Joined: 05 Jan 2010 Posts: 12 Location: Houston, TX
Posted: Tue Jan 05, 2010 8:34 pm Post subject: Improving tone quality with software
Hi guys. I've lurked around the boards for a while but decided to let you know about a computer program I made to help people study and improve their tone quality (timbre).
Basically, the software shows onscreen all the unique overtone and harmonic frequencies generated when you play. You can instantly see how changes to your embouchure, airstream, pressure, etc impact the picture/model of your sound. It solves that challenge of being able to visualize your sound, and it also helps teachers explain what the students are actually playing.
Even better, when you play you have a unique pattern and combination. But when a professional plays, their pattern/combination is much different. You can use the software to learn to match the professional example, by modifying your techniques until your patterns match.
The students I've worked with using the software have learned to produce professional sounds within a few weeks of practicing with the software. Please take a look at the site... I have some videos detailing it more.
www.audi-graph.com _________________ www.audi-graph.com - visually see your sound, and learn to improve your sound with comparisons to professional examples
Good effort, makes sense. Harmonics spectrum of pro players' long tones and scales overlaid with the harmonics spectrum of your own long tones. Learn by adjusting the visual part.
Joined: 04 May 2005 Posts: 1139 Location: Essex, England
Posted: Wed Jan 06, 2010 5:21 am Post subject:
Very interesting and potentially very valuable!
However, it relies on you shaping your natural sound to that of the "pro" player that has been recorded and prepared. So, essentially, the student has his sound 'cloned' to that of the pro recording. This ultimately lowers, if not destroys, much of the value of this device.
Additionally, there is very likely to be a cultural deficit. Were the list of pro players to include Maurice Murphy or Richard Marshal then I would have already ordered!
Joined: 14 Jan 2005 Posts: 7770 Location: Des Moines, IA area
Posted: Wed Jan 06, 2010 1:06 pm Post subject:
A kid from China that was in my college studio a couple of semesters would bring his laptop to lessons and occasionally record me, then himself. I'm not certain what app he was using (PC, not Mac) but it showed a visual above and below a center or median which would fill both directions with the volume and tone. He always pointed out to me the broader spectrum of my sound on the same thing as compared with his. I'd almost forgotten about it. Maybe someone can post what he might have been using. I really liked that kid, it was just before China cut a lot of the 'net from their people. He would always laugh when someone would talk about what downloads cost here and reply: "Evely ting flee on the inellnet in China!" Not sure about now. I always hope he remained safe when he returned.
Joined: 04 May 2005 Posts: 1139 Location: Essex, England
Posted: Wed Jan 06, 2010 1:46 pm Post subject:
Craig,....
Could that software have been called something like 'Wavlab' or 'Wavelab'?
When I had my first PC at home about 12 years back it came with a lot of software ready installed and one of the progs was, I think, something like 'Wavelab lite'. It was remerkably like that which you have described but being very new to computers at the time i could never get into it although it looked tantalisingly useful to a musician.
Joined: 05 Jan 2010 Posts: 12 Location: Houston, TX
Posted: Wed Jan 06, 2010 2:12 pm Post subject:
Craig Swartz wrote:
it showed a visual above and below a center or median which would fill both directions with the volume and tone. He always pointed out to me the broader spectrum of my sound on the same thing as compared with his.
I don't know what he was using, but when I was recording videos, I was using audacity (on mac). There's a horizontal median reference line and when you play a long note you get a type of big fat solid line, if you zoom out far enough. By reading the data as he must have, you could interpret it as being a broadness etc in sound. Audacity is free, but the educational benefits are limited.
What audi-graph shows is the intensity and combination of the harmonics you're playing. It can show you exactly how much brighter or darker a mouthpiece or embouchure change has to your sound, and how it compares to a symphony professional's sound. _________________ www.audi-graph.com - visually see your sound, and learn to improve your sound with comparisons to professional examples
Additionally, there is very likely to be a cultural deficit. Were the list of pro players to include Maurice Murphy or Richard Marshal then I would have already ordered!
Yup - No disrespect for the pros that supplied the templates - but having samples of many different top players in many different genres of performance would surely help move this product.
