IP using Lights

#166499

Can you use the "Lights" function using a IP address and using DMX (e1.31)??
Not the Lights for signal heads, just the lights function.

Dennis

Re: Sluggish native panel response time compared to web

Randall Wood

#166498

The rendering (and therefor visual responsiveness) of the native panel is performed in the Event Dispatch Thread (EDT) within JMRI (this is a hard requirement within Java). Because threading is very complex, JMRI also performs many other operations in the EDT.

It appears to me that the GNOME (Ubuntu graphical application API) / Java interaction means that rendering is controlled by the EDT, but is on a GNOME-specific thread, which may mean that all EDT processing has to be completed and then the native interface updates smoothly (this implies that a macOS or Windows user may see different behavior than you do).

Randall

Sluggish native panel response time compared to web

#166497

Recently I've spent some time comparing the web presentation (as shown in a web browser) for a layout editor panel, with that of the 'native' panel (that shown directly on the computer). What I noticed surprised me; the web panel was significantly more responsive than the native panel.

The native panel gives the impression of being 'polled', where sometimes it responds quickly, and other times there is a noticeable delay. I've not been able to actually measure the delays, but would put them all at sub-second, so these are by no means serious delays, but noticeable nevertheless! I'm just wondering if other people have noticed the effect and might know what is causing it.

I've tried experimenting with browsers on the 'server' machine (i.e. where JMRI is running), on a separate machine and on a pretty old smartphone. I've had the server running on a cheap laptop and higher spec desktop and the effect is similar. In the case of the smartphone there was a slight delay in the browser response (but this corresponded in timing to a browser on the server machine), but the panel again lagged behind. The effects are similar whether the change is in response to a trigger from the browser or from the native panel itself.

I'm not aware of this being a recent phenomenon, but for information I've tried it on the latest production release of 4.16 as well as the latest test (4.17.6) and development releases (4.17.7ish) and am not aware of a difference between them. From which I conclude it's not a recent effect, but cannot conclude how long it's been present.

I've been doing my tests using a simple panel I created to demonstrate web differences: /g/jmriusers/files/ProblemsBeingWorkedOn/Andy%20Brown%20-%20161045/CrossoverBlockOccupancyIssue03.xml

Desktop:

Ubuntu 19.10
Firefox 70.0.1 (64-bit)
JMRI 4.16 and 4.17.6 and later
Java 11.0.5-ea (en_GB)
Intel(R) Core(TM) i3-8100 CPU @ 3.60GHz (Quad-core)
16GB memory
240GB SSD

Laptop:

Ubuntu 19.10
Firefox 70.0.1 (64-bit)
JMRI 4.16 and 4.17.6 and later
Java 11.0.5-ea (en_GB)
Intel(R) Celeron(R) CPU N2840 @ 2.16GHz (Dual-core)
8GB memory
240GB SSD

Smartphone:

Motorola Moto G (2013)
Android 5.1
8GB
Browser: Lightning v5.1.0 (open source)

Regards
Andy

Re: Making VSD Files and Config - Volume question #vsdecoder

#166496

Phil,

Am 28.11.2019 um 18:45 schrieb Phil in gorgeous Young Harris via Groups.Io:
Howdy all. In my latest rabbit hole, I am putting a mini wifi camera (SQ23) on a train and plan to have an Engineer in another room run the train with the help of local Switchmen and Conductor (on the ground at the train; no radios, just hand signals like 1950). My friend said it would be cool if there was audio of the train too, so I embarked on a VSD trail.
The train will be using an Alco S-1, and while I have gotten the examples to work, I wanted an Alco, so I am making a vsd file for that.? I have the sounds and steps for Idle, Notch 1 and Notch 2 and for notching up and down, and I also have the Bell (F1), the Horn (F2), added a Shorthorn (F3).
It's a good decision to make a try with VSD! Sound enriches our hobby.
I think you're very close to a good VSD file. I'm just gonna make a few recommendations.


The problem I am having is the volume.? I can edit the sound files and manually change any I need (they are all normalized at -1db), but I thought (apparently incorrectly) that the "gain" in the config file would adjust the volume of the various pieces.
Right, the "gain" parameter is used to adjust the volume of the different sound files like the Bell, Horn, and so on. For engine sound however "engine-gain" is used instead of "gain". Engine sound covers the start and the stop sound, the idle sound and all the notch sounds. Default engine-gain is 0.8 (from a range of 0.0 to 1.0). Engine-gain = 0.0 would mute the engine sound.


