Author Topic: Thanks for being at the Academy guys.  (Read 4922 times)

Offline JerryPlak

  • Supporting Member
  • ******
  • Join Date: Jul 2013
  • Location: Sunbury, OH
  • Posts: 608
  • Kudos: 2
Re: Thanks for being at the Academy guys.
« Reply #15 on: July 15, 2014, 12:15:58 PM »
Very cool stuff.  ;D


Echo  Very cool stuff. 8)
Jerry Plak

Offline JerryPlak

  • Supporting Member
  • ******
  • Join Date: Jul 2013
  • Location: Sunbury, OH
  • Posts: 608
  • Kudos: 2
Re: Thanks for being at the Academy guys.
« Reply #16 on: July 15, 2014, 12:22:34 PM »
I know I was excited. Whole house effects are now possible. The three pi players could have been located in different states. Falcon put player, the most amazing thing I have seen.

 Sean then all of them will need set up using GMT time  ;)

Offline David Pitts

  • Administrator
  • *****
  • Join Date: Mar 2013
  • Location: Falcon, CO
  • Posts: 3,943
  • Kudos: 80
Re: Thanks for being at the Academy guys.
« Reply #17 on: July 15, 2014, 12:24:20 PM »
I know I was excited. Whole house effects are now possible. The three pi players could have been located in different states. Falcon put player, the most amazing thing I have seen.

 Sean then all of them will need set up using GMT time  ;)

Well... No. The remotes do not use the scheduler. They are sent data from master as to when to start and so forth.
PixelController, LLC
PixelController.com

Offline JerryPlak

  • Supporting Member
  • ******
  • Join Date: Jul 2013
  • Location: Sunbury, OH
  • Posts: 608
  • Kudos: 2
Re: Thanks for being at the Academy guys.
« Reply #18 on: July 15, 2014, 12:53:50 PM »
I know I was excited. Whole house effects are now possible. The three pi players could have been located in different states. Falcon put player, the most amazing thing I have seen.

 Sean then all of them will need set up using GMT time  ;)

Well... No. The remotes do not use the scheduler. They are sent data from master as to when to start and so forth.
  I was thinking about different states time zone  ;)

Offline David Pitts

  • Administrator
  • *****
  • Join Date: Mar 2013
  • Location: Falcon, CO
  • Posts: 3,943
  • Kudos: 80
Re: Thanks for being at the Academy guys.
« Reply #19 on: July 15, 2014, 12:55:42 PM »
I know I was excited. Whole house effects are now possible. The three pi players could have been located in different states. Falcon put player, the most amazing thing I have seen.

 Sean then all of them will need set up using GMT time  ;)

Well... No. The remotes do not use the scheduler. They are sent data from master as to when to start and so forth.
  I was thinking about different states time zone  ;)

I realized that but if remotes do not use the time to fire the sequence then they can be on their local timezones. All on GMT is not needed. :)

Offline CaptainMurdoch

  • Administrator
  • *****
  • Join Date: Sep 2013
  • Location: Washington
  • Posts: 9,856
  • Kudos: 214
Re: Thanks for being at the Academy guys.
« Reply #20 on: July 15, 2014, 02:33:08 PM »
None of the FPP instances at the academy had the right time if I recall and they stayed in perfect sync because the sync protocol is not time based.  It is frame based, the master says "I am on frame 55" and the slave reads that sync packet and does what it needs to speed up or slow down to match the master's playback.  They are not in perfect sync down to the microsecond but they are probably within a few milliseconds of each other normally I estimate.  I need to do some timing tests sometime but that is just out of curiosity because to the eye I think they are in near-perfect sync.


-
Chris

Offline JerryPlak

  • Supporting Member
  • ******
  • Join Date: Jul 2013
  • Location: Sunbury, OH
  • Posts: 608
  • Kudos: 2
Re: Thanks for being at the Academy guys.
« Reply #21 on: July 15, 2014, 02:48:52 PM »
Great info

Offline drlucas

  • Developer
  • ******
  • Join Date: Sep 2013
  • Location: Whitby, Ontario - Canada
  • Posts: 1,424
  • Kudos: 3
Re: Thanks for being at the Academy guys.
« Reply #22 on: July 15, 2014, 06:20:14 PM »
None of the FPP instances at the academy had the right time if I recall and they stayed in perfect sync because the sync protocol is not time based.  It is frame based, the master says "I am on frame 55" and the slave reads that sync packet and does what it needs to speed up or slow down to match the master's playback.  They are not in perfect sync down to the microsecond but they are probably within a few milliseconds of each other normally I estimate.  I need to do some timing tests sometime but that is just out of curiosity because to the eye I think they are in near-perfect sync.
If you introduce 25ms or so of latency by connecting houses together do you start to run into timing issues? I'd be willing to plug open some ports on the firewall here at home if someone wants to set a Guinness book of records on the furthest synched xmas light display.
Ryan

