<html><head></head><body><div>Le jeudi 23 février 2023 à 09:38 +0100, Israel Díaz via gstreamer-devel a écrit :</div><blockquote type="cite" style="margin:0 0 0 .8ex; border-left:2px #729fcf solid;padding-left:1ex"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><span style="color:rgb(55,65,81);font-family:Söhne,ui-sans-serif,system-ui,-apple-system,"Segoe UI",Roboto,Ubuntu,Cantarell,"Noto Sans",sans-serif,"Helvetica Neue",Arial,"Apple Color Emoji","Segoe UI Emoji","Segoe UI Symbol","Noto Color Emoji";font-size:16px;white-space:pre-wrap;background-color:rgb(247,247,248)">I'm currently working with an SDI video input that includes ancillary data. I'm using the v4l2src plugin in my pipeline to capture this video input, but I'm unsure whether the plugin will discard the extra data and only output the video data. Can anyone confirm whether the v4l2src plugin ignores ancillary data, or if it passes it through to the output as well? </span></div></div></div></div></blockquote><div><br></div><div>While we don't have GStreamer support for these, device maybe be providing a VBI or a Medata nodes (like UVC camera driver does provide metadata node). A video node, will only deliver the video.</div><div><br></div><blockquote type="cite" style="margin:0 0 0 .8ex; border-left:2px #729fcf solid;padding-left:1ex"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><span style="color:rgb(55,65,81);font-family:Söhne,ui-sans-serif,system-ui,-apple-system,"Segoe UI",Roboto,Ubuntu,Cantarell,"Noto Sans",sans-serif,"Helvetica Neue",Arial,"Apple Color Emoji","Segoe UI Emoji","Segoe UI Symbol","Noto Color Emoji";font-size:16px;white-space:pre-wrap;background-color:rgb(247,247,248)"><br></span></div><div dir="ltr"><span style="color:rgb(55,65,81);font-family:Söhne,ui-sans-serif,system-ui,-apple-system,"Segoe UI",Roboto,Ubuntu,Cantarell,"Noto Sans",sans-serif,"Helvetica Neue",Arial,"Apple Color Emoji","Segoe UI Emoji","Segoe UI Symbol","Noto Color Emoji";font-size:16px;white-space:pre-wrap;background-color:rgb(247,247,248)">For context, here's the pipeline I'm currently using: '<b>gst-launch-1.0 v4l2src device=/dev/video0 io-mode=4 ! video/x-raw,width=1920,height=1080,format=NV16_10LE32,framerate=60/1 ! omxh264enc target-bitrate=8000 ! video/x-h264, alignment=au ! fakesink</b>'. My ultimate goal is to introduce this pipeline into an appsrc element so that I can access both the ancillary data and the encoder-encoded data. </span><span style="color:rgb(55,65,81);font-family:Söhne,ui-sans-serif,system-ui,-apple-system,"Segoe UI",Roboto,Ubuntu,Cantarell,"Noto Sans",sans-serif,"Helvetica Neue",Arial,"Apple Color Emoji","Segoe UI Emoji","Segoe UI Symbol","Noto Color Emoji";font-size:16px;white-space:pre-wrap;background-color:rgb(247,247,248)"><br></span></div><div dir="ltr"><span style="color:rgb(55,65,81);font-family:Söhne,ui-sans-serif,system-ui,-apple-system,"Segoe UI",Roboto,Ubuntu,Cantarell,"Noto Sans",sans-serif,"Helvetica Neue",Arial,"Apple Color Emoji","Segoe UI Emoji","Segoe UI Symbol","Noto Color Emoji";font-size:16px;white-space:pre-wrap;background-color:rgb(247,247,248)"><br></span></div><div dir="ltr"><span style="color:rgb(55,65,81);font-family:Söhne,ui-sans-serif,system-ui,-apple-system,"Segoe UI",Roboto,Ubuntu,Cantarell,"Noto Sans",sans-serif,"Helvetica Neue",Arial,"Apple Color Emoji","Segoe UI Emoji","Segoe UI Symbol","Noto Color Emoji";font-size:16px;white-space:pre-wrap;background-color:rgb(247,247,248)">Assuming that the v4l2src plugin does indeed filter out ancillary data, is there any way to use the v4l2 driver directly (without using v4l2src) to access this extra data? Any suggestions or advice would be greatly appreciated. Thank you!</span><span style="color:rgb(55,65,81);font-family:Söhne,ui-sans-serif,system-ui,-apple-system,"Segoe UI",Roboto,Ubuntu,Cantarell,"Noto Sans",sans-serif,"Helvetica Neue",Arial,"Apple Color Emoji","Segoe UI Emoji","Segoe UI Symbol","Noto Color Emoji";font-size:16px;white-space:pre-wrap;background-color:rgb(247,247,248)"><br></span></div></div></div></div></blockquote><div><br></div><div>No, v4l2src is not responsible for this. The SDI driver should expose both a /dev/videoX node for the video frames, and a /dev/vbiX node (the device is exposed as /dev/mediaX and you can walk its topology to find the associated memory nodes). I'm guessing this is Xilinx ZynMP (they are the only one using <b style="color: rgb(55, 65, 81); font-family: Söhne, ui-sans-serif, system-ui, -apple-system, "Segoe UI", Roboto, Ubuntu, Cantarell, "Noto Sans", sans-serif, "Helvetica Neue", Arial, "Apple Color Emoji", "Segoe UI Emoji", "Segoe UI Symbol", "Noto Color Emoji"; white-space: pre-wrap; background-color: rgb(247, 247, 248);">NV16_10LE32). </b>You should raise this question on their github issue tracker. Note that vbi and video buffer will share the same sequence, so you can match them.</div><div><br></div><div>regards,</div><div>Nicolas</div><div><span></span></div></body></html>