I have posted my project in the ProblemsBeingWorkedOn/Phil Abernathy files section if someone wants to look. It sounds really cool, but the prime mover is far too loud, while the horn is too soft, and I don't have many notches (although it will be used for Yard switching, and may not use many notches).Thanks for the upload! That was very helpful.

I recommend to reduce the volume of the prime mover sound with your audio editor. Yes, the horn is too soft. You could try to increase the volume in config.xml, e.g. gain = 1.2 (or reduce the engine sounds, e.g. engine-gain = 0.6).

You mentioned that "16 bit, 47khz" is not working for you. JMRI supports 16 bit, 44100Hz. Please check your rate value of 47khz! Class64.vsd is working with 16 bit, 44100Hz chuff sounds.

Klaus

Re: jynstrument error for USB throttle in 4.17.6 #scripting

#166495

Allen,

Great!

That change to the script is now in the main JMRI repository so will be correct in future releases.

Thanks for your help confirming this.

Best regards,

Matt H

Re: jynstrument error for USB throttle in 4.17.6 #scripting

#166494

Matt,
That did it!
Thanks for the help,

Allen

Re: speed matching

#166493

开云体育

Bachmann decoders are a good reason for using DecoderPro. It’s easy to get confused when manually programming these decoders, especially when dealing with CV29. Why? Because the NMRA, most books, and other manufacturers number the bits of a CV as 0-7, but Bachmann numbers them 1-8.

?

Mark Granville

Re: Integrating JMRI to my MERG CBUS Layout #layouteditor

#166492

开云体育

Good evening Steve. I am going to have another marathon run tonight. I take your points but will read through them again carefully. I said to PeteB tonight, the help you all have given has been tremendous. I was not sticking with using Internal Turnouts but thought it was necessary because I could not establish what system names I should use for the turnouts. Pete pointed to 3 tools in JMRI that may help, CBUS Node Manager, CBUS Event Capture Tool to list 2. I shall play with these tonight.

Best wishes

Peter

Show quoted text

From: [email protected] <[email protected]> On Behalf Of steve young via Groups.Io
Sent: 28 November 2019 23:44
To: [email protected]
Subject: Re: [jmriusers] Integrating JMRI to my MERG CBUS Layout #panelpro #layouteditor

?

Hi Peter,

I'm still unsure as to why you're using Internal Turnouts rather than MERG CBUS Turnouts.

If you look at the JMRI CBUS Turnout Table, click on show Feedback Information.

If you now operate the Turnout via your non-JMRI panel, you'll notice that only the "Feedback" ( state ) is updated, not the ( commanded ) "State".
If you setup the Turnout as a CBUS Turnout, the ( commanded ) state will also update when the Turnout event is heard incoming to JMRI.

This also enables the physical Turnout to be set via clicking on the Layout Editor Panel or in the CBUS Turnout Table.
Your system will be a lot more in-sync if it knows there's a discrepancy between the Commanded and Feedback states.

There is absolutely no harm in JMRI knowing what the Commanded State is, so would recommend changing from Internal to CBUS Turnouts.
You don't HAVE to use JMRI for sending Turnout events, but if you did in future it's a lot easier to set it up now.

Consider the following CBUS events,

Event 1 is taught to operate a Servo
Event 2 is the Servo Closed Feedback Event
Event 3 is the Servo Thrown Feedack Event

23:39:39.252 JMRI < < CBUS | Short Event On EN:1 Command Turnout Thrown 23:39:39.334 JMRI < < CBUS | Short Event Off EN:2 Closed Off
23:39:41.531 JMRI < < CBUS | Short Event On EN:3 Thrown On
23:39:56.905 JMRI < < CBUS | Short Event Off EN:1 Command Turnout Closed 23:39:57.695 JMRI < < CBUS | Short Event Off EN:3 Thrown Off 23:39:59.896 JMRI < < CBUS | Short Event On EN:2 Closed On

In this common scenario, light the panel LEDs by teaching the LED module events 2 and 3.

If you require a single Sensor to trigger a Route, have you looked at Routes?

When the Route is confirmed with all Turnouts feedback set correctly, you can have Sensors / Turnouts triggered to illuminate LEDs on Control Panels.

These are a lot easier to configure than Logix imho!
As you are using Internal Turnouts, the commanded state would not be known, so you would not be able to trigger routes, only monitor them.

Routes can be also triggered on JMRI Startup, handy for sending a Start of Day CBUS Event when your panel has been initialised and tables loaded,

You could specify multiple SOD events with a time gap between them by selecting an additional delay between Turnout commands.

