Skip to content

hi3516cv200 + OV2710: no tuned IQ profile (ov2710.ini is an ov9750 copy, default.ini is ar0130) - needs PQTools tuning #304

Open

Description

@zx97

What works, and what does not

Hi3518EV200 (hi3516cv200) board, OV2710 on its MIPI connector, 1 data lane,
24 MHz XVCLK, libsns_ov2710_mipi.so from this repo, RAW10 at 1920x1080/30.
The link, the interface mode, the lane count and the frame rate are settled:
IntfM = MIPI, IntCnt 30/s, IntGapT 33424 us, TmgErr 0, ccErrN 0, and
/image.jpg returns a 1920x1080 JPEG.

The picture is per-pixel noise. At 1:1 there is no recoverable detail; only the
coarse average follows the scene.

The test pattern isolates it

Writing 0x503d = 0x80 (colour bars; the Dafang table has the line commented
out as //[7] [5:4] color bar) produces a perfect frame - seven clean bars,
correct colours, sharp edges, local 8x8 block noise of 0.0, JPEG 51 971 bytes
against 1 146 997 for the live scene.

That rules out, on this board: the CSI-2 link, the pads, the lane count, the
RAW10 unpacking, the ISP, DRC, sharpening, the JPEG path, and the Bayer order
(the bars are the right colours). The digital path is clean, so the noise is
generated in the analogue readout.

What has been eliminated since

Every measurement below is /proc/umap/vi, /proc/umap/isp, ipctool i2cget
and plain JPEG snapshots on the same board.

Candidate Test Result
integration time 2 / 550 / 1101 lines frame mean 109 / — / 171, block noise 72 / 72 / 92 - it scales with the signal, not against it
analogue gain 0x350b 0x00 (1x) and 0x10 (2x) 71.8 vs 92.3, again proportional
data_type RAW8 / RAW10 / RAW12 / RAW14 rejected / 30 fps / ~100 counts per s / rejected
CFA phase Isp_Bayer 0-3 mean RGB identical to a tenth
sharpening, DRC SharpenD_* = 0, DRCEnable = 0 79.8 vs 80.4, no effect
the AE isp.aeMode: manual, exposure forced to 1101 lines, gain 2x same relative noise (~0.55 x mean)
gain encoding cmos_gains_update() revisited against Allwinner's sensor_s_gain() the encoding in main is correct; a patch I proposed was wrong and has been closed (#303)
temporal or fixed two consecutive frames noise uncorrelated (0.097), scene correlated (0.911)

The noise is multiplicative and temporal: sigma/mean sits at 0.54-0.66
whatever the exposure or gain, and every frame differs. A working sensor at 1x
and a 30 ms integration does not do that.

What was checked against other drivers

Two published OV2710 MIPI drivers agree with each other and with what this one
writes:

  • allwinner-zh/linux-3.4-sunxi, ov2710_mipi.c - gain as range bits plus
    mantissa, exposure split 0x3500 bits 16-19 / 0x3501 bits 8-15 / 0x3502
    bits 0-7, 0x3503 = 0x07 (AEC manual, AGC manual, VTS manual);
  • espressif/esp-video-components, ov2710_mipi_1lane_24Minput_1920x1080_raw10_25fps.h
    • which is this board's exact operating point. Its window registers match what
      the sensor reports here (0x3800..0x3813 = 01c4 000a 0780 0438 0780 0438),
      HTS 0x0974 = 2420, and the same pixel rate (2420 x 1322 x 25 against
      2420 x 1104 x 30 here).

The one value every reference writes and this driver's init does not is
0x4800 = 0x24 (Allwinner and Espressif both); the part reads 0x04 here.
Writing 0x24 at runtime changed nothing in the image, and writing 0x20 or
0x3000 = 0x02 stops the stream outright, so this board wants the values the
shipped table leaves.

The questions

  1. Is there anything in the init that shapes the analogue path and that the
    84-write table lacks? The table is well short of the 220 rows in the Ingenic
    Dafang driver, and the parts that differ (PLL 0x3034..0x3037, 0x4000 /
    0x401d black level, 0x5000/0x5001, 0x4600) are exactly the ones that
    decide the readout rather than the interface.
  2. Could this be a damaged or non-standard part? The board is a retail
    VStarcam C23S-class camera whose sensor was identified by ipctool and by
    the I2C address (0x6c), not by reading a chip ID against a datasheet
    (ov2710_cmos.c has no ID register path). A noise figure that ignores
    exposure and gain is also what a failing sensor looks like.

I can test any patch, register sequence or library on the board - it is on the
bench and reachable. What I would rather not do is keep poking registers blind:
two forced-exposure sweeps put it into a spin that the watchdog had to reboot.

For completeness, the two things already reported elsewhere: the shipped
ov2710_mipi_1080p.ini selected the parallel interface (fixed in
OpenIPC/firmware#2869), and my own gain-encoding patch was wrong and is closed
(#303) - the encoding in main matches Allwinner's driver.

Activity

  1. zx97 commented on Nov 16, 2026

    @zx97
    Author

    Two corrections and one new fact, from a further hour on the board.

    1. The magenta cast in the pictures above was not the sensor. majestic
    eventually named it: the board has no IR-cut filter wired to ir_cut_pin1/2, so
    the filter sat in whichever position it powered up in, and the picture carried
    its cast. With night_mode: { enable: false, ir_cut: off }:

    before (IR-cut driven)   R=111.3 G=104.1 B=129.1   B +12.4 %   magenta
    after  (ir_cut: off)     R=115.1 G=107.4 B=102.5   R  +6.3 %   warm, plausible
    

    The frame becomes a recognisable room - furniture in warm lamp light, a window
    on the right showing the night sky in blue. Any colour measurement taken before
    this point in my report carries that cast, including the Isp_Bayer sweep.

    2. The noise is unchanged by it: same scene, local 8x8 block noise 71.8
    before and 69.6 after. So the report still stands, on the noise alone.

    3. The AE is ruled out as well. majestic accepts isp.aeMode: manual; with
    exposure forced the part took 1101 lines (the VTS maximum) and aGain 2x:

    1101 lines, 2x   frame mean 171.5   block noise 92.3   sigma/mean 0.54
    AE auto (2-7 lines) frame mean 109.1   block noise 71.8   sigma/mean 0.66
    

    More integration gives more signal and more noise in the same proportion, so the
    figure tracks the signal rather than the operating point. That also confirmed the
    gain encoding a second time: asking for 2x put 0x10 in 0x350b, exactly as
    Allwinner's sensor_s_gain() says.

    What is left is the thing the test pattern already pointed at: the analogue
    readout produces ~55-65 % relative noise at any exposure and gain, while the
    sensor's own digital test pattern arrives with none at all.

  2. openipc-ai commented on Nov 16, 2026

    @openipc-ai
    Contributor

    Thanks for writing this up so carefully. The colour-bar result and the table of what you've ruled out make this easy to pick up. Some answers, one correction, and a way to stop guessing at registers.

    The ini fix is now OpenIPC/firmware#2871, not #2869. The shared ov2710_mipi_1080p.ini stays as it is, and your four values go into a second profile, ov2710_mipi_1lane_1080p.ini. majestic tries that file only on boards where the default captures nothing. Details and the test we need from you are in the review on #2869.

    Q2, is it a genuine OV2710? ipctool does read the chip ID. Its OmniVision probe reads 0x300A/0x300B and reports OV2710 only when they return 0x27 0x10. You can confirm it on this unit directly:

    ipctool i2cget 0x6c 0x300a   # 0x27
    ipctool i2cget 0x6c 0x300b   # 0x10

    The right ID tells you the die answers as an OV2710. It doesn't tell you the die is healthy, so this alone can't separate "damaged part" from "missing init".

    Q1, what the table lacks. The 84 writes already include the readout registers you list: 0x4000 = 0x05, 0x401d = 0x02, 0x5000 = 0x5f, 0x5001 = 0x4f. The clock setup for MIPI is 0x300e/0x300f/0x3010/0x3011/0x3030. So the difference from other tables is mostly values, not missing registers, and comparing table against table register by register is the slow way to find it.

    The test that answers both questions: the camera's own firmware. The C23S ships a vendor firmware that streams from this exact sensor. If you have the flash backup you made before installing OpenIPC:

    1. Boot the stock firmware once and look at the picture.
      • Clean: the part is fine and the difference is in the init.
      • Same noise: the part is the problem, and no driver change will fix it.
    2. If it's clean, dump the sensor while the stock firmware streams:
      ipctool i2cdump --script 0x6c 0x3000 0x3fff > stock_3xxx.sh
      ipctool i2cdump --script 0x6c 0x4000 0x4fff > stock_4xxx.sh
      ipctool i2cdump --script 0x6c 0x5000 0x5fff > stock_5xxx.sh
      Then take the same three dumps on OpenIPC and diff them. The diff is the exact set of registers our init gets wrong for this module.
    3. ipctool trace on the stock streamer records its init writes in order. Use that if the dump diff turns up registers where the order matters, such as the PLL.

    The dump costs nothing and you can't damage the camera by reading. It also avoids the blind sweeps that sent the board into the watchdog spin. Please attach the three diffs (or the raw dumps) here. If they show a short, generic delta, it can go into ov2710_sensor_ctl.c. If the delta looks specific to this module, it will need the same handling as the lane count.

