Re: [SlimDevices: Plugins] Softsqueeze And 48K Flacs
philippe_44 wrote: > Agreed, I should have been more precised - for self-describing formats, > the size and rate is left to '?' in strms and only for pcm it is > explicitely set. What I was saying is that for flac, you find the sample > rate in the STREAMINFO header that is the first thing that is > transmitted (https://xiph.org/flac/format.html). > > Please not that this is not a criticism of the way you've implemented > that, just adding information if it helps :) Got it. I'll save my work, revert, and see if I can come at it 100% from the client side by extracting the requisite sample rate from the flac data itself. Perfect weekend project. :D bitspiel's Profile: http://forums.slimdevices.com/member.php?userid=63033 View this thread: http://forums.slimdevices.com/showthread.php?t=103688 ___ plugins mailing list plugins@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/plugins
Re: [SlimDevices: Plugins] Softsqueeze And 48K Flacs
philippe_44 wrote: > As a curiosity, why adding sample rate in the strms ? Normally, flac is > self describing and you find it in the block header (that's what I'm > doing on my client). I feel the mod should rather be making sure that > soft squeeze correctly reads the header. They are variants in how the > rate is coded In flac, but they are not difficult to handle and would > give you the result without mods on the server side. Strm records already includes a sample rate, so I am not technically adding one. What my tweak does is replace the server's hard coded, for flac tracks, value of '?' with the track's corresponding value as documented for strm sample rate. bitspiel's Profile: http://forums.slimdevices.com/member.php?userid=63033 View this thread: http://forums.slimdevices.com/showthread.php?t=103688 ___ plugins mailing list plugins@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/plugins
Re: [SlimDevices: Plugins] Softsqueeze And 48K Flacs
ralphy wrote: > I always find it interesting at the number of softsqueeze downloads that > still happen each week, 241 for this past week alone. > > The windows version has been downloaded more than 71000 times since I > released it 4 years ago. > > Not bad for a long dead app! > > Can the server side flac changes be applied only to softsqueeze client > versions > 2, so it doesn't screw up all the current players? > Then if you change Softsqueeze.properties:firmwareversion=3 for the new > client, nothing *should* be affected by the changes. > > If we can satisfy these two conditions, with working code, I think it > would be possible to ask Michael to have the changes applied to lms > 7.9. > > Perhaps the jflac class could also be updated? Maybe not to the newest, > but newer. > > Unfortunately, I really can't take this on right now, but it'd be > willing to help get the code changes into the softsqueeze cvs repository > and eventually create new builds, but that's not likely before the fall. I'm puzzled about why you're worried that changing the server to include the sample rate for flac streams would affect legacy softsqueeze clients. As I read softsqueeze's code, the "sudden" appearance of a non '?' sample rate would be simply ignored by legacy softsqueezes as it uses a hard code sample rate for flac "streams." bitspiel's Profile: http://forums.slimdevices.com/member.php?userid=63033 View this thread: http://forums.slimdevices.com/showthread.php?t=103688 ___ plugins mailing list plugins@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/plugins
Re: [SlimDevices: Plugins] Softsqueeze And 48K Flacs
bpa wrote: > Softsqueeze (i.e. The Java emulator of IP3K products) is effectively > dead - there is no point making a change especially as nobody has > noticed before this although I find it odd as WMA radio stream which are > typically 48k used to be transcoded into Flac 48K and nobody complained. > > > What version of LMS are you using ? > The server LMS is being updated by mherger as a community support - so > if you think it is a real problem with latest 7.9 then file a bug report > with your patch. Server version in play: 7.8.0 With respect to WMA radio stream being converted to 48K flac without issues: my physical slim devices did NOT have an issue with 48K flac files. Rather, only softsqueeze was dropping the ball. I think it important to note that softsqueeze decodes flac on the fly (via a stale version of jflac). The correction I implemented drops softsqueeze's hard coded 44100 sample rate parameter when creating the javax.sound.sampled.AudioFormat object. In order to tell AudioFormat which sample rate to utilize, I changed the server to pass that information, via strm, based on the sample rate being reported for the track. bitspiel's Profile: http://forums.slimdevices.com/member.php?userid=63033 View this thread: http://forums.slimdevices.com/showthread.php?t=103688 ___ plugins mailing list plugins@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/plugins
[SlimDevices: Plugins] Softsqueeze And 48K Flacs
Added some new music to my library and noticed that a couple of the purchased flac tracks were sampled at 48K. My classics have no problem playing the 48K variants, but softsqueeze had issues (sort of like pitch shifting down). So I pulled the code on the two respective products and tweaked the strm protocol. I had to change not only softsqueeze but also the server as well. (Specifically I needed the server to not always send '?' as the sample rate for all flac files and modify softsqueeze to react to the given sample rate.) The question I have is there any point in re-contributing these changes back to the repositories? Both products are essentially dead. Further complicating the decision is the question of is my "solution" the best practice? Anyone have some constructive thoughts here? bitspiel's Profile: http://forums.slimdevices.com/member.php?userid=63033 View this thread: http://forums.slimdevices.com/showthread.php?t=103688 ___ plugins mailing list plugins@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/plugins
