Re: ADV7180 Capture with i.MX53

[Steve Longerbeam <slongerbeam@gmail.com>]

On 10/8/19 10:20 AM, Ian Arkver wrote:
> On 08/10/2019 18:14, Steve Longerbeam wrote:
>>
>>
>> On 10/8/19 9:55 AM, Tim Harvey wrote:
>>> On Tue, Oct 8, 2019 at 4:21 AM Fabio Estevam <festevam@gmail.com> 
>>> wrote:
>>>> On Mon, Oct 7, 2019 at 10:07 PM Fabio Estevam <festevam@gmail.com> 
>>>> wrote:
>>>>
>>>>> So now I need to see if I can get Gstreamer to accept a pipeline 
>>>>> like:
>>>>>
>>>>> gst-lauch-1.0 v4l2src ! kmssink
>>>> Ok, so now I decided use the hardware video deinterlacer:
>>>>
>>>> media-ctl -l "'adv7180 1-0021':0 -> 'ipu1_csi0':0[1]"
>>>> media-ctl -l "'ipu1_csi0':1 -> 'ipu1_vdic':0[1]"
>>>> media-ctl -l "'ipu1_vdic':2 -> 'ipu1_ic_prp':0[1]"
>>>> media-ctl -l "'ipu1_ic_prp':2 -> 'ipu1_ic_prpvf':0[1]"
>>>> media-ctl -l "'ipu1_ic_prpvf':1 -> 'ipu1_ic_prpvf capture':0[1]"
>>>>
>>>> media-ctl -V "'adv7180 1-0021':0 [fmt:UYVY2X8/720x480 
>>>> field:alternate]"
>>>> media-ctl -V "'ipu1_csi0':1 [fmt:AYUV32/720x480]"
>>>> media-ctl -V "'ipu1_vdic':2 [fmt:AYUV32/720x480 field:none]"
>>>> media-ctl -V "'ipu1_ic_prp':2 [fmt:AYUV32/720x480 field:none]"
>>>> media-ctl -V "'ipu1_ic_prpvf':1 [fmt:AYUV32/720x480field:none]"
>>>>
>>>> And then Gstreamer can be launched:
>>>>
>>>> # gst-launch-1.0 v4l2src device=/dev/video2 ! kmssink --verbose
>>>> Setting pipeline to PAUSED ...
>>>> Pipeline is live and does not need PREROLL ...
>>>> /GstPipeline:pipeline0/GstKMSSink:kmssink0: display-width = 800
>>>> /GstPipeline:pipeline0/GstKMSSink:kmssink0: display-height = 480
>>>> Setting pipeline to PLAYING ...
>>>> New clock: GstSystemClock
>>>> /GstPipeline:pipeline0/GstV4l2Src:v4l2src0.GstPad:src: caps =
>>>> video/x-raw, format=(string)YUY2, width=(int)720, height=(int)480,
>>>> framerate=(fraction)25/1, colorimetry=(string)bt601,
>>>> interlace-mode=(string)progressive
>>>> /GstPipeline:pipeline0/GstKMSSink:kmssink0.GstPad:sink: caps =
>>>> video/x-raw, format=(string)YUY2, width=(int)720, height=(int)480,
>>>> framerate=(fraction)25/1, colorimetry=(string)bt601,
>>>> interlace-mode=(string)progressive
>>>>
>>> Fabio,
>>>
>>> Yes, you need to use the vdic to capture from adv7180 with gstreamer
>>> as it can't handle alternate.
>>>
>>>> However the video looks like a broken old TV scrolling the image 
>>>> horizontally:
>>>> https://www.dropbox.com/s/2yef8egn6s8z7ff/mx53_adv7180_capture.mp4?dl=0 
>>>>
>>>>
>>> This would be because of the initial corrupt frames that this and many
>>> other decoders produce while waiting for proper sync. I added
>>> 'g_skip_frames' support in 9483a3f8e1b58ba1d7cd21687d8d0a63a015c36b
>>> but I'm not sure how to get gstreamer to use it?
>>>
>>> I still carry around a patch from Steve for imx-csi that drops first
>>> few frames from BT656 sources:
>>> https://github.com/Gateworks/linux-imx6/commit/959fbd42ee6433f49ef4a04fb1abe8f8c78db5ad 
>>>
>>> to deal with this.
>>
>> Yes, that's likely the issue, from a look at Fabio's video. The patch 
>> referenced by Tim hard-codes the number of frames to skip, instead of 
>> calling the adv7180's g_skip_frames op. I still don't have an answer 
>> as to how to call the adv7180 from the CSI subdev.
>
> Seems to me initial corrupt frames would produce a fixed offset of 
> some kind. A rolling video like that looks more like the number of 
> lines being captured is wrong.

Nope, rolling video is one of the symptoms of initial corrupt frames, 
from my own experience. I don't really have an explanation for it, but 
IIRC the IPU will insert lines on its own to recover from an initial 
wrong # lines captured, to regain vertical sync. That should mean the 
rolling should eventually stop once vertical sync is re-established, but 
I've seen many instances where rolling video continues, and skipping the 
initial corrupt frames fixes it.

Steve