[Libav-user] QImage(RGB32) to AVFrame(RGB32)

[Valeriy Shtoma]

Hmm… I don’t think so… I tried few variants of do it.

Kind Regards,
Valeriy V Shtoma

> On Apr 18, 2018, at 18:54, salsaman <salsaman at gmail.com> wrote:
> 
> Maybe you have a bug when you create the QImage ?
> 
> 
> http://lives-video.com
> https://www.openhub.net/accounts/salsaman
> 
> On Wed, Apr 18, 2018 at 11:00 AM, Valeriy Shtoma <shtomavaleriy at gmail.com> wrote:
> Thanks for your reply!
> 
> Okay, it’s clear. For now I have another code:
> AVFrame *frame = av_frame_alloc();
> AVBufferRef *buffer = av_buffer_create(img->bits(), img->sizeInBytes(),
>                                                        [](void *pOpaque, uint8_t *pData) {
>                                                                pData = nullptr;
>                                 const auto p = static_cast<QImage *>(pOpaque);
>                                                                 delete p;
>                                                        }, img, 0);
> frame->buf[0] = buffer;
> frame->data[0] = buffer->data;
> 
> But QImage::bits() always returned nullptr.
> buffer.data = ‘\0'
> 
> Kind Regards,
> Valeriy V Shtoma
> 
> > On Apr 18, 2018, at 16:54, salsaman <salsaman at gmail.com> wrote:
> > 
> > Actually av_image_alloc() should set it for you so you shouldn't need to set it yourself.
> > 
> > 
> > Gabriel.
> > 
> > 
> > http://lives-video.com
> > https://www.openhub.net/accounts/salsaman
> > 
> > On Wed, Apr 18, 2018 at 10:49 AM, salsaman <salsaman at gmail.com> wrote:
> > linesize should be w * 4 for ARGB I think.
> > 
> > Gabriel.
> > 
> > 
> > http://lives-video.com
> > https://www.openhub.net/accounts/salsaman
> > 
> > On Wed, Apr 18, 2018 at 2:04 AM, Valeriy Shtoma <shtomavaleriy at gmail.com> wrote:
> > Hi to all dear developers,
> > 
> > I want to convert QImage to AVFrame and than push it to
> > filter. But when I pushed AVFrame, I got an error:
> > 
> >  [in @ 0x11c71c200] Changing frame properties on the fly
> > is not supported by all filters.
> > 
> > and than
> > 
> >  [libx264 @ 0x10ea52c00] Input picture width (640) is greater
> >  than stride (0)
> > 
> > But in this case my picture 1280x720...
> > I know, that my code wrong(not correct pointers and linesize)
> > What I’m doing wrong?
> > 
> > My code:
> > 
> > Varian 1:
> > AV_PIX_FMT_ARGB or AV_PIX_FMT_RGB32
> > 
> > QImage img = ...;
> >  AVFrame *frame = av_frame_alloc();
> >  avpicture_fill((AVPicture*)frame, img.bits(), AV_PIX_FMT_ARGB,
> >  img.width(), img.height());
> > 
> > Variant 2:
> > AVFrame *frame = av_frame_alloc();
> > av_image_alloc(frame->data, frame->linesize, w, h, AV_PIX_FMT_ARGB, 1);
> > frame->width = w;
> > frame->height = h;
> > frame->format = AV_PIX_FMT_ARGB;
> > frame->linesize[0] = w;
> > av_image_fill_arrays(frame->data, frame->linesize, (uint8_t*)img.bits(),
> >  AV_PIX_FMT_ARGB, w, h, 1);
> > 
> > Thanks for attention. :)
> > 
> > Kind Regards,
> > Valeriy V Shtoma
> > _______________________________________________
> > Libav-user mailing list
> > Libav-user at ffmpeg.org
> > http://ffmpeg.org/mailman/listinfo/libav-user
> > 
> > 
> > _______________________________________________
> > Libav-user mailing list
> > Libav-user at ffmpeg.org
> > http://ffmpeg.org/mailman/listinfo/libav-user
> 
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user
> 
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/libav-user