• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 97 subscribers
  • Views 214 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

DM365: CGAMMAWD and MSB position on data bus

Derek Walters
Prodigy 205 points

Can anyone help explain the use of CGAMMAWD[4:1] in plain english? It seems to be different than the way I understand it from the datasheet (VPFE Programmer's Guide).

I have a 12-bit sensor attached to C_DATA[15:4], therefore sensor[11] is attached to C_DATA[15]. I read the setting as "the bit of the bus (C_DATA) that is attached to the MSB of the sensor". Thus, the value should be set up to 0x0 so that the MSB is 15. 

However, it seems that a setting of 4 (MSB == C_DATA[11]) provides better results (i.e. CGAMMAWD[4:1] = 0x0 is much darker than when set to 0x4). Is this correct? I'm not sure if this setting is somehow in line with another datashift down in the pipleline somewhere, and/or if I am  losing the four upper bits from this (apparent) initial shift.

Regards,

Derek

  • 0
    TI__Expert 6850 points

    Derek,

     

    Sorry for the delay.  Evidently there was a technical gltich that prevented my response on 4/5 at 10:08am from posting.  Anyway, here it is again:

     

    CGAMMAWD (GWDI) is the setting where the MSB is connected. In your case, it should be zero.  After passing this setting, the H/W takes only upper 12bit and the data stored in memory is 12bit.

    In general, Raw data is not very bright, but you can do some testing.

    With normal exposure (in sensor setting), please capture the light bulb scene (very close to light bulb) and check whether you see 1 in the most significant bit.

    For verification, I assume you can capture the Raw data (the output of ISIF).  

    It would be difficult to say something based on IPIPE result since there would be more parameters which can change overall scene brightness.

    In case, you comment based on IPIPE output, then you may try the same testing. If you lose MSB based on wrong data shift, then the brightest part (light bulb) can be darker or show weird color (instead of white).

    Please refer to Table 6-32. Gamma Correction Settings (CGAMMAWD) Field Descriptions, bits 4-1, for additional info.