m2v decoder issue in DM6467T.

Bhushan,

Can you please mention the complete release version the MPEG2 decoder, which you have downloaded.

I have also downloaded the stream, we will verify and  let you know about the same.

Regards,

Mahantesh

Hi Mahantesh,

Following are the required details of MPEG2 Decoder.

1)Version 1.10.00

2)Build ID 1.10.00.07

Thanks & Regards,

Bhushan.

Hi Bhushan,

Strean verification is in progress at my end, decoded few frames, but still I haven't observed any error code returned from the process call.

Can you please mention at which frame you have observed the error code 4 returned from decoder.

Regards,

Mahantesh

Hi Bhushan,

I have verified with the stream up to 212 frames, the decoder is able to decode. However the decoder has returned extended error code as 0xA00 ( D-2560) at frame number 72, which indicates the corrupted data, but still the decoder has continued by applying the concealment. Can you please let me know whether it's a valid stream or erronious. And also mention the complete 16 bit error code observed at your end and the frame number.

Regards,

Mahantesh

hi Mahantesh,

Yes your are right its a erronious stream.The file is captured from live streaming and  error got introduced in transmission.

Regarding the error code and frame number i will let you know soon..

Have you seen the display thread ouput on TV??....it displays the same frame repeatedly.. as if  the decoder is giving the same frame to the display thread everytime..

Thanks & Regards,

Bhushan.

hi Mahantesh,

I have attached the log file giving the frame number and the return value given by Vdec2_process()  for that frame.

Till frame number 185 the decoder returns zero , but after that  returns 4 for the rest frames.    8272.log.txt

Thanks & Regards,

Bhushan.

Hi Bhushan,

It seems the error has returned at frame # 72 itself  as I have seen in the log you sent as below, might be natural because of error in the bit stream,

"DEBUG MSG : Vdec2_process() returned 4  for FrameNo :72"

Even at my end got the error at same frame #72, but I got -1. I dont know how decoder is returning 4 at youe end.

Can you please check the extended error code returned through "status.viddecStatus.extendedError"  at your end. I am getting  2560 (0xA00) at frame # 72, indicates the corrupted data.

It means decoder is able to conceal from frame # 72 itself and is able to decode up to 180 frames at your end. I think something is going wrong here because of value 4 what your getting.

Regards,

Mahantesh

hi Mahantesh,

i am getting extended error code as 0x0 .. for the frames returning Dmai_EBITERROR ....i am printing the following code  in Vdec2_process() function ..

if (status != VIDDEC2_EOK) {
        if (XDM_ISFATALERROR(outArgs.decodedBufs.extendedError)) {
            Dmai_err2("VIDDEC2_process() failed with error (%d ext: 0x%x)\n",
                      (Int)status, (Uns) outArgs.decodedBufs.extendedError);
            return Dmai_EFAIL;
        }
        else {
            //Dmai_dbg1("VIDDEC2_process() non-fatal error 0x%x\n",
            //          (Uns) outArgs.decodedBufs.extendedError);
            printf("VIDDEC2_process() non-fatal error 0x%x\n",
                      (Uns) outArgs.decodedBufs.extendedError);
            ret = Dmai_EBITERROR;
        }
    }

is it what you were expecting  me to check??

i have attached the log file.

Thanks&Regards,

Bhushan.

4213.log_error.txt

Hi Mahantesh,

any updates??

Thanks & Regards,

Bhushan.

Hi Bhushan,

It seems we are alomst reproduced the issue whatever you have mentioned for this stream, we will cross verify and let you know about the conclusion.

Regards,

Mahantesh

Bhushan,

Identifying the root cause is still in progress, we will update this in the next one week time frame.

Regards,

Mahantesh

Bhushan,

It's been verified that in the latest release version "01_10_00_10" the issue hasn't been reproduced. Can you please download this from web page and verify at your end as well.

You can verify for both the following streams,

6305[1].receiver_2frames.m2v -  for hang issue

3113[1].receiver_ccv.m2v - for assert failure at app level

Regards,

Mahantesh

Mahantesh,

okay i will check and let you know....thanks a lot.

Regards,

Bhushan.

Mahantesh,

can you provide me the link for the above mentioned release version?

Regards,

Bhushan.

Mahantesh,

We are still facing the same problem with the latest library version "01_10_00_10",  can you just look it again its deadlock for us.

we need to complete the product by mid May and we are struck with this critical issue .

Regards,

Bhushan.

Bhushan,

Thanks for the information...

We will cross check once again on the latrest library (1.10.00.10) and let you knwo about the same.

Regards,

Mahantesh

Mahantesh,

i have verified that the number of bytes consumed by decoder and its never zero.

i am providing you the log (receiver_2frames.txt) which shows number of bytes consumed each time when the Vdec2_process() is called and as stated earlier its never zero.

Regarding the issue whether same input is given to the decoder , i have dumped the 16bit hex data at a given offset for each frame to be decode as

suggested by feroz. , it is clear from the file that input is indeed changing.

i have attached both the files . All the above experiments i have done with TI's standalone decode application with the corrupt input file and new mpeg2dec

codec.

What are your our futher obersvations on this issue?

Regards,

Bhushan.

5483.Dump.txt

8321.receiver_2frames.txt