On 22/06/2007 6:12 Am, "Martin Redington" <email@hidden> wrote:
> On 22 Jun 2007, at 14:49, Mike Kluev wrote:
>
>> I figured to get uncompressed data I can use:
>> SGSetVideoCompressorType(chan, kRawCodecType);
>
> I'm pretty sure WhackedTV never calls this - I think you can skip
> that call.
If I skip it, I'm getting yuv2 format by default, while I want to
get 32bpp xrgb. Anyone knows what exactly is kRawCodecType format?
Any guarantees about it? Here I make an ungrounded assumption that
KRawCodecType = 32bpp xrgb which doesn't feel correct, but I do
not see other appropriate constant for 32bpp xrgb to use it instead.
>> Also, I'm not sure I get the rowBytes value correctly.
>> Please comment on the following source code.
>
> I think that's wrong. You're rowBytes = len/height assumption is not
> correct, at least if I understand rowBytes correctly.
Under what circumstances could it fail? Is there other way to get
rowBytes without dealing with "decompression"? (See below, I hope
to get uncompressed data initially.)
> As I mentioned in an earlier post, do your processing in the
> decompression session callback, and call CVPixelBufferGetBytesPerRow
> on the pixelbuffer there to get bytes per row ..
>
> As far as I can tell, everything you need to know is in the sample
> code ...
Maybe. I just want to avoid unnecessary compression/decompression
steps (where compression is first done by some codec and then I
have to "undo" its work and decompress data with either
DecompressSequence API or ICMDecompressionSessionCreate or whatnot).
Besides anything else this wastes CPU cycles. Why not to skip the
whole compression/decompression step altogether? Is that not reasonable?
Mike
_______________________________________________
Do not post admin requests to the list. They will be ignored.
QuickTime-API mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden
