VPIF cannot capture from a specific 720p60 source

Nir Aloni,

I think you are referring to BT.1120. 720p is described in BT.1120. 

Nir Aloni said:

I tried to check with logic analyzer if there is difference between the BT.656 codes of this source and the other working sources, but couldn't find any significant difference.

what differences you see?

Nir Aloni said:

I checked EAV-to-SAV, SAV-to-EAV and  number of lines during vertical blanking.

I feel you need 2 check lines and pixels per line(width) in active area a also.

Nagabhushana,

Yes, I wrote BT.656 since the codes are the same.

The differences I see are in EAV-to-SAV - 360 clocks when it works and 358 when it doesn't.

Also, a whole frame time is 16.666 when it works and 16.646 when it doesn't.

Regards,

Nir

Nir Aloni said:

The differences I see are in EAV-to-SAV - 360 clocks when it works and 358 when it doesn't.

I feel this is the problem. Since ISIF register LPFR set 720, ISIF waits for 720 lines. But, due to clock difference, I dont think 720 lines are coming into ISIF. I feel you need 2 check the source and fix.

Nagabhushana,

The EAV-to-SAV parameter is not related to number of lines, so this cannot be the problem you mentioned.

Regards,

Nir Aloni

Is this issue handled? no reply for long time.

Thanks,

Nir

Nir,

I feel you need to check registers which program vertical size of the frame(spruer9d, section 3.28) or number of lines in a frame. VPIF might expect this even in case of BT.656.  I feel since the source has slight change in terms of clock cycles, lines in frame also change.

Sorry for reference to ISIF in last reply, I somehow considered this thread as related to dm365.

Nagabhushana,

I already tried that with no success - I tried many values for the vertical size.

Can you think of something I can send, like signals from the VPIF bus or something else which can help you get into this?

Thank you,

Nir

Nir,

I don't have setup with me, otherwise I would have tried what you suggested.

Somebody else in TI?

Nir