Home of the Squeezebox™ & Transporter® network music players.
Page 94 of 96 FirstFirst ... 44849293949596 LastLast
Results 931 to 940 of 960
  1. #931
    Senior Member
    Join Date
    Apr 2005
    Location
    UK/London
    Posts
    4,352
    I don't think so.
    All of the "Stream" commands have an asterisk beside them meaning subscription required
    https://mediaserver.smartskills.tech...mAlbum.html#75
    Paul Webster
    http://dabdig.blogspot.com
    Author of "Now Playing" plugins covering Radio France (FIP etc), PlanetRadio (Bauer - Kiss, Absolute, Scala, JazzFM etc), KCRW, Supla Finland, ABC Australia, CBC/Radio-Canada and RTE Ireland

  2. #932
    Senior Member
    Join Date
    Oct 2010
    Posts
    168

    MediaServer skill on piCorePlayer

    Hi hoping you can help
    This is a second setup for me, for my son, my setup works perfectly on same hardware
    Have got the skill working for him on a pi4 with piCorePlayer 7.0.1 and works very well.
    However if I reboot the pi the ngrok tunnel is lost, running the startup script manually then shows a new tunnel and I get the 'Skill(s) successfully updated with (new) tunnel details --> OK 1' message in the terminal but this still does not fix the problem.
    To get things working again I have to go to the Alexa app remove the skill and then relink with the updated tunnel url .

    Any help appreciated.

  3. #933
    Senior Member philchillbill's Avatar
    Join Date
    Jan 2019
    Location
    The Netherlands
    Posts
    705
    Quote Originally Posted by gegen View Post
    Hi hoping you can help
    This is a second setup for me, for my son, my setup works perfectly on same hardware
    Have got the skill working for him on a pi4 with piCorePlayer 7.0.1 and works very well.
    However if I reboot the pi the ngrok tunnel is lost, running the startup script manually then shows a new tunnel and I get the 'Skill(s) successfully updated with (new) tunnel details --> OK 1' message in the terminal but this still does not fix the problem.
    To get things working again I have to go to the Alexa app remove the skill and then relink with the updated tunnel url .

    Any help appreciated.
    Are you sure that the uuid is correct/unique? If you re-used your setup script from your first installation, it will still have 'your' uuid in it and that's then the one you would be updating via the script. The uuid is stored on your machine in your browser's localStorage cache until you deliberately delete it (if ever). If you visit the configurator page for the skill in the future, it will retrieve your 'known' uuid and re-use it - if it's gone or was never there, a new one is generated for you. If you want to download a setup script for a different user then you can either do it from a different PC or browser, do it in incognito mode, or reset the uuid by pressing on the tiny (i) symbol at the top of the configurator page.

    You'll see the uuid at the top of the setup.sh script that you downloaded.

    Just make sure that each instance of the skill at different locations has its own uuid.

    Let me know if that solves it.

  4. #934
    Senior Member
    Join Date
    Oct 2010
    Posts
    168
    Quote Originally Posted by philchillbill View Post
    Are you sure that the uuid is correct/unique? If you re-used your setup script from your first installation, it will still have 'your' uuid in it and that's then the one you would be updating via the script. The uuid is stored on your machine in your browser's localStorage cache until you deliberately delete it (if ever). If you visit the configurator page for the skill in the future, it will retrieve your 'known' uuid and re-use it - if it's gone or was never there, a new one is generated for you. If you want to download a setup script for a different user then you can either do it from a different PC or browser, do it in incognito mode, or reset the uuid by pressing on the tiny (i) symbol at the top of the configurator page.

    You'll see the uuid at the top of the setup.sh script that you downloaded.

    Just make sure that each instance of the skill at different locations has its own uuid.

    Let me know if that solves it.
    Hi philchillbill

    Thanks very much for the reply, hoping to catch up with my son today so will checkout the UUID. Pretty sure we created a new setup script for his location with different ip and ngrok token but did do it on the same machine........will check.

  5. #935
    Senior Member philchillbill's Avatar
    Join Date
    Jan 2019
    Location
    The Netherlands
    Posts
    705

    V4 adds extensive display of Cover Art and Metadata

    Amazon has freshly certified V4 of MediaServer which is now live (no need to update anything on your end - you automatically use the newest version always).

