TH Sonic Signatures (Information)

Derek Reaban · Posted: Fri Aug 19, 2016 8:24 am Post subject:

Mike, Richard, and mm55,

Thank you all for your comments. While I know very little about recording, I think this discussion will allow me to consider some stumbling blocks that can be resolved with a list of best practices / lessons learned. The hint about changing the Project Rate from 44,100 Hz to 32,000 Hz is a technical detail that would have allowed me to compare the built-in mic against the better microphones in the Zoom H2. I will be interested in seeing how that affects the Plot Spectrum, but my general impression is that my sound is my sound regardless of the quality of the microphone. Looking forward to being able to record myself on the MacBook Pro with the good mics, though (thanks Richard)! And the suggestion about having a tuner available before recording is also valuable. However, dialing in the best sound (letting the horn find where it wants to ring best), would be better if “tuning” was done before the actual recording, and the player was not watching the needle while recording (I want players to realize that getting the best sound is the goal of this project), and Audacity is a way to visual see that best sound. Floating down a few cents may be required to get that great sound. Once a player discovers that best sound, then “fine tuning” to get the slide in alignment with A440 AND keeping that great sound would be the goal.

I’m excited to share the rest of this project and get feedback. I’m learning along with everyone and think this can be an easy tool to use for anyone that has a computer. I want to make it as simple as possible, so once some of these “growing pains” have been addressed, this technical discussion regarding clipping, pitch offsets, comb filtering (that few players are interested in) will change to “How can I use this to get more vibrancy in my sound?”. Our language will change at that point to things that trumpet players understand, and then they can see how specific exercises lead to a tangible change regarding the resonance in their sound.

I will definitely keep sharing until I get all of my ideas out there!

Thanks very much.
_________________
Derek Reaban
Tempe, Arizona
Tempe Winds / Symphony of the Southwest

mm55 · Heavyweight Member Joined: 01 Jul 2013 Posts: 1412

I wasn't suggesting that you should tune using the tuner. I meant that a tuner would give a more accurate frequency than clicking on the frequency plot, which is very inaccurate.
_________________
'75 Bach Strad 180ML/37
'79 King Silver Flair
'07 Flip Oakes Wild Thing
'42 Selmer US
'90 Yamaha YTR6450S(C)
'12 Eastman ETR-540S (D/Eb)
'10 Carol CPT-300LR pkt
'89 Yamaha YCR2330S crnt
'13 CarolBrass CFL-6200-GSS-BG flg

RichardCampbell · Posted: Sat Aug 20, 2016 5:10 am Post subject:

How did you limit your Audacity plot to 11 kHz? Mine went up to 15 kHz.

Richard Campbell

Derek Reaban · Posted: Sat Aug 20, 2016 5:52 am Post subject:

Hi Richard,

First of all, thanks for participating! I'm glad that my information post full of instructions was understandable.

When I take a screen shot on my iMac, I make sure that the window containing the plot is rectangular (basically the shape of my iPhone on its side), and I mark exactly what I want to display. Since there wasn't much activity above 11,000 Hz, that seemed like a good cutoff point. I really like the fact that you included the peak frequency content from Audacity.

Looking at the plot, I think the best place to "crop" the window would be 9,500 Hz. This way we can see all of the Audacity drop downs and eliminate the white space at the higher frequencies.

Thanks Richard.
_________________
Derek Reaban
Tempe, Arizona
Tempe Winds / Symphony of the Southwest

Honkie · Veteran Member Joined: 20 Apr 2013 Posts: 245

Derek, this is a really interesting idea, thanks for your work on it.

I worked as an audio engineer for some years, and I think you may be underestimating the degree to which microphones effect the frequency response of the recorded signal. Do you know how Trumpet Herald has thousands of threads discussing the different sounds/responses of various horns? There are several audio recording forums containing hundreds of thousands of threads about the sound/responses of various microphones. (I'm mentioning this, just to suggest that the way the microphone exaggerates or downplays aspects of the sound is a a factor that would need to be controlled for, in a scientific setting.)

Derek Reaban · Posted: Mon Aug 22, 2016 10:57 am Post subject:

mm55 · Heavyweight Member Joined: 01 Jul 2013 Posts: 1412

It's not an issue of sensitivity, that's a completely different quantity.

And it's not something that can be reduced to a simple number indicating microphone quality. Flatness of frequency response in the frequency band of interest, however, is at the heart of the issue.

If you're trying to detect differences between harmonic levels on the order of +/-3 dB, and the microphone varies by +/-6 dB in the passband, then the results you get may not mean much. For this kind of analysis, a low-noise microphone may not make much difference, although that may be a quality that's important for recording, for instance. And directivity and sensitivity may not be critical. But the frequency response of a built-in microphone in a laptop could easily vary by +/-6 dB. And in the "family room" setup in your photo, the room effects could easily yield errors of +/-12 dB or more, under the cabinets, over the counter, and so close to the wall.

The frequency response of the microphone, in the room where you're recording, should be flatter than the variations in harmonic level that you're trying to measure, and I'm not sure that you've achieved that. Just as you couldn't really measure valve alignment errors of tenths of a millimeter with a carpenter's tape measure, I'm not sure you can reliably detect the relative levels of harmonics in a trumpet's tone with your setup.
_________________
'75 Bach Strad 180ML/37
'79 King Silver Flair
'07 Flip Oakes Wild Thing
'42 Selmer US
'90 Yamaha YTR6450S(C)
'12 Eastman ETR-540S (D/Eb)
'10 Carol CPT-300LR pkt
'89 Yamaha YCR2330S crnt
'13 CarolBrass CFL-6200-GSS-BG flg

Don Herman rev2 · Posted: Mon Aug 22, 2016 6:29 pm Post subject:

Derek, you can buy a calibrated UMIK-1 from Cross Spectrum Labs (http://cross-spectrum.com/measurement/calibrated_umik.html) for about $100 that would well in your setup. I have been using one for audio measurements (with REW) in place of my Earthworks M30 and it is comparable in frequency response (though not in noise floor or max SPL, but those are not a big issue for this project).

FYI - Don
_________________
"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley

Derek Reaban · Posted: Tue Aug 23, 2016 8:48 am Post subject:

mm55,

Don Herman rev2 · Posted: Tue Aug 23, 2016 10:38 am Post subject:

Unfortunately I am working 70-80+ hours/week and have been for a while, and just found out we need to step up the pace. Meaning I'll be working this weekend. Blah. I post whilst tests are running.

I have not installed Audacity since changing notebooks but I may be able to do something in the next week or two. My media room is very heavily damped and would work well for this, and I could record using my Zoom H2 and my Earthworks M30 mic for comparison. The M30 is a true measurement mic so perfect for this exercise. I don't have a good preamp for it now (sold, one of those "taking up space and I'll never do pro recording again" things) but do have a cheap m-audio preamp that should be good enough for this.

Hauling it off to David's would be a bigger challenge though an excellent excuse to see him again as it has been a few years.

The trumpet's range is not all that difficult to capture. My guess is the room is the biggest player. If you could play into a closet filled with clothes or similarly heavily-damped space with the mic fairly close you may alleviate some of the concerns.
_________________
"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley

mm55 · Heavyweight Member Joined: 01 Jul 2013 Posts: 1412

Don Herman rev2 · Posted: Tue Aug 23, 2016 11:00 am Post subject:

I already own an M30, and a preamp with phantom power, and a Zoom H2, so for a relative comparison I have what we need. Except for time...

I sold my "good" preamp a few years ago, alas, but the little one I have should be good enough for this, and my room is heavily treated plus I have an extra floating panel I can use behind the mic so can get close to anechoic response.
_________________
"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley

mm55 · Heavyweight Member Joined: 01 Jul 2013 Posts: 1412

Yeah, I was thinking more of what equipment would be feasible for many other people to use for home recording of their own trumpet sound. Earthworks might not be the optimal choice for many people.
_________________
'75 Bach Strad 180ML/37
'79 King Silver Flair
'07 Flip Oakes Wild Thing
'42 Selmer US
'90 Yamaha YTR6450S(C)
'12 Eastman ETR-540S (D/Eb)
'10 Carol CPT-300LR pkt
'89 Yamaha YCR2330S crnt
'13 CarolBrass CFL-6200-GSS-BG flg

Don Herman rev2 · Posted: Tue Aug 23, 2016 6:25 pm Post subject:

Agreed, probably not many want to invest in a $1k mic and $2k preamp... My guess is the Zoom's mics are fine for this (they include a compensatory network to provide at least decent response). Proper recording techniques and reasonable room is more likely an issue. That's why I suggested playing into a full closet, to help take the room out of the picture (sound), and watching the levels.

Make sure to not engage AGC in the Zoom or other recorder, and watch the levels and mic placement to avoid overdriving the mic.

I've never compared the Zoom's mics to my Earthworks, be an interesting test.
_________________
"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley

beagle · Posted: Sat Aug 27, 2016 7:11 am Post subject:

Derek,

I just posted my plot for you. It occured to me that if your aim is to try to show the difference between the plot of players at various levels of ability, then you really need to collect information about the playing ability of players who created the plot. For example you at least need something like this:
How would you best describe your level? (beginner, intermediate, or advanced)
and perhaps also:
How many years have you been playing the trumpet?

Of course, it would be particularly useful to have the plots of a number of different professional players.

Good luck with the project.

Regards,
Rob
_________________
Schagerl Sig. James Morrison
Selmer Paris Chorus 80J

Derek Reaban · Posted: Sun Aug 28, 2016 7:18 am Post subject:

Hi Rob,

Thanks very much for participating in this project. I like your suggestion about requesting each persons playing level (if someone goes through all of the plots that we get to do a correlation study, this additional information would certainly be important).

How would you best describe your playing level? (beginner, intermediate, or advanced)

I like your categories. I think we could add a few examples beside each for clarity.

Beginner (Comeback player - 6 months of playing)
Beginner (6th Grade - 2 years of playing)
Intermediate (Comeback player - 3 years of playing)
Intermediate (Sophomore in High School - 4 years of playing)
Advanced (Full Time Music Student - 12 years of playing)
Advanced (Professional - 25 years of playing)
Advanced (Non-Music Day Job - 37 years of playing)

And looking at your plot, you have lots of activity up to the 3,500 Hz range!

Thanks again.
_________________
Derek Reaban
Tempe, Arizona
Tempe Winds / Symphony of the Southwest