Barwa dźwięku i jej podcechy II rok reżyserii dźwięku 8.10.17 AM_3_2017
Plan wykładu Podcechy barwy dźwięku Ostrość Chropowatość Siła fluktuacji Tonalność Przykłady wtyczek audio Mag, mcdrum, mc FM, mcsnare, pspmixsat, RLH Subiektywna ocena 17 skrzypiec
Sharpness: Weighted first moment of distribution of critical band rates of specific loudness, proportion of high-frequency spectral components to low-frequency ones Roughness/: Fluctuation strength: Time structure of the sound signal, modulation factor and level difference determine roughness amplitudeand frequency modulation Tonality: Share of tonal, narrow-band components of a sound signal, depending on frequency, level difference and bandwidth
Ostrość 1 acum 1kHz, 60dB, wąskopasmowy szum, 1 acum Środek ciężkości masy
Ostrość Przykład dźwiękowy wąskopasmowy szum o częstotliwości środkowej f=1 khz szum jednakowo pobudzający wysoko pasmowy szum z częstotliwością odcięcia f=3 khz
white noise Sharpness: Comparison von Bismark Aures of Bismark / Aures white noise and amplification of 8 khz von Bismark Aures
Siła fluktuacji Przykład dźwiękowy amplitudowo modulowany szerokopasmowy szum amplitudowo modulowane tony częstotliwościowo modulowane tony z częstotliwościami modulacji:1 Hz, 4 Hz,16 Hz Maksimum siły fluktuacji dla 4 Hz
Przykład dźwiękowy Chropowatość ton o f=1 khz modulowany amplitudowo z częstotliwością 70 Hz zmieniać się będzie stopień modulacji : 1.0, 0.7, 0.4, 0.25, 0.125, 0.1, 0. Trzy różne dźwięki są prezentowane amplitudowo modulowany szerokopasmowy szum, amplitudowo modulowane tony, częstotliwościowo modulowane tony (+-700) z różnymi częstotliwościami modulacji: 20 Hz, 70 Hz i 200 Hz
Chropowatość
74,2 db 75,1 db 74,7 db Same A-level, Same Third Octave Spectrum
Critical band
Marta -skrzypce
Expert and Non-expert Judgments of Musical Instruments: Subjective Evaluation vs Acoustical Characteristics of Musical Sound Anna Preis, Marta Chudzicka Institute of Acoustics, Adam Mickiewicz University, Poznan, Poland apraton@main.amu.edu.pl
Objectives of the study Comparison of the subjective rankings of instruments obtained for musicians and non-musicians Search for an acoustic criterium corresponding to the subjective ranking of the instruments
Method The experienced, professional violinist played J. S. Bach's "Partita d-moll" on 17 instruments It was a mono recording with a Neuman microphone, type U-87, and DAT Professional PCM 2600 Only the 15s excerpt of this performance was subject to acoustical analysis and to subjective assessment performed by musicians and nonmusicians subjects
Psychoacoustic Experiments Experiment I - sounds from 17 instruments were presented in 136 pairs. Each pair was constructed in the following way: 15s - first instrument, 0.5s pause, 15s - second instrument, 4s pause. Experiment II - sounds of 8 instruments investigated in the Experiment I were selected. These were sounds judged as very good (4 instruments) and as not so good (4 instruments). They were artificially equalized in loudness and assessed.
Apparatus Both experiments were controlled by the PC computer Subjects were listening to the sounds diotically through the headphones TONSIL type 5D-524 Sounds were presented randomly from the playlists created in Sound Forge 4.5
Subjects Experiment I - 100 subjects participated: 48 musicians and 52 non-musicians with normal-hearing They were between 20 and 50. Experiment II - 5 subjects participated: 2 musicians and 3 non-musicians, from 20 to 23.
