This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TDA4VM: AE Brightness instability problem

Part Number: TDA4VM

In the DNP 1000lux environment(In fact, it will also appear in any test scenarios), we continuously capture a certain number of pictures .  we found that the brightness of the pictures fluctuated back and forth, as shown in the figure below. We want to know what causes this problem and How can we optimize and solve it. Thank you!

  • Could you please provide more information (camera, sensor, sensor dynamic range, AE settings, tone mapping, who does the ISP tuning, ...)?

  • Sorry for the late reply, the camera we use is ox03c,sensor dynamic range:140 dB HDR4;AE settings is default, target is 40,   sensor tone mapping is 24bit -> 12bit; and we tuning the isp by  ourselves; 

  • I cannot tell what could be possibly causing the issue in your system.

    The first step is to test and make sure your AE output (sensor exposure time and gain) is stable for the 1000lux environment.

    Then you may move to check ISP tuning.

  •    

    The above figure is AE output for exposure and  agian、dgain  for the 1000lux environment.

    From the log, we can see that the gain value  is not  stable ;the log is the AE output. thanks!
  • If your test scene is stable (no lighting changes), AE shall not adjust sensor exposure.

    You will have to debug AE and sensor driver to find out the cause of the exposure changes.

  • the scene is stable and 100% confirmation; The current problem is that the environment is stable and AE is adjusting; The sensor driver   according to Ae  adjusts the exposure  and gain; My question is What are the possible reasons of the AE problem  and  other customers feedback his similar problem,Thanks!

  • AE won't change sensor exposure if its input from VISS-H3A is stable.
    There is a control loop between AE and sensor.
    I cannot tell what is wrong in your system with your sensor driver and VISS-H3A settings.

    You will have to debug AE input to see if and how it is changing and then trace down to sensor driver and VISS-H3A settings.

  • BTW, there is a known bug in "vx_aewb_target.c" related to AE (but I am not sure if it is related to your issue).

    You need to (1) swap line 4 and line 5 below and (2) swap line 12 and line 13 below.

                    if(NULL == nodePrms->dcc_output_params)
                    {
                        h3a_buf = decode_h3a_header((uint8_t *)h3a_aew_af_target_ptr->data,
                                                 &nodePrms->p_awb_params->ti_awb_data_in.frame_data.h3a_data_x,
                                                 &nodePrms->p_awb_params->ti_awb_data_in.frame_data.h3a_data_y,
                                                 &nodePrms->p_awb_params->ti_awb_data_in.frame_data.pix_in_pax);
                    }
                    else
                    {
                        h3a_buf = decode_h3a_header_dcc((uint8_t *)h3a_aew_af_target_ptr->data,
                                                 &nodePrms->dcc_output_params->ipipeH3A_AEWBCfg,
                                                 &nodePrms->p_awb_params->ti_awb_data_in.frame_data.h3a_data_x,
                                                 &nodePrms->p_awb_params->ti_awb_data_in.frame_data.h3a_data_y,
                                                 &nodePrms->p_awb_params->ti_awb_data_in.frame_data.pix_in_pax);
                    }

  • thank you for your reply, and I will try it!

  • Please enable AE_DEBUG and catch related log, and I will check for the stability of H3A.

  • AEDEBUG_1000Lux_LOG.txt

    The attachment is AE  Debug  log, thanks!

  • The values of R/G/B change regularly from your log. So I think something happened in the sensor side, please check this with sensor vendor. Thanks!

    Line 21800: [MCU2_0] 8072.610031 s: (1)AE debug: R = 66, G = 47, B = 40
    Line 23051: [MCU2_0] 8073.321886 s: (1)AE debug: R = 33, G = 24, B = 21
    Line 23614: [MCU2_0] 8073.853587 s: (1)AE debug: R = 67, G = 48, B = 42
    Line 24763: [MCU2_0] 8074.205935 s: (1)AE debug: R = 67, G = 48, B = 42
    Line 25603: [MCU2_0] 8074.381945 s: (1)AE debug: R = 34, G = 24, B = 21
    Line 26445: [MCU2_0] 8074.576976 s: (1)AE debug: R = 33, G = 24, B = 21
    Line 27371: [MCU2_0] 8074.774426 s: (1)AE debug: R = 66, G = 47, B = 40
    Line 28330: [MCU2_0] 8074.994195 s: (1)AE debug: R = 33, G = 24, B = 21
    Line 29288: [MCU2_0] 8075.209994 s: (1)AE debug: R = 67, G = 48, B = 42
    Line 30247: [MCU2_0] 8075.430492 s: (1)AE debug: R = 34, G = 24, B = 21
    Line 30337: [MCU2_0] 8075.656180 s: (1)AE debug: R = 33, G = 24, B = 21
    Line 31327: [MCU2_0] 8075.901774 s: (1)AE debug: R = 66, G = 47, B = 40
    Line 32350: [MCU2_0] 8076.164870 s: (1)AE debug: R = 33, G = 24, B = 21
    Line 33372: [MCU2_0] 8076.429181 s: (1)AE debug: R = 67, G = 48, B = 42
    Line 34379: [MCU2_0] 8076.669397 s: (1)AE debug: R = 34, G = 24, B = 22
    Line 35401: [MCU2_0] 8076.932732 s: (1)AE debug: R = 33, G = 24, B = 21
    Line 36391: [MCU2_0] 8077.174168 s: (1)AE debug: R = 66, G = 47, B = 40
    Line 37414: [MCU2_0] 8077.438351 s: (1)AE debug: R = 33, G = 24, B = 21
    Line 38500: [MCU2_0] 8077.726394 s: (1)AE debug: R = 67, G = 48, B = 42

  • carry and gang, we modify the  TIAE_adj_exposure paramter, and the brightness of picture is stable . thank you for your support !

     

  • Which parameters did you modify, and could you share the solution, thanks!

  • we Increase the delta value, and add the  adjRatio point.