  3. zx97 commented on Nov 16, 2026

    @zx97
    Author

    Chip ID, as you suggested:

    ipctool i2cget 0x6c 0x300a   -> 0x27
    ipctool i2cget 0x6c 0x300b   -> 0x10
    

    So the die answers as an OV2710, which is one thing settled.

    The test you propose cannot be run on this unit: there is no stock-firmware backup.
    That was my mistake at the very start of the project, and I should have said it in
    the report instead of letting you propose a path I cannot take. This U-Boot has no
    tftpput, so the only export route was the 115200 serial console, and 16 MB that
    way was not something I attempted. The stock image is gone; the vendor OTA
    (api4.eye4.cn) wants the exact version string, which went with the flash.

    So the "boot the stock firmware and look" branch is closed here, and with it the
    ipctool i2cdump --script diff against a stock stream - both need the stock
    firmware running.

    What I can still do, if it helps, in order of how much I trust it:

    1. Point me at a C23S-class stock image for hi3518ev200 that can be downloaded
      (vendor mirror, a dump from another unit, a rescue image). I will flash it,
      look at the picture, and take the three dumps exactly as you wrote them. That
      settles "damaged part" versus "missing init" in one step.
    2. Dump this unit now and publish it. It is only reading, it costs nothing, and
      the three --script dumps over 0x3000-0x5fff would give anyone with a known
      good OV2710 module something to diff against. Say the word and I will attach
      stock_3xxx.sh-style dumps taken from OpenIPC (naming them openipc_*).
    3. ipctool trace on our own streamer records what majestic + the sensor
      library write, in order. That is not the vendor's sequence, but it makes the
      order of our 84 writes explicit, which is the half of the problem a table
      diff cannot see.

    One piece of good news for the other half of the report: your reviewer's plan for
    the ini works on this board. PR OpenIPC/firmware#2871, tested with the default file
    restored and isp.sensorConfig cleared -

    matched sensor config: /etc/sensors/ov2710_mipi_1080p.ini
    matched sensor config: /etc/sensors/ov2710_mipi_1lane_1080p.ini
    Several sensor configs match 'ov2710_mipi'; trying each until one delivers frames
    Using /etc/sensors/ov2710_mipi_1080p.ini sensor configuration file
    Using /etc/sensors/ov2710_mipi_1lane_1080p.ini sensor configuration file
    Sensor config /etc/sensors/ov2710_mipi_1lane_1080p.ini delivers frames
    

    IntfM MIPI, IntCnt 30/s, TmgErr 0, ccErrN 0, 1920x1080 JPEG served. The
    default is tried first and falls through on its own, which is exactly the design
    they described. The noise is unchanged by it, as expected.

  4. zx97 commented on Nov 16, 2026

    @zx97
    Author

    Read-only dumps of this unit, as promised - option 2 from my previous comment. No stock firmware is involved, so this is not a reference sequence: it is the state of a running camera, useful for diffing against another OV2710 module that works.

    Taken while streaming with the one-lane profile from OpenIPC/firmware#2871:

    ipctool i2cdump --script 0x6c 0x3000 0x3fff > openipc_3xxx.sh
    ipctool i2cdump --script 0x6c 0x4000 0x4fff > openipc_4xxx.sh
    ipctool i2cdump --script 0x6c 0x5000 0x5fff > openipc_5xxx.sh

    The stream survived all three (IntCnt kept climbing, IntfM MIPI unchanged), so reading is safe on this part as you said. Only non-zero registers are listed: 173 in 0x3xxx, 78 in 0x4xxx, 432 in 0x5xxx, out of 4096 each. Zeros add nothing to a diff, and the full 4096-line files are available if you want them.

    Registers that decide the readout

    reg running value what
    0x3000 0x20 system
    0x3008 0x2 streaming
    0x300e 0x4 MIPI
    0x300f 0xc3 MIPI
    0x3011 0xa PLL
    0x3030 0x2b MIPI
    0x3034 0x10 PLL
    0x3035 0xff PLL
    0x3036 0x00 PLL
    0x3037 0x3 PLL
    0x3038 0x00 PLL
    0x3501 0x00 exposure mid
    0x3502 0xa0 exposure low
    0x3503 0x7 AEC/AGC
    0x350a 0x00 gain high
    0x350b 0x00 gain low
    0x380c 0x9 HTS H
    0x380d 0x74 HTS L
    0x380e 0x4 VTS H
    0x380f 0x50 VTS L
    0x4000 0x5 BLC
    0x401d 0x2 BLC update
    0x4800 0x4 MIPI ctrl
    0x4801 0xf MIPI lanes
    0x5000 0x5f lens shading
    0x5001 0x4f AWB

    The PLL block reads 0x3034 = 0x10, 0x3035 = 0xff, 0x3036 = 0, 0x3037 = 0x3, 0x3038 = 0 - power-on values, since neither this driver nor
    Allwinner's writes them. If a working module of yours reads something else there,
    that is the answer to Q1: the difference is a value, and it would explain a
    readout that is noisy while the digital test pattern is clean.

