Enginespec to Enginesound relationship for geared locos?

I am building a Class A Special Order (2 Cylinder) Shay, which was at the top of the power output for the 2 cylinder Lima versions. Everything is working visually, but I am having trouble getting the piston stroke sounds to come anything close to the actual cylinder movements. I have had lots of experience with conventional drive Americans and Moguls, but this is my first attempt at a geared loco.

The design calls for a gear ratio of 37 engine revolutions to 17 Driver revolutions. I do not know how the enginesound picks up revolutions from the enginespec, and therefore do not know how to mess with the revs to get to the appropriate sound cycles. To further increase the problem, I cannot find current code for either class, and the Wiki is very vague for revision levels above 2.5 regarding steam engines since the steam-locomotive was done away with.

Looking for some guidance as to how to get the visual and sound cycles close, if not actually synced.

Try this Forum Member,

He is very knowledgeable......Give a little time for more ideas from other members who can help you.......

https://forums.auran.com/trainz/member.php?412211-shebashetan7

Steam pulses are related to drive wheel circumference and are modified by speed over the ground. Work back to get the notional wheel circumference to match your drive gear ratio. It is possible that you could assume that in your config your wheel diameter is smaller than the drive wheels.

Narrowgauge

Create an invisible bogey with the correct size to give you double the number of revs of the engine and specify it as direct-drive 1 This method was used by some creators in TRS2004 to give four chuffs per revolution of the drive wheels. However with a geared loco the sound may come out as a continual hiss with that gearing, i.e. there will be very little gap between chuffs. You may have to crop the ends of the existing sound.

Cheers,
Bill69

narrowgauge said:

Steam pulses are related to drive wheel circumference and are modified by speed over the ground. Work back to get the notional wheel circumference to match your drive gear ratio. It is possible that you could assume that in your config your wheel diameter is smaller than the drive wheels.

Narrowgauge

Click to expand...

Thanks, that is usefull information. Are the script sources publically available for the enginespec and soundspec code at the current revision levels? I think I know the tags, but I am not sure that I know the full implications of each. The code would be of great help to see what is expected. All I can find on-line, other than examples, is the old and outdated steam-engine class. The Wiki is not very helpfull and contains little information on the steam contaner and related tags of the engine spec.

Bill69 said:

Create an invisible bogey with the correct size to give you double the number of revs of the engine and specify it as direct-drive 1 This method was used by some creators in TRS2004 to give four chuffs per revolution of the drive wheels. However with a geared loco the sound may come out as a continual hiss with that gearing, i.e. there will be very little gap between chuffs. You may have to crop the ends of the existing sound.

Cheers,
Bill69

Click to expand...

Thanks, got it. The invisible bogey was the missing link. I created the invisible one onto a center attach point with a gear radio equivalent reduction in the circumference of the wheel and made it the direct-drive. That with the retained distance to travel per frame resulted in an increase in engine RPM up to very near the actual. Then I doubled the power strokes in the engine spec to simulate the closed dual stroke pistons. That provided a very pleasing short, sharp chuff that is almost a hiss but not quite. It synced nicely with the drive shaft, which is turning at 1/2 the rpm that the invisible bogey thinks it is turning at. The pressure relief valve is only engaging very infrequently, so I am very happy with the result. Shay Shop Number 2392, a Class A20 - 36 is now fully operational with the Pacific Timber Company and using the Tacoma & Eastern rail under contract.

Further development. Does anybody know how to change the sound mix for the wave files in enginesound? All of the files are working, but the "steam_loop_***rpm.wav" files do not reproduce with enough volume and are in the background behind the safety and compressor .wav sounds. I have been unable to find tags or script examples that deal with volume mix, and I have reservations about just increasing the amplitude of the .wav files in Audacity. Anyone have any ideas?

Robert704 said:

Further development. Does anybody know how to change the sound mix for the wave files in enginesound? All of the files are working, but the "steam_loop_***rpm.wav" files do not reproduce with enough volume and are in the background behind the safety and compressor .wav sounds. I have been unable to find tags or script examples that deal with volume mix, and I have reservations about just increasing the amplitude of the .wav files in Audacity. Anyone have any ideas?

