• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 98 subscribers
  • Views 261 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.

Original question:

TDA4VM: New findings on the purple edge problem on TDA4

TDA4VM: The purple edge problem around the street lamp on TDA4

wang zhili
Intellectual 885 points
Part Number: TDA4VM

Hi Gang,

For the purple circle problem, we found it is very obvious around the street lamp at night, as shown in the screenshot below:

  

while at the same place in raw, there is a gray area that is less white than the center of the lamp, as shown in the screenshot below (between the two red circle):

the raw&yuv image is attached here:

YUV&RAW.zip

the pixel value of green channel in YUV is arround 90-210, but the value is around 220-230 (converted), the value at the center of lamp is 255, it looks like that TDA4 makes the gray area closer to the black, which caused the purple edge become more obvious. What do you think about this phenomenon?

What can cause this phenomenon in TDA4 VPAC? Or is there a way to improve this phenomenon in TDA4?

Please help me check. Thanks.

  • 0
    TI__Guru 59410 points

    Hi,

    Could you please confirm which 2 sets of the knee points are used for the raw image?

    I cannot tell for sure about this.
    It looks like tone mapping makes it much worse than it appears in the raw image.
    There is also some saturation/color shift about that area, i.e., within that dark circle raw pixels are saturated and on that circle color changes very fast because of the lights.

    The color issue also gets worse while the light goes away from image center with a similar behavior as chromatic aberration.