timecode beutify enhancement

[warrentc3]

          @niksedk if all that's really needed is pkt_dts_time... could be even more precise on the cmd.

Although, part of me thinks that pts_time or best_effort_timestamp_time might be more fit for purpose.

ffprobe.exe -hide_banner -loglevel warning -err_detect ignore_err -threads 1 -select_streams v -show_optional_fields auto -show_entries frame=pts_time,pkt_dts_time,best_effort_timestamp_time -output_format csv=print_section=0:nokey=1 -i "inputfile"

I'd be curious, in cases of a variance or non-existence of pts||dts stamps, what might be achievable.

The above command spits out something like this...
0.000000,0.000000,0.000000,
0.042000,0.042000,0.042000,
0.083000,0.083000,0.083000,
0.125000,0.125000,0.125000,
0.167000,0.167000,0.167000,
0.209000,0.209000,0.209000,
0.250000,0.250000,0.250000,
0.292000,0.292000,0.292000,
0.334000,0.334000,0.334000,
0.375000,0.375000,0.375000,
0.417000,0.417000,0.417000,
0.459000,0.459000,0.459000,
0.501000,0.501000,0.501000,
0.542000,0.542000,0.542000,

Originally posted by @warrentc3 in #9097 (comment)

[niksedk]

What do you mean by this?

[warrentc3]

After the last change, I realized that -show_frames was part of why the side_data was being displayed. With specifying the requisite entries, that parameter isn't required. There are some edge cases where pts time and dts time are not 1:1, hence the "best_effort_timestamp_time".

At a minimum, should drop the "-show_frames" to mitigate superflous optional fields

[niksedk]

So, why the edge cases where pts time and dts time are not 1:1?
Could you test if best_effort_timestamp_time" gives the correct seconds when different from pts time?

Can you make a PR?