    0x3000 to 0x3fff

    173 non-zero registers
    ipctool i2cset 0x6c 0x3000 0x20
    ipctool i2cset 0x6c 0x3004 0xdf
    ipctool i2cset 0x6c 0x3005 0xff
    ipctool i2cset 0x6c 0x3006 0xff
    ipctool i2cset 0x6c 0x3007 0x3f
    ipctool i2cset 0x6c 0x3008 0x2
    ipctool i2cset 0x6c 0x3009 0x1
    ipctool i2cset 0x6c 0x300a 0x27
    ipctool i2cset 0x6c 0x300b 0x10
    ipctool i2cset 0x6c 0x300c 0x1
    ipctool i2cset 0x6c 0x300e 0x4
    ipctool i2cset 0x6c 0x300f 0xc3
    ipctool i2cset 0x6c 0x3011 0xa
    ipctool i2cset 0x6c 0x3012 0x1
    ipctool i2cset 0x6c 0x3022 0xff
    ipctool i2cset 0x6c 0x3023 0x9
    ipctool i2cset 0x6c 0x3025 0x1
    ipctool i2cset 0x6c 0x302d 0x90
    ipctool i2cset 0x6c 0x302e 0x8
    ipctool i2cset 0x6c 0x302f 0x2
    ipctool i2cset 0x6c 0x3030 0x2b
    ipctool i2cset 0x6c 0x3033 0x3
    ipctool i2cset 0x6c 0x3034 0x10
    ipctool i2cset 0x6c 0x3035 0xff
    ipctool i2cset 0x6c 0x3037 0x3
    ipctool i2cset 0x6c 0x3040 0x5
    ipctool i2cset 0x6c 0x3041 0xa
    ipctool i2cset 0x6c 0x3100 0x6c
    ipctool i2cset 0x6c 0x3101 0x3
    ipctool i2cset 0x6c 0x3103 0x93
    ipctool i2cset 0x6c 0x3104 0x1
    ipctool i2cset 0x6c 0x3109 0x3
    ipctool i2cset 0x6c 0x3200 0x40
    ipctool i2cset 0x6c 0x3201 0x4a
    ipctool i2cset 0x6c 0x3202 0x54
    ipctool i2cset 0x6c 0x3203 0x5e
    ipctool i2cset 0x6c 0x3204 0x68
    ipctool i2cset 0x6c 0x3205 0x72
    ipctool i2cset 0x6c 0x3206 0x7c
    ipctool i2cset 0x6c 0x3208 0xf0
    ipctool i2cset 0x6c 0x3209 0xf0
    ipctool i2cset 0x6c 0x320a 0x4c
    ipctool i2cset 0x6c 0x320b 0x2
    ipctool i2cset 0x6c 0x3400 0x4
    ipctool i2cset 0x6c 0x3402 0x4
    ipctool i2cset 0x6c 0x3404 0x4
    ipctool i2cset 0x6c 0x3406 0x1
    ipctool i2cset 0x6c 0x3502 0xa0
    ipctool i2cset 0x6c 0x3503 0x7
    ipctool i2cset 0x6c 0x3600 0x4
    ipctool i2cset 0x6c 0x3601 0x9
    ipctool i2cset 0x6c 0x3602 0xe4
    ipctool i2cset 0x6c 0x3603 0xa7
    ipctool i2cset 0x6c 0x3604 0x60
    ipctool i2cset 0x6c 0x3605 0x5
    ipctool i2cset 0x6c 0x3606 0x3f
    ipctool i2cset 0x6c 0x3610 0x48
    ipctool i2cset 0x6c 0x3611 0xc4
    ipctool i2cset 0x6c 0x3612 0xac
    ipctool i2cset 0x6c 0x3613 0x36
    ipctool i2cset 0x6c 0x3620 0x37
    ipctool i2cset 0x6c 0x3621 0x4
    ipctool i2cset 0x6c 0x3622 0x8
    ipctool i2cset 0x6c 0x3630 0x6d
    ipctool i2cset 0x6c 0x3631 0x26
    ipctool i2cset 0x6c 0x3700 0x4
    ipctool i2cset 0x6c 0x3701 0x4c
    ipctool i2cset 0x6c 0x3702 0x9e
    ipctool i2cset 0x6c 0x3703 0x5c
    ipctool i2cset 0x6c 0x3704 0x40
    ipctool i2cset 0x6c 0x3705 0xda
    ipctool i2cset 0x6c 0x3706 0x61
    ipctool i2cset 0x6c 0x3707 0x52
    ipctool i2cset 0x6c 0x370b 0x40
    ipctool i2cset 0x6c 0x370c 0x20
    ipctool i2cset 0x6c 0x370d 0xf
    ipctool i2cset 0x6c 0x370f 0x40
    ipctool i2cset 0x6c 0x3710 0x9e
    ipctool i2cset 0x6c 0x3711 0x4f
    ipctool i2cset 0x6c 0x3712 0xc
    ipctool i2cset 0x6c 0x3713 0x9f
    ipctool i2cset 0x6c 0x3714 0x4c
    ipctool i2cset 0x6c 0x3715 0x58
    ipctool i2cset 0x6c 0x3716 0x31
    ipctool i2cset 0x6c 0x3717 0x1
    ipctool i2cset 0x6c 0x3718 0xa6
    ipctool i2cset 0x6c 0x3800 0x1
    ipctool i2cset 0x6c 0x3801 0xc4
    ipctool i2cset 0x6c 0x3803 0xa
    ipctool i2cset 0x6c 0x3804 0x7
    ipctool i2cset 0x6c 0x3805 0x80
    ipctool i2cset 0x6c 0x3806 0x4
    ipctool i2cset 0x6c 0x3807 0x38
    ipctool i2cset 0x6c 0x3808 0x7
    ipctool i2cset 0x6c 0x3809 0x80
    ipctool i2cset 0x6c 0x380a 0x4
    ipctool i2cset 0x6c 0x380b 0x38
    ipctool i2cset 0x6c 0x380c 0x9
    ipctool i2cset 0x6c 0x380d 0x74
    ipctool i2cset 0x6c 0x380e 0x4
    ipctool i2cset 0x6c 0x380f 0x50
    ipctool i2cset 0x6c 0x3810 0x10
    ipctool i2cset 0x6c 0x3811 0x6
    ipctool i2cset 0x6c 0x3814 0x1
    ipctool i2cset 0x6c 0x3815 0x1
    ipctool i2cset 0x6c 0x3816 0xa
    ipctool i2cset 0x6c 0x3817 0x24
    ipctool i2cset 0x6c 0x3818 0x80
    ipctool i2cset 0x6c 0x3819 0x80
    ipctool i2cset 0x6c 0x381a 0x1a
    ipctool i2cset 0x6c 0x381c 0x21
    ipctool i2cset 0x6c 0x381d 0x50
    ipctool i2cset 0x6c 0x381e 0x1
    ipctool i2cset 0x6c 0x381f 0x20
    ipctool i2cset 0x6c 0x3822 0x3
    ipctool i2cset 0x6c 0x3823 0x3
    ipctool i2cset 0x6c 0x3824 0x1
    ipctool i2cset 0x6c 0x3825 0x7c
    ipctool i2cset 0x6c 0x3827 0xa
    ipctool i2cset 0x6c 0x3829 0xf
    ipctool i2cset 0x6c 0x382d 0x50
    ipctool i2cset 0x6c 0x382e 0xf
    ipctool i2cset 0x6c 0x3a00 0x78
    ipctool i2cset 0x6c 0x3a01 0x4
    ipctool i2cset 0x6c 0x3a02 0x2
    ipctool i2cset 0x6c 0x3a03 0x28
    ipctool i2cset 0x6c 0x3a05 0x30
    ipctool i2cset 0x6c 0x3a06 0x10
    ipctool i2cset 0x6c 0x3a07 0x18
    ipctool i2cset 0x6c 0x3a08 0x14
    ipctool i2cset 0x6c 0x3a09 0xc0
    ipctool i2cset 0x6c 0x3a0a 0x11
    ipctool i2cset 0x6c 0x3a0b 0x40
    ipctool i2cset 0x6c 0x3a0c 0xe4
    ipctool i2cset 0x6c 0x3a0d 0x4
    ipctool i2cset 0x6c 0x3a0e 0x3
    ipctool i2cset 0x6c 0x3a0f 0x40
    ipctool i2cset 0x6c 0x3a10 0x38
    ipctool i2cset 0x6c 0x3a11 0x90
    ipctool i2cset 0x6c 0x3a13 0x54
    ipctool i2cset 0x6c 0x3a14 0x1
    ipctool i2cset 0x6c 0x3a15 0xf2
    ipctool i2cset 0x6c 0x3a17 0x89
    ipctool i2cset 0x6c 0x3a19 0x7a
    ipctool i2cset 0x6c 0x3a1a 0x6
    ipctool i2cset 0x6c 0x3a1b 0x48
    ipctool i2cset 0x6c 0x3a1c 0x6
    ipctool i2cset 0x6c 0x3a1d 0x18
    ipctool i2cset 0x6c 0x3a1e 0x30
    ipctool i2cset 0x6c 0x3a1f 0x10
    ipctool i2cset 0x6c 0x3a20 0x20
    ipctool i2cset 0x6c 0x3a21 0x78
    ipctool i2cset 0x6c 0x3a25 0x10
    ipctool i2cset 0x6c 0x3a37 0x2
    ipctool i2cset 0x6c 0x3a40 0x7
    ipctool i2cset 0x6c 0x3a41 0x4
    ipctool i2cset 0x6c 0x3a43 0xe0
    ipctool i2cset 0x6c 0x3a47 0x9
    ipctool i2cset 0x6c 0x3a48 0x60
    ipctool i2cset 0x6c 0x3a49 0x8
    ipctool i2cset 0x6c 0x3a4c 0xa0
    ipctool i2cset 0x6c 0x3a4f 0xf0
    ipctool i2cset 0x6c 0x3a51 0x7
    ipctool i2cset 0x6c 0x3a52 0x80
    ipctool i2cset 0x6c 0x3a53 0x4
    ipctool i2cset 0x6c 0x3a60 0x2
    ipctool i2cset 0x6c 0x3e02 0xf
    ipctool i2cset 0x6c 0x3e03 0xdb
    ipctool i2cset 0x6c 0x3e04 0x14
    ipctool i2cset 0x6c 0x3e06 0x3
    ipctool i2cset 0x6c 0x3e07 0x20
    ipctool i2cset 0x6c 0x3e08 0xff
    ipctool i2cset 0x6c 0x3e09 0xfe
    

    0x4000 to 0x4fff

    78 non-zero registers
    ipctool i2cset 0x6c 0x4000 0x5
    ipctool i2cset 0x6c 0x4003 0x40
    ipctool i2cset 0x6c 0x4007 0x10
    ipctool i2cset 0x6c 0x4009 0x20
    ipctool i2cset 0x6c 0x400b 0xed
    ipctool i2cset 0x6c 0x400d 0xcd
    ipctool i2cset 0x6c 0x400f 0xcd
    ipctool i2cset 0x6c 0x4011 0xed
    ipctool i2cset 0x6c 0x4013 0xee
    ipctool i2cset 0x6c 0x4015 0xce
    ipctool i2cset 0x6c 0x4017 0xcf
    ipctool i2cset 0x6c 0x4019 0xee
    ipctool i2cset 0x6c 0x401a 0xff
    ipctool i2cset 0x6c 0x401b 0xff
    ipctool i2cset 0x6c 0x401c 0x8
    ipctool i2cset 0x6c 0x401d 0x2
    ipctool i2cset 0x6c 0x401e 0xff
    ipctool i2cset 0x6c 0x4300 0xff
    ipctool i2cset 0x6c 0x4301 0xff
    ipctool i2cset 0x6c 0x4600 0x80
    ipctool i2cset 0x6c 0x4602 0x4
    ipctool i2cset 0x6c 0x4604 0x3
    ipctool i2cset 0x6c 0x4606 0x2
    ipctool i2cset 0x6c 0x4607 0x10
    ipctool i2cset 0x6c 0x4609 0x2
    ipctool i2cset 0x6c 0x460b 0x37
    ipctool i2cset 0x6c 0x460c 0x20
    ipctool i2cset 0x6c 0x4610 0x4
    ipctool i2cset 0x6c 0x4611 0x10
    ipctool i2cset 0x6c 0x4617 0x9
    ipctool i2cset 0x6c 0x4619 0x8
    ipctool i2cset 0x6c 0x461c 0x1f
    ipctool i2cset 0x6c 0x461d 0xa4
    ipctool i2cset 0x6c 0x4700 0x4
    ipctool i2cset 0x6c 0x4701 0x1
    ipctool i2cset 0x6c 0x4708 0x1
    ipctool i2cset 0x6c 0x4800 0x4
    ipctool i2cset 0x6c 0x4801 0xf
    ipctool i2cset 0x6c 0x4803 0x50
    ipctool i2cset 0x6c 0x4804 0x8d
    ipctool i2cset 0x6c 0x4805 0x10
    ipctool i2cset 0x6c 0x4806 0x28
    ipctool i2cset 0x6c 0x4807 0xa0
    ipctool i2cset 0x6c 0x4808 0x2
    ipctool i2cset 0x6c 0x4809 0x1
    ipctool i2cset 0x6c 0x4810 0xff
    ipctool i2cset 0x6c 0x4811 0xff
    ipctool i2cset 0x6c 0x4814 0x2a
    ipctool i2cset 0x6c 0x4819 0x96
    ipctool i2cset 0x6c 0x481b 0x3c
    ipctool i2cset 0x6c 0x481c 0x1
    ipctool i2cset 0x6c 0x481d 0x86
    ipctool i2cset 0x6c 0x481f 0x3c
    ipctool i2cset 0x6c 0x4821 0x56
    ipctool i2cset 0x6c 0x4823 0x3c
    ipctool i2cset 0x6c 0x4825 0x32
    ipctool i2cset 0x6c 0x4827 0x32
    ipctool i2cset 0x6c 0x4829 0x64
    ipctool i2cset 0x6c 0x482a 0x5
    ipctool i2cset 0x6c 0x482b 0x4
    ipctool i2cset 0x6c 0x482e 0x34
    ipctool i2cset 0x6c 0x4831 0x4
    ipctool i2cset 0x6c 0x4835 0xff
    ipctool i2cset 0x6c 0x4836 0xff
    ipctool i2cset 0x6c 0x4837 0x15
    ipctool i2cset 0x6c 0x4838 0x2
    ipctool i2cset 0x6c 0x483a 0x8
    ipctool i2cset 0x6c 0x483b 0x33
    ipctool i2cset 0x6c 0x483c 0x4f
    ipctool i2cset 0x6c 0x483d 0x10
    ipctool i2cset 0x6c 0x483e 0x6
    ipctool i2cset 0x6c 0x483f 0x14
    ipctool i2cset 0x6c 0x4842 0x1
    ipctool i2cset 0x6c 0x4844 0x1
    ipctool i2cset 0x6c 0x4846 0xf
    ipctool i2cset 0x6c 0x4847 0xff
    ipctool i2cset 0x6c 0x4866 0xff
    ipctool i2cset 0x6c 0x4867 0xfe
    

