DCC Impulses Column - 3
3: Blackstone HOn3 Locomotive and cars.
3
within the company – we have two former locomotive
engineers here on staff! Some of us are modelers, too, and
we don’t believe in putting out anything that we would not be
proud to have on our own layouts. Of course, the 16-bit sound
(we were the first to do this!) is huge. Anyone who has listened
to 8-bit sound decoders can instantly hear the difference.
MRH:
Do you "synthesize" sounds that are not available for
live recording?
ST:
No. On rare occasions, we have manipulated digital
recordings to replicate something that no longer exists, i.e., a
tonal shift perhaps to make one exhaust chuff into another, but
they are not ‘made-up’.
MRH:
So you have to find a functioning entity for your
recording (motor, whistle, air pump, etc.) and send folks there
to record it.
Yes. In general, it takes from a half day to two days to collect
the raw recordings for each engine. It also typically takes two
people.
MRH:
I've visited your facility several times and have always
been impressed with your sound editing facility. Give our
readers a tour of it, please.
ST:
When we had our building designed, we included a special
room (5) that would eliminate unwanted echoes and isolate
the sound engineer from other sounds within the building.
We use a Mac computer, with the latest in hardware and
software tools.
MRH:
Explain what goes into taking a sound file from your
raw recording and turning it into a sample ready for inclusion
in a decoder.
ST:
The raw sounds of course, need to be put into the digital
format that we use for editing. There is a lot of cleanup –
regardless of how careful you are – sometimes you can’t
control sounds occurring around you during the recording
process. Most of what happens next is to make the sounds fit
with our software design. Lots of testing, lots of tweaks!
MRH:
Can you estimate the number of man-hours that goes
into a new sound set? Starting from the field recording and
going through your sound room and making the files ready for
inclusion in a Tsunami decoder.
ST:
Well this certainly varies. Obviously a trip to England or
Japan to record takes longer to collect the raw recording than
popping in down the street at the Durango & Silverton Narrow
Gauge Railroad! So there is the recording session investment
that we discussed previously: one to four man-days per loco
plus travel time.
Once we have the sound recorded, it takes several weeks to
massage the sound into the format we need and then again
into our software. So, it takes about 100 man-hours after the
sound is captured before the sound set is ready for one version
of decoder.
MRH-Oct 2013
