Ioanna

Age: 27

Nov
6
awarded Notable Question
Oct
28
awarded Popular Question
Oct
18
awarded Popular Question
Aug
21
awarded Critic
Sep
19
awarded Revival
Sep
19
awarded Teacher
Sep
19
answered Remove lines ending with a specific format in Notepad++
Sep
13
awarded Revival
Apr
11
awarded Teacher
Apr
3
accepted What is the DTW result (applied to the MFCCs of an audio file) proportional to?
Apr
3
comment What is the DTW result (applied to the MFCCs of an audio file) proportional to?
Thank you. So the fact that for similar smaller files the distance from DTW is smaller than for similar large files is irrelevant for this labelling process...
Apr
3
comment What is the DTW result (applied to the MFCCs of an audio file) proportional to?
I was talking about "similar", not "smaller" files :). Anyway, in your words, my goal seems to be to "verify" the sounds, to assign "correct" or "incorrect" labels. I'm sorry, but after all this, it's still not clear what comes next after applying DTW...
Apr
3
comment What is the DTW result (applied to the MFCCs of an audio file) proportional to?
No, my goal was not different... "You can do that, it returns the distance number." - why then, in some cases, the distance is larger for two similar files than for two different ones?
Apr
3
comment What is the DTW result (applied to the MFCCs of an audio file) proportional to?
let us continue this discussion in chat
Apr
3
comment What is the DTW result (applied to the MFCCs of an audio file) proportional to?
I am simply trying to see if I could use the MFCCs and DTW to compare recordings of the same sound, it's just a simple experiment. I assumed it would be far less complicated than speech recognition. In example 1. I recorded "dad" two times. In example 3. I recorded "dad" and "Dan", but when playing "Dan" I started recording a little earlier, so the silence at the beginning of the file is a little longer (where from the difference in size between 1. and 3.). In example 2. I recorded "mother" two times. In example 4. I recorded "mother" and "father".
Apr
3
comment What is the DTW result (applied to the MFCCs of an audio file) proportional to?
Ok, I don't seem to be so good at explaining things so let me try again: I'm playing a small .wav file and at the same time I record it to another .wav file (using NAudio). Let's call the recorded sound S1. Then I play the same .wav file again and record it again => S2. Compare the two (DFT + MFCC + DTW) => Diff1. Play a .wav file that differs only by one letter from the first one (say "dad" vs. "Dan") and record it => S3. Compare S1 and S3 => Diff2. Compare Diff1 with Diff2 to find a threshold. Now please see again the examples in the question - that is my trouble.Thank you for your patience.
Apr
2
awarded Commentator
Apr
2
comment What is the DTW result (applied to the MFCCs of an audio file) proportional to?
I'm not writing at the same time the same template: what I'm doing more exactly is that I record a sound and use it as a reference file. Then I record exactly the same sound (same voice, approx. same background noise) again and compare the two. That is what I meant by "same sound" in the examples 1 and 2 from my question. Then I record a slightly different sound (but same voice, ~same background noise) and compare it to the same reference file - examples 3 and 4. What I meant in the comment was that examples 1 & 3 used the same reference file, and so did examples 2 & 4.
Apr
2
awarded Supporter
Apr
2
comment What is the DTW result (applied to the MFCCs of an audio file) proportional to?
My two pairs for which I compared the DTW were [one template + a similar file] and [same template + a slightly different file] for pairs 1&3, 2&4 in my example. So according to what you are writing, I try to match the same sample => I should be allowed to compare DTW. However, it is true that I use the same threshold. That leads me to another dependency question: what does the threshold depend on? You are writing that it must depend on every template... but that for two templates of the same length it should be different. Does it depend on the magnitudes computed by the FFT? On their sum?
1 2 3 4