<div dir="ltr">Tim,<div style>Love to use 1.0 but this is eventually for an embedded target so I'm stuck with 0.10.</div><div style><br></div><div style>At any rate, if I don't set the appsrc/appsink caps with gst_app_sink/src_set_caps then I get a GST_MESSAGE_ERROR</div> <div style>"Error in pipeline: Internal data flow error."</div><div style>That is why I initially started heading down the road of explicitly setting the caps. I like your idea of setting src caps off the first frame but that gives me a pipeline error "Could not negotiate format"</div> <div style><br></div><div style>But, then I found that even when setting the caps the color channels were mixed up, but your great tip of setting the masks greatly helped.</div><div style>If I set the caps for the appsink (videotestsrc) to this</div> <div style><div> "blue_mask", G_TYPE_INT, 65280,</div><div> "green_mask", G_TYPE_INT, 16711680,</div><div> "red_mask", G_TYPE_INT, -16777216,</div><div style>And the caps for the appsrc (ximagesink) to this</div> <div style><div> "red_mask", G_TYPE_INT, 65280,</div><div> "green_mask", G_TYPE_INT, 16711680,</div><div> "blue_mask", G_TYPE_INT, -16777216,</div><div style>I get RGBA on the sink and BGRA on the src.</div> <div style>What's sorta fascinating/annoying is that ximagesink really wants BGR ordering. If I try to use the RGB ordering from the videotestsrc caps the pipeline bails and says it "Could not negotiate format"</div> <div style><br></div><div style>So, to summarize, my pipeline works but has BGRA ordering, which is fine but weird. Whoever wrote the ximagesink filter must have been an OpenCV developer ;)</div><div style>-rich</div><div style> <br></div><div style><br></div><div style><br></div></div></div><div style><br></div><div style><br></div><div style><br></div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Sun, May 26, 2013 at 7:37 AM, Tim-Philipp Müller <span dir="ltr"><<a href="mailto:t.i.m@zen.co.uk" target="_blank">t.i.m@zen.co.uk</a>></span> wrote:<br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On Sat, 2013-05-25 at 21:04 -0700, Richard Cagley wrote:<br> <br> Hi Richard,<br> <div class="im"><br> > Thanks for the hint about endianness. The below gets me BGRA with B<br> > being the LSB. This mask stuff is confusing. How do I get RGBA? What's<br> > with the negative sign?<br> ><br> <br> </div>Yes, it's all very confusing, that's why we changed it in 1.0 :)<br> <div class="im"><br> > caps = gst_caps_new_simple ("video/x-raw-rgb",<br> > "endianness", G_TYPE_INT, 4321,<br> > "red_mask", G_TYPE_INT, 65280,<br> > "green_mask", G_TYPE_INT, 16711680,<br> > "blue_mask", G_TYPE_INT, -16777216,<br> > "width", G_TYPE_INT, 640,<br> > "height", G_TYPE_INT, 480,<br> > "framerate", GST_TYPE_FRACTION, 30000, 1001,<br> > NULL);<br> <br> </div>You may find<br> <a href="http://gstreamer.freedesktop.org/data/doc/gstreamer/0.10.36/gst-plugins-base-libs/html/gst-plugins-base-libs-gstvideo.html#gst-video-format-new-caps" target="_blank">http://gstreamer.freedesktop.org/data/doc/gstreamer/0.10.36/gst-plugins-base-libs/html/gst-plugins-base-libs-gstvideo.html#gst-video-format-new-caps</a> and other utility functions useful. Ignore the negative signs for the mask, it's basically the unsigned integer cast to a signed integer.<br> <br> But again, why do you create the caps yourself?<br> <br> As I see it, if you're just passing the data through you should be doing<br> something like this:<br> <br> buffer = gst_app_sink_pull_buffer (appsink)<br> if (first_buffer) {<br> gst_app_src_set_caps (appsrc, GST_BUFFER_CAPS (buffer));<br> }<br> gst_app_src_push_buffer (appsrc, buffer);<br> <br> Cheers<br> <span class="HOEnZb"><font color="#888888"> -Tim<br> </font></span><div class="HOEnZb"><div class="h5"><br> > On Sat, May 25, 2013 at 4:00 PM, Tim-Philipp Müller <<a href="mailto:t.i.m@zen.co.uk">t.i.m@zen.co.uk</a>><br> > wrote:<br> > On Sat, 2013-05-25 at 15:49 -0700, Richard Cagley wrote:<br> ><br> > Hi,<br> ><br> > > I'm using videotestsrc and ximagesink and changing the<br> > patterns on<br> > > videotestsrc and looking at the output. I also have some<br> > print<br> > > statements to read the raw data pointers.<br> > ><br> > ><br> > > What would cause this? I'm setting the caps like this...<br> > ><br> > ><br> > > caps = gst_caps_new_simple ("video/x-raw-rgb",<br> > > "format", G_TYPE_STRING, "RGB",<br> ><br> ><br> > Are you using 1.0 or 0.10 ? Those caps look like a mix of<br> > both.<br> ><br> ><br> > In 0.10 it's<br> > video/x-raw-rgb,endianness=...,{red,blue,green,alpha}_mask=..<br> ><br> > In 1.0 it's video/x-raw,format=(string)xyz<br> ><br> > Why are you creating the caps yourself anyway? Why not just<br> > forward/set<br> > the caps you get on the appsink?<br> ><br> > Cheers<br> > -Tim<br> ><br> ><br> > _______________________________________________<br> > gstreamer-devel mailing list<br> > <a href="mailto:gstreamer-devel@lists.freedesktop.org">gstreamer-devel@lists.freedesktop.org</a><br> > <a href="http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel" target="_blank">http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel</a><br> ><br> ><br> > _______________________________________________<br> > gstreamer-devel mailing list<br> > <a href="mailto:gstreamer-devel@lists.freedesktop.org">gstreamer-devel@lists.freedesktop.org</a><br> > <a href="http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel" target="_blank">http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel</a><br> <br> <br> _______________________________________________<br> gstreamer-devel mailing list<br> <a href="mailto:gstreamer-devel@lists.freedesktop.org">gstreamer-devel@lists.freedesktop.org</a><br> <a href="http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel" target="_blank">http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel</a><br> </div></div></blockquote></div><br></div>