PDA

View Full Version : Testing Ukulele Sound - Any good scientific ways at home?



SweetWaterBlue
02-01-2010, 04:10 AM
In the beginner uke buying tips section there was a discussion this morning of the Fluke and whether it was not as loud as more conventional tenors. I originally posted this analysis there, but thought it more appropriate here, and moved it. To my ear my Flea is loud, but in a different way. It seems to have ore mid to low range energy and less on the high end than a good soprano. High frequencies are generally percieved to pierce better, but low notes are necessary too.

I thought it might be interesting to download an MP3 of a few ukes being compared and use Audacity to do a frequency comparison. I did that with a four tenor comparison UkiSociety did (http://www.youtube.com/watch?v=tMpXJVU_3-U), including a Fluke. I wondered what people thought about the validity of this type of analysis?


Here are the results, with the Fluke first:

http://farm3.static.flickr.com/2725/4321687455_a6c1dd3bcd_o_d.jpg


I then selected what appeared to me to be the loudest wooden tenor he played. Here is its frequency analysis:

http://farm3.static.flickr.com/2788/4322421092_e2d29d35be_o_d.jpg

I am not an expert in this, but the fluke does appear to have more energy in the lower end of the spectrum, and the chosen conventional wooden uke has a more distinct higher frequency. The higher notes are quite interesting to compare in the plots. Some of the sound differences may also be due to how Alan fretted the notes on the different fretboards.

SweetWaterBlue
02-01-2010, 04:30 AM
There is a comparison of a Mainland and Collings on YouTube. The Collings sounds better, for course, but the Mainland definitely puts up a good fight against it, at about 1/4 the price. I think I will try this analysis on that video.

SweetWaterBlue
02-01-2010, 09:00 AM
Just so no one relies on these plots, I am starting to think they are useless. Middle C (C4) has a frequency of 265 Hz, and even C6 two octaves above it, has a frequency of only 1046 Hz. Therefore the highest and lowest notes on the standard tuning uke really don't fit the graphs well, since everything is so bunched up near 1000 Hz.

Back to the drawing board, or my ear. I think the solution will be to have Audacity plot them on a log scale, rather than a linear one.

Skrik
02-01-2010, 09:05 AM
"Science" and "tone" do not belong in the same sentence. I thought MGM proved that with his blind test.

paraclete
02-01-2010, 09:10 AM
My failproof home uke test: play uke. If it turns me on and I can't put it down, it's amazing. If I have to play it for awhile before I'm convinced, it's borderline. If I immediately start thinking about how much money I can get for it on craigslist, it's not very good. If I drop it and run screaming from the room, it's a lost cause. :D

SweetWaterBlue
02-01-2010, 10:12 AM
Problem is, a lot of us live in uke-backward communities, and can't play em all. We rely on YouTube and other internet sound samples as a screening tool. I am lucky now that I have discovered UkeRepublic in the next town over, which I guess means we are no longer backward. Mike always brings a nice sampling of new ukes to the SEUkers bi-weekly meetings. Others are not so lucky, and its still fun to listen to them on the net though. I agree that there is no real substitute to actually playing them. Even if you like the sound, you may like the way one neck is shaped better, etc. etc.

dktoller
02-02-2010, 04:13 AM
plot them on a log scale, rather than a linear one.

Neat idea SWB. Would be interesting to see this on a log scale.
Ultimately it comes down to whatever your ear hears and likes best, but it's great when you can connect this to a quantitative explanation.

Dino
02-02-2010, 05:17 AM
Here's a way. Hold your uke in your hand facing towards you. Strum across the fret board. The sound you hear is what your uke will sound like to....basically.....to everyone you play for. If you like the sound, great. If volume is an issue and you cant hear the uke to well, then install a pickup and have the amp facing toward you!

dktoller
02-02-2010, 06:37 AM
-geek advisory ON- Everybody else stop reading and go play your uke!

Started thinking about this and it bugged me so I dug a bit deeper. I took four ~8 seconds clips from UkiSociety's posting. Put them through Audacity and graphed them in Excel.
http://farm5.static.flickr.com/4067/4325005663_f5eb57ce64.jpg

I offset the curves so they were easier to view.
http://farm5.static.flickr.com/4010/4325005673_deb37d5650.jpg

The fundamental frequency range for a uke is roughly A3/220 to A5/880. Everything above this is overtones. Zooming in on 100-1000 Khz shows the peak response is around 200-400 Hz for each instrument. Asserting anything more gets dicey. The clips are taken from different passages of a single song, so the notes played (and volume and expression) vary. The dropoff at 10000 may be a microphone limitation(?) and I wouldn't be surprised if much of the 1000-10000 range is just noise(?). Anybody want to chime in on this?
http://farm5.static.flickr.com/4025/4325005669_91e631f9b1.jpg


Conclusions? None really, except that it's gonna be hard to shop for a uke this way ; )

-geek advisory OFF-

SweetWaterBlue
02-02-2010, 07:21 AM
Nice work dk. Saves me the trouble. Now, all I have to do is study them and see if they lead me anywhere. I think it does show that the all wood instruments have more high frequency response than say a Fluke, which seems to be everyone's hearing observation. The Fluke has a strong response from about 100Hz to the middle of your graph which probably accounts for the the lower frequency resonant overtone I sometimes think I hear on my Flea. Maple and mahogany appear to be "brightest" in that order. Too bad there were no Koas, spruce or cedars in the mix.

scottie
02-02-2010, 07:29 AM
This is a good topic. I'm just getting into home recording meaning I bought recording gear within the last year or so and I'm trying to figure out what to do with it.

A frequency spectrum analysis is going to be only one aspect of the larger analysis of the tone of any given instrument. There are going to be waveform differences as well, especially between ukes of widely differing materials such as, say, a solid wood instrument and a fluke or flea, and perhaps a Dolphin.

Then there are going to be aspects of the instrument which will also be quantifiable such as resonant frequencies, main top resonant frequency, main air resonant frequency. . . there's a lotta stuff here.

There's a really good book called "Critical Listening Skills for Audio Professionals" by F. Alton Everest. It's a good intro to acoustics and perceptual phenomena for people who want to get into home recording and/or become better critical listeners. . . something which would include the ability to critically evaluate the sonic quality of musical instruments. The book comes with a companion CD which is best listened to along with reading the corresponding chapters. It's best read and listened to with a pair of usable monitor speakers (by usable, I don't mean you have to go pop 2k++ for monitors) correctly positioned in a reasonably well treated room will make these things easily apparent. In fact, until the monitors are correctly positioned in a rectilinear (not square) room with at least some acoustic treatment, you will not be able to judge the quality of monitors.

One thing that becomes apparent is that (given that your instrument and your audio gear are of reasonable quality) the quality of your listening environment plays a much bigger role in the overall sound quality of an instrument than many of us appreciate. . . and this also applies in terms of evaluating the qualities of musical instruments. It's why I made the earlier points about sound samples recorded via YouTube and posted on the internet as being inherently suspect.

dktoller
02-02-2010, 07:59 AM
SWB -- glad you enjoyed. And again, I'd be wary of drawing conclusions based on our little analysis. To many unaccounted factors at play.


the quality of your listening environment
Absolutely. I roamed the house last night playing and was amazed at the different 'sound' I got. The room I normally play in is large and probably the deadest room in house. Amazing what a naturally reverb can do to make things sound fuller.

cpatch
02-02-2010, 08:17 AM
My failproof home uke test: play uke. If it turns me on and I can't put it down, it's amazing. If I have to play it for awhile before I'm convinced, it's borderline. If I immediately start thinking about how much money I can get for it on craigslist, it's not very good. If I drop it and run screaming from the room, it's a lost cause. :D
Ding, ding, ding...we have a winner! :)