Steve.

Re: Integrating JMRI to my MERG CBUS Layout #layouteditor

#166491

Hi Peter,

I'm still unsure as to why you're using Internal Turnouts rather than MERG CBUS Turnouts.

If you look at the JMRI CBUS Turnout Table, click on show Feedback Information.

If you now operate the Turnout via your non-JMRI panel, you'll notice that only the "Feedback" ( state ) is updated, not the ( commanded ) "State".
If you setup the Turnout as a CBUS Turnout, the ( commanded ) state will also update when the Turnout event is heard incoming to JMRI.

This also enables the physical Turnout to be set via clicking on the Layout Editor Panel or in the CBUS Turnout Table.
Your system will be a lot more in-sync if it knows there's a discrepancy between the Commanded and Feedback states.

There is absolutely no harm in JMRI knowing what the Commanded State is, so would recommend changing from Internal to CBUS Turnouts.
You don't HAVE to use JMRI for sending Turnout events, but if you did in future it's a lot easier to set it up now.

Consider the following CBUS events,

Event 1 is taught to operate a Servo
Event 2 is the Servo Closed Feedback Event
Event 3 is the Servo Thrown Feedack Event

23:39:39.252 JMRI < < CBUS | Short Event On EN:1 Command Turnout Thrown 23:39:39.334 JMRI < < CBUS | Short Event Off EN:2 Closed Off
23:39:41.531 JMRI < < CBUS | Short Event On EN:3 Thrown On
23:39:56.905 JMRI < < CBUS | Short Event Off EN:1 Command Turnout Closed 23:39:57.695 JMRI < < CBUS | Short Event Off EN:3 Thrown Off 23:39:59.896 JMRI < < CBUS | Short Event On EN:2 Closed On

In this common scenario, light the panel LEDs by teaching the LED module events 2 and 3.

If you require a single Sensor to trigger a Route, have you looked at Routes?

When the Route is confirmed with all Turnouts feedback set correctly, you can have Sensors / Turnouts triggered to illuminate LEDs on Control Panels.

These are a lot easier to configure than Logix imho!
As you are using Internal Turnouts, the commanded state would not be known, so you would not be able to trigger routes, only monitor them.

Routes can be also triggered on JMRI Startup, handy for sending a Start of Day CBUS Event when your panel has been initialised and tables loaded,

You could specify multiple SOD events with a time gap between them by selecting an additional delay between Turnout commands.

Steve.

Re: ESU Select Question

#166490

That's the one.

Seriously, sometimes I don't know what or where I am.

John

Show quoted text

---------- Original Message ----------
From: Dave Heap <dgheap@...>
Date: November 28, 2019 at 4:44 PM

Marc and John,

On 29 Nov 2019, at 9:36 AM, forfoum@... wrote:

On Thu, Nov 28, 2019 at 04:58 PM, John wrote:
In the files section there should be either a pdf or Word document listing
folks
willing to help other ESU users, someone may live close to you...make a new
friend?
Maybe in the ESU forum but not in here. I searched for it out of curiousity.The list is in the Files section of the LokSound group, recently moved from
Yahoo to groups.io

Dave in Australia

Re: ESU Select Question

#166489

开云体育

Marc and John,

On 29 Nov 2019, at 9:36 AM, forfoum@... wrote:

On Thu, Nov 28, 2019 at 04:58 PM, John wrote:

In the files section there should be either a pdf or Word document listing folks
willing to help other ESU users, someone may live close to you...make a new
friend?
Maybe in the ESU forum but not in here. I searched for it out of curiousity.

The list is in the Files section of the LokSound group, recently moved from Yahoo to

Dave in Australia

Re: speed matching

[email protected]

#166488

I thought mallet sounded a little less violent : -)

Marc

Re: ESU Select Question

[email protected]

#166487

On Thu, Nov 28, 2019 at 04:58 PM, John wrote:

In the files section there should be either a pdf or Word document listing folks
willing to help other ESU users, someone may live close to you...make a new
friend?

Maybe in the ESU forum but not in here. I searched for it out of curiousity.

Marc

Re: ESU Select Question

#166486

Hello Brad,

There are Select micros and V4 micros but there are not Select V4 micros. The
Select line is a "dumbed down" version of the V4. The selct can only be loaded
with ESU sound files, with the V4 the modeller can make their own sound files
and load them into whatever sound slot they may desire.