Click to expand...

I think Audacity is the only way to do that. As well as slightly increasing the amplitude of the steam sounds decrease the amplitude of the safety and compressor sounds until you get the required ballance.

Cheers,
Bill69

When I did the sounds for this project, I created a sound script within the KIND vehicle that was based on bogie velocity for the direct-drive bogie. I worked to get the sounds I wanted, but I could not shift the time period (tempo) between chuffs based on velocity without some programming on the wave file contents. So I tried to use the KIND enginesound, and was surprised that when I moved the wave files here, the audio went away for the rpm based steam_loops.

I then discovered from the Wiki that these files are not used for any speed under 40kph (about 24 mph). Heck, they will never be used for any geared loco with a top speed well under that value, virtually all geared logging locomotives. The Wiki may be out of date here, but from my own experience, this appears to be valid information.

Does anyone know if there is a way to override the 40kph minimum in KIND enginesound? There is no .gs file for this, so I assume it is somewhere in the encrypted stuff and will require a developer to get at it. If so, it would useful to have a tag in KIND enginesound for Geared_Loco or Slow_Loco that would lower this 40kph threshold.

Robert I checked this wiki page here and couldn't find any reference to the 40kph limit you're referring to in your post above. In fact it sounds like you looking at something that is describing the old system used before TRS09 IIRC.

The new system uses wav files keyed to RPM from 30 up to 2000RPM. The rpm is keyed to the bogey with the direct-drive 1 tag. In the case of geared locos like Shays this should be the bogey that represents the drive shaft not any of the drive wheels.

It should be noted that unlike the earlier steam loop, which was based on track speed, rather than wheel rotation speed, this system enables sound files to be shared between locomotives with different wheel sizes. 60RPM (which is one revolution per second) on a 2 cylinder loco will produce four chuffs per second irrespective of the wheel size - the only difference this makes is what speed the train will be doing when the drive wheels are rotating at 60RPM.

Click to expand...

For the A class your working with - I assume an A class 18-2 (ABLE) Shay:
drive wheel dia is 29 inches
std gear ratio is 1 to 3.07
alt gear ratio is 1 to 2.17

Using the alt gear ratio of 2.17 since you mentioned it in the OP:

Drive wheel makes 5280x12/(29xpi) = 695.45 revs per mile

At the lowest wav - 30RPM - the drive wheels make 30x60/2.17 = 829.49 revs per hour

With the drive shaft turning at 30 rpm the Shay is traveling at = 829.49/695.45 = 1.193 mph.

The looping wav files should cover the full range of speed of the Shay with no problems using wavs up to 400RPM (15.9 mph).

Bob pearson

Thanks Bob for taking the time to look at this. I think I found the problem. The engine is an A20-2 (class code ABLER). Shop No 2392 was a "Special" (8" Cylinders) and had only 26" drivers with the 3.071 gear set. The Pacific Timber Company was a Mexican concern that operated it in the Pacific Northwest USA to harvest logs for export from Tacoma, WA to their sawmills in Mexico (often on wood rails). They trans-shipped harvested logs from the Cascades south of Mt. Rainer via the Tacoma Eastern RR to the Port of Tacoma. When Weyerhaeuser moved in and bought the TE, they took their equipment back to Mexico, and it became the property of the [FONT=&quot]Colima Lumber Company.[/FONT]

https://www.shaylocomotives.com/data/dataframe.htm

This type was typical for backwoods power where operating cost was king and speed was not an issue. That means a top speed of only 10.4 MPH. All things converted, at 1 MPH this gives Meters/Sec of 0.447, A wheel RPM of 25.85 RPM and a driveshaft RPM of 79.41 (direct-drive). At top speed that means that the driveshaft is at 825 RPM, over twice the frequency in the standard set of .wav files for the geared locos sound files downloads. I did not have sound files that went above 400RPM in the asset. I cannot find the .zip download that has the full set of .wav files at the moment, but am looking for them in other enginesound downloads.