Just think of the names that could be on the list - half of them are probably members here on TH
For me cornet wise - Russell Gray, Richard Marshall, Roger Webster, David Dawes, Guy Touvron, Danial Doyan, etc. OK - I'll even throw in Phillip McCann for good measure. Now "THATS" a spectrum I want to see - Get a lineup like that and I'll take two copies of the software _________________ Smith Watkins "Soloist" cornet with K2 and T4 leadpipes and Denis Wick RW4 mpc
Last edited by Flattergrub on Wed Jan 06, 2010 5:36 pm; edited 1 time in total
Joined: 21 Oct 2003 Posts: 704 Location: Pennsylvania
Posted: Wed Jan 06, 2010 5:31 pm Post subject:
i had this question too . i would like to hear/see the trumpet sampled before purchasing. unless it's there and i missed it.
Bob Stevenson wrote:
Very interesting and potentially very valuable!
However, it relies on you shaping your natural sound to that of the "pro" player that has been recorded and prepared. So, essentially, the student has his sound 'cloned' to that of the pro recording. This ultimately lowers, if not destroys, much of the value of this device.
Additionally, there is very likely to be a cultural deficit. Were the list of pro players to include Maurice Murphy or Richard Marshal then I would have already ordered!
Mmmmm.....couldn't you do the same thing with any software oscilloscope and a microphone? The point of difference for your product would be having a library of sound samples of great players to compare your own sound to. _________________ Cannonball 789RL
KMI Golden Flair
Conn Director Coprion Bell
Besson Sovereign Flugel
www.jazzinspired.blogspot.com
Posted: Wed Jan 06, 2010 7:10 pm Post subject: Improving tone quality with software
brassjunkie wrote:
Mmmmm.....couldn't you do the same thing with any software oscilloscope and a microphone?
No an oscilloscope will not do for this. This is a Spectrum analyzer showing the amplitude of the various harmonics present in the tone together with a template overlay of the tone one is trying to emulate.
I tried this 'bio feedback' technique using a spectrum analyzer and microphone but simply tried to increase the amplitude of the upper harmonics while listening to my tone and looking at the display. I can't say that it did much for me in the way of tone improvement but I believe that such a system with a clearer display of the 'picture' one is aiming at may have merit. I do agree with the other posters, that a larger 'pro' library would be desirable.
Joined: 14 Aug 2009 Posts: 2156 Location: Jackson NC 27845
Posted: Wed Jan 06, 2010 7:12 pm Post subject:
This seems to be part of a forensic audio comparator ... now all one needs is a link to an international data base of ALL the musical performances so one could be selective as to what and who they wanted to emulate / clone.
Yes, a freeze frame oscilloscope, coupled to an analyzer would produce this imagery, that output to a computer would provide a print-out.
Personnel in forensic audio labs today could possibly discern on what instrument brand a 488hz A and mouthpiece combination was played if such were in their data base. It's publicly known they match speech patterns now.
The downside of such that I perceive is that it could work it's way into the education system and become a standard upon which achievement is marked.
While it might be interesting as a learning device, I personally don't want to sound like anyone else ... I want to do my own interpretation ... just as arrangers modify existing works to create a different SOUND that is unique and their own.
Joined: 11 Oct 2005 Posts: 415 Location: Fort Smith, AR
Posted: Thu Jan 07, 2010 9:43 am Post subject:
Very interesting. For those of you considering this but wanting some more ideas, look at a couple of threads from a few years ago. A number of folks contributed some very good things (and then there was me muddling the waters). Anyhow, if you're interested, look at the following threads. BTW, on the second page of the "Visualizing Resonance" thread I pasted some e-mail correspondence from Joe Wolfe who is a PhD acoustician/trumpet player from Australia who felt that the idea of "biofeedback" from such a program would not be useful. I'll probably purchase the Audi-Graph program just to evaluate it.
Excellent for those who want to work on theyr tone and mimic somebody elses sound...
but i dont want to sound like anybody else and have my own unique sound, dont need the software to tell me when i resonate the best, everyone can do this by using theyr ears.
Joined: 03 May 2005 Posts: 8951 Location: Monument, CO
Posted: Thu Jan 07, 2010 10:14 am Post subject:
Disclaimer: I have not had time to review the website thoroughly, just a quick glance. It would be interesting to try the software with my measurement mic.