To install the Full Throttle features you will need a LokProgrammer. In the
files section there should be either a pdf or Word document listing folks
willing to help other ESU users, someone may live close to you...make a new
friend?

ESU uses a proprietary communications method to talk to their decoders, Decoder
Pro can only manipulate CVs within the decoder, it cannot load sound files or
firmware updates into the decoder, again the LokProgrammer is required for this.

Hope this helps.

John

Show quoted text

---------- Original Message ----------
From: "Brad Kobielusz via Groups.Io" <bkobielusz@...>
Date: November 28, 2019 at 1:42 PM



Hi all
    I am have been looking into the possibility of using decoder pro to
upgrade Micro Select V4 decoders to the FUll Throttle version.   My guess is
that it can’t be done I had considered trying to use a Full Throttle version
to clone to an older version with the same prime mover type. My guess is the
sound file is different enough that it needs a full new sound file loaded from
ESU.   However knowing there are some pretty smart people on this list I
figured I would ask.
Thanks Brad

Re: Webserver layout editor panel showing incorrect block occupation for a crossover

#166485

Andy, I'd suggest a new thread for the performance issue. Include details such as your OS, processor and amount of memory. I'd also include testing with a browser on the server machine.
--SteveT

Re: Fatal crash - JMRI V 4.16

#166484

Hi John,

No need for a system log unless you spot another error !!.
Thanks for confirming sorted,

Steve.

Re: speed matching

#166483

开云体育

PMSL

There is no engineering problem that cannot be resolved by the use of a bigger hammer!

Dave

On 28/11/2019 11:37, Dave Audley wrote:

Show quoted text

I like the last sentence!

?

Dave Audley

?

From: forfoum@...

Sent: Wednesday, November 27, 2019 11:06 PM

To: [email protected]

Subject: Re: [jmriusers] speed matching

?

Googlng, I came upon this instruction sheet for the Bachmann 36-550, 4 function decoder.

? :

Bit 5 of? CV29 must be set too enable " USER defined Speed table "?? When you go into DecoderPro and Bachmann 36-550, Speed Table Tab, you must select the? "Use Speed Table" button? so that CV29. bit 5? gets set to proper value. Fail to do this and the User Speed Table will have no effect.

If DecoderPro can not IDENTIFY the decoder and match it to an existing roster entry,? you have issues with your program setup or the decoder in question.
When you read CV7 and CV8? you should get? 46 and 101, which you are getting from somewhere.? These are also used by DecoderPro to match with the roster.?

If you are still unsuccessful, use Program on Main? and fire out the User Speed Table and? CV29 value to the decoder..

If this still does not work. Give up, give in,? Remove decoder and place on table. Take mallet in hand and " convince " the decoder.?

Marc

Re: Decoder Pro & Tsunami Decoders #tsunami #pr3

#166482

PS
7 - and toggling back and forth unnecessarily on DecoderPro between Service mode programming(Programning track) and Programming on main subsequently to make more changes and test when toggling the DPDT switch was all that was needed.

Re: ESU Select Question

#166481

Brad:

You need the ESU/Lok programmer to download the full project to a Lok decoder. Then you can program the CV with DecoderPro.

If your particular locomotive has an FT project, I think most (US) do.

If you don't have a programmer available check that the sound project you would want is available and let us know with an approximate location and if there's someone who might be able to help we might be able to put folks together.

Happy Thanksgiving!

Jim Albanowski

Re: Decoder Pro & Tsunami Decoders #tsunami #pr3

#166480

Marc, Yes and no, both.

1 - Opening DecoderPro and setting it for Service mode programming.
2 - Opening the ROSTER entry for programming.
3 - Programming/Setting CV values.
4 - Issuing a Write changes to sheet, still getting 308.? Sheet write operations would come back 308 and change to red, even when the actual functionality became enabled upon testing with throttle.
5 - Toggling to Program on Main (Didn't need to do that, just needed to reset the DPDT to main track) - just needed to test with throttle with DPDT switch set to main and not reset to POM.
6 - Testing adjusted functionality and finding that values actually saved as expected, except on ditch light programming(multiple values had to be in sync, so had a conflict of two fx keys calling the same command, my fault. Now ditch lights are alternating).

So basically the programming is working, but I just tested again, by adjusting bulb brightness and although I received a 308 on Write, the change took.? Track was set to Service mode programming.? I can totally live with the 308, if somehow the decoder isn't acking properly.? If there's anything else I can try or adjust my sequence of operations, that would be great.

Thanks again for all of the assistance.? I've worked in support and I know that it can be "interesting".??

Russ