In tuomaseerola/onsetsynch: Analysis and Visualisation of Synchronisation of Music Onset Data

What is the minimal amount of representation for the common functions? It depends on the function and the purpose of the analysis, but here we document three minimal cases: One with just pair of onsets, the same with a reference time (mean onset time), and an example with an external timing of the cycles.

Minimal example: Onsets of two instruments and a beat sub-division

We assume the minimal case will be two instruments playing together. Here we create 8 joint onsets tapping in isochrony around twice a second with a random variation. We also assume that there is a beat division representing common time (4 beats/cycle), although the existence of the beat sub-division (SD) does not need to reflect any particular cycle, it could also be just a sequence of numbers.

library(onsetsync)
set.seed(12345) # to create the same random variation every time
ex <- data.frame(SD = rep(1:4,2), 
                 instr1 = rep(0:7/2)+runif(8,min = -0.02,max=0.02), 
                 instr2 = rep(0:7/2) + runif(8,min = -0.02,max=0.02)
                 )
print(knitr::kable(ex))

Here we have two instruments (instr1 and instr2) and a simple beat sub-division (SD). Note that beat sub-division do not need to be specific, it could be a column of numbers (you could create such a column with ex$SD <- 1:nrow(ex) if you only have onsets, nothing else). For the basic analysis of synchrony, this is already sufficient for the calculation of asynchronies.

d <- sync_sample_paired(ex,'instr1','instr2', beat = 'SD')
print(paste('Mean absolute asynchrony:', round(mean(abs(d$asynch*1000)),0),'ms'))

Minimal example with a reference time

Let's add a reference time for a minimal representation that is required for plotting. Here we create a reference time which is the mean of the instrument onsets using add_isobeats() function.

ex1 <- add_isobeats(ex, instr = c("instr1","instr2"), beat = "SD")
print(knitr::kable(head(ex1,4)))

This representation (two instruments, beat sub-divisions, and a reference timing) is the minimal representation for many plotting functions (e.g., plot_by_beat()). The explicit reference time (Mean.Time) is used in plotting.

library(dplyr)
library(ggplot2)
g1 <- plot_by_beat(ex1, instr = c("instr1","instr2"), beat = "SD",virtual = 'Mean.Time')
g1

Minimal example with an external reference time (annotations of cycles)

Sometimes you might have independent information (through annotations) about the timing of the sub-divisions and how they relate to cycles. Let's create this scenario and add Cycle information and timing (CycleTime).

ex2 <- cbind(ex, Cycle = c(rep(1,4),rep(2,4)), CycleTime=NA) # add cycle
ex2$CycleTime[seq(1, 8, by = 4)] <- seq(0, 2, by = 2) # add cycle time for beginnings
ex2 <- rbind(ex2,data.frame(Cycle = 3, SD = 1, # add beginning of the 3rd cycle
                            instr1 = 4, instr2 = 4, CycleTime = 4)) 
print(knitr::kable(ex2))

Now we add the timing of the onsets across beat sub-divisions (SD) based on this external CycleTime. The outcome of the add_isobeats() will now be called Iso.Time, although you can rename it with beatlabel option.

ex2 <- add_isobeats(ex2, instr = "CycleTime", beat = "SD")
print(knitr::kable(head(ex2,4)))

g2 <- plot_by_beat(ex2, instr = c("instr1","instr2"), beat = "SD", virtual = 'Iso.Time')
g2

tuomaseerola/onsetsynch documentation built on Jan. 17, 2024, 10:58 a.m.