UK Train Horns

[Armchair_Engo]

I am definitely beginning to suspect the sound handling as I experimented with the MEPUltWhistlersnd using three tests with different frequencies, 11025, 22050 and 44100. The horn sounded exactly the same in all three, one high and one foghorn tone. External again was high pitched and fast. I get the feeling that OR is playing the internal sounds at half speed and the externals at double speed.

Looking at this as a separate issue, I tried changing the frequency in the FrequencyCurve() the TRAINS\TRAINSET\common.MEP\MEPUltWhistlersnd\MEPWh istlerCab01.sms file and was able to distinguish a difference each time although I used the as-supplied 5512, then 11025 and 22050.
This was tested on two different systems, one running XP Pro SP3 32bit and one running Win 7 Pro 32bit, using OR 1.0.1.3095 stable release in both systems. Both systems also run on-motherboard sound chips, no fancy add-on sound cards.

Code snippet from my MEPWhistlerCab01.sms:

					Stream (
				Priority ( 6 )
				Triggers ( 1
					Discrete_Trigger ( 8				PlayOneShot	( 1 File ( "x_horn1.wav" -1 )	SelectionMethod ( SequentialSelection )	)	)
				)
				FrequencyCurve(
					Variable2Controlled
					CurvePoints ( 2
					      0		5512
						1		5512
					)
					Granularity ( 1000 )
				)
			)

If I change 5512 to 11025, I get the horn consistent between inside and outside. How pitch-perfect the x_horn1.wav is when played by, for example, Windows Media Player or GoldWave I don't know, but it is consistent with OR running 11025 in that FrequencyCurve(), as near as I can hear.

[baldwin]

Did you edit the horn filenames as well?

Yes I did, but I have found one MEP loco that sounds fine in the cab and outside, the Class 56, so it is not Open Rails at fault, it is the crappy wav files that are causing the problems. I am switching all my locos over to the MEP class 56 horn, which is actually using the same file for in cab and external.

[Armchair_Engo]

Yes I did, but I have found one MEP loco that sounds fine in the cab and outside, the Class 56, so it is not Open Rails at fault, it is the crappy wav files that are causing the problems.

Sorta kinda. That single horn file only works inside to your liking because the 56 class cab sms is not set as Stereo(), so all internal sounds are the same as external sounds by being 176kbps 11025Hz mono, and the 56 class cab horn stream does not use a FreqencyCurve().

I am switching all my locos over to the MEP class 56 horn, which is actually using the same file for in cab and external.

That may be a lot of work. From what I have seen, I guess the FrequencyCurve() kluge is there to enable the use of "external-spec" horn files to be used in a cab .sms which is set to be Stereo().
So you will either have to retain the FrequencyCurve(), or remove any Stereo() set and change out all the other files inside each of the other cab .sms's to the external versions.

BTW, I'm done for the night. I'll be back into this tomorrow night.

[OTTODAD]

Is having to do all this for use in ORTS then likely to change how these horns sound in MSTSBin ?

[NorthernWarrior]

I tried to post the sms file content from the ELR Class 40 but unfortunately the length exceeded the permitted post length!

Which file(s) and which sections do I need to copy and paste from to condense it down a bit?

[Armchair_Engo]

From the top to the end of the horn stream/s should be enough.
Failing that, maybe zip the cab sms file in Windows and add that as an attachment. Right click on the cab sms, Send to... Compressed (zipped) folder

[Armchair_Engo]

Hi Vern !

Here I am comparing the horn sounds of a Mid-East-Plus BR-Class-40 in MSTSBin and ORTS.

https://youtu.be/XPmDW-J2ze8

As NorthernWarrior noticed, when the horn sounds in that MEP 40 class cab in OR, the horn starts correctly then about halfway through it slows.
I get that as well. Also, if the horn is run more than once while still in the cab, the entire .wav runs slow.

There seems to be a difference in the way that MSTS and ORTS cope with the combination of a Stereo() flag, a mono .wav and a FrequencyCurve().

