------------------------------------------------------------------------ A poll associated with this post was created, to vote and see the results, please visit http://forums.slimdevices.com/showthread.php?t=19736 ------------------------------------------------------------------------ Question: "gapless" crossfade type would be useful - Yes - I'd use it - No - not interested ------------------------------------------------------------------------
I'm not convinced that slimserver is the way to solve this problem. To understand why, understand that there are (at least) 4 sources of gaps in mp3 playback. 1. Silence in the original material (which should be left there). 2. Silence that the encoder put there (either out of a misguided sense of "needing it" or because of an idea that "padding" was harmless). In some subcases there, there are also tags attached to the file to show how much was "incorrect" silence. 3. Silence that the SB hardware imposes in a song transition. It's my understanding that this is quite small. 4. Silence that occurs because of a network problem just as the buffer for the old song runs down and the buffer hasn't started to fill yet for the new song. 5. Silence that occurs because, in a synched playback environment, all the players stop and wait for them all to partly fill their buffers before starting the next track. There may be other sources I don't know about, but let's start with these 5. Which one are you asking us to vote on? Because my answer isn't the same for all. I believe slim should work on 3-5, in order to make the hardware output as faithful to the original as possible. I don't think anyone should automatically be removing silence from (1). The problem is, it is indistinguishable from (2) without intensive scrutiny. I would agree to Slim interpreting the silence tags if the encoder put them in. But I don't think Slim should act automatically, at run time, as if it knows better. This sounds like the right place for an off-line program, that lets you listen to the tips and tails of a song and decide where the song should begin and end playback and tag the files appropriately. In fact, I'd go further and say that support should be added to SB2/3s to send them a file and a set of start and end timestamps. That could be used for trimming silence (in this case) but also for playing a file fragment. For instance, if you don't like the applause at the beginning or the end of a live track, you could say by tagging that you don't want in general to hear it. That would not remove the sound from the track, it's still there in case you want to hear it, but in general, when it's played, it's left out. My $0.02 -- Michaelwagner ------------------------------------------------------------------------ Michaelwagner's Profile: http://forums.slimdevices.com/member.php?userid=428 View this thread: http://forums.slimdevices.com/showthread.php?t=19736 _______________________________________________ Discuss mailing list Discuss@lists.slimdevices.com http://lists.slimdevices.com/lists/listinfo/discuss
