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.

TVP5150AM1 PAL-M Problem

Other Parts Discussed in Thread: TVP5150AM1, TVP5150A, TVP5160, TVP5150, TVP5146

Hello,

I am having some difficulties with the TVP5150AM1 and PAL-M input video. The resulting digital video output from TVP5150AM1 shows some serious anomalies on sharp transitions of the pictures. I am attaching some screenshots to help illustrate the problem. I am currently using the TVP5150A_v049002 patch (it was the only one that correctly recognized the PAL-M color setting).

Here is the input reference video:

And here are the waveforms for lines 142 and 143 of the input reference video (this is where the transition between colorbars and sweep occurs):

Everything looks good as it should. Now these are the resulting lines of the digital video encoded by the TVP5150AM1:

Line 142:

Line 143:

Line 144:

Line 145:

As you can see, there are some serious interference happening in this transition and only by line 145 the output is ok. This results in a low quality and distorted image.

I'm not sure why this is happening and what could be done to improve the results. I've already tried changing most of the register settings with no luck until now.

Can anyone help me?

Thank you!

Kind regards.

Thiago

  • With composite video you will never be able to re-constrict the original signal entirely correctly.

    Fundamentally sharp transitions and/or monochrome edges will become distorted.

    This is a fundamental issue with CVBS where the chroma and luma information are encoded within the same bandwidth.

    There are some things which can be done ti mitigate some of the issues for certain image types but these require more complex video decoders.

    The TVP5150 is the lowest quality, cheapest video decoder from TI. If you need better quality you might need to look at the TVP5146/7 or possibly even the TVP5160 devices.

    These devices have incremental quality improvements but are more expensive/complex.

    Have you actually re-constructed the video signal back into an image?

    Are you 100% sure that the artifacts you are seeing are not caused by the re-encoding back to CVBS?

    Are you using anti-alias filters on the analog inputs?

    BR,

    Steve

  • Hi Steve,

    Thank you for your replies and suggestions. I will have a look at both TVP5146/7 and TVP5160. From these, which one would have the best image quality?

    "Have you actually re-constructed the video signal back into an image?

    Are you 100% sure that the artifacts you are seeing are not caused by the re-encoding back to CVBS?"

    My test setup is

    CVBS Input -> TVP5150 -> FPGA -> SDI -> Tektronix WFM7120

    The FPGA is acting as a serializer only, no video processing is being done.

    The waveforms I attached are from the Tektronix and to display them it will convert the SDI to CVBS internally. I do not think that this process is causing the artifacts because they are also visible on the picture image displayed by the equipment (and this is obtained from the SDI directly).

    We are using this anti-alias filters configuration:

    Best regards,

    Thiago

  • In the reconstructed video signal does it look like 'hanging dots' or 'dot crawl' on the transitions? If yes, then this is a fundamental limitation of a lower quality CVBS video decoder. Try Googling "[NTSC/PAL] dot crawl" or [NTSC/PAL] hanging dots" for an explanation of why these transitions cause problems.
    e.g.
    en.wikipedia.org/.../Dot_crawl

    www.extron.com/.../article.aspx

    books.google.com/books

    BR,
    Steve