[baldwin]

Something strange going on. I am now using Win 10, don't know if that would affect this. I changed the horn for the MEP class 40 to that used in the default Dash 9, renamed the originals and moved them into a different folder as well. I also changed the cab sms to mono. When I run the class 40 in OR I still get the two tone horn in the cab and outside. I have also just loaded the East Lancs route mentioned by NorthernWarrior and the class 40 in there is even worse.

[Armchair_Engo]

I am now using Win 10, don't know if that would affect this.

I suspect it does.

I changed the horn for the MEP class 40 to that used in the default Dash 9, renamed the originals and moved them into a different folder as well. I also changed the cab sms to mono. When I run the class 40 in OR I still get the two tone horn in the cab and outside.

Hmm, a Vista-onward OS, edits that don't seems to have effect or stick... If I didn't know any better then I would say it's a \Program Files(x86) issue, but you have been around long enough to know better.
Can I guess that Win10 was an upgrade and not a clean install?
Just for kicks, where are you shortcuts pointing to?
Can you see your edits in the cab sms?
Did the "Date Modified" for the cab sms change?

I have also just loaded the East Lancs route mentioned by NorthernWarrior and the class 40 in there is even worse.

I don't have that route. At a guess, the class 40 in that bundle used an earlier iteration of the sound pack used in the MEP class 40.

[NorthernWarrior]

For some strange reason I couldn't find any reference to the horn at all in the ELR Class 40 common sound entries. The following found for the Class 37 however:

Class37Cab01.sms:

)
Stream (
Priority ( 6 )
Triggers ( 1
Discrete_Trigger ( 8 PlayOneShot ( 1 File ( "x_horn1.wav" -1 ) SelectionMethod ( SequentialSelection ) ) )
)
FrequencyCurve(
Variable2Controlled
CurvePoints ( 2
0 5512
1 5512
)
Granularity ( 1000 )
)
)

And for Class37Eng01.sms:

)
Stream (
Priority ( 6 )
Triggers ( 1
Discrete_Trigger ( 8 PlayOneShot ( 1 File ( "x_horn1.wav" -1 ) SelectionMethod ( SequentialSelection ) ) )
)
)

Hope this helps - all the sms files seem to have been created by a chap called Stuart Williamson - not sure if he's still about and could shed some light on whether there is a conflict between his scripting and OR, or whether OR itself is broken?

Edit - with regard to the Win 10 issue, the problem was manifesting itself under Win 7 so that's possibly a red herring.

[baldwin]

I suspect it does.


Hmm, a Vista-onward OS, edits that don't seems to have effect or stick... If I didn't know any better then I would say it's a \Program Files(x86) issue, but you have been around long enough to know better.
Can I guess that Win10 was an upgrade and not a clean install?
Just for kicks, where are you shortcuts pointing to?
Can you see your edits in the cab sms?
Did the "Date Modified" for the cab sms change?

All trainsim items are on two non OS drives and do not even reside in a program files folder and I have full permissions on both drives.
Win10 was an upgrade.
All shortcuts point to the correct locations.
All edits in cab and eng sms and eng files are visible.
Date modified for all altered files changed to the change date.
I can find nothing related in the VirtualStore either.

Never had this problem with Win 7. It is almost as though OR itself is storing something somewhere.

[Armchair_Engo]

For some strange reason I couldn't find any reference to the horn at all in the ELR Class 40 common sound entries. The following found for the Class 37 however:

Class37Cab01.sms:

)
Stream (
Priority ( 6 )
Triggers ( 1
Discrete_Trigger ( 8 PlayOneShot ( 1 File ( "x_horn1.wav" -1 ) SelectionMethod ( SequentialSelection ) ) )
)
FrequencyCurve(
Variable2Controlled
CurvePoints ( 2
0 5512
1 5512
)
Granularity ( 1000 )
)
)

And for Class37Eng01.sms:

)
Stream (
Priority ( 6 )
Triggers ( 1
Discrete_Trigger ( 8 PlayOneShot ( 1 File ( "x_horn1.wav" -1 ) SelectionMethod ( SequentialSelection ) ) )
)
)

It looks quite familair.
I'll assume that the cab .sms has Stereo() somewhere near the top of the file, and I'll assume that both the cab .sms and the eng .sms
use that same external-spec horn .wav.

