On Tue, Jul 26, 2005 at 12:14:13AM +0200, Michael Banck wrote: > On Mon, Jul 25, 2005 at 11:53:18PM +0200, Robert Millan wrote: > > Package: jack > > Version: 3.1.1-3 > > Severity: normal > > Tags: patch > > > > This patch includes fixes for Flac support. Basicaly: > > > > - Use only a command for best performance, and mark it as such by naming > > it > > "cmd" instead of "vbr-cmd". > > Is flac incapable of doing VBR?
No, but note: - If I understood correctly what VBR means, it requires recieving the quality parameter from jack, which is done via "%q". The previous command had no %q so I guess this means it doesn't really do VBR. - The point of using Flac is that it is a lossless codec. Using a quality parameter other than --best (currently -8) defeats this point. > > - Remove the "otf" command which includes invalid flags (probably was > > added > > for an old version of flac). > > I asked the flac maintainer: > > <mdz> azeem: if jack is using raw audio, you need all those flags. if > not, you don't > <mdz> azeem: they have nothing to do with the compression; they're > setting the attributes of the raw stream > <mdz> azeem: if it's encoding a wav file, it gets that information from > the header instead > > So this sounds consistent with on-the-fly to me, are you sure the > options are bad? The command in question gives syntax errors. Perhaps it was added before a change in the CLI. When attempting to encode a raw stream, this is what flac says, which is different from the command we had before: ERROR: for encoding a raw file you must specify a value for --endian, --sign, --channels, --bps, and --sample-rate Asides from this, can't jack pipe a wav stream instead? (I know about the fixed length field, but..) I added Matt to CC. -- Robert Millan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
