Jump to content

Spectrum Analysis - Steven

Rate this topic


six20aus

Recommended Posts

Hi Steven,

Directing this question specifically to you though others are free to chime in.

I read your 'analysis' posts with a lot of interest (as I'm sure do others) and it makes me wonder how I might use spectrum analysis in my own training ?

Is it of any use in self training or do people see it as more of a research tool ?

Link to comment
Share on other sites

Hi Steven,

Directing this question specifically to you though others are free to chime in.

I read your 'analysis' posts with a lot of interest (as I'm sure do others) and it makes me wonder how I might use spectrum analysis in my own training ?

Is it of any use in self training or do people see it as more of a research tool ?

six20aus: Hi!

I use it myself for both. A number of voice teachers at the collegiate level use a professional version of it in their studios and for their own practice.

Though I use it a fair amount for research and analysis, and for visualizing things that I could only previously hear, I find myself leaving it on for my own practice sessions to help me keep my resonances fine-tuned. I've used it this way for a few years now.

There are some other folks on the list that I have gotten started with it. Perhaps they will chime in.

Link to comment
Share on other sites

Thanks Steven,

And would you be kind enough to share which software you are using ?

I had a few lessons with an Estill teacher earlier this year and he was using Estil Voiceprint which looked fairly informative. I would like to follow up on it and see if there is some way to incorporate this in my training to identify areas I can improve.

Link to comment
Share on other sites

I started using the software about 2 months ago, so I am still learning how to interpret what I am seeing. But I use it the same way, working on note and vowel resonance. So far I’ve gone thru my basic vowels up and down the heart of my usable scale. The first time I practiced using it; later when I sang my songs, I noticed an immediate change of how my voice seemed to automatically adjust and seek out more ease and resonance. That pretty much sold me. I use it about 10-15 minutes a day most days during my workout and I’m currently using it to do messa di voce exercises. It just gives you a great visual representation of your dynamics and how your various harmonics are lining up with the amplifying formants. It has made me focus a lot more to the details of what is taking place at the fold level. As you make slight adjustments, you can visually see its impact.

I would highly recommend anyone to at least experiment with it. Being able to develop all the different shades of vowel sounds and dipthongs throughout your range and at various volumes is such a huge part of developing vocal freedom. This tool gives you a unique perspective to develop that.

Link to comment
Share on other sites

Thanks Steven,

And would you be kind enough to share which software you are using ?

I had a few lessons with an Estill teacher earlier this year and he was using Estil Voiceprint which looked fairly informative. I would like to follow up on it and see if there is some way to incorporate this in my training to identify areas I can improve.

six20aus: I am using a freeware program called 'spectrogram 16'. Its by Richard Horne, the same guy that wrote the Estill Voiceprint.

If you are interested to try it out, I will send you a zip copy with some installation instructions. Just send me an email (link on the left) and I will send as an attachment. It works under Windows.

If you have questions on how to use it, I can even set up a skype session to run you through it.

Link to comment
Share on other sites

six20aus: I am using a feeware program called 'spectrogram 16'. Its by Richard Horne, the same guy that wrote the Estill Voiceprint.

If you are interested to try it out, I will send you a zip copy with some installation instructions. Just send me an email (link on the left) and I will send as an attachment. It works under Windows.

If you have questions on how to use it, I can even set up a skype session to run you through it.

steve has set me up with this program. it will definitely help. when you are singing freer (not easier, but freer) you see it on the graph.

Link to comment
Share on other sites

six20aus: I am using a freeware program called 'spectrogram 16'. Its by Richard Horne, the same guy that wrote the Estill Voiceprint.

If you are interested to try it out, I will send you a zip copy with some installation instructions. Just send me an email (link on the left) and I will send as an attachment. It works under Windows.

If you have questions on how to use it, I can even set up a skype session to run you through it.

Hi Steven, I am new on the forum, could you please do a quick explanation of what this software is for? I think it would be very helpful from what I've read on this post!

Link to comment
Share on other sites

Hi Steven, I am new on the forum, could you please do a quick explanation of what this software is for? I think it would be very helpful from what I've read on this post!

blackstar: Below is a link to an article on the main site that should help in understanding what the software is for, and introduces some of the concepts involved. I suggest you read through it after you finish what I have written here.

In the simplest way of describing it, a spectragraph is a picture which represents the sounds present in the air. Sounds have frequency, expressed in cycles per second, and intensity, expressed in decibels (dB). The spectragraph shows you each sound's frequency and intensity.

Sounds which seem musical to us have some particular characteristics. In musical sounds, frequencies have a mathematical relationship, what we call the 'harmonic series'. Our minds interpret the harmonic series into 'pitch', 'loudness' and 'tone quality' based on the patterns and intensity of the sounds. In a human voice, some of the the tone quality information is also experienced by us as 'vowel sounds'.

Using the spectragraph, we can see the intensity of the harmonics which are involved in the vowel sounds, and all the others, too. By being able to see a representation of them, while we are also hearing them, we create a firmer understanding of the relationships between the sounds.

I hope this helps. Here is the link:

http://www.themodernvocalist.com/profiles/blogs/introduction-to-resonance-and

Link to comment
Share on other sites

  • 2 weeks later...

Thanks for your reply Steven, I'm going to try to find out more about this because it sounds very interesting! :D

blackstar: Cool. Just send me a private e-mail, and I will send you back the stuff in an attachment, and walk you through a simple orientation. Its what I have done for all the others that have had an interest.

Link to comment
Share on other sites

blackstar: Cool. Just send me a private e-mail, and I will send you back the stuff in an attachment, and walk you through a simple orientation. Its what I have done for all the others that have had an interest.

Thanks for the offer Steven, I just sent you an email! :)

Link to comment
Share on other sites

Hi Steven,

Any chance you could take a look at these spectragraphs? It's me (the green line) and another singer (the blue one) singing a note at same pitch (E5) and vowel (AH). His sounds way cooler so I wonder if you can give me some clues based on the analysis on how to come closer to his sound.

The blue one may have some audience sounds in the background. Take care!

Edit:

Sorry to intrude in this thread, I didn't want to put up a new one just to give an example (although maybe I should have:/)

Link to comment
Share on other sites

Any chance you could take a look at these spectragraphs? It's me (the green line) and another singer (the blue one) singing a note at same pitch (E5) and vowel (AH). His sounds way cooler so I wonder if you can give me some clues based on the analysis on how to come closer to his sound.

DJDeth: There is so much noise in the blue-traced signal that its impossible to see anything distinguishable as harmonics. The sort of analysis you are looking for is much better done with two audio samples accompanying the spectrogram. Can you post those?

Link to comment
Share on other sites

Thank you for the reply Steven. I tried to get a better signal this time. The green one is mine - the blue one is the one I'm aiming for - Tim Ripper Owens:).

Here's the sound clip:

http://www.box.net/shared/6s7opamqe2

The first voice is mine of course:) I clicked the 'capture' button after I had added a little bit of distortion (in the 2nd second of the recording). You can hear that there is a great difference between the tone of our voices, while the signals look quite similar (at least to me)

THe only thing I could say basing on my amateur analysis is that Tim Ripper Owens adds more distorion - It is obvious, but I think you can say that by looking at the 'valleys' between the peeks. I've noticed they become shallow when rasp is added. Am I on the right track? What about the 1st, 3rd, 6th, 7th and 8th harmonics? Why would mine be stronger? Can you see any other important nuances?

Link to comment
Share on other sites

DJDeth:

Just so I could be sure, I split the two examples, and overlayed them in time, you on the left channel, and him on the right. I used direct feeds from my audio 'out' to my 'in', so as to take my own speakers and room acoustics out of consideration. Also, to take the effects of vibrato out of the equation, I set the averaging value to 400 ms, which is 2 tenths of a second. That makes the display change much more slowly during the courses of the notes, and helps analysis of this type.

The contrast between the noise floor in your recording and the other is noticable, but IMO not important. Its likely that the other recording is from a noisy environment, and is being amped with compression.

Far more interesting is the comparative intensity of the individual harmonics. These two recordings have been balanced (by you?) so that the fundamentals are the same volume. Your 2nd, 4th and 6th are about the same, or a bit better than the other recording. The only harmonic that is softer when averaged is the 5th harmonic.

Your vibrato's are about the same width, and about the same speed.

To my ear, there are two differences, very minor, when comparing your tone quality:

1) the vowel you are singing is shaded toward /A/ (as in 'hat'), and his is shaded toward /a/ (as in 'father'). This causes a very slight difference in the formant alignment with the 2nd harmonic. You cannot see this in your spectrogram, but you can see it in the one I captured, below. Your note is traced in white, and his in blue.

Notice in the 2nd harmonic trace of your voice that there is a strong peak on the 'right' hand side, while his has a strong peak on the left side. That means, in the course of the vibrato cycle, that you have tuned the formant (via your vowel choice) so that it aligns when you are getting more resonance on the 'high' side of the vibrato cycle, and he has tuned his vowel so that it aligns best when it is on the 'low' side. Neither one of you is singing it spot-on, but the difference is so small when compared to overall vocal power that it takes very good ears to hear it. The difference to the audience is minimal.

Harmonic 5 is interesting, as your voice has a measurably softer (10dB) volume on it. This could be from a different recording environment, something as simple as some EQ boost in his, or some curtains in your room attenuating that frequency band.

To my ear, the biggest difference in these recordings which affects the perception of vocal quality, is the reverb. Both have it, but his has the sense of being in a 'larger' room.

Sorry to disagree with you, but I do not believe your voices sound as different as you seem to think. To my ear, they sound very much alike on this note, with the exceptions of the things I have mentioned already. I doubt that if you were recorded with the same PA as he was that any audience member would be able to hear the difference.

That is why the spectrograhic traces look so similar... because your voices are!

I hope this helps.

Link to comment
Share on other sites

Steven, thank you so much for sharing your incredible knowledge! One more question: which resonators are responsible for amplifying these frequencies that my fifth peek aligns with? Is it possible to tell at all?

DjDeth: Its probably in need of just a smidge more twang :) In addition, you might play around with some other vowels, for example, /U/ (uh, as in 'cup') is worthy of some experimentation, and even the schwa /oe/ as in 'foot'.

Which vowel shade will be most resonant in your voice depends somewhat on your native tone quality. To find the particular one that works best, you do some experimentation down an octave, singing what will be the E5, but down an octave at E4. When you do that, the H2 (harmonic 2) frequency of the E5 is now occupied by H4 of the E4, so it will align identically with the lower formant that is responsible for your prominent H2.

Thinking about the harmonic series for a moment, the frequency range of your E5's 5th harmonic (which is 2 octaves and a major 3rd above the fundamental when singing the E5,) is now occupied by several harmonics in a cluster centered around H10 of E4. If you sing the E4 without vibrato, you can see the individual harmonics very clearly.

As to the actual experimentation, here are the principles: 1) find all the vowels in your voice which emphasize H4 of the E4. There will be several. Start with /A/ (as in hat) /a/ (as in father), and the bright and dark versions of /a/ too, and the /oe/ (foot) and /U/ (hut). If a vowel is close, you can tune the resonance by raising or lowering your lower jaw slightly. For each vowel that gives you good resonance on the H4, write down the vowel shade, and a word that reminds you of it. Perhaps 1/2 hour or so of playing around like this, and you should have your short list.

2) Now, to the upper resonances. From your short list, try each of them, and see which ones emphasize the various harmonics in the mid and upper ranges, particularly the H10. Add and remove twang until you get the sound the way you want it, where H10 is as prominent as you can make it. Write that particular vowel from your short list down.

To test your result, sing that vowel on the E4, center it, and then leap up to the E5 with the same vowel. Your resonance peaks from the E4 should align well with your E5's harmonics.

I hope this helps.

Link to comment
Share on other sites

Amazing, Steven. I never thought about where vibrato centers. I had always assumed that it centered on the pitch and varied above and below. But now that you mention it, I think I possibly line up on the high side, where the variation of pitch descends and then back up to the actual pitch, especially on some songs.

Link to comment
Share on other sites

Steven, thank you. I'll start the vowel experiments as soon as I can (it's half past midnight here in Poland:) It seems like You know pretty much everything when it comes to voice science!

DJDeth: Let me know how else I can help. While I appreciate your kind words, I don't know all there is to know about the voice, but I am happy to write about what I have researched. When it comes to twang and some of the distortion techniques, there are others better on this board better equipped to make suggestions.

I should have said it in an earlier post, but that 10dB difference in harmonic 5 of the note is audible as 'ring', and interpreted subjectively by many as 'vocal quality'. It happens to align very well with the most sensitive part of human hearing, so what you do there with your own singing will make a difference.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...