almost all the time we select a reliable, tried-and-tested tool. One that we (and many other engineers before us) have used before, and that we understand. It will have limitations, but we know what they are and can live with them.

Very rarely, none of our tools will do the job, and so we need to invent a new tool.

[Simon]

I’ve created a forum for reporting errors in the Jupyter notebooks.

Most readings have a forum already which is where to report errors or ask for clarification.

For anything else, just post on any forum and we’ll see your message.

[Simon]

Please try the machine scp1.ppls.ed.ac.uk and report back (instructions now updated).

[Simon]

Please try the machine scp1.ppls.ed.ac.uk and report back (instructions now updated).

[Simon]

No, that’s not correct. The last N/2 values are identical to the first N/2 – they are just copies of the same values (their ordering is mirrored around the Nyquist frequency).

There are only N/2 magnitudes. No more.

The N numbers (samples) in the time domain (waveform) have been transformed into N/2 magnitudes and N/2 phases in the frequency domain. So, in the frequency domain (= magnitude spectrum & phase spectrum) there are also exactly N numbers.

In other words, the transform to the frequency domain has preserved all of the information in the waveform. That means the inverse transform will perfectly reconstruct the waveform.

[Simon]

Yes, that’s all correct.

[Simon]

You obtained the notebooks from a git repository and that created a folder called uoe_speech_processing_course somewhere on your machine.

Navigate to that folder in the terminal before running jupyter notebook and then your browser will open in the right place.

Otherwise, just navigate in your browser to wherever you have uoe_speech_processing_course (but noting that Jupyter won’t allow you to navigate upwards from the folder it was started in – this might be your problem?).

[Author]

Posts