The Smurfs - Dreams (PPSA21607) creates its raw M4AAC decoder with up to
8 channels and surround channel interleave order 2, and decodes stereo.
Since 8de11f73ee the order threw at creation, so the title aborted before
its first frame. The order only places surround channels, so accept
orders 1 and 2 and throw only when a frame with more than 2 channels is
decoded under one of them.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
On the console only libSceAudiodec.native exports sceAudiodecGetAjmContext
and sceAudiodecUnregisterCodec; libSceAudiodec does not. They move out of
the source shared by both libraries into the native module, and a new
test checks that libSceAudiodec no longer exports them.
sceAudiodecGetAjmContext takes a context and two optional outputs, as a
firmware 13.42 run of libSceAudiodec.native shows: a null context
returns 0x807F0002, otherwise the 32-bit ids at offsets 0 and 4 go to
the outputs that are not null and the call returns 0.
The iBwPcm entry was measured on a PS5 (firmware 13.60). A PS4 Pro
(firmware 10.71) confirms it for MP3 and M4AAC over iBwPcm -1 to 3:
MP3 accepts only 16-bit (1), M4AAC accepts 16-bit and float (1 and 2),
and every other value returns ERROR_INVALID_WORD_LENGTH.
ATRAC9's word length is not reachable with invalid config data: the
console returns ERROR_AT9_INVALID_CONFIG_DATA for every word length, so
it validates the AT9 config before the word size.
Measured with a homebrew probe calling the console's libSceAudiodec
(valid ctrl, no null pointer to an sce API); source and results can be
attached on request.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
The 28- and 32-byte M4AAC parameter blocks carry two fields at offsets
24 and 28 whose range errors fpPS4 names
SCE_AUDIODEC_ERROR_M4AAC_INVALID_ENABLE_NONDELAY_OUTPUT (0x807F0305) and
SCE_AUDIODEC_ERROR_M4AAC_INVALID_SURROUND_CHANNEL_INTERLEAVE_ORDER
(0x807F0306). shadPS4's AJM AAC decoder outputs every decoded frame from
the first one on when non-delay output is enabled, which is what the
decoder here does, so a decoder created with non-delay output 1 is
accepted and decodes as before. Surround channel interleave orders 1
and 2 still throw, now naming the order: no source says which order
each value selects. Name both errors and test that every access unit
yields its frame from the first one with non-delay output enabled.
ui_enable_heaac was ignored and FFmpeg applied Parametric Stereo, so an
HE-AAC v2 stream decoded to 48 kHz stereo whatever the flag: 2048
samples in 2 channels per frame, and ERROR_INVALID_PCM_SIZE for a PCM
buffer sized for the console's output. The console returns the 24 kHz
mono core with HE-AAC disabled, and SBR output at 48 kHz in one channel
with it enabled.
AAC access units now reach FFmpeg without their ADTS header, with an
AudioSpecificConfig built from the header or from the raw parameters.
With HE-AAC disabled it signals SBR as absent. With HE-AAC enabled SBR
stays implicit, and a mono stream that FFmpeg decodes to two channels
is reopened with SBR present and Parametric Stereo absent. The new test
decodes fdk-aac HE-AAC v2 and HE-AAC stereo streams, ADTS and raw, with
HE-AAC on and off; it aborts on main.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
Adrià