Hope this helps - all the sms files seem to have been created by a chap called Stuart Williamson - not sure if he's still about and could shed some light on whether there is a conflict between his scripting and OR, or whether OR itself is broken?

Mr Williamson seems to have been somwhat of a powerhouse in the UK MSTS sound scene, at least in the early years.
Why there would have been any benefit in using an external, mono horn sound inside a Stereo() cab .sms, and slowing down that sound with a FrequencyCurve() so it sounds normal, I do not know. I have to assume there was some reason for that back around 2003 (possibly long before ORTS was even a pipe-dream, and years before MSTSBIN existed). As hardware has improved and OS's evolved, that reason may not be quite as valid as it used to be.
Mr Williamson uses the handle "scefhwil" on UKTS.

I would hesitate to call OR "broken". I would say "There seems to be a difference in the way that MSTS and ORTS cope with the combination of a Stereo() flag, a mono .wav and a FrequencyCurve()."

[slipperman]

Hi Gentlemen,
Although I am not currently running Open Rails under Win 10, I am following this thread with interest
May I throw in a shot in the dark? I notice that OpenRails.exe doesn't have the "run as administrator" indicator set, although, from what little C# experience I have, this can be set in the compiler. If this indicator is "manually" set for all the relevant executables, do the problems still occur?

Cheers,
Ged

[baldwin]

I come offering 'humble pie'

After a more thorough investigation it turns out that I have two installs of the MEP on two different drives, and I was modding the one NOT linked to Open Rails. Rectified that this morning and mods are working. I changed the class 40 cab sms to mono, and the in cab horn sounds the same as the exterior sound. Without actually looking at the wav in detail, it looks like the in cab horn wav is recorded as mono and having a stereo cab gives the mournful hooter.
So, I returned the cab to stereo and changed the frequency to 11025 which again gave identical inside and outside sounds. The stream now looks like

 	Stream (
				Priority ( 6 )
				Triggers ( 1
					Discrete_Trigger ( 8				PlayOneShot	( 1 File ( "x_horn1.wav" -1 )	SelectionMethod ( SequentialSelection )	)	)
				)
                                FrequencyCurve(
					Variable2Controlled
					CurvePoints ( 2
					      0		11025
						1	11025
					)
					Granularity ( 0 )
				)
				)

[Armchair_Engo]

After a more thorough investigation it turns out that I have two installs of the MEP on two different drives, and I was modding the one NOT linked to Open Rails.

To quote a Corn Flakes advert from years ago: "The simple things are often the best".

Without actually looking at the wav in detail, it looks like the in cab horn wav is recorded as mono and having a stereo cab gives the mournful hooter.

The horn .wav you are using is, by MSTS naming convention, an external sound, confirmed by the mono status. The properties as stated by Windows Explorer should confirm that it is a bitrate of 176kbps. XP gives more useful information such as channels, 16bit PCM etc. That's progress for you.

So, I returned the cab to stereo and changed the frequency to 11025 which again gave identical inside and outside sounds. The stream now looks like

 	Stream (
				Priority ( 6 )
				Triggers ( 1
					Discrete_Trigger ( 8				PlayOneShot	( 1 File ( "x_horn1.wav" -1 )	SelectionMethod ( SequentialSelection )	)	)
				)
                                FrequencyCurve(
					Variable2Controlled
					CurvePoints ( 2
					      0		11025
						1	11025
					)
					Granularity ( 0 )
				)
				)

If your {whatever}\TRAINS\TRAINSET\common.MEP\MEPUltWhistl ersnd folder is like mine and contains a horn1.wav file from 2003 with a bitrate of 352kbps, that is the "correct" stereo horn sound to use in the cab .sms. If you use horn.wav, you can dispense with the FrequencyCurve() and the horn should sound correct in both MSTS and ORTS.

I'm a little surprised that Granularity(0) works.