The change item is recording.gatf which could be found in the ZIP file.
for recording.gatf
change parameter for MBDRC in all scenarios
Mackup_gain_dB 24 -> 27.0
Min_Gain_dB -25 -> -29.0
Threadhold_dB[1] -44 -> -38
CompressRatio [1]0.8 -> 1.0
CompressRatio [2]0.2 -> 0.0
KneeWidth[2] 0 -> 6
Limiter_gain_dB -0.2 -> -2
Change IIR2 8kHz to 2dB
Test: verified by HW wjacky@
Bug: 264746360
Change-Id: Icb53d0429fe8adab43a1743bedb71ee81eafd947
Signed-off-by: yixuanjiang <yixuanjiang@google.com>
Bug: 243078049
Bug: 243079260
Test: long press gesture is registered when UDFPS is enabled.
Change-Id: Iaf5553978cdb79a92a24b674f837ae694e187d48
Signed-off-by: davidycchen <davidycchen@google.com>
No cooling device is added yet. The limits will be determined and added
once validation completes.
Bug: 244348786
Test: Confirm boot and IRQ triggered mitigation logger
Change-Id: I46c86d09d5b69620bc6c2f25bd6820507b168ca3
Signed-off-by: George Lee <geolee@google.com>
//SB3
Handset: Duncan
1: Modify NB/WB/SWB gain setting and IIR setting to meet 3GPP requirements
HAC: Owen/Duncan
1. Modify NB/WB/SWB DL FDEQ slightly to meet Y-axis magnetic FR reqs of HAC v2019
2. Copy DL settings above to Handset_HAC TMOUS / RESERVE2
Handsfree: Ricky
1. All band first tuning release for Proto stage
BT: Gene
1. Fine tune BT EL for NB WB SLR
2. BT HE table version from 1.3.1 to 1.3.3
3. Make HK BT LE EASSA to be same with SB (SB is correct originally)
Files from the latest attachment on b/264499596
Bug: 264499596
Test: verified by rickycheng@
Change-Id: I09730d343f27cec4ae4a285373f2b05daa6b2423
//HK3
Handset: Owen
1. Modify NB/WB/SWB DL FDEQ, TDDRC Gain, and SPK_Vol Gain(NB only) for 3GPP regs.
2. Modify NB/WB/SWB UL PreEQ MIC1 based on HK3 MIC Acoustic performance to match MIC0 FR
3. Modify MIC distance DIST2REF1 to 166mm based on HK3 MIC placement
4. Set NB/WB/SWB UL PGA_0/_1 gain as 20dB(from 6dB) to cover codec gain issue as a short term
5. Copy the settings above to Handset TMOUS / RESERVE2
6. Modify NB/WB/SWB Tx NS para from FM simulation to improve S-MOS(Handset Generic only)
HAC: Owen
1. Modify NB/WB/SWB DL FDEQ slightly to meet Y-axis magnetic FR reqs of HAC v2019
2. Copy DL settings above to Handset_HAC TMOUS / RESERVE2
Handsfree: Ricky
1. All band first tuning release for Proto stage
BT: Gene
1. Fine tune BT EL for NB WB SLR
2. BT HE table version from 1.3.1 to 1.3.3
3. Make HK BT LE EASSA to be same with SB (SB is correct originally)
Files from the latest attachment on b/264499619
Bug: 264499619
Test: verified by rickycheng@
Change-Id: I709998c8aa108ef21211176b6e8c50b983b2e5a0
For all the use-cases, restrict camera daemon to use
4 small and 4 mid cores by default.
Bug: 261760700
Test: GCA video recording
Change-Id: Idf373a98f13e066361cc5144a14cbfefa27e7ca1
There are recording.gatf and template.xml in the ZIP file.
for recording.gatf
Camcorder path : IIR1(high pass), SUR(calibrated filter), IIR2(high pass), AGC (-24 ideal RMS), MBDRC ( -44 threshold)
MIC path: correct MBDRC
for template.xml
correct SUR module ID
Files from the latest attachment on b/263448513
Bug: 263448513
Test: verified by wjacky@
Change-Id: I7d7dc2496ee3cf37b35b30f34b9f72febc512189
There are recording.gatf and template.xml in the ZIP file.
for recording.gatf
Camcorder path : IIR1(high pass), SUR(calibrated filter), IIR2(high pass), AGC (-24 ideal RMS), MBDRC ( -44 threshold)
MIC path: correct MBDRC
for template.xml
correct SUR module ID with AZ1.0 parameters
Files from the latest attachment on b/263449687
Bug: 263449687
Test: verified by darrenyen@
Change-Id: I5a31582ecf6b8f8b81be7e61079e172bddb09ba0
Update with v2 virtual coefficients
Update trigger sensors
Align hot thresholds with P22
Use VIRTUAL-SKIN as combination where applicable
Bug: 259450323
Test: Build and boot on device. Check thermal service
adb shell dumpsys thermalservice
Change-Id: I5f738a1cc1fe3735f5fae41dd42d27e6bd305e05
Signed-off-by: Sayanna Chandula <sayanna@google.com>
Align control temperature and switch on temperature thresholds
with current kernel configuration
Bug: 259450323
Test: Build and boot on device. Quick thermal check
Change-Id: I423a68762ae5b4df5b09b058322f40888aae8fce
Signed-off-by: Sayanna Chandula <sayanna@google.com>
1. Set the LE Audio unicast offload property
2. Add the offload capability file for unicast
Bug: 257377483
Test: Manual
Change-Id: If6233bfcefa8894f52d0a8c7799ffa5b6b786f53
- Full tuning includes volume curve to match previous pixel
at volume 18 to 1 perceived loudness.
- Passes all power targets
- Pass top speaker heat budget for 40C skin
- Fails bottom speaker heat budget for 40, 41, 42C skin. Passes 43C skin.
- Test acoustics shows superior performance to past premium.
Bug: 262787252
Test: manual test
Change-Id: Iec4664f0f1ad7499caf5e3cde2158151ca849111
Signed-off-by: Jasmine Cha <chajasmine@google.com>
The 2nd attempt (ag/20294157) was reverted due to b/259732944. To avoid
the bug, we explicitly define the "high" quality encoder as a 4k
encoder. Note (1) below explains the reasons.
8k encoder specs:
Codec: h264
bitRate: 80000000 (80M); source: b/241178681#comment3
Size: 7680x4320 (8K UHD)
Frame rate: 30fps
Notes:
(1) Regarding b/259732944 (the bug caused by the 2nd attempt):
If CamcorderProfile.QUALITY_HIGH isn't explicitly specified, the highest
resolution available will be used.
Therefore, when the 8k profile was added, apps requesting QUALITY_HIGH
would ask HAL to perform 8k recording.
This created the following issues:
(a) As 8k support in HAL isn't ready, existing tests using QUALITY_HIGH
will fail (such as CTS, PTS, FATP, etc.).
(b) 3rd-party apps currently using QUALITY_HIGH will get 8k, which might
not be their intention.
So we chose to keep QUALITY_HIGH at 4k, which was the highest resolution
on P22 devices.
Detailed discussions are in b/259599407.
----------------
(2) Regarding the "quality" attribute in EncoderProfile:
It must match what's defined in MediaProfiles::sCamcorderQualityNameMap
(see frameworks/av/media/libmedia/MediaProfiles.cpp).
----------------
(3) Regarding the camera IDs:
The camera IDs in media_profiles_{device_name}.xml correspond to the
IDs inside vector ModelSpecificDataControllerBase::visible_camera_ids_.
The vector is populated in
lyric/silo/hardware/devices/{device_name}/data/model_specific_data_controller.cc
According to the documentation of "visible_camera_ids_", "the first ID
for each facing will be used as the default camera ID". As
visible_camera_ids_[0] is always the first (default) ID for the rear
camera, and only the rear camera supports 8k, we add 8k encoder
profiles for camera 0 in media_profiles.
The explanation above isn't obvious. You'll need to read
model_specific_data_controller.cc to understand it.
----------------
(4) Regarding tests:
We can't test the 8k encoder since we haven't brought up the 8k support
in ISP.
We can use CTS, PTS, and EECoexer to test this CL for existing use cases.
Note that RecordingTest#testBasicRecording currently fails at TOT due to
a different reason. For this CL, we only need to make sure no 8k stream
is requested by observing logs with the "GCH_HalUtils" tag.
Bug: 232811715
Bug: 262423942
Test: build & boot Android
Test: atest RecordingTest#testBasicRecording
Test: adb shell /data/sysval/scripts/eecoexer/eecoex.sh -s CAM,22,0,5,true,HIGH
Test: manually testing GCA video recording
Change-Id: Ie6c054e641810da9c231f3839961cbe8fcc16fdc
