Virtual Sound Decoders

[Ed Kapuscinski]

Yesterday, on my way home after the Timonium Show, I had what I thought was a brilliant idea about how to easily bring sound to N scale operations: use JMRI to monitor Loconet, and map that to sound files that, essentially, perform the functions of a sound decoder, but through a computer.

So, I get home, and start googling about some of the parts I'd need to glue together to make it work.

Turns out, someone else not only already had the idea, but already implemented it directly in JMRI!
http://jmri.org/manual/3-6_DecoderPro/Main_VSD.shtml

It looks like it's only in its infancy, but I'd like to know if any of you have tried it, how you like it, and where you've gotten your VSD files?

[sirenwerks]

That's kinda like the idea I brought forward years ago, except the sound decoder wasn't virtual, it was throttle-based (along with the speaker).  So one sound decoder per throttle, rather than one per loco.  Of course, with no real DCC experience beyond operating others' layouts, I couldn't resolve the practicality of the idea.   

[Philip H]

I can see this having a lot of utility on smaller switching layouts where a couple of good quality speakers attached to a lap top make the sound manageable in the space. but how do you think it will work with larger layouts - even with yours - where I would expect to need multiple speakers to get some sense of movement from the sounds as the loco moves around?

[GaryHinshaw]

This has been on my watch list.  Note that there are a number of methods being developed to enable location following so that sound can be directed to remote speakers based on train location (see halfway down the web page).  These are either hardware based (transponding decoders) or software based (through jmri Operatons).  This has a lot of potential for things like ambient sounds as well (flange or brake squeal, etc.).

[C855B]

... ambient sounds as well (flange or brake squeal, etc.).

Or, with a subwoofer or two, the gut-thumping thunder of a brace of SD45s clawing their way through Woodford and up the west approach to Walong.

[...sigh...]

[Ed Kapuscinski]

This has been on my watch list.  Note that there are a number of methods being developed to enable location following so that sound can be directed to remote speakers based on train location (see halfway down the web page).  These are either hardware based (transponding decoders) or software based (through jmri Operatons).  This has a lot of potential for things like ambient sounds as well (flange or brake squeal, etc.).

Yep, this, exactly, is the answer to Phil's concerns. It understands geography.

I really want to get more into this. The problem I see right now is mapping the VSD files. However, if someone can find a way to convert Loksound files, the world is our Oyster...

[wm3798]

It definitely makes more sense to use speakers larger than are practical in N scale if sound is your cup of tea.  But I find the cacophony of a bunch of sound equipped locos to be distracting at best, and annoying at worst.

Optimal for a small switching layout for those yummy back and forth rumbles of a single switcher, but a room full of speakers locos and sounds would make an operating session a bit hard to bear. At least for very long.

Lee

[ednadolski]

This is very intriguing. Sound at Tehachapi Loop is a major, almost visceral part of the experience:

Not a valid vimeo URL
I have been recently trying a setup with some speakers installed under the layout and running audio clips from the computer.  This works to a point, but keeping the sound in sync with the moving train is one of the key challenges at this point.

I'm also considering a Surroundtraxx, but that seems to have its own limitations.  One concern is, while it can present the prime mover sounds, it does not seem to have a way to reproduce the 'ambience'.   For example, if you listen to a proto train at the real Loop,  the sounds of the loco's alternators and traction motors are a distinctive component.

I find the cacophony of a bunch of sound equipped locos to be distracting at best, and annoying at worst

Very true, esp. with the tinny sound generated my most on-board speakers, and/or if there are a lot of sound locos operating independently but in close proximity.  (Most on-board sound decoders do not sound very good at Notch 8.)  I'm a bit more fortunate with a locale like the Loop, since in general there will only ever be one or maybe two trains in operation at once, and they are all part of the same scene. 


Does the VSD support multiple channels and crossfading, like the Surroundtraxx?   Does it use transponding or some other method to locate trains?

Ed

[svedblen]

Hefty Ed! A great show 

But I don't think it will be easy to replicate that kind of sound experience with a VSD based system. As you say it would require multiple channels and crossfading. And even if you manage to get everything correct for a passing train, would that not only be from a specific view (listening) point. If you, as a spectator, move along the layout, would you then not expect to hear sounds from another part of the train? And what about several listeners at different locations?

I do not say someone will not eventually pull that trick off, but I guess it will be a real challenge. I will be eagarly listenning 

[peteski]

I for one do not expect a gut-shaking deep Diesel rumble to emanate from a 5" long model observed from 4 real feet away.  My brain just cannot associate that this kind of sound with a tiny model.  Also, low frequency sounds are not very directional so this sound would be heard loudly in the entire layout room.  This wouldn't be so bad if it was am HCD layout and it had only one locomotive running on the layout, but on a room-size layout with half a dozen operators running trains, this type of setup is not very realistic.