    0x5000 to 0x5fff

    432 non-zero registers
    ipctool i2cset 0x6c 0x5000 0x5f
    ipctool i2cset 0x6c 0x5001 0x4f
    ipctool i2cset 0x6c 0x5002 0xe0
    ipctool i2cset 0x6c 0x5004 0x8
    ipctool i2cset 0x6c 0x5005 0xdc
    ipctool i2cset 0x6c 0x5006 0xc
    ipctool i2cset 0x6c 0x5007 0xc0
    ipctool i2cset 0x6c 0x5008 0x9
    ipctool i2cset 0x6c 0x5009 0x90
    ipctool i2cset 0x6c 0x500a 0xc
    ipctool i2cset 0x6c 0x500b 0xc0
    ipctool i2cset 0x6c 0x500c 0x9
    ipctool i2cset 0x6c 0x500d 0x90
    ipctool i2cset 0x6c 0x5018 0x8
    ipctool i2cset 0x6c 0x501a 0x6
    ipctool i2cset 0x6c 0x501f 0x3
    ipctool i2cset 0x6c 0x503b 0x9
    ipctool i2cset 0x6c 0x503c 0x90
    ipctool i2cset 0x6c 0x504e 0x10
    ipctool i2cset 0x6c 0x5050 0x7
    ipctool i2cset 0x6c 0x5051 0xa0
    ipctool i2cset 0x6c 0x5052 0x4
    ipctool i2cset 0x6c 0x5053 0x44
    ipctool i2cset 0x6c 0x5180 0x40
    ipctool i2cset 0x6c 0x5181 0x20
    ipctool i2cset 0x6c 0x5182 0x4
    ipctool i2cset 0x6c 0x5183 0x8
    ipctool i2cset 0x6c 0x5184 0x1
    ipctool i2cset 0x6c 0x5185 0xe0
    ipctool i2cset 0x6c 0x5186 0x4
    ipctool i2cset 0x6c 0x5188 0x4
    ipctool i2cset 0x6c 0x518a 0x4
    ipctool i2cset 0x6c 0x518c 0xf0
    ipctool i2cset 0x6c 0x518d 0xf0
    ipctool i2cset 0x6c 0x518e 0xf0
    ipctool i2cset 0x6c 0x5190 0x4
    ipctool i2cset 0x6c 0x5192 0x4
    ipctool i2cset 0x6c 0x5194 0x4
    ipctool i2cset 0x6c 0x5196 0x4
    ipctool i2cset 0x6c 0x5197 0x4
    ipctool i2cset 0x6c 0x5198 0x4
    ipctool i2cset 0x6c 0x519b 0x12
    ipctool i2cset 0x6c 0x519d 0x13
    ipctool i2cset 0x6c 0x519f 0x11
    ipctool i2cset 0x6c 0x51a1 0x7
    ipctool i2cset 0x6c 0x51a2 0xa0
    ipctool i2cset 0x6c 0x51a3 0xe0
    ipctool i2cset 0x6c 0x51a5 0x9
    ipctool i2cset 0x6c 0x51a6 0xc5
    ipctool i2cset 0x6c 0x51a7 0x8b
    ipctool i2cset 0x6c 0x51a9 0xb
    ipctool i2cset 0x6c 0x51aa 0x19
    ipctool i2cset 0x6c 0x51ab 0x68
    ipctool i2cset 0x6c 0x51ad 0x3c
    ipctool i2cset 0x6c 0x51ae 0x53
    ipctool i2cset 0x6c 0x51af 0x18
    ipctool i2cset 0x6c 0x51b1 0x7a
    ipctool i2cset 0x6c 0x51b2 0x53
    ipctool i2cset 0x6c 0x51b3 0x2d
    ipctool i2cset 0x6c 0x51b5 0x7e
    ipctool i2cset 0x6c 0x51b6 0xf7
    ipctool i2cset 0x6c 0x51b7 0x60
    ipctool i2cset 0x6c 0x51b8 0x4
    ipctool i2cset 0x6c 0x51ba 0x4
    ipctool i2cset 0x6c 0x51bc 0x4
    ipctool i2cset 0x6c 0x51be 0x1
    ipctool i2cset 0x6c 0x51c0 0x40
    ipctool i2cset 0x6c 0x51c1 0x20
    ipctool i2cset 0x6c 0x51c2 0x4
    ipctool i2cset 0x6c 0x51c3 0x8
    ipctool i2cset 0x6c 0x51c4 0x1
    ipctool i2cset 0x6c 0x51c5 0xe0
    ipctool i2cset 0x6c 0x51c6 0x4
    ipctool i2cset 0x6c 0x51c8 0x4
    ipctool i2cset 0x6c 0x51ca 0x4
    ipctool i2cset 0x6c 0x51cc 0xf0
    ipctool i2cset 0x6c 0x51cd 0xf0
    ipctool i2cset 0x6c 0x51ce 0xf0
    ipctool i2cset 0x6c 0x51d0 0x4
    ipctool i2cset 0x6c 0x51d2 0x4
    ipctool i2cset 0x6c 0x51d4 0x4
    ipctool i2cset 0x6c 0x51d6 0x4
    ipctool i2cset 0x6c 0x51d7 0x4
    ipctool i2cset 0x6c 0x51d8 0x4
    ipctool i2cset 0x6c 0x51db 0x12
    ipctool i2cset 0x6c 0x51dd 0x13
    ipctool i2cset 0x6c 0x51df 0x11
    ipctool i2cset 0x6c 0x51e1 0x20
    ipctool i2cset 0x6c 0x51e2 0x68
    ipctool i2cset 0x6c 0x51e3 0xa4
    ipctool i2cset 0x6c 0x51e6 0x7a
    ipctool i2cset 0x6c 0x51e7 0x9c
    ipctool i2cset 0x6c 0x51e9 0x3
    ipctool i2cset 0x6c 0x51ea 0x5b
    ipctool i2cset 0x6c 0x51eb 0xaf
    ipctool i2cset 0x6c 0x51ed 0x77
    ipctool i2cset 0x6c 0x51ee 0x29
    ipctool i2cset 0x6c 0x51ef 0x6
    ipctool i2cset 0x6c 0x51f1 0x87
    ipctool i2cset 0x6c 0x51f2 0x49
    ipctool i2cset 0x6c 0x51f3 0xe2
    ipctool i2cset 0x6c 0x51f5 0x85
    ipctool i2cset 0x6c 0x51f6 0x5b
    ipctool i2cset 0x6c 0x51f7 0x8b
    ipctool i2cset 0x6c 0x51f8 0x4
    ipctool i2cset 0x6c 0x51fa 0x4
    ipctool i2cset 0x6c 0x51fc 0x4
    ipctool i2cset 0x6c 0x51fe 0x1
    ipctool i2cset 0x6c 0x5684 0x7
    ipctool i2cset 0x6c 0x5685 0xa0
    ipctool i2cset 0x6c 0x5686 0x4
    ipctool i2cset 0x6c 0x5687 0x43
    ipctool i2cset 0x6c 0x5688 0x3
    ipctool i2cset 0x6c 0x5689 0x1
    ipctool i2cset 0x6c 0x5690 0x4
    ipctool i2cset 0x6c 0x5780 0x7f
    ipctool i2cset 0x6c 0x5781 0x20
    ipctool i2cset 0x6c 0x5782 0x18
    ipctool i2cset 0x6c 0x5783 0x8
    ipctool i2cset 0x6c 0x5784 0x4
    ipctool i2cset 0x6c 0x5785 0x40
    ipctool i2cset 0x6c 0x5786 0x18
    ipctool i2cset 0x6c 0x5787 0x8
    ipctool i2cset 0x6c 0x5788 0x4
    ipctool i2cset 0x6c 0x5789 0x8
    ipctool i2cset 0x6c 0x578a 0x20
    ipctool i2cset 0x6c 0x578b 0x7
    ipctool i2cset 0x6c 0x578d 0x40
    ipctool i2cset 0x6c 0x578e 0x40
    ipctool i2cset 0x6c 0x5791 0x8
    ipctool i2cset 0x6c 0x5793 0x18
    ipctool i2cset 0x6c 0x5795 0x80
    ipctool i2cset 0x6c 0x5796 0x1
    ipctool i2cset 0x6c 0x579b 0x20
    ipctool i2cset 0x6c 0x579c 0x40
    ipctool i2cset 0x6c 0x579f 0x8
    ipctool i2cset 0x6c 0x57a0 0x7f
    ipctool i2cset 0x6c 0x57a1 0x20
    ipctool i2cset 0x6c 0x57a2 0x18
    ipctool i2cset 0x6c 0x57a3 0x8
    ipctool i2cset 0x6c 0x57a4 0x4
    ipctool i2cset 0x6c 0x57a5 0x40
    ipctool i2cset 0x6c 0x57a6 0x18
