[gpz]

I'm afraid a test case longer than some minutes would be over a kind of an annoyance margin. ;)

dennisat, on 26 December 2013 - 10:03 AM, said:

If you don't do this you get the occasional SystemException

That is exactly the reason why it is there: so you shouldn't remove that piece.

dennisat, on 26 December 2013 - 10:03 AM, said:

I have moved the bulk of the code change in X1849 back to Sound.cs

Reverting changes is not a solution, that is nothing more, just reverting changes. :) (I know which part of the code causes the problem, the challenge is to find the critical line or character in it. And the problem is so rare, that the hardest thing is just to find a test case.)

[cjakeman]

Hi BB25187,

gpz, on 26 December 2013 - 02:24 PM, said:

And the problem is so rare, that the hardest thing is just to find a test case.)

Hope I'm not "teaching my grandmother to suck eggs" by saying this, but once you've repeated the problem, I hope you'll be using F2 Save and then exporting an ORSavePack which you can send to gpz so he can replay (from the start) to reproduce the problem as often as needed.

Just my thoughts,

[dennisat]

gpz, on 26 December 2013 - 02:24 PM, said:

...That is exactly the reason why it is there: so you shouldn't remove that piece....
Reverting changes is not a solution....

I phrased myself badly, I haven't just restored the original code. My solution (depends how you look at it I suppose) achieves the purposes of X1849 (as I understand them) without courting thread serialization issues. As I said previously, I'm teaching myself C# as I go along.

Incidentally, I think sometimes you get the impression that sounds have vanished when you are "head out" because the AL Listener seems to fade sounds "behind" you. This seems to be more marked as the speed increases. You can try this out easily by rotating your point of view when "head out" to look rearwards. When I do this I can hear the train sounds increasing as I rotate but still not as loud as I'd expect. Perhaps the AL Listener has this behaviour to produce a good 3D sound and Doppler effect? Or perhaps I'm imagining it all, I have no sound meter to make genuine measurements.

Dennis

[BB25187]

Hi,

cjakeman, on 27 December 2013 - 01:23 AM, said:

Hope I'm not "teaching my grandmother to suck eggs" by saying this, but once you've repeated the problem, I hope you'll be using F2 Save and then exporting an ORSavePack which you can send to gpz so he can replay (from the start) to reproduce the problem as often as needed.

When Peter told about saving the activity, I had the save/resume stuff in mind. But you are right: a replay may be helpful.

gpz, on 26 December 2013 - 02:24 PM, said:

I'm afraid a test case longer than some minutes would be over a kind of an annoyance margin. :lol2:

The question is: did anyone observe the problem on a very short activity? Would it be acceptable for you to use the replay mechanism to save your time?
Otherwise, we may have another option: if you send me some svn patches, I can test them for you. This way, you won't spend time in testing. On my side, I will save the effort of preparing the test setup. Just tell me your preference.

Regards

[Csantucci]

One idea to reduce testing time could also be that of trying with accelerated time speed (debug speed up).

[gpz]

Could I ask you all to test, if the issue disappeared in r1916, please? (Downloadable from here.)

[BB25187]

Hi,

gpz, on 03 January 2014 - 05:24 AM, said:

Could I ask you all to test, if the issue disappeared in r1916, please? (Downloadable from here.)

Thanks for the modification.
I will test this revision tomorrow. I let you know once done.

Regards

[BB25187]

Hi,

I tested revision #1917. I could complete the activity which exposed the issue without observing any loss of sound. So, the fix seems to be a good candidate. Nevertheless, I got an exception when exiting the activity (thus, not a real trouble, but maybe a symptom that something is still wrong):

Error: System.NullReferenceException: La référence d'objet n'est pas définie à une instance d'un objet.
   à ORTS.Processes.UpdaterProcess.Update()
   à ORTS.Processes.UpdaterProcess.DoUpdate()
   à ORTS.Processes.UpdaterProcess.UpdaterThread()
   à System.Threading.ThreadHelper.ThreadStart_Context(Object state)
   à System.Threading.ExecutionContext.Run(ExecutionContext executionContext, ContextCallback callback, Object state)
   à System.Threading.ThreadHelper.ThreadStart()

On my side, I will try to run longer activities which are likely to expose the issue, and let you know what is going on.

Regards

[dcarleton]

gpz, on 03 January 2014 - 05:24 AM, said:

Could I ask you all to test, if the issue disappeared in r1916, please? (Downloadable from here.)

It is an improvement, but the player locomotive sounds still do eventually disappear. The train keeps moving but the sounds stay behind.

[Sid P.]

gpz, on 03 January 2014 - 05:24 AM, said:

Could I ask you all to test, if the issue disappeared in r1916, please? (Downloadable from here.)

Running Surfliner 2 "Autoracks to San Diego" in X1919 - looks OK - the horn sound does not vanish either.

(Oops! A little later - ) After 50 minutes of activity time, the exterior sounds in 2 view were low again, and horn sounds were only heard in cabview (1). Exit and resume fixed it.