If I want to experience a gut-shaking rumble of a locomotive I'll stand next to the 1:1 item.  If I want sound-realism on an N scale layout I will go for very low-volume sound coming directly from a model.  To me it is sounds like the horn and bell which give realism to a model during an operating session.  Those sounds are reproduced fairly well by small speakers and at low volume levels, and can they are directional (so you can actually tell which model they are coming from).  Again, you don't want eardrum-busting sound of the bell or horn - just a faint auditory experience is the best.  At least for me.

The other thing too is that most people will quickly tire of the loud and full sound of a diesel locomotive.  About 25 years ago one of our NTRAK members built a diesel loco sound synthesizer. It could be hooked up to a DC throttle.  He brought it to one of the model RR shows and hooked it up to a throttle which he used on his engine service facility module.  The audio output was hooked up to a stereo amplifier and a pair of large HiFi speakers. The sound effect was amazingly realistic with very deep bass sound.  The venue was a fairly large high arena (Shriner's Auditorium) bit the sound could be heard clearly on the far end of the arena.  Ifter few hours we all begged him to lower the volume - we just couldn't take it anymore!

[C855B]

...  About 25 years ago one of our NTRAK members built a diesel loco sound synthesizer. It could be hooked up to a DC throttle.  He brought it to one of the model RR shows and hooked it up to a throttle which he used on his engine service facility module.  The audio output was hooked up to a stereo amplifier and a pair of large HiFi speakers. The sound effect was amazingly realistic with very deep bass sound.  The venue was a fairly large high arena (Shriner's Auditorium) bit the sound could be heard clearly on the far end of the arena.  Ifter few hours we all begged him to lower the volume - we just couldn't take it anymore!

Similar situation recently at local shows. We had a vendor selling [something], "Realistic Diesel Sound!", in a rather impressive display. He had a rock-concert speaker system probably like your fellow did. Shook the entire place, and thankfully he cranked it up only about once an hour. It was interesting and quite accurate, but quickly tiring, especially that it was the same loop every time. He was absent from the most recent show, tho' I do recall it was a lot more subdued at the one previous, after several years of being pushed back into a corner of the show floor. It would be easy to surmise complaints to the show manager from other vendors and maybe even the public were a factor.

[GaryHinshaw]

But I find the cacophony of a bunch of sound equipped locos to be distracting at best, and annoying at worst.

I used to think that too, but as I've attended more and more sound-equipped sessions, I have started to change my tune (pun intended).  For example, I recently ran a sound-equipped through freight over Farron Summit on Mark's Columbia & Western: coming down the east slope with the dynamic breaks whining was a visceral experience, even though it was barely above audible.  The session as a whole was far from cacophonous.  On-board sound has come a long way in N scale!

If I could effectively add ambient train sounds to that experience, I'd be hooked.  It seems eminently doable with a scheme like VSD: train tracking is fairly easy if you have transponding decoders, block detection, or some of the schemes noted in the OP's link.  And the volume is adjustable.  Set it low for a session, and crank it for some straight-up train watching. 

Great video Ed!

[Scottl]

I'm thinking about this as well and how to implement it.  Once I started adding sound to videos, the absence was notable- they seemed almost unrealistic.  I did this video with a few iTunes sound files, but I think in person the squeal and ambient noises would add a tremendous sense of place.  Virtual decoder functions are the way to make this happen, so I'm watching these developments with interest.

[u18b]

Chris.... I guess I have been doing something at least slightly like this for years.

One of the very best electronic projects I ever did was the ones by Kieth Gutierez published in Model Railroader around 1990.

He had a Programmable Diesel Horn (called a PDH-1) and a digital sound generator for an SD40-2 (called the SDX-1).

I built both of those and then modified them.  So they are now inside one enclosure.  Also, I added a plain old decoder and connected it to the DCC system.

This sound system detects voltage from the track in DC.  So as volts go up, the sound of the loco ramps up.  Other sound systems work the same way.

So what I did was install a decoder inside the sound system BETWEEN the track and the sound system board...... where the orange and gray motor wires now become the inputs to the sound system.
This allows the system to work on DCC.

So what I do is this.  when I program a loco address, I program the sound system at the same time.  So the sound system and the loco have the same address.

Obviously, I can only have the sounds of one loco, but man, with external speakers, it will make the windows rattle.

In fact, here is a video I made to show the complex decoder wiring/features of a U18B I built for a guy.
The sound you hear is my sound system- not a sound decoder in the loco.  I used a 6 function decoder for class lights, warning beacons, headlights, and ditchlights.

/>

So I like the idea of having a speaker in a throttle.  For someone who does not want to go to the cost or effort of sound decoders in all locos, I think it is a good compromise.

[Ed Kapuscinski]

That's kinda exactly how I got started thinking about it. At first my thought was just to program a sound decoder to the same address, and mount it on the layout, but then I thought "wait, a computer can do this a lot cheaper than a decoder". I thought I might have to try to build it, but luckily, someone else already has.