    ipctool i2cset 0x6c 0x57a7 0x8
    ipctool i2cset 0x6c 0x57a8 0x4
    ipctool i2cset 0x6c 0x57a9 0x8
    ipctool i2cset 0x6c 0x57aa 0x20
    ipctool i2cset 0x6c 0x57ab 0x7
    ipctool i2cset 0x6c 0x57ad 0x40
    ipctool i2cset 0x6c 0x57ae 0x40
    ipctool i2cset 0x6c 0x57b1 0x8
    ipctool i2cset 0x6c 0x57b3 0x18
    ipctool i2cset 0x6c 0x57b5 0x80
    ipctool i2cset 0x6c 0x57b6 0x1
    ipctool i2cset 0x6c 0x57bb 0x20
    ipctool i2cset 0x6c 0x57bc 0x40
    ipctool i2cset 0x6c 0x57bf 0x8
    ipctool i2cset 0x6c 0x57c0 0x7f
    ipctool i2cset 0x6c 0x57c1 0x20
    ipctool i2cset 0x6c 0x57c2 0x18
    ipctool i2cset 0x6c 0x57c3 0x8
    ipctool i2cset 0x6c 0x57c4 0x4
    ipctool i2cset 0x6c 0x57c5 0x40
    ipctool i2cset 0x6c 0x57c6 0x18
    ipctool i2cset 0x6c 0x57c7 0x8
    ipctool i2cset 0x6c 0x57c8 0x4
    ipctool i2cset 0x6c 0x57c9 0x8
    ipctool i2cset 0x6c 0x57ca 0x20
    ipctool i2cset 0x6c 0x57cb 0x7
    ipctool i2cset 0x6c 0x57cd 0x40
    ipctool i2cset 0x6c 0x57ce 0x40
    ipctool i2cset 0x6c 0x57d1 0x8
    ipctool i2cset 0x6c 0x57d3 0x18
    ipctool i2cset 0x6c 0x57d5 0x80
    ipctool i2cset 0x6c 0x57d6 0x1
    ipctool i2cset 0x6c 0x57db 0x20
    ipctool i2cset 0x6c 0x57dc 0x40
    ipctool i2cset 0x6c 0x57df 0x8
    ipctool i2cset 0x6c 0x57e0 0x7f
    ipctool i2cset 0x6c 0x57e1 0x20
    ipctool i2cset 0x6c 0x57e2 0x18
    ipctool i2cset 0x6c 0x57e3 0x8
    ipctool i2cset 0x6c 0x57e4 0x4
    ipctool i2cset 0x6c 0x57e5 0x40
    ipctool i2cset 0x6c 0x57e6 0x18
    ipctool i2cset 0x6c 0x57e7 0x8
    ipctool i2cset 0x6c 0x57e8 0x4
    ipctool i2cset 0x6c 0x57e9 0x8
    ipctool i2cset 0x6c 0x57ea 0x20
    ipctool i2cset 0x6c 0x57eb 0x7
    ipctool i2cset 0x6c 0x57ed 0x40
    ipctool i2cset 0x6c 0x57ee 0x40
    ipctool i2cset 0x6c 0x57f1 0x8
    ipctool i2cset 0x6c 0x57f3 0x18
    ipctool i2cset 0x6c 0x57f5 0x80
    ipctool i2cset 0x6c 0x57f6 0x1
    ipctool i2cset 0x6c 0x57fb 0x20
    ipctool i2cset 0x6c 0x57fc 0x40
    ipctool i2cset 0x6c 0x57ff 0x8
    ipctool i2cset 0x6c 0x5800 0x3
    ipctool i2cset 0x6c 0x5801 0xc
    ipctool i2cset 0x6c 0x5802 0x3
    ipctool i2cset 0x6c 0x5803 0x6
    ipctool i2cset 0x6c 0x5804 0x22
    ipctool i2cset 0x6c 0x5805 0x7
    ipctool i2cset 0x6c 0x5806 0xc2
    ipctool i2cset 0x6c 0x5807 0x8
    ipctool i2cset 0x6c 0x5808 0x3
    ipctool i2cset 0x6c 0x5809 0xc
    ipctool i2cset 0x6c 0x580a 0x3
    ipctool i2cset 0x6c 0x580b 0x6
    ipctool i2cset 0x6c 0x580c 0x22
    ipctool i2cset 0x6c 0x580d 0x7
    ipctool i2cset 0x6c 0x580e 0xc2
    ipctool i2cset 0x6c 0x580f 0x8
    ipctool i2cset 0x6c 0x5810 0x3
    ipctool i2cset 0x6c 0x5811 0xc
    ipctool i2cset 0x6c 0x5812 0x3
    ipctool i2cset 0x6c 0x5813 0x6
    ipctool i2cset 0x6c 0x5814 0x22
    ipctool i2cset 0x6c 0x5815 0x7
    ipctool i2cset 0x6c 0x5816 0xc2
    ipctool i2cset 0x6c 0x5817 0x8
    ipctool i2cset 0x6c 0x5818 0x4
    ipctool i2cset 0x6c 0x5819 0x80
    ipctool i2cset 0x6c 0x581a 0x6
    ipctool i2cset 0x6c 0x581b 0xc
    ipctool i2cset 0x6c 0x581c 0x80
    ipctool i2cset 0x6c 0x5820 0x3
    ipctool i2cset 0x6c 0x5821 0xc
    ipctool i2cset 0x6c 0x5822 0x3
    ipctool i2cset 0x6c 0x5823 0x6
    ipctool i2cset 0x6c 0x5824 0x22
    ipctool i2cset 0x6c 0x5825 0x7
    ipctool i2cset 0x6c 0x5826 0xc2
    ipctool i2cset 0x6c 0x5827 0x8
    ipctool i2cset 0x6c 0x5828 0x3
    ipctool i2cset 0x6c 0x5829 0xc
    ipctool i2cset 0x6c 0x582a 0x3
    ipctool i2cset 0x6c 0x582b 0x6
    ipctool i2cset 0x6c 0x582c 0x22
    ipctool i2cset 0x6c 0x582d 0x7
    ipctool i2cset 0x6c 0x582e 0xc2
    ipctool i2cset 0x6c 0x582f 0x8
    ipctool i2cset 0x6c 0x5830 0x3
    ipctool i2cset 0x6c 0x5831 0xc
    ipctool i2cset 0x6c 0x5832 0x3
    ipctool i2cset 0x6c 0x5833 0x6
    ipctool i2cset 0x6c 0x5834 0x22
    ipctool i2cset 0x6c 0x5835 0x7
    ipctool i2cset 0x6c 0x5836 0xc2
    ipctool i2cset 0x6c 0x5837 0x8
    ipctool i2cset 0x6c 0x5838 0x4
    ipctool i2cset 0x6c 0x5839 0x80
    ipctool i2cset 0x6c 0x583a 0x6
    ipctool i2cset 0x6c 0x583b 0xc
    ipctool i2cset 0x6c 0x583c 0x80
    ipctool i2cset 0x6c 0x5840 0x3
    ipctool i2cset 0x6c 0x5841 0xc
    ipctool i2cset 0x6c 0x5842 0x3
    ipctool i2cset 0x6c 0x5843 0x6
    ipctool i2cset 0x6c 0x5844 0x22
    ipctool i2cset 0x6c 0x5845 0x7
    ipctool i2cset 0x6c 0x5846 0xc2
    ipctool i2cset 0x6c 0x5847 0x8
    ipctool i2cset 0x6c 0x5848 0x3
    ipctool i2cset 0x6c 0x5849 0xc
    ipctool i2cset 0x6c 0x584a 0x3
    ipctool i2cset 0x6c 0x584b 0x6
    ipctool i2cset 0x6c 0x584c 0x22
    ipctool i2cset 0x6c 0x584d 0x7
    ipctool i2cset 0x6c 0x584e 0xc2
    ipctool i2cset 0x6c 0x584f 0x8
    ipctool i2cset 0x6c 0x5850 0x3
    ipctool i2cset 0x6c 0x5851 0xc
    ipctool i2cset 0x6c 0x5852 0x3
    ipctool i2cset 0x6c 0x5853 0x6
    ipctool i2cset 0x6c 0x5854 0x22
    ipctool i2cset 0x6c 0x5855 0x7
    ipctool i2cset 0x6c 0x5856 0xc2
    ipctool i2cset 0x6c 0x5857 0x8
    ipctool i2cset 0x6c 0x5858 0x4
    ipctool i2cset 0x6c 0x5859 0x80
    ipctool i2cset 0x6c 0x585a 0x6
    ipctool i2cset 0x6c 0x585b 0xc
    ipctool i2cset 0x6c 0x585c 0x80
    ipctool i2cset 0x6c 0x5860 0x3
    ipctool i2cset 0x6c 0x5861 0xc
    ipctool i2cset 0x6c 0x5862 0x3
    ipctool i2cset 0x6c 0x5863 0x6
    ipctool i2cset 0x6c 0x5864 0x22
    ipctool i2cset 0x6c 0x5865 0x7
    ipctool i2cset 0x6c 0x5866 0xc2
    ipctool i2cset 0x6c 0x5867 0x8
    ipctool i2cset 0x6c 0x5868 0x3
    ipctool i2cset 0x6c 0x5869 0xc
    ipctool i2cset 0x6c 0x586a 0x3
    ipctool i2cset 0x6c 0x586b 0x6
    ipctool i2cset 0x6c 0x586c 0x22
    ipctool i2cset 0x6c 0x586d 0x7
    ipctool i2cset 0x6c 0x586e 0xc2
    ipctool i2cset 0x6c 0x586f 0x8
    ipctool i2cset 0x6c 0x5870 0x3
    ipctool i2cset 0x6c 0x5871 0xc
    ipctool i2cset 0x6c 0x5872 0x3
    ipctool i2cset 0x6c 0x5873 0x6
    ipctool i2cset 0x6c 0x5874 0x22
    ipctool i2cset 0x6c 0x5875 0x7
    ipctool i2cset 0x6c 0x5876 0xc2
    ipctool i2cset 0x6c 0x5877 0x8
    ipctool i2cset 0x6c 0x5878 0x4
    ipctool i2cset 0x6c 0x5879 0x80
    ipctool i2cset 0x6c 0x587a 0x6
    ipctool i2cset 0x6c 0x587b 0xc
    ipctool i2cset 0x6c 0x587c 0x80
    ipctool i2cset 0x6c 0x5880 0x3
    ipctool i2cset 0x6c 0x5881 0xc
    ipctool i2cset 0x6c 0x5882 0x3
    ipctool i2cset 0x6c 0x5883 0x6
    ipctool i2cset 0x6c 0x5884 0x22
    ipctool i2cset 0x6c 0x5885 0x7
    ipctool i2cset 0x6c 0x5886 0xc2
    ipctool i2cset 0x6c 0x5887 0x8
    ipctool i2cset 0x6c 0x5888 0x3
    ipctool i2cset 0x6c 0x5889 0xc
    ipctool i2cset 0x6c 0x588a 0x3
    ipctool i2cset 0x6c 0x588b 0x6
    ipctool i2cset 0x6c 0x588c 0x22
    ipctool i2cset 0x6c 0x588d 0x7
    ipctool i2cset 0x6c 0x588e 0xc2
    ipctool i2cset 0x6c 0x588f 0x8
    ipctool i2cset 0x6c 0x5890 0x3
    ipctool i2cset 0x6c 0x5891 0xc
    ipctool i2cset 0x6c 0x5892 0x3
    ipctool i2cset 0x6c 0x5893 0x6
    ipctool i2cset 0x6c 0x5894 0x22
    ipctool i2cset 0x6c 0x5895 0x7
    ipctool i2cset 0x6c 0x5896 0xc2
    ipctool i2cset 0x6c 0x5897 0x8
    ipctool i2cset 0x6c 0x5898 0x4
    ipctool i2cset 0x6c 0x5899 0x80
    ipctool i2cset 0x6c 0x589a 0x6
    ipctool i2cset 0x6c 0x589b 0xc
    ipctool i2cset 0x6c 0x589c 0x80
    ipctool i2cset 0x6c 0x58a0 0x3
    ipctool i2cset 0x6c 0x58a1 0xc
    ipctool i2cset 0x6c 0x58a2 0x3
    ipctool i2cset 0x6c 0x58a3 0x6
    ipctool i2cset 0x6c 0x58a4 0x22
    ipctool i2cset 0x6c 0x58a5 0x7
    ipctool i2cset 0x6c 0x58a6 0xc2
    ipctool i2cset 0x6c 0x58a7 0x8
    ipctool i2cset 0x6c 0x58a8 0x3
    ipctool i2cset 0x6c 0x58a9 0xc
    ipctool i2cset 0x6c 0x58aa 0x3
    ipctool i2cset 0x6c 0x58ab 0x6
    ipctool i2cset 0x6c 0x58ac 0x22
    ipctool i2cset 0x6c 0x58ad 0x7
    ipctool i2cset 0x6c 0x58ae 0xc2
    ipctool i2cset 0x6c 0x58af 0x8
    ipctool i2cset 0x6c 0x58b0 0x3
    ipctool i2cset 0x6c 0x58b1 0xc
    ipctool i2cset 0x6c 0x58b2 0x3
    ipctool i2cset 0x6c 0x58b3 0x6
    ipctool i2cset 0x6c 0x58b4 0x22
    ipctool i2cset 0x6c 0x58b5 0x7
    ipctool i2cset 0x6c 0x58b6 0xc2
    ipctool i2cset 0x6c 0x58b7 0x8
    ipctool i2cset 0x6c 0x58b8 0x4
    ipctool i2cset 0x6c 0x58b9 0x80
    ipctool i2cset 0x6c 0x58ba 0x6
    ipctool i2cset 0x6c 0x58bb 0xc
    ipctool i2cset 0x6c 0x58bc 0x80
    ipctool i2cset 0x6c 0x58c0 0x3
    ipctool i2cset 0x6c 0x58c1 0xc
    ipctool i2cset 0x6c 0x58c2 0x3
    ipctool i2cset 0x6c 0x58c3 0x6
    ipctool i2cset 0x6c 0x58c4 0x22
    ipctool i2cset 0x6c 0x58c5 0x7
    ipctool i2cset 0x6c 0x58c6 0xc2
    ipctool i2cset 0x6c 0x58c7 0x8
    ipctool i2cset 0x6c 0x58c8 0x3
    ipctool i2cset 0x6c 0x58c9 0xc
    ipctool i2cset 0x6c 0x58ca 0x3
    ipctool i2cset 0x6c 0x58cb 0x6
    ipctool i2cset 0x6c 0x58cc 0x22
    ipctool i2cset 0x6c 0x58cd 0x7
    ipctool i2cset 0x6c 0x58ce 0xc2
    ipctool i2cset 0x6c 0x58cf 0x8
    ipctool i2cset 0x6c 0x58d0 0x3
    ipctool i2cset 0x6c 0x58d1 0xc
    ipctool i2cset 0x6c 0x58d2 0x3
    ipctool i2cset 0x6c 0x58d3 0x6
    ipctool i2cset 0x6c 0x58d4 0x22
    ipctool i2cset 0x6c 0x58d5 0x7
    ipctool i2cset 0x6c 0x58d6 0xc2
    ipctool i2cset 0x6c 0x58d7 0x8
    ipctool i2cset 0x6c 0x58d8 0x4
    ipctool i2cset 0x6c 0x58d9 0x80
    ipctool i2cset 0x6c 0x58da 0x6
    ipctool i2cset 0x6c 0x58db 0xc
    ipctool i2cset 0x6c 0x58dc 0x80
    ipctool i2cset 0x6c 0x58e0 0x3
    ipctool i2cset 0x6c 0x58e1 0xc
    ipctool i2cset 0x6c 0x58e2 0x3
    ipctool i2cset 0x6c 0x58e3 0x6
    ipctool i2cset 0x6c 0x58e4 0x22
    ipctool i2cset 0x6c 0x58e5 0x7
    ipctool i2cset 0x6c 0x58e6 0xc2
    ipctool i2cset 0x6c 0x58e7 0x8
    ipctool i2cset 0x6c 0x58e8 0x3
    ipctool i2cset 0x6c 0x58e9 0xc
    ipctool i2cset 0x6c 0x58ea 0x3
    ipctool i2cset 0x6c 0x58eb 0x6
    ipctool i2cset 0x6c 0x58ec 0x22
    ipctool i2cset 0x6c 0x58ed 0x7
    ipctool i2cset 0x6c 0x58ee 0xc2
    ipctool i2cset 0x6c 0x58ef 0x8
    ipctool i2cset 0x6c 0x58f0 0x3
    ipctool i2cset 0x6c 0x58f1 0xc
    ipctool i2cset 0x6c 0x58f2 0x3
    ipctool i2cset 0x6c 0x58f3 0x6
    ipctool i2cset 0x6c 0x58f4 0x22
    ipctool i2cset 0x6c 0x58f5 0x7
    ipctool i2cset 0x6c 0x58f6 0xc2
    ipctool i2cset 0x6c 0x58f7 0x8
    ipctool i2cset 0x6c 0x58f8 0x4
    ipctool i2cset 0x6c 0x58f9 0x80
    ipctool i2cset 0x6c 0x58fa 0x6
    ipctool i2cset 0x6c 0x58fb 0xc
    ipctool i2cset 0x6c 0x58fc 0x80
    ipctool i2cset 0x6c 0x5900 0x1
    

