Table 1.

Results of the comparison between recorded undirected songs and model songs for all 15 birds.

Bird	Model R (random sequence)				Model C (consistent sequence)
	Deviation	p < 0.05	Deviation	p < 0.05	Deviation	p < 0.05	Deviation	p < 0.05
	bird < model		model < bird		bird < model		model < bird
2994	100	100	0	0	95	73	5	0
3534	98	70	2	0	98	50	2	0
4042	100	100	0	0	93	82	7	0
4052	100	100	0	0	84	66	16	0
4217	100	100	0	0	100	98	0	0
4295	100	100	0	0	77	57	23	11
4306	100	100	0	0	98	86	2	2
4344	100	60	0	0	95	50	5	0
4427	100	100	0	0	82	50	18	2
4462	100	100	0	0	77	43	23	7
4502	84	0	16	0	59	5	41	0
4523	100	98	0	0	98	77	2	0
4635	100	100	0	0	66	39	34	14
4669	100	94	0	0	70	39	30	9
g13r8	100	100	0	0	18	9	82	70
Mean:	99	88	1	0	81	55	19	8

Artificial song creation was repeated 50 times for each song chunk with different pseudorandom values for each repetition. Values are the percent of repetitions in which the pulse^S deviation (FRMSD) was lower in natural songs compared to model songs and vice versa (first and third column in each block) and the percentage in which this difference was statistically significant (LMM; p < 0.05; second and fourth column in each block). Column mean is given at the bottom.