    There are 2 new commands (to play/stream a random playlist from your library). However, the most noticeable change is that the skill now uses Alexa Presentation Language (APL) to provide visual feedback for all commands. Obviously, your Echo will need to have a display for this to work In the past, the 'Stream' commands always displayed an image (which recently also became cover-art instead of just a placeholder) on Echo's with screens, but that was done via a separate Amazon feature especially for the AudioPlayer interface. This new APL stuff is something I have specifically/custom designed using Amazon's JSON-based markup language.


    In this first release with APL, I've defined 4 templates to best match the nature of a command. A now-playing type screen shows large Cover Artwork of the track together with several categories of metadata from your tags:


    Name:  nowplaying.png
Views: 128
Size:  178.8 KB


    When it is appropriate to display more than one cover, a sideways scrollable list is shown with smaller images and less metadata (for example, when you ask "what's up next?" or to "list albums by Snow Patrol"):


    Name:  listalbums.jpg
Views: 125
Size:  30.5 KB


    If a list containing cover art would be very long, or for a list like favorites/genres/playlists, a vertically-scrollable text list is shown instead:


    Name:  listalbumstext.png
Views: 125
Size:  148.6 KB


    For certain sparse-info categories like "List my players" a single, larger, line of text is displayed in the middle of the screen:


    Name:  discovery.jpg
Views: 123
Size:  24.4 KB


    A point worth remembering is this. Material or iPeng have the luxury of retrieving artwork via your LAN (usually Gb/s or at worst 100's Mb/s) which always feels 'snappy'. Amazon is retrieving artwork via the upload speed of your ISP and this is much more likely to be in the 10's Mb/s range at best. The good news is that Amazon actually fetches it after the skill has responded so any slowness does not cause timeout issues. Also, the Echo displays APL for approx. 30 sec after a skill session or one-shot closes so there's time to 'paint' if needed. With decent broadband the overall experience is very fluid, but with rural internet YMMV. For this reason, you can use a new Inhibit command to "disable artwork display" if your internet is not up to it. It's all in the docs. Note that if your Echo does not have a display the artwork is never retrieved wastefully.

    Also, while APL is being displayed after a one-shot command, the session is held open but with the Echo's microphone closed. This means that for your next one-shot you can drop the "... ask MediaServer to ..." but you will have to put an "Alexa, ..." in front of it to open the mic. Example:

    "Alexa, ask MediaServer what's playing?" [one-shot]
    --> spoken response, with APL on screen
    "Alexa, what's coming up?" [partial one-shot] rather than "Alexa, ask MediaServer what's coming up?"

    This does mean that while APL is displayed, any command starting with "Alexa, ..." is directed to the skill and not to Alexa in general. MediaServer does not know 'what time is it ?', so saying 'Exit' or Home Screen will resolve.

    Sessions are not affected by this new behavior and continue to work as they always did (the docs explain in detail what is meant by one-shots and sessions, but in a nutshell a session results from the open/launch command and allows back-to-back commands).

    By the way, I can trap what item (cover/metadata) you click on when you scroll a list on the screen so this could actually be used to play/select items in LMS. That would mean that MediaServer becomes a hybrid voice/touch controller. I'd be interested in hearing comments on the desirability of that.

    Enjoy!
    Last edited by philchillbill; 2021-05-16 at 02:56. Reason: Replaced screen photos by Alexa APL simulator screenshots

  6. #936
    Senior Member
    Join Date
    Apr 2005
    Location
    UK/London
    Posts
    4,352
    Very clever.
    Paul Webster
    http://dabdig.blogspot.com
    Author of "Now Playing" plugins covering Radio France (FIP etc), PlanetRadio (Bauer - Kiss, Absolute, Scala, JazzFM etc), KCRW, Supla Finland, ABC Australia, CBC/Radio-Canada and RTE Ireland

  7. #937
    Senior Member philchillbill's Avatar
    Join Date
    Jan 2019
    Location
    The Netherlands
    Posts
    705

    More natural/shuffled results for certain Alexa intents

    Some of MediaServer's commands, such as e.g.

    Play something by The Corrs
    Play songs about 'Girls'
    Play some Classic Rock

    would always result in exactly the same generated playlist each time you used the intent, unless player shuffle was enabled. That's because under the hood they use the CLI's [ 'playlist, 'loadtracks', '...' ] construct which was (until now) deterministic in its returns.

    From LMS version 8.2.0_1620395888 onwards, either the new 'Balanced' shuffle (if enabled) or the traditional Fischer-Yates shuffle is applied to the generated playlist (thanks Michael!), even if the player being targeted has shuffle disabled. This results in a far more natural user experience with the skill, as these commands by their very nature have no inherent expectation of order-preservation. Asking to play something by Tangerine Dream and getting exactly the same 2700-song playlist every time gets long in the tooth

    The change also applies to e.g. "Play the song 'Run Like Hell' by Pink Floyd", which will now shuffle all 24 variants of that song found across your library from all your albums before playing them (here, there's also no natural expectation of any order). Don't worry, the "Play the track 'Run Like Hell' from the album 'Pulse'" intent is still there to drill down if that's what you want.

    Specifically, this applies to the PlaySong, PlayGenre, PlaySomethingBy, and PlaySomethingAbout intents, plus their 'Stream' counterparts.

    Enjoy !

  8. #938
    Senior Member
    Join Date
    Jun 2005
    Location
    The South, UK
    Posts
    316
    Quote Originally Posted by philchillbill View Post
    From LMS version 8.2.0_1620395888 onwards, either the new 'Balanced' shuffle (if enabled) or the traditional Fischer-Yates shuffle is applied to the generated playlist (thanks Michael!), even if the player being targeted has shuffle disabled.
    Works a treat, a great improvement, thanks Phil/Michael. Now I'm off to play "some" Calypso...
    Location 1: LMS 8.2 on Win 10 Brix Server, x2 SB Radios, x1 Touch, x1 Controller : Location 2: LMS 8.2 on Win 10 Brix Server, x2 SB Radios, x1 Duet Receiver, x1 Controller : Alexa Mediaserver Smart Skill, Material Android, SqueezeliteX control

  9. #939
    Senior Member philchillbill's Avatar
    Join Date
    Jan 2019
    Location
    The Netherlands
    Posts
    705
    Name:  helpscreen.jpg
Views: 136
Size:  35.4 KB

    I've added a 5th APL template to MediaServer which is displayed when you use the Help intent.

    A (long) scrollable list with all the example commands from the website is displayed if your Echo has a screen. When you touch the screen to start scrolling, the spoken response stops.

    This should hopefully get some of you using the more esoteric commands as it's now trivial to find them

    Enjoy !
    Last edited by philchillbill; 2021-05-09 at 10:51.

  10. #940
    Senior Member philchillbill's Avatar
    Join Date
    Jan 2019
    Location
    The Netherlands
    Posts
    705

    Lossless Internet Radio Streaming

    Skills are now permitted to send lossless streams to Echo devices - yaay! Under the hood, MediaServer uses the stream.mp3 mechanism to setup a network client whose audio is directed to the Echo. This always transcodes the stream to - you guessed it - mp3, so it's lossy. I have some ideas for how to get lossless directed to the Echo natively, but short of a miracle like the devs creating stream.flac for us, all would involve no pseudo-player in LMS showing up for the Echo in question. Is that a show-stopper? Full access to your local LMS library, but bypassing the GUI when streaming lossless files?

    To give you a flavour of what to expect, I've modified MediaServer's code for the StreamFavorite intent. When you ask to Stream an internet radio favorite whose URL starts with https:// and which has the word 'FLAC' or 'lossless' in the URL string itself, that URL is passed directly to the Echo without starting up stream.mp3 in LMS. The sound quality is noticeably better, even on a Dot 3 but certainly on an Echo Show 2 which is a very decent device. I've even succeeded in sending 24-bit 88.2KHz stuff to the Echo in testing.

    To give it a try with e.g. RP, create a favorite from one of the flac/flacm URLs at https://www.hiresaudio.online/radio-paradise/ and make sure to make it https. Call it Radio Paradise. Now, when you say "Alexa, stream favorite 'Radio Paradise'" you'll get the lossless stream with no transcoding by stream.mp3. You'll notice this worked correctly if the background image on an Echo Show says 'LMS DirectStream' instead of 'LMS Stream', the favorite's name is shown with [Lossless] added, and a different image is displayed. The now-playing metadata from the station will not display (yet).

    Enjoy !

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •