Personal Qin w/ Yuesheng M-N Strings

Here you will find various datasets for Yuesheng metal nylon (M-N) strings that I have tested on my qin. This string set is my own set, and is the second set of strings I have bought and used, prior to obtaining the many generously donated composite and silk strings. This set also happens to be the second set of m-n strings that I have used with my qin – the first set was a set of cheap Dunhuang brand strings. The Dunhuang strings I first bought I used for the buzz-checking phase of when I was constructing my qin, and was before I ever had the inclination or thought of anaylizing data for various string sets, and as of present, I currently do not have data for the Dunhuang string set (several strings in that set have also broken due to fatigue and wear of constant stringing and restringing during the buzz checking phase.) Of the two sets however, I will say that the Yuesheng set is far better, in my opinion, than the Dunhuang set. Both sets are very cheap, low quality sets, but the Yuesheng seems better. That being said, both sets still have a characteristic metallic tone to them, and while quite loud on my particular qin, I find them much less preferable to the silk and composite qin strings I have tried.

The graphs below are the harmonic content data, spectrograms, and autocorrelation graphs for this set of Yuesheng M-N strings, with the tuning and technique specified for each data set. You can enlarge the images by clicking on the thumbnails. At the bottom of the page is a brief description of each set of data.


1. Linear Spectrum Harmonic Content Graphs


2. Autocorrelation


3. Spectrograms (Window 4096)


4. Spectrograms (Window 2048)


5. Spectrograms (Window 512)


  1. Linear Spectrum Harmonic Content Graphs – Shows the harmonic content of each string, graphed along the linear spectrum in terms of frequency to intensity. A very accurate way to easily visualize the harmonics and overtones of each string.
  2. Autocorrelation – Shows the periodic nature or trends from a given set of data. Autocorrelation can provide a unique look at data, and can reveal repeating patterns from seemingly random datapoints. For this application, it is derived from the original signal and more clearly shows the decaying oscillatory nature of the plucked string.
  3. Spectrograms (Window 4096) – Shows the spectrogram of each string, with a window setting of 4096. This setting allows one to clearly view all of the harmonics by showing the frequency, intensity, and duration of each harmonic. This graph can be most easily cross-correlated to the linear spectrum harmonic content graphs to compare durations and intensities of harmonics in a string.
  4. Spectrograms (Window 2048) –  Shows the spectrogram of each string, with a window setting of 2048. For this application, I have found that this setting is ideal in viewing the oscillatory instabilities of the guqin string more clearly, which cannot be seen as well in higher window settings. These are seen as wavering lines, which are most noticeably present in the mid-upper harmonics.
  5. Spectrograms (Window 512) – Shows the spectrogram of each string, with a window setting of 512. For this application, I have found that this setting, while having the lowest frequency band resolution of the three settings, allows one to zoom out on the entirety of the harmonic spectrum, and see how the overall power level and intensity shifts from one string to another, and where the harmonic content is overall most present for a given string.