Offline Steve Gase

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Georgetown, TX (near Austin)
  • Posts: 1,037
  • Kudos: 5
    • WinterLightShow in Georgetown, TX
Re: Thanks for being at the Academy guys.
« Reply #23 on: July 15, 2014, 06:32:03 PM »
None of the FPP instances at the academy had the right time if I recall and they stayed in perfect sync because the sync protocol is not time based.  It is frame based, the master says "I am on frame 55" and the slave reads that sync packet and does what it needs to speed up or slow down to match the master's playback.  They are not in perfect sync down to the microsecond but they are probably within a few milliseconds of each other normally I estimate.  I need to do some timing tests sometime but that is just out of curiosity because to the eye I think they are in near-perfect sync.
maybe a round-trip calculation can be added to the protocol -- to allow each remote to adjust for transmission delays.
http://WinterLightShow.com  |  110K channels, 50K lights  |  Nutcracker, Falcon, DLA, HolidayCoro

JET

  • Guest
Re: Thanks for being at the Academy guys.
« Reply #24 on: July 15, 2014, 07:39:54 PM »
Just a quick thanks to Falcon team. Was great to meet you and discuss the product. Look forward to long term relationship.

Jimmy

Offline CaptainMurdoch

  • Administrator
  • *****
  • Join Date: Sep 2013
  • Location: Washington
  • Posts: 9,856
  • Kudos: 214
Re: Thanks for being at the Academy guys.
« Reply #25 on: July 15, 2014, 08:46:53 PM »
maybe a round-trip calculation can be added to the protocol -- to allow each remote to adjust for transmission delays.

I thought about this originally but decided that if the systems were far enough apart to need something like that then you probably wouldn't notice them out of sync slightly anyway.  I'll think about it and may revisit sometime soon.  Getting the time to offset is easy and the sync code is already in my head, so I may go ahead and do it anyway.  I'll put a note back on my TODO so I don't forget about it among other things.

Offline drlucas

  • Developer
  • ******
  • Join Date: Sep 2013
  • Location: Whitby, Ontario - Canada
  • Posts: 1,424
  • Kudos: 3
Re: Thanks for being at the Academy guys.
« Reply #26 on: July 15, 2014, 09:03:45 PM »
may come in handy for wifi networks - not sure how to deal with wifi networks and my wife making microwave popcorn...but that's a different issue

Offline CaptainMurdoch

  • Administrator
  • *****
  • Join Date: Sep 2013
  • Location: Washington
  • Posts: 9,856
  • Kudos: 214
Re: Thanks for being at the Academy guys.
« Reply #27 on: July 15, 2014, 09:11:49 PM »
may come in handy for wifi networks - not sure how to deal with wifi networks and my wife making microwave popcorn...but that's a different issue

The MultiSync code does not need sync packets to operate, it can handle missed packets.  It uses the packets to determine whether it needs to speed up or slow down.  If the master were to stop sending packets, the remotes would keep playing at whatever rate they were playing at when they last received a sync packet until they reached the end of the sequence and then they would stop.

Still doesn't handle the popcorn scenario totally, but we can handle packet loss as best we can.  :)

Offline JerryPlak

  • Supporting Member
  • ******
  • Join Date: Jul 2013
  • Location: Sunbury, OH
  • Posts: 608
  • Kudos: 2
Re: Thanks for being at the Academy guys.
« Reply #28 on: July 15, 2014, 10:28:19 PM »
this info is Cool info and look forward playing with this  8)  ;D   

Offline drlucas

  • Developer
  • ******
  • Join Date: Sep 2013
  • Location: Whitby, Ontario - Canada
  • Posts: 1,424
  • Kudos: 3
Re: Thanks for being at the Academy guys.
« Reply #29 on: July 16, 2014, 06:21:56 AM »

may come in handy for wifi networks - not sure how to deal with wifi networks and my wife making microwave popcorn...but that's a different issue

The MultiSync code does not need sync packets to operate, it can handle missed packets.  It uses the packets to determine whether it needs to speed up or slow down.  If the master were to stop sending packets, the remotes would keep playing at whatever rate they were playing at when they last received a sync packet until they reached the end of the sequence and then they would stop.

Still doesn't handle the popcorn scenario totally, but we can handle packet loss as best we can.  :)

Now I get how the sync works. I haven't got my second pi up yet because of the minor issue I'm having with encoding. Got a good discussion going back on the raspberrypi forums. Hope to resolve that issue in a day or two (damn day job keeps getting in the way).

Anyways if you can handle packet loss that's great. If it can handle popcorn lost in cushions that would be a decent workaround to the microwave issue - I think :scratch head:

 

Back to top