    For the record, option 3 is still open on my side: ipctool trace on the streamer would give the order of our 84 writes, which no dump shows. Say if you want it rather than the raw dumps.

  5. openipc-ai commented on Nov 16, 2026

    @openipc-ai
    Contributor

    Thanks. The chip ID settles that it is an OV2710. And thanks for being upfront about the backup: no stock image means we need a different reference, and there is a good one.

    We don't have a C23S-class stock image we can give you, so option 1 is out from our side too. On your other two options:

    • Option 2: yes, take the OpenIPC-side dumps. Take them before and after the test below so the two can be compared.
    • Option 3: skip ipctool trace on our own streamer. Our 84 writes are a straight list in ov2710_sensor_ctl.c, so their order is already visible there.

    A reference for your exact operating point. Espressif ships ov2710_mipi_1lane_24Minput_1920x1080_raw10_25fps.h (esp-video-components @ 3620887, Apache-2.0). It is 1 lane, 24 MHz, RAW10. I diffed it against our init, comparing the final value of each register for the MIPI branch. It comes from the same OmniVision reference: the analogue block (0x36xx/0x37xx), the clock setup (0x3103, 0x300e/0x300f/0x3011) and 0x401d match value for value. Three groups differ:

    registers
    only ours manual AWB at 1x (0x3406 = 0x01, 0x3400..0x3405), manual AEC/AGC (0x3503 = 0x07) with its starting exposure and gain, 0x4000 = 0x05, 0x5000 = 0x5f, 0x5001 = 0x4f, 0x302c = 0x00, 0x3030 = 0x2b, 0x3010 = 0x00
    only Espressif's the explicit 1080p window 0x3800..0x3813, 0x3818 = 0x80, VTS 0x052a (1322, 25 fps), frame control 0x4201/0x4202, 0x4800 = 0x24
    both everything else

    The table starts with a soft reset (0x3008 = 0x82), so one run replaces every value our init wrote. That makes one A/B of the whole table, not more blind pokes.

    The test, with majestic running and streaming as now (do it from the serial console, given the earlier watchdog spin):

    1. Take a snapshot and record the frame mean and 8x8 block noise, as before.
    2. Run the script below. It is Espressif's table converted 1:1 to ipctool i2cset, followed by the stream-on writes their driver uses (0x4201/0x4202 = 0, 0x3008 = 0x02).
    3. In /proc/umap/vi, IntCnt should keep climbing, now at about 25/s with IntGapT near 40000 us. If it stops, reboot restores our init.
    4. Take a snapshot and measure the same two numbers.

    With 0x3503 back at its default, the sensor runs its own AEC/AGC and majestic's exposure writes may be ignored. That is fine for this test; just report the frame mean alongside the noise.

    Reading the result:

    • Noise drops: one of our extra writes causes it. There are eight candidates. Re-apply them one at a time on top of Espressif's table, and the culprit goes into a patch to ov2710_sensor_ctl.c.
    • Noise unchanged: the register set now matches a reference table at this exact operating point, so the init is ruled out. What remains is hardware: the module, its analogue supply (AVDD 2.8 V, which a scope on the module connector can show), or the part itself.
    esp_1080p_1lane.sh (90 writes)
    #!/bin/sh
    # Espressif esp-video-components @3620887 (Apache-2.0),
    # ov2710_mipi_1lane_24Minput_1920x1080_raw10_25fps.h, written at runtime.
    i2c() { ipctool i2cset 0x6c "$1" "$2" >/dev/null; }
    i2c 0x3008 0x82
    usleep 4000
    i2c 0x4800 0x1
    i2c 0x3008 0x42
    usleep 4000
    i2c 0x4201 0x00
    i2c 0x4202 0x0f
    i2c 0x3103 0x93
    i2c 0x3017 0x7f
    i2c 0x3018 0xfc
    i2c 0x3706 0x61
    i2c 0x3712 0x0c
    i2c 0x3630 0x6d
    i2c 0x3800 0x01
    i2c 0x3801 0xb4
    i2c 0x3802 0x00
    i2c 0x3803 0x0a
    i2c 0x3818 0x80
    i2c 0x3804 0x07
    i2c 0x3805 0x80
    i2c 0x3806 0x04
    i2c 0x3807 0x38
    i2c 0x3808 0x07
    i2c 0x3809 0x80
    i2c 0x380a 0x04
    i2c 0x380b 0x38
    i2c 0x3810 0x10
    i2c 0x3811 0x06
    i2c 0x3812 0x00
    i2c 0x3813 0x00
    i2c 0x3621 0x04
    i2c 0x3604 0x60
    i2c 0x3603 0xa7
    i2c 0x3631 0x26
    i2c 0x3600 0x04
    i2c 0x3620 0x37
    i2c 0x3623 0x00
    i2c 0x3702 0x9e
    i2c 0x3703 0x5c
    i2c 0x3704 0x40
    i2c 0x370d 0x0f
    i2c 0x3713 0x9f
    i2c 0x3714 0x4c
    i2c 0x3710 0x9e
    i2c 0x3801 0xc4
    i2c 0x3605 0x05
    i2c 0x3606 0x3f
    i2c 0x302d 0x90
    i2c 0x370b 0x40
    i2c 0x3716 0x31
    i2c 0x3707 0x52
    i2c 0x380d 0x74
    i2c 0x5181 0x20
    i2c 0x518f 0x00
    i2c 0x4301 0xff
    i2c 0x4303 0x00
    i2c 0x3a00 0x78
    i2c 0x300f 0x88
    i2c 0x3011 0x28
    i2c 0x3a1a 0x06
    i2c 0x3a18 0x00
    i2c 0x3a19 0x7a
    i2c 0x3a13 0x54
    i2c 0x382e 0x0f
    i2c 0x381a 0x1a
    i2c 0x401d 0x02
    i2c 0x5688 0x03
    i2c 0x5684 0x07
    i2c 0x5685 0xa0
    i2c 0x5686 0x04
    i2c 0x5687 0x43
    i2c 0x3011 0x0a
    i2c 0x300f 0x8a
    i2c 0x3017 0x00
    i2c 0x3018 0x00
    i2c 0x300e 0x04
    i2c 0x4801 0x0f
    i2c 0x300f 0xc3
    i2c 0x3a0f 0x40
    i2c 0x3a10 0x38
    i2c 0x3a1b 0x48
    i2c 0x3a1e 0x30
    i2c 0x3a11 0x90
    i2c 0x3a1f 0x10
    i2c 0x380c 0x09
    i2c 0x380d 0x74
    i2c 0x380e 0x05
    i2c 0x380f 0x2a
    i2c 0x4800 0x24
    usleep 2000
    # stream on, as ov2710_set_stream() does
    i2c 0x4201 0x00
    i2c 0x4202 0x00
    i2c 0x3008 0x02
  6. zx97 commented on Nov 16, 2026

    @zx97
    Author

    A/B done, exactly as you specified. The noise did not change: the init is ruled out.

    Method: with majestic running and streaming (the one-lane profile from
    OpenIPC/firmware#2871, auto-probed), snapshot and measure, run your
    esp_1080p_1lane.sh in full (90 writes, usleep present on this image), wait
    six seconds, snapshot and measure again.

    frame mean Y 8x8 block noise sigma/mean
    before, our init 108.8 67.9 0.62
    after, Espressif's table 154.7 94.4 0.61

    The picture is brighter under Espressif's table, and the noise scaled with it:
    sigma/mean is 0.62 before and 0.61 after. Same figure, same scene, same lens.

    The writes took: after the run the part reads 0x3503 = 0x00 (their table does
    not touch it, so it is back to its reset default - AEC and AGC automatic),
    0x4800 = 0x24, 0x4000 = 0x01, 0x3030 = 0x0b. The stream never stopped:
    IntCnt kept climbing across the run, and GET /image.jpg answered throughout.

    So, by your own reading of the result: the register set now matches a reference
    table for this exact operating point (1 lane, 24 MHz, RAW10, OV2710), and the
    noise is unchanged. The init is ruled out. What is left is what you named:
    the module, its analogue supply, or the part.

    Two things I can still do on this side, if either is useful to you:

    • Take the three --script dumps in this post-Espressif state and publish
      them next to the pre-test ones I posted above. They would be a reference of
      "this part, configured exactly as Espressif configures it, still noisy" -
      which is a useful thing for someone to diff against a healthy module.
    • Measure AVDD. I do not have a scope here, but I can report what the module
      connector gives with a multimeter, and the supply rails are also readable in
      the sensor's own black-level registers if a wrong rail shows up there.

    I have put the board back on its own init (a S95majestic restart re-runs
    ov2710_sensor_ctl.c), so it is in the state it was before the test.

    One operational note for anyone repeating this: writing a table at runtime while
    the streamer runs is safe, but do not follow it with a burst of S95majestic
    restarts. After about ten of them the VPSS to VENC binding went stale here,
    majestic logged Timeout from venc channel 0 every two seconds, and
    /cgi-bin/live.cgi answered 500 while /image.jpg and /mjpeg stayed 200 - the
    H.264 channel was dead, the JPEG paths were fine. A reboot rebuilt the binding;
    no configuration change was needed.

