ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
AY-YM / ZX-Speccy emulations special cases test results (2020-09-08 - 2021-11-28) vs real
2
freq rangenoiseenveloperetrig reinittonecomment
3
randomnessperiod 0/1period 31 freqhalf-amp 0half-amp 1harmonicsperiod 0period 1periods 0-2
harmonics
4
chipAY-8912fullOKsame3400cleanclean5OKcleancleansame5real chip
5
YM2149ffullOKsame3500cleanclean5OKcleancleansame5real chip
6
high qualityay-emul2.916 KHzOKsame3500cleanclean7OKcleancleansame6exemplary implementation
7
zxtune5000fullOKsame3500cleanclean5OKcleancleandifferent3mediocre lowpass filter
8
specemu3.220 KHzOKsame3530cleanclean4OKcleancleansame4mediocre lowpass filter
9
xpeccy0.6.2022012419 KHzOKsame3500cleanclean3OKcleancleansame6mediocre lowpass filter
10
ZX-Next19 KHzOKsame3500cleanclean7OK3.4KHz -53db3.4KHz -53dbsame3mediocre lowpass filter
11
mame0.225fullOKsame3500cleanclean4OK3.5KHz -53db3.5KHz -53dbdifferent3mediocre lowpass filter
12
unreal 39.0(2021)19 KHzOKsame3500cleanclean4OK11KHz -43db11KHz -43dbdifferent1mediocre lowpass filter
13
ulax emul b26fullOKsame3500cleanclean6OKcleancleansame6No stereo, regular clicks with 44.1KHz on integrated audio
14
wrong noise freqszx-spin0.719 KHzOKsame7100cleanclean3OKcleancleansame3bad lowpass filter, wrong noise freq
15
gleck0.0519 KHzOKsame7200cleanclean3OKcleancleandifferent3wrong noise freq, low process priority, bad stereo
16
cspect2.12.3616 KHzOKdifferent3500cleanclean3OK2.6KHz -46db2.6KHz -46dbdifferent1mediocre lowpass filter, wrong noise freq 0
17
zxmak2 2.9.3.820 KHzOKdifferent3500cleanclean5OK4KHz -36db4KHz -36dbdifferent1bad lowpass filter, wrong noise freq 0
18
spectaculator8.020 KHzOKsame7100cleanclean2OK10KHz -42db10KHz -42dbdifferent1bad lowpass filter, wrong noise freq
19
no retrig, bad half-amp env, clicks, short noise loopayumifullOKsame3500cleanclean7FAILcleancleansame7no retrig
20
ay-play13.619 KHzOKsame7000cleanclean6FAILcleancleansame6no retrig, wrong noise freq
21
emuzwin2720 KHzOKsame7100cleanclean2FAIL11KHz -51db11KHz -51dbdifferent3bad lowpass, wrong noise freq
22
z80stealth1.1720 KHzOKsame3500FAILFAIL6OKcleancleansame6first 5 freqs volume is flat-half
23
fuse1.6.018KHzOKsame3500cleanclean7OKclicksclickssame4clicks when low periods (<8)
24
emuzgl18 KHzshort loopsame7200cleanclean2FAIL20KHz -35db20KHz -35dbdifferent3bad lowpass, wrong noise freq, bad random
25
zx-poly2.1.020 KHzOKsame7100harmonicsharmonics0OK3.6KHz -25db3.6KHz -25dbdifferent1wrong noise freq, bad lowpass filter, clicks, low process priority, mono
26
zesarux9-108 KHzshort loopsame3500garbageclicks7FAILclicksclicksgarbageclicks16khz/8bit, total garbage sound, clicks with periods 0..7
27
Sprinterto be added...
28
29
Таблица отсортирована по качеству звучания (наверху эталонные чипы) / Table is sorted by quality (with standard hardware at the top), by comparing with standard
30
Wrong noise - небольшие проблемы с шумом (неправильные частоты), ниже - хуже: отсутствие ретригов, щелчки, кривой шум, кривой звук огибающей с тоном 0
real original chips
31
Real YM chip has been captured with 192000Hz
stand-alone players (not emulators)
32
All emu sounds were captured with 44100Hz stereo with separation ACB, and then processed (monofied) - substracted "empty" channel B from "main" channel A (to cancel wire noise), and then normalized to 75%software emulators
33
hardware emulators
34
https://volutar.myds.me/zx/ayymtest/ayphastest.7zsna/hobeta/ay test itselfmild results deviations
35
https://volutar.myds.me/zx/ayymtestcaptured resulting samples (real chips + emulations)hard results deviations
36
37
freq response - с какой частоты начинается фильтрация верхов / low-pass filter value
38
noise randomness - нет ли слышимых повторов (сэмплированность) / does it have hearable noise loop
39
noise period 0/1 - совпадают ли спектры делителя 0 и 1 (должны совпадать) / does it have same noise with 0/1 periods (should be the same)
40
noise period 31 freq - самая низкая частота "выреза" при периоде шума 31 / lowest noise frequency notch for period 31 (should be 3300-3700)
41
env half-amp 0 - насколько чисто звучит огибающая при делителе тона =0 / how clear it renders half-volume envelope with tone period=0
42
env half-amp 1 - то же самое при делителе тона 1 / same for tone period 1
43
env harmonics - с какого делителя уже становятся заметны высокочастотные гармоники-частоты (лучше 4 или больше) / which period causes hearable high harmonics-pitch (better >=4)
44
retrig reinit - есть ли сбивка цикла при переключении на делитель 0 / is there tone wave phase "break" with using of period 0
45
tone period 0 - какие гармоники слышны при делителе 0 (должна быть тишина) / what's harmonics are generated with tone period 0 (must be silence)
46
tone period 1 - что выводится при делителе 1 / what's generated with tone period 1
47
tone periods 0-2 - одинаковые ли звуки при делителях 0-2 (в смеси с огибающей) (должны быть одинаковы) / are tone periods 0-2 with envelope should sound the same?
48
tone harmoics - с какого делителя видны высокочастотные гармоники (должны быть отфильтрованы как минимум до 5) / starting with which period high pitch harmonics appear (should be 5 or more)
49
50
Test contents:
51
noise 36 x 25int
Noise freq 0..35
freq=0,1,32,33 should sound the same, freq=31 should have first notch at 3500hz
52
53
54
env half 12 x cycles
Tone freq 0..11
each cycle uses full and half amplitude of envelope (by using tone+envelope with extreme low tone period). at least first 4 periods shouldn't affect on high harmonics
55
56
57
retrig fragment
Note lengths 6,12,24
each note is initialized with tone cycle being "reset" by putting 0 as period before actual note period. this provokes only two phase of T+N versions
58
59
60
tone 12 x 25int
pure tone periods 0..11
checking decimation quality, filtering of higher [mirrored] harmonics
61
62
tone+env 12 x 25int
tone periods 0..11 with env
checking decimation quality, along with envelope with period=0 (type=C)
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100