[FFmpeg-devel] [PATCH] avdevice/decklink_dec: fix no warning message if no input signal detected

[Marton Balint]

On Sat, 20 Jun 2020, lance.lmwang at gmail.com wrote:

> On Sat, Jun 20, 2020 at 01:31:43AM +0200, Marton Balint wrote:
>> 
>> 
>> On Sat, 20 Jun 2020, lance.lmwang at gmail.com wrote:
>> 
>> > On Fri, Jun 19, 2020 at 10:46:33PM +0200, Marton Balint wrote:
>> > > 
>> > > 
>> > > On Fri, 19 Jun 2020, Limin Wang wrote:
>> > > 
>> > > > On Thu, Jun 18, 2020 at 07:37:50PM +0200, Marton Balint wrote:
>> > > > > > > > > On Thu, 18 Jun 2020, lance.lmwang at gmail.com wrote:
>> > > > > > > >From: Limin Wang <lance.lmwang at gmail.com>
>> > > > > >
>> > > > > >Signed-off-by: Limin Wang <lance.lmwang at gmail.com>
>> > > > > >---
>> > > > > >libavdevice/decklink_dec.cpp | 2 +-
>> > > > > >1 file changed, 1 insertion(+), 1 deletion(-)
>> > > > > >
>> > > > > >diff --git a/libavdevice/decklink_dec.cpp b/libavdevice/decklink_dec.cpp
>> > > > > >index 82106aa..90569dc 100644
>> > > > > >--- a/libavdevice/decklink_dec.cpp
>> > > > > >+++ b/libavdevice/decklink_dec.cpp
>> > > > > >@@ -751,7 +751,7 @@ HRESULT decklink_input_callback::VideoInputFrameArrived(
>> > > > > >                }
>> > > > > >            }
>> > > > > >
>> > > > > >-            if (!no_video) {
>> > > > > >+            if (no_video) {
>> > > > > >                av_log(avctx, AV_LOG_WARNING, "Frame received (#%lu) - No input signal detected "
>> > > > > >                        "- Frames dropped %u\n", ctx->frameCount, ++ctx->dropped);
>> > > > > >            }
>> > > > > > > No, this is good as is, the !no_video condition ensures that
>> > > it is
>> > > > > only logged once when the input actually disappears, and not for
>> > > > > every consecutive frame with no input.
>> > > > > It's good point, but if no input signal after run, no any
>> > > warning message.
>> > > 
>> > > I don't understand what you are saying here.
>> > > 
>> > > > I don't think it's helpful to check the signal status,
>> > > 
>> > > What do you mean?
>> > 
>> > If no_video is status machine of signal, then the first status is 1(no signal).
>> > So I think it's wrong to initialized to 0. So if the input signal is
>> > lost, no any warning message then. I'll submit a patch for your review.
>> 
>> initially no_video = 0, so we assume there is valid signal. When there is no
>> signal, we log the warning if no_video==0, and we set no_video to 1. So we
>> do give warning message when the signal is lost.
>
> If you don't use autodetect for the format, you'll get signal lost after runn, no
> warning message for the signal lost. I'm not sure it's expected behavior.

Did you actually test this? Because for me that is simply not the case.

Regards,
Marton