I've been working on the steam loco "chuffs .....

[krswen]

The spacing between *2.wav, *3.wav and *4.wav chuffs were quite non-linear. With the help of GoldWave I was able to snip and insert "quiet" passages and get the progression more acceptable. No, it's not directly linked to wheel motion as we would all like, but at least the frequencies are matched periodically.

Then I reworked the *cab.sms and *eng.sms files to "center" each of the 8 volume/speed curves on 4 chuffs per wheel revolution for a two cylinder loco with 63" drivers (Mikado). So now the motion and sound are correct eight times as you go up or down in speed. Above about 40mph you're maxed out, but the human eye can no longer distinguish the motion anyway (or maybe it's the video card). And other than crack passenger, most steam trains rarely spent much time over 40 anyway.

I'm going to write an excel file to make the job easier for other cylinder/driver combinations and tune up the hand-off from each volume curve to the next before I upload anything. I'm still not completely happy with a couple of my transitions.

Does anybody know if eight interleaved curves are the maximum you can have in the .sms file? Or did KUJU just get tired of coding (it IS a rather tedious job to plot the curves and synchronize the sounds). They also don't have matching numbers of stereo (in the cab) and mono (outside) files. Fortunately, with Goldwave you can convert back and forth.

The other thing I noticed, one of the 380 .sms files calls for a *8.wav file and there isn't one! Did anyone ever get the train going fast enough to need that file and crash the sim? Or, what does happen when it can't find a file?

By the way, the numbers in the tables that look like they would be mph are off by about 2.2, but not even that is consistent. No wonder the default chuffs sound goofy.

Should have more to report after the weekend.

krswen

[SP9800]

>Or, what
>does happen when it can't
>find a file?

It simply won't play it.

[timken]

Sounds like you are working up a pretty good compromise solution!

I've redone the sound files for my engines and gotten them so they have 4 chuffs at 4, 10, 18 and about 24 mph. I also happen to have access to nice recordings of the real thing and have spent a great deal of time around steam. I've got about 600 hours of firing, 60 hours of running, and even about 100 hours of operting a steam logging donkey engine! It really helps to know what they should sound like!

Another item in steam engines, They sound different in the cab than outside. In the cab, the sound comes partly through the firebox, and has a deep, gut whacking chug, at least when working hard. There's also a lot of misc. rattling and other racket, steam hissing, and various clanking.
Outside, the exhaust is sharper, with less of a deep chug.

I've been changing my cab sounds with Sound Forge's EQ, humping up the bass, and slightly depressing the treble. It gives a nice effect, when you go from in cab to cab out view. The result isn't perfect, but much closer to my experience.

Double heading 2 identical steam locos in MSTS sounds very strange, with a flange type echo when you are between them. Rarely would 2 different engines of the same class have identical diameter drivers. They replace the tires after a certain loss of material, from turning them down. New tires were even sometimes applied that were 4" thick instead of 3". I think the minimum was 1 1/2" when they had to be replaced. So double headers rarely stay in sync with each other. To get this effect, I've renamed a 2nd engine, and shortened the chug files by 4% on this one. The effect is just great!

Steve

[scefhwil]

For faster sounds take a look here:-

http://members.ozemail.com.au/~captmoo/

The numbers in the tables are in metres per second, the coversion factor from mph being 2.25 (ish).

Note: In my experience mapping the .wav period to the wheel rotation is straight forward (although a spreadsheet certainly helps) when using the Speed variable, but not so when using Variable1, which seems to have a different magnitude to Speed. Have you found this? It seems to vary also from loco to loco.

[moward]

Re your web page, the Scotsman is indeed a 3 cylinder engine, so 6 beats per wheel rev is correct. My first correction to the slow beats was to backup the cruise *1 and *2 files. Then I renamed *3 as *1, *4 as *2 etc. The last files are simply duplicates of *9. This gave great synch at approx 5 - 10 mph, particularly around 8 mph in runby mode. Higher speeds seem better as well, but there is a big jump on each change.

Also, the chuffs don't start until 4mph. They should really start on the slow chuffs (original files *1 and *2) after half a wheel rev - when the first cylinder exhausts. The original *1 and *2 wav files are only for 1 to 4 mph.

I too would like to be able to work on the problem of cutting the chuffs when regulator is closed or say less than 5%. How is the SMS file viewed or edited? I only have a slow system - 300MHz PII so I have to run on lower settings.

[scefhwil]

Hi moward,

>Re your web page,

Sorry if I have misled, but the web page belongs to Murray Grant, not me.

>the Scotsman is indeed a 3 cylinder engine, so 6 beats per
>wheel rev is correct. My first correction to the slow beats
>was to backup the cruise *1 and *2 files. Then I renamed *3 as
>*1, *4 as *2 etc. The last files are simply duplicates of *9.
>This gave great synch at approx 5 - 10 mph, particularly around
>8 mph in runby mode. Higher speeds seem better as well,
>but there is a big jump on each change.
>
>Also, the chuffs don't start until 4mph. They should really >start on the slow chuffs (original files *1 and *2) after
>half a wheel rev - when the first cylinder exhausts.
>The original *1 and *2 wav files are only for 1 to 4 mph.

Do your comments relate to Murray's extra files? You need to use them with the original sounds and using his revised .sms files. The period of the beats in each .wav file determines the wheel rev they relate to dependant on the wheel diameter and the number of beats per rev required. So using the correct .wav and .sms files together is important.

>I too would like to be able to work on the problem of cutting >the chuffs when regulator is closed or say less than 5%. How
>is the SMS file viewed or edited? I only have a slow system - >300MHz PII so I have to run on lower settings.

I have developed a means of controlling the chuff volume, including quiet at zero regulator, but so far only for external sound. It uses Murray's extra .wav files and some edited versions of the sim's originals. It's a compromise solution though because you have to loose the way the .wav files are changed over using the fade out/fade in method. This results in a much more distinct, instantanious change over. I have posted more details on it on previous (pre sound forum) threads. I think it's overall better than the original. It also has as good a 6 beats per rev timing as the MSTS sound system will allow, has a wheel slip effect and some speed-varying rail joint sound. It was available for download from trainsimfiles.co.uk before that site closed down. It's currently not available whilst I'm working on other changes that will improve the cross-over issue.

The .sms files are in unicode format so a compatible editor is needed. Open them up from within WordPad.

Bad news if you have to run the sounds at anything less than High settings. You will only hear some of the wav files and not all of them.

Stuart Williamson