More later,

Maybe I'm missing something but I get some different numbers. BTW you don't have to consider meters per sec here just the rpm of the drive wheel and drive shaft. That assumes you have the correct animdist for the drive shaft in the config and drive wheels in the bogey configs.

26 in dia wheel has circumference of 26 x pi = 81.68 inches so animdist of drive wheel is 2.075 meters
in 1 mile the drive wheel turns 5280 x 12 / 81.68 = 775.7 revs
at 1 mph the drive wheel turns at 775.7 rev per hour
rpm of the drive wheel at 1 mph is 775.7 / 60 = 12.93 rpm

Drive shaft turns at 3.071 times the drive wheel rpm so:
at 1 mph the drive shaft turns at 12.93 x 3.071 = 39.71 rpm
at 10.4 mph the drive shaft turns at 39.71 x 10.4 = 412.90 rpm

This is basically within the rpm ranges I posted above.

The animdist of the drive shaft bogey should be drive wheel circumference / gear ratio = 2.075 / 3.017 = 0.688 meters.

Bob Pearson

PS. When you 1st started this thread I was going to reply but didn't finish a response. I've played with geared locos a bit and had no problems with the sounds or especs. It looked like the replies you got were getting you there so to speak. Ben Neal's B class 28-2 Shay uses a drive shaft animation that includes 2 revs of the drive shaft just to confuse things a bit - I guess. All the figures above assume the animations are done over 1 rev per drive wheel and 1 rev per drive shaft. To add insult to injury in past version of Trainz the animation distance was defined differently for direct-drive 1 steam bogeys in CAB control mode and DCC/AI control mode. Whether it affected the visual rpm of the drive wheels and sound was dependent on how content creators set up the animation sequences. I don't know if N3V has cleaned up that mess or not and frankly have given up caring. Fortunately it didn't effect all steam locos.

PPS. You can use metric through out the calcs and get similar results
26 inch dia is 0.660 meters and the circumference is 0.660 x pi =2.075 m

1 mile is 1609.34 meters and 1 mph is 0.4470 m/s
at 1 mph the drive wheel rpm is 0.4470 m/s x 60 s/min / 2.075 m/rev = 12.93 rev per min
same as above

PPPS. Great site for Shay data I've used it a lot myself. Glad to see it's still going and adding info. As for the rpm range of steam loco engines railway engineers used "piston speed" as a metric in the design process. Calculated as 2 x stroke(ft) x rpm in units of ft/min it was not a max since the travel is basically a modified sinusoid. In the 19th and early 20th century in the US I think 1000 ft/min was considered an optimum value but not the maximum an engine could operate at. In the golden age steam loco engines were under 1800-2000 ft/min at top speed.

Your Lima # 2392 at 10.4 mph would have a piston speed of 2 x (12 / 12) x 412.9 = 826 ft/min and based on this I'd say it could probably do 12-13 mph.

Stupid mistake, mine! Early in my spreadsheet, radius instead of diameter got used to calculate circumference. The RPM came out 2x the actual. Fixed now.

Also made a few interesting discoveries regarding engine sound. After making the change above, the engine was still not producing any sort of relative sound, but there was a faint background audio. On a hunch I boosted the audio (with peak clipping allowed) in the wave files by 4db with a 7db treble gain and all of a sudden I had engine sounds. Some further tweaking by track combining all of the stationary sounds into the slow, med and fast files got me the compressor and dynamo into the engine idle.

I suspect that the engine sound playback uses amplitude peak detection to set the volume for the playback. Steam Engine sounds are rich in peak transients, so using the above amplification with clipping got me decent amplitudes for the playback. Still tweaking on that, but the result is very pleasing and much like my recorded Shay sounds from a field trip to Mt. Rainer Scenic RR at Mineral Lake, WA. It is sort of like a hissing sewing machine at speed or perhaps a hissing machine gun, a very rapid, broken hiss.

None of the above is documented in the Wiki, but it should be. I would do it, but I simply do not have enough confirmed information.

Thanks again, everyone, for your comments and assistance.