MOTM

John Laudicini writes:
>>It could not be better.  How many tracks does it take to record 
something like this?<<

Actually it could be a bit better.  It is really just a test to see 
how well something like this could be done in a little over a day.  
I'd like to spend a few days and do some things differently.

As far a track count (all my tracks are stereo), I usually use more 
than I need to because I have plenty available and a fast enough 
computer.  For instance if I want to use the same wheel sound twice 
and they will overlap each other, I use a second track.  Sometimes I 
use addition tracks for the same sounds because I'm too lazy to draw 
volume envelopes on all the sounds on a single track, so I just put 
them in other tracks and just set the mixer slider levels instead.  
Not all of the sounds are playing at the same time.  So it's hard to 
put a number on it.  You can hear the horn, bell, track clicks 
(barely audible), a faked general noise filler, plus some a couple of 
train sounds ontop of that, so there may be a half dozen sounds at 
any one time.

John Blacet writes:
>>Needs some good bass rumble.<<

That's something I'd like to add more of too.  On some sounds I 
already have EQ bass set to max, and it still wasn't enough.  If 
available, the listener can also crank up the bass or hit the 
loudness button to add a bit more.

groovyshamen writes:
>>You can hear all of the parts moving, the squeaky wheel on a 
particular car going by, different wheel-different squeak, etc. and 
they raise and lower in volume and pan as expected with appropriate 
depth added.  Even a touch of doppler effect on the whistle - cool.<<

Glad the doppler effect is audible.  It was more obvious before I 
added the train sound ontop of it.  Some of the panning is a bit 
off.  One sound I'm using a pedal for that, another I'm manually 
turning a pot, another I tried using an Encore UEG to control 
panning.  What I really needed was a simulataneous up ramp in sync 
with a triangle, both positive voltages only, the ramp needing to go 
from 0 to 10 volts (to control the dotcom panner), and they shouldn't 
return back to their staring point when done.  That's almost 
impossible on an analog synth.  I still need to come up with a decent 
solution to that or perhaps just do panning on the synth and volume 
envelopes on the computer, though that usually cuts off the reverb.

>>I'm curious how many different patches this took?<<

Since quite a few of the sounds were modifications of one patch, it's 
hard to answer completely.  But it is something like the Horn (and 
modifications of it for distance), Bell, Engine, and then everything 
else being a modification of the same general chaotic metalic noise 
patch.  So it's really only about 4 base patches plus knob tweaks.

>>And this was completely MOTM stuff?<<

MOTM plus stereo chorus (to stereofy mono sounds), parametric EQ (for 
more bold formants and resonant sounds), Aphex aural exciter (used on 
horn sound), digital delay (used for some comb filtering, and also to 
delay the horn sound so it reached your ears at different times 
depending on its location as in real life), and reverb.

>>And one other thought, did you start with a recording of a train 
going by and work from that, or did you work from a set of sounds in 
your head?<<

It was from sounds in my head.  I've had to listen to real trains 
around here many times, so I've paid attention to some of the sounds 
they make.  I didn't want to use a real recording as a referrence 
because that tends to cause one to get bogged down trying to get as 
close as possible as opposed to just getting it done.  Actually there 
was just one exception.  In a real train recording I filtered out all 
of the sound except what I believed were the first harmonics of the 
two horn notes.  Turned out I already had the intervals correct and 
was only off by a semitone.  I made the change to my horn sound 
before recording it.

Andrew Sanchez writes:
>>How did you go about getting the sounds of the wheels (or whatever 
the source is of that sound in real life)? Sorry if that is a little 
vague.<<

I don't quite know what you are considering the wheel sounds.  When a 
train passes by me, I'm hearing a lot of high pitched metalic 
screeches that I'm never quite sure whether it's the wheels scraping 
on the track or brakes being applied.  There is also the bouncing and 
rattling sound of the train cars.  Both of those sounds used noise 
modulated VCO's going through filters in parallel, into a ring mod, 
into more filtering, into parametric EQ, into stereo chorus, and I 
can't remember which sounds exactly, but possibly a bit of comb 
filtering (or not).

>>Another question I have is, did you do this solely by your memory 
of the train's sound<<

As per above, it was from memory.

Richard Brewster writes:
>>A very realistic simulation, Elhardt!  Thanks for sharing it.  I am 
interested in the train horn sounds and have wondered about their 
tunings.<<

There are probably a few different tunings.  I used pitches that fall 
on the keys of E4 and A#4.  I noticed in some train pictures that 
some trains have a small third horn pointing backwards.  Whether it's 
a duplicated of the forward facing horn or at another pitch, I don't 
know.

KA4HJH writes:
>>Now, speaking as a longtime railfan who's stood alongside the 
tracks at 3:00AM and just drank in this sort of thing...there should 
be much more of a roar as the locomotives pass.<<

Yes, I agree.  But as a rushed synth job to guage an idea and not 
wanting to over power the horn sound, there were comprimises.  If I 
go back to make changes, I'll see what I can do.

>>Also, the "song" of the horns is close but not quite right. The 
pattern for approaching a grade crossing goes like this:  long long 
short looooonnnnngggg (with the last one whizzing by center stage 
just as it's about to fade out for maximum dramatic effect).<<

As a layman like myself, I didn't know the pattern.  And as somebody 
who can hear trains in the distance from my house, the horn patterns 
here seem so random.  Some trains pass with a few toots, and others 
are just constantly (and rudely, especially at 4:30 in the morning) 
blasting very long and large numbers of sounds.

-Elhardt