The idea is intriguing... I think it may be more useful for general guidance than actually trying to match a particular player's tone. E.g., add more highs or lows to your sound, see what happens; vice-versa, change your sound from bright to dark, and observe the spectra.
As for matching specifics, that'd be a challenge... There are a lot of variables both in the playing and the measurement, any one of which (let alone several in combination) could skew the results. For example, time-averaged FFTs don't show the modulation of tonal structure (e.g. amplitude and phase variations of the frequencies in the signal) that is critical to what we hear and perceive. On the measurement side, sampling rate, sample size (and corresponding bin width), weighting, etc. etc. etc. all play a part in what you see vs. what you get.
There are various programs that will generate a spectrum from a mic hooked up to your PC. You wouldn't get the role models, but it could be fun to play around.
Interesting idea - Don _________________ "After silence, that which best expresses the inexpressible, is music" - Aldous Huxley
Posted: Thu Jan 07, 2010 3:01 pm Post subject: Improving tone quality with software
It is probably unlikely that one would be able to emulate the sound of the role model.
However, just having the richer harmonics displayed should provide benefit as there is a visual representation of a richer audio spectrum (than one's own) to strive for.
Joined: 05 Jan 2010 Posts: 12 Location: Houston, TX
Posted: Fri Jan 08, 2010 11:21 am Post subject:
Wow! I didn't anticipate this level of discussion! There's been a great exchange of information! I'm here answer question or concern because I feel this tool can really help. I've found this has drastically turned around my own playing, and I'd love to help others improve their own sounds.
PiCK Kanstul wrote:
i had this question too . i would like to hear/see the trumpet sampled before purchasing. unless it's there and i missed it.
As requested, I uploaded a quick video comparison of two of the 4 trumpet players contained in Audi-Graph (all playing with a classical sound). Audi-Graph does not contain audio recordings of the musicians, only the graphical representation for you to match against. The best way to determine which is a preferred sound is to look into the bio of the musicians, or try to match each and see which you prefer.
I am currently working on a mac version. I have a mac myself, so this is of interest to me. It somewhat works with an emulator, but not straightforward enough for general release. I made it on PC because I'm more familiar with the development environment.
Flattergrub wrote:
having samples of many different top players in many different genres of performance would surely help move this product.
Can do! This started as part of a NASA research project, and I live in Houston, so I recorded a certain type of musician. Any suggestions you may have are greatly appreciated. I record musicians weekly, and more musicians are being included future releases.
Jim-Wilson wrote:
Joe Wolfe who is a PhD acoustician/trumpet player from Australia who felt that the idea of "biofeedback" from such a program would not be useful.
Wolfes website has a wealth of great information, and are a great resource for acoustic details. Audi-Graph is similar, but very different than the tools he's worked with in the past. It provides biofeedback in the same way a tuner helps teach/learn. But instead of addressing just pitch, it displays sound quality.
Jetrang wrote:
but i dont want to sound like anybody else and have my own unique sound
Excellent concern. Finding your unique sound is important as a musician. Audi-Graph is geared towards student musicians seeking to develop a good foundation for a core sound by providing a reference point and something to learn with, before you begin tweaking and customizing your sound.
Audi-Graph doesn't take into account things like vibrato, style, or rhythm, which also helps distinguish the individual musician. But I do have a high school freshman user now replicating a seasoned symphony sound after only a few weeks of using it. And her finding the proper embouchure to match in the lower octaves suddenly expanded her upper range almost immediately.
Don Herman rev2 wrote:
There are a lot of variables both in the playing and the measurement, any one of which could skew the results.
jimh46 wrote:
It is probably unlikely that one would be able to emulate the sound of the role model.
Microphone placement, instrument differences, and room acoustics make it unlikely to perfectly match the role model. But it is extremely precise and can help you develop a good core sound faster than traditional methods, by providing an objective, consistent, and accurate analysis of your sound.
There are a couple more videos online with me playing saxophone demonstrating functionality and the corresponding feedback when adjusting your embouchure.
http://www.audi-graph.com/resources.php _________________ www.audi-graph.com - visually see your sound, and learn to improve your sound with comparisons to professional examples
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
