JMRI Virtual Sound Decoder for Small N Scale layout

BigJake Jun 14, 2020

  1. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    Search does not appear to be working on this site, so maybe the title subject has been addressed before, but...

    For small N scale layouts (single hollow core door), I'm wondering if JMRI's Virtual Sound Decoder would be a better way to go than N scale sound decoders in the locomotives?

    The small layout means direction of the sound following the locomotive location is not as important as for a room-sized layout, and I don't want to use location detection (or the sound system) to project the sound near the lotomotive(s). I'm thinking of a typical bookshelf speaker under the layout.

    The sound quality should be much better than is possible in N scale, especially at the lower frequencies. I'm not as much concerned that a given locomotive sounds exactly like its 1:1 counterpart, as I am that it sounds more like _a_ real locomotive.

    I only have DC now, but am planning my upgrade to a DCC system. I understand that not all DCC systems are compatible with VSD (e.g. NCE Powercab is incompatible, surprisingly, which I was leaning towards). I'm now leaning towards a Pi Sprog system, with the throttle app running on my Android phone. The VSD idea has me leaning towards conventional (non-sound) decoders for my engines.

    Thoughts, or better yet, experiences?
     
  2. FlightRisk

    FlightRisk TrainBoard Member

    548
    237
    14
    Everything you said is true ;) I use DCC++ EX. I also use sound decoders, but I like playing with VSD for some reason. I know a way to get spacial sound out of it, tracking the loco, but it just wasn't worth it. It sounds just fine. Ambient railroad sounds are better than no sounds. Small speakers have come a long way, But they still give "miniature sound". You can get more base out of bigger speakers and low frequencies are non directional. I say just try it and see. Non sound decoders are cheap. You can always upgrade later.
     
  3. IronMan1963

    IronMan1963 TrainBoard Member

    161
    173
    9
    I have just started into figuring out the VSD in JMRI myself. I run a Raspberry Pi 3B+ running either my DCC++ or my Digitrax system. I think with some decent speakers around my layout I will be happy with it. I have a small 6X3. Info is available at the JMRIGROUPS.IO. Not a lot of different Loco sounds but guys are taking other decoder sound files apart and making their own files. The example file has a generic steam, an Emd 567 and a GE44 tonner I believe. there is also an ALCO RS3 file available. Working in N Scale unless the loco is made for a DCC Sound decoder it gets pretty tight. Obviously if you buy Sound equipped locos no worries. I am working on a sound decoder install in an old Atlas/Kato diesel mechanism. However even with my mini mill it is a challenging install. Not much room to work with in N Scale.

    Later Richard
     
  4. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    This wasn't clear in the documentation I've read so far, but does VSD play multiple locos' sound if multiple locos are active, or just one at a time?
     
  5. IronMan1963

    IronMan1963 TrainBoard Member

    161
    173
    9
    I just had this working last night. It can have 4 running at once. I had tried listing all 9 of my roster and it gave an error message that it maxed at 4.
    Later Richard
     
  6. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    Thanks; I don't think that will be a serious problem.

    Does a consist count as one in that list?
     
  7. IronMan1963

    IronMan1963 TrainBoard Member

    161
    173
    9
    I can't say. Haven't tried it yet.
    Later Richard
     
  8. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    According to documentation, VSD uses the throttle commands to tell it what's going on (which for a consist would be just the consist, not the individual loco).

    Therefore, I assume an entire consist counts as only one of the maximum 4 locos in the list, but I've been proven wrong on assumptions before.

    Actual test data trumps assumptions almost every time.

    Come to think of it, the incompatibility of the PowerCab with VSD may only be when the PowerCab is used as the throttle, not if a JMRI/WiFi throttle is used...
     
  9. rray

    rray Staff Member

    8,277
    9,277
    133
    I used to run VSD at train shows on out Z Scale modules. We used a 2.1 sub sat speaker system under the module tables. It was a lot of fun, and the bass was amazing. Sound is too hard to do in Z, and even when they fit it in, the quality is less than cell phone sound quality.

    I just purchased a Digitrax system and the Surroundtraxx system for my new Z Scale modules, in hopes of under layout sound that follows the trains.
     
  10. S t e f a n

    S t e f a n TrainBoard Member

    167
    93
    6
    Any updates on availability of sound files? I did a google search, but nothing really comes up. It doesn't help that microsoft has some other .vsd thing going. Is there a group or repository somewhere for .vsd sound files, beyond the five or so files on github?
     
  11. BigJake

    BigJake TrainBoard Member

    3,259
    6,173
    70
    Nothing that does not take significant sound file manipulation to merge together. There is a website for sound files for one of the sound decoders (ESU?) that has fragments which need to be "stitched" together, and I found a few bits of information here and there, but nothing "ready to run".
     
  12. IronMan1963

    IronMan1963 TrainBoard Member

    161
    173
    9
    Try the JMRI Groups.IO
    Later Richard
     
  13. rray

    rray Staff Member

    8,277
    9,277
    133
    The VSD was written by Trainboard member TwinDad at least 10 years ago, and he was only messing with it for a few months before moving on to other aspects of the hobby. It's the same thing as the guy who wrote DCC++ four years ago, after doing that project he too, moved on to other interests.

    Someone else needs to pick up the ball and run with VSD if you want to keep it going, just like what several guys are doing with DCC++ these days.
     

Share This Page