Procedure You will listen to the same piece of music played on two instruments. Please mark which of the two instruments presented in a pair you would choose if you had a choice The order of the stimuli in pair was randomized The subject listened to the first stimulus, then, after a 500ms pause, to the second stimulus, and next, the subject had 4s to mark her/his judgment in a special form
Procedure cont. Experiment I - each of the 100 subjects made 136 preference judgments, and each instrument occurred 16 times. Each subject judged each pair of instruments only once Experiment II - each of 5 subjects made 28 preference judgments and each instrument occurred 7 times. The number of subjects was reduced but each subject judged 28 pairs of sounds in 5 repetitions
Results It was calculated how many times each instrument was chosen by each subject Calculated numbers were added for all subjects and divided by the number of subjects Thus, the averaged subjects choice of the preferred instrument was calculated
11 12 13 14 15 16 17 19 Subjects' choices of preferred instrument [%] 100 80 60 40 20 musicians non-musicians 0 1 34 5 67 8 9 10 Number of the instrument
A1, A4
A16, A9
A16, A5
Experiment I - results The most frequently chosen was instrument No. 1 The least chosen was instrument No. 16 The differences in the results of musicians and non-musicians are not statistically significant
Musicians' choices of preferred instrument [%] 100 80 60 40 20 0 R 2 = 0,9305 0 20 40 60 80 100 Non-musicians' choices of preferred instrument [%]
Experiment I - results cont. Three groups of instruments were identified: 1, 3, 4, 5, 6 were judged as very good 7, 8, 10, 17 were judged as good, 9, 11, 12, 13, 14, 15, 16, 19 were judged as not so good
Experiment I results cont. The subjective ranking obtained in Experiment I highly correlated with the values of total loudness (R 2 =0.84) It indicates that the preference choices were primarily influenced by loudness Experiment II - 4 loudest and 4 softest violins were chosen. The total loudness of these 8 instruments (Table 1, column 5) was artificially equalized.
Violin Jansson,Niewczyk Present study N Ex.I [sone] N Ex.II [sone] 1 Good Very good 49.9 49.9 3 Good Very good 43.8-4 Very good Very good 46.6 50.3 5 Good Very good 48.4 49.3 6 Not so good Very good 48.3 49.3 7 Not so good Good 44.4-8 Not so good Good 45.0-9 Good Not so good 35.5 46.7 10 Good Good 41.2-11 Not so good Not so good 39.4-12 Very good Not so good 35.1 46.7 13 Not so good Not so good 37.0-14 Good Not so good 35.4 50.9 15 Not so good Not so good 40.3-16 Not so good Not so good 35.0 49.5 17 Good Good 43.3-19 Not so good Not so good 37.7 -
Experiment II results cont. The subjective ranking obtained in Experiment II is almost identical with the results obtained in Experiment I, (with the exception of the instrument No.14) However, the differences between the instruments 5, 6 and 9, 12, 14, 16 were smaller in Experiment II than in Experiment I
Subjects' choices of preferred instrument [%] Musicians I Non-musicians I Non-musicians II 100 90 80 70 60 50 40 30 20 10 0 1 2 3 4 5 6 7 8 Number of the instrument 1 4 5 6 9 12 14 16
Experiment II results cont. The time patterns of the instruments belonging to the same group are strongly similar. This is true for the pairs 1-4, 9-16. In the case of pair 5-6 they are identical The time patterns of the instruments belonging to the different groups are shifted both in time and in the amplitude (total loudness versus time) eg. pair 5-16
Total loudness [sone] No. 1 No.4 90 80 70 60 50 40 30 20 10 0 Time [s]
Total loudness [sone] No. 5 No. 16 90 80 70 60 50 40 30 20 10 0 Time [s]
Conclusions The preference judgments of the violins obtained for musicians do not differ from the preference judgments obtained for nonmusicians It is easy, even for a layman to make a preference judgment of the instruments presented in a pair. Judgment of a single instrument demands experience. This explains why judgments of musicians and non-musicians did not differ significantly. In the original recordings of the different violins the total loudness correlates best with the subjective judgments of the instrument
Subjects' choices of preferred instrument [%] 11 12 13 14 15 16 17 19 Subjects' choices of preferred instrument [%] 100 80 60 40 20 musicians non-musicians 0 1 34 5 67 8 9 10 Number of the instrument Musicians I Non-musicians I Non-musicians II 100 90 80 70 60 50 40 30 20 10 0 1 2 3 4 5 6 7 8 Number of the instrument