From a133e419b0cf430e3023d7a86e80e77e75041fc9 Mon Sep 17 00:00:00 2001 From: Hector Martin Date: Sat, 28 Oct 2023 02:53:56 +0900 Subject: [PATCH] docs: Add audump.py and speakers.txt Signed-off-by: Hector Martin --- docs/audump.py | 297 ++++++++++++++++++++++++++++++++++++++++++++++ docs/speakers.txt | 20 ++++ 2 files changed, 317 insertions(+) create mode 100644 docs/audump.py create mode 100644 docs/speakers.txt diff --git a/docs/audump.py b/docs/audump.py new file mode 100644 index 0000000..8514769 --- /dev/null +++ b/docs/audump.py @@ -0,0 +1,297 @@ +import struct, sys, os.path, plistlib, pprint + +LABELS = { + "spp3": { + 0: { + 0: "thermal protection enabled", + 1: "displacement protection enabled", + 2: "thermal/power control gain attack time (s)", + 3: "thermal/power control gain release time (s)", + 4: "ambient temperature", + 5: "SafeTlim", + 6: "SafeTlimTimeMin", + 7: "SafeTlimOffset", + 8: "LookaheadDelay_ms", + 9: "peak attack time (s)", + 10: "peak decay time (s)", + 11: "feedback integration time", + 12: "thermal gain (dB)", + 13: "displacement gain (dB)", + 14: "spk pwr averaging window time (s)", + 15: "modeled speaker power", + 16: "measured speaker power", + 17: "power control gain", + 18: "CPMS power control enabled", + 19: "CPMS power control closed loop", + }, + 4: { + 0: "temperature limit", + 1: "hard temp limit headroom", + 2: "T_sett_vc", + 3: "T_sett_mg", + 4: "tau_Tvc", + 5: "tau_Tmg", + 6: "ThermalFFSpeedupFactor", + 7: "temperature", + 8: "OL temperature", + 9: "Reb_ref", + 10: "Rshunt", + 11: "Rampout", + 12: "mt", + 13: "ct", + 14: "kt", + 15: "ag", + 16: "g_bw", + 17: "Q_d", + 18: "phi", + 19: "x_lim", + 20: "ThermalMeasurementMethod", + 21: "pilot tone enabled", + 22: "CL thermal feedback enabled", + 23: "TlimErrDecayTime", + 24: "TempSenseWindowTime", + 25: "TempSenseSmoothingTau", + 26: "a_t_inv", + 27: "PilotAmplHi_dB", + 28: "PilotAmplLo_dB", + 29: "PilotUpperThres", + 30: "PilotLowerThres", + 31: "PilotDecayTime", + 32: "PilotFreq", + 33: "LPMLSPreGain", + 34: "LPMLSPostGain", + 35: "LPMLSLowerCorner", + 36: "LPMLS pre clip level", + 37: "mu_Re", + 38: "mu_Le", + 39: "mu mechanical (PU)", + 40: "Max relative displacement", + 41: "abs(Min relative displacement)", + 42: "DisplacementProtectionType", + 64: "thermal gain", + 65: "displacement gain", + 66: "power control gain", + 67: "PilotDecayTimeStage2", + 68: "PilotEnableThres", + }, + }, + "atsp": { + 0: { + 0: "Bypass", + 40: "Gain link all audio channels", + 1: "speakerType A: Amplifier sensitivity [V/Fs]", + 2: "speakerType A: VoiceCoil: DC resistance [Ohms]", + 3: "speakerType A: VoiceCoil: thermal resistance [C/Watt]", + 4: "speakerType A: Voice Coil: thermal time constant [s]", + 5: "speakerType A: Magnet: thermal resistance [C/Watt]", + 6: "speakerType A: Magnet: thermal time constant [s]", + 7: "speakerType A: Ambient temperature, [C]", + # The target temperature of the speakers + 8: "speakerType A: Temperature limit [C]", + 9: "speakerType A: Attack time (ms)", + 10: "speakerType A: Release time (ms)", + 11: "speakerType A: Temperature hard limit headroom [C]", + 12: "speakerType A: Gain link", + 13: "speakerType A: Audio channel assignment", + 14: "speakerType B: Amplifier sensitivity [V/Fs", + 15: "speakerType B: VoiceCoil: DC resistance [Ohms]", + 16: "speakerType B: VoiceCoil: thermal resistance [C/Watt]", + 17: "speakerType B: Voice Coil: thermal time constant [s]", + 18: "speakerType B: Magnet: thermal resistance [C/Watt]", + 19: "speakerType B: Magnet: thermal time constant [s]", + 20: "speakerType B: Ambient temperature, [C]", + 21: "speakerType B: Temperature limit [C]", + 22: "speakerType B: Attack time (ms)", + 23: "speakerType B: Release time (ms)", + 24: "speakerType B: Temperature hard limit headroom [C]", + 25: "speakerType B: Gain link", + 26: "speakerType B: Audio channel assignment", + 27: "speakerType C: Amplifier sensitivity [V/Fs]", + 28: "speakerType C: VoiceCoil: DC resistance [Ohms]", + 29: "speakerType C: VoiceCoil: thermal resistance [C/Watt]", + 30: "speakerType C: Voice Coil: thermal time constant [s]", + 31: "speakerType C: Magnet: thermal resistance [C/Watt]", + 32: "speakerType C: Magnet: thermal time constant [s]", + 33: "speakerType C: Ambient temperature, [C]", + 34: "speakerType C: Temperature limit [C]", + 35: "speakerType C: Attack time (ms)", + 36: "speakerType C: Release time (ms)", + 37: "speakerType C: Temperature hard limit headroom [C]", + 38: "speakerType C: Gain link", + 39: "speakerType C: Audio channel assignment", + } + } +} + +""" +ATSP protection behavior: + +Max gain reduction is 20dB. +"Temperature limit" is the target temperature +If temperature exceeds limit + "Temperature hard limit headroom", +protection goes into panic mode and triggers 20dB reduction. + +For settings: + amp = 12 r = 4 + rVc = 50 aVc = 2 rMg = 1 aMg = 1 + Ta = 50 Tlim = 150 Theadroom = 5 + +We see this limiter behavior: +In Out +0 -9.7 +-8 -9.7 +-9 -9.6 +-9.5 -9.7 +-9.8 -9.9 +-10 -10 + +In other words, it behaves like a hard limit / compressor with infinite ratio. + +Theadroom has no influence on the gain reduction, it just affects stability +(temperature does exceed Tlim transiently, if the transient is > Theadroom + it panics). Too low a Theadroom leads to unstable behavior. +""" + + +def dump_audata(labels, data): + top = {} + while data: + hdr = data[:0xc] + data = data[0xc:] + typ, grp, cnt = struct.unpack(">III", hdr) + d = {} + for i in range(cnt): + blk = data[:0x8] + data = data[0x8:] + key, val = struct.unpack(">If", blk) + if typ in labels: + if key in labels[typ]: + key = labels[typ][key] + d[key] = val + top[(typ, grp)] = d + pprint.pprint(top, stream=sys.stderr) + return top + +def process_spp3(e): + # Grab the plist file, which is mostly redundant but contains + # some details not in the au preset + for i in prop["Boxes"]: + if i["Name"] == e["displayname"]: + for p in i["Properties"]: + if p["Number"] == 64003: + path = os.path.join(base, "DSP", p["Path"].split("/DSP/")[1]) + + pl = plistlib.load(open(path, "rb")) + + d = dump_audata(LABELS["spp3"], e["aupreset"]["data"]) + spkrs = "" + channels = 0 + gbl = d[(0, 0)] + for (typ, ch), p in sorted(d.items()): + if typ != 4: + continue + chp = pl["ChannelSpecificParams"][f"Channel{ch}"] + channels += 2 + spkrs += f""" + +[Speaker/{chp["SpeakerName"]}] +group = {chp["SpeakerGroup"]} +tr_coil = {p["T_sett_vc"]:.2f} +tr_magnet = {p["T_sett_mg"]:.2f} +tau_coil = {p["tau_Tvc"]:.2f} +tau_magnet = {p["tau_Tmg"]:.2f} +t_limit = {p["temperature limit"]:.1f} +t_headroom = {p["hard temp limit headroom"]:.1f} +z_nominal = {p["Reb_ref"]:.2f} +z_shunt = {p["Rshunt"]:.2f} +is_scale = 3.75 +vs_scale = 14 +is_chan = {2 * ch} +vs_chan = {2 * ch + 1}""" + + print(f"""\ +[Globals] +visense_pcm = 2 +t_ambient = {gbl["ambient temperature"]} +t_hysteresis = 5.0 +t_window = 20.0 +channels = {channels} +period = 4096 +link_gains = True + +[Controls] +vsense = VSENSE Switch +isense = ISENSE Switch +amp_gain = Amp Gain Volume +volume = Speaker Volume{spkrs}""") + +def process_atsp(e): + # print(e) + d = dump_audata(LABELS["atsp"], e["aupreset"]["data"])[(0,0)] + t_ambient = None + + spkrs = "" + channels = 0 + for gid, gn in enumerate("ABC"): + p = f"speakerType {gn}: " + ch = int(d[p + "Audio channel assignment"]) + if not ch: + continue + if ch == 0xffff: + ch = 1 + + ambient = d[p + "Ambient temperature, [C]"] + assert t_ambient is None or t_ambient == ambient + t_ambient = ambient + + for i in range(16): + if ch & (1 << i): + channels += 2 + spkrs += f""" + +[Speaker/{gn}_ch{i}] +group = {gid} +tr_coil = {d[p + "VoiceCoil: thermal resistance [C/Watt]"]:.2f} +tr_magnet = {d[p + "Magnet: thermal resistance [C/Watt]"]:.2f} +tau_coil = {d[p + "Voice Coil: thermal time constant [s]"]:.2f} +tau_magnet = {d[p + "Magnet: thermal time constant [s]"]:.2f} +t_limit = {d[p + "Temperature limit [C]"]:.1f} +t_headroom = {d[p + "Temperature hard limit headroom [C]"]:.1f} +z_nominal = {d[p + "VoiceCoil: DC resistance [Ohms]"]:.2f} +is_scale = 3.75 +vs_scale = 14 +is_chan = {2 * i} +vs_chan = {2 * i + 1}""" + + print(f"""\ +[Globals] +visense_pcm = 2 +t_ambient = {t_ambient} +t_hysteresis = 5.0 +t_window = 20.0 +channels = {channels} +period = 4096 +link_gains = {bool(d["Gain link all audio channels"])} + +[Controls] +vsense = VSENSE Switch +isense = ISENSE Switch +amp_gain = Amp Gain Volume +volume = Speaker Volume{spkrs}""") + +if __name__ == "__main__": + base = sys.argv[1] + + au = plistlib.load(open(os.path.join(base, "DSP/Strips/builtin_speaker_out_general.austrip"), "rb")) + try: + prop = plistlib.load(open(os.path.join(base, "DSP/Strips/builtin_speaker_out_general.propstrip"), "rb")) + except: + prop = None + + for s in au["strips"]: + for e in s["effects"]: + if e["unit"]["subtype"].to_bytes(4) == b"spp3": + process_spp3(e) + if e["unit"]["subtype"].to_bytes(4) == b"atsp": + process_atsp(e) + diff --git a/docs/speakers.txt b/docs/speakers.txt new file mode 100644 index 0000000..2ed509b --- /dev/null +++ b/docs/speakers.txt @@ -0,0 +1,20 @@ +Speaker configs + +Model ID Amp Gain Speakers Sense Prot +j180 AID19 sn012776 10 1× 1W+1T no atsp +j274 AID6 tas5770 20 1× 1W no atsp +j293 AID3 tas5770 15 2× 2W no atsp +j313 AID4 tas5770 10 2× 1W no atsp +j314 AID8 sn012776 15 2× 2W+1T yes spp3 +j316 AID9 sn012776 15 2× 2W+1T yes spp3 +j375 AID10 sn012776 15 1× 1W no atsp +j413 AID13 sn012776 15 2× 1W+1T yes spp3 +j414 AID14 sn012776 15 2× 2W+1T yes spp3 +j415 AID27 sn012776 15 2× 2W+1T yes spp3 +j416 AID15 sn012776 15 2× 2W+1T yes spp3 +j456 AID5 ssm3515 15 2× 1W+1T no atsp +j457 AID7 ssm3515 15 2× 1W+1T no atsp +j473 AID12 sn012776 20 1× 1W no atsp +j474 AID26 sn012776 20 1× 1W no atsp +j475 AID25 sn012776 20 1× 1W no atsp +j493 AID18 sn012776 15 2× 2W unused? atsp