  7. openipc-ai commented on Nov 16, 2026

    @openipc-ai
    Contributor

    Thanks, that's a clean result. Same sigma/mean before and after on a register set matching the reference table rules the init out. And the read-back (0x3503, 0x4000, 0x3030, 0x4800) shows the table really landed.

    A correction to my last comment first: the supply voltage I gave was wrong. The OV2710 datasheet gives the analogue supply (VDD-A) as 3.0–3.6 V, 3.3 V typical, not 2.8 V. The other rails are I/O 1.7–3.6 V (1.8 V typical) and core 1.425–1.575 V (1.5 V typical; the part has a built-in 1.5 V regulator for the core). That makes the multimeter reading worth taking, scope or not:

    • AVDD on the module connector, measured while streaming. Many sensors in this class run their analogue rail at 2.8 V. If this board was laid out for one of them and gives the OV2710 2.8 V, the part is under spec. That would fit what you see: noise generated only in the analogue readout, with a perfectly clean test pattern.
    • AVDD within 3.0–3.6 V: the supply is ruled out as far as a DC reading can tell. What's left is the module or the part, and no software change will fix that.

    A multimeter is enough for the DC level, which is the question here. I wouldn't try to infer the rails from the black-level registers: they reflect the readout, not the supply.

    The dumps in the post-Espressif state aren't needed. With the init ruled out they don't add anything to diff against.

    The datasheet backs up your earlier measurements. It lists the MIPI output as one lane (up to 800 Mbps) and RAW output as 8/10-bit only, which matches lane 0 and RAW10 being the only settings that streamed.

    The stale VPSS→VENC binding after repeated S95majestic restarts is a streamer bug, not a sensor one. Please file it on OpenIPC/majestic. Include the number of restarts, Timeout from venc channel 0 every two seconds, /cgi-bin/live.cgi answering 500 while /image.jpg and /mjpeg stay at 200, and that a reboot clears it.

    Once you have the AVDD reading I'll close this either way: as a supply problem on this board, or as a faulty module or part.

  8. zx97 commented on Nov 16, 2026

    @zx97
    Author

    Correction taken, and confirmed in the same datasheet before I put a probe on anything:

    spec summary     : analog: 3.0~3.6V (3.3V typical)
    note 3           : "AVDD is 3.0~3.6V of sensor analog power (clean). 3.3V is typical."
    power section    : "Image sensor analog power is provided from a 3.3V (typical)
                        power circuit coming from the system."
    reference BOM    : XC62FP3302-SOT89 3.3V
    

    So the reading I will take is against 3.0-3.6 V, and a 2.8 V rail would be
    the answer rather than a pass. Worth saying plainly because 2.8 V is exactly what
    a board laid out for a neighbouring sensor would give, and I would otherwise have
    read it as "correct".

    No post-Espressif dumps then.

    The stale VPSS to VENC binding is filed as OpenIPC/majestic#377, with the restart
    count, the two-second Timeout from venc channel 0, the 500-versus-200 split, the
    venc0_encoded_frames_total evidence, and the fact that a reboot clears it. I
    also noted there what happens if isp.sensorConfig is set explicitly instead of
    letting the probe run - load average 8.78 and both endpoints dead - and offered to
    file that separately if you would rather keep #377 narrow. The Issues link in
    majestic's README points at OpenIPC/majestic_issues, which returns 404, which is
    why it is filed on the repository.

    I will come back with the AVDD figure, and if it is in range I will say so
    plainly so this can be closed as a faulty module.

  9. zx97 commented on Nov 16, 2026

    @zx97
    Author

    AVDD measured, with a multimeter, on the module board, camera streaming. Good
    ground established first (the mounting screws are not all grounded on this
    board - the one next to CD1 is the one that reads 0 ohm to the plane; worth
    knowing for anyone repeating this).

    regulator output rail spec (datasheet) verdict
    U2 3.33 V AVDD (analogue) 3.0-3.6 V in spec
    U3 1.48 V DVDD (core) 1.425-1.575 V in spec
    U1 1.78 V DOVDD (I/O) 1.7-3.6 V in spec

    So the analogue rail is at 3.33 V, comfortably inside 3.0-3.6 V, and nothing on
    this board is under-volting the part. By your reading of the result, the supply
    is ruled out and what remains is the module or the part.

    For the record, that is now the complete set:

    • link, lanes, packing, CFA, ISP and encoder: clean, proven by the part's own
      test pattern;
    • init table: ruled out, the A/B against Espressif's table at this exact
      operating point gave an identical sigma/mean;
    • supply: ruled out, 3.33 V measured at the regulator;
    • chip ID: answers as OV2710 (0x300a/0x300b = 0x27/0x10).

    Nothing in software is left that I can test on this unit. I will take the module
    off and put a replacement in when I can get one, and come back with the result -
    if a fresh module is clean on the same board, that closes it as a faulty part,
    and if it is noisy too, the board itself is the last suspect and that is a
    different conversation.

    Close it as you offered. Thank you for the time - the test-pattern idea and the
    Espressif reference are what made the search converge this fast.

  10. zx97 commented on Nov 16, 2026

    @zx97
    Author

    I need to withdraw the "module or part" inference I closed on, because the board's
    owner pointed out something I had not weighed: the stock firmware produced a
    clean picture on this exact hardware.
    The camera streamed properly before
    OpenIPC was installed. So the part was healthy at flash time, and a conclusion of
    "faulty module" does not follow from my measurements.

    What the evidence actually supports is narrower, and I should have said it this
    way:

    • everything from the sensor's digital output onwards is clean (test pattern);
    • the init table is not the difference (A/B against Espressif, identical
      sigma/mean);
    • the analogue rail is not the difference (3.33 V, in spec);
    • the image tuning is: the stock firmware shipped an IQ profile made for this
      module, and OpenIPC drives it with a generic one.

    Concretely, on this image majestic loads /etc/sensors/iq/default.ini, which is
    a symlink to ar0130.ini - an APTINA sensor's profile. The ov2710.ini in the
    same directory is not the answer either: its md5 is identical to ov9750.ini, so
    it is a copy of another OmniVision part's tuning, and loading it made the picture
    worse (sigma/mean 0.62 -> 1.14).

    So the remaining gap is a missing IQ profile for this sensor on this SoC, not a
    broken component. That is a different problem, and one that could be solved by
    tuning rather than by soldering.

    Two measurements that bear on it, in case they are useful to someone:

    downscaling the same frame   1920x1080 sigma/mean 0.62
                                 1280x720           0.50
                                  960x540           0.39
    

    so the noise is what a generic profile leaves visible at full resolution, and

    [3dnr] on this profile at the ISO step in force:
      s32YSFStr = 84   s32YTFStr = 64   s32TFStrMax = 12
    

    which is already close to its ceiling, which is why raising the spatial or
    temporal strength is not a lever I can pull further from the profile alone.

    I will leave the issue open rather than asking you to close it as hardware: if
    anyone has an IQ profile for an OV2710 on hi3516cv200 - from a vendor firmware, or
    one they tuned - that is the missing piece. Otherwise the honest state is "streams
    correctly, noise is the generic tuning's, no evidence of a faulty part".

  11. openipc-ai commented on Nov 16, 2026

    @openipc-ai
    Contributor

    Agreed, and thanks for correcting it: the stock firmware's clean picture changes the reading. What you have now is a missing tuning, not a broken part.

    Both observations hold in the firmware tree:

    • On cv200, /etc/sensors/iq/default.ini is a link to ar0130.ini, set at build time (general/package/hisilicon-osdrv-hi3516cv200/hisilicon-osdrv-hi3516cv200.mk). So a camera without a profile under its sensor's name runs the AR0130 tuning.
    • iq/ov2710.ini is byte-identical to ov9750.ini. That was deliberate when OV2710 support went in (#2035 on OpenIPC/firmware): it was described as a first cut from the closest OmniVision sibling, to be tuned with PQTools later. Nobody has tuned it since, and your 0.62 → 1.14 shows it is no better a fit than the AR0130 one.

    We have no OV2710 camera in the lab, so we cannot tune this ourselves. The tuning has to come from someone with the hardware, and here that is you or the board's owner. The path:

    1. Tune with HiSilicon PQTools against this module on hi3516cv200: noise reduction, sharpening, gamma, CCM and AWB at a few gain steps.
    2. Try it on the camera with isp.iqProfile: /etc/sensors/iq/<your file>.ini in majestic.yaml, so you are not depending on which file is picked by default.
    3. Open a PR on OpenIPC/firmware that replaces general/package/hisilicon-osdrv-hi3516cv200/files/sensor/iq/ov2710.ini with the tuned file, with before/after numbers like the ones you have been posting. PRs are welcome.

    One thing a PR cannot take: values copied out of the vendor firmware's own tuning or driver. The repository does not accept tables lifted from a camera vendor's binaries. Using the stock picture as the reference you tune towards is fine; it is the visual target, not a source.

    I'll leave this open as the place for that tuning, retitled to match what it is now.

  12. changed the title [-]hi3516cv200 + OV2710 over MIPI: test pattern clean, live scene carries ~60% relative noise at any exposure and gain[/-] [+]hi3516cv200 + OV2710: no tuned IQ profile (ov2710.ini is an ov9750 copy, default.ini is ar0130) - needs PQTools tuning[/+] on Nov 16, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions