Hi Claudia...
I got carried away with a train of thought in section 1, and
ended up addressing some of the concepts in 2 and 3 there.
I hope this irregularity of format doesn't matter with respect
to the overall presentation:
1 - The majority of the basic instruments (piano, sax, guitar, flute)
on the Kurzweil, as with many synths, are sampled, which is to say
they are a series of several audio waveforms which have been
digitally recorded and are laid out across the keyboard, with each
sample being used for several consecutive keys - the pitch of the
sample being raised as it moves up the keys. Because of this, the
'root' sample for each group of keys is an exact duplicate of the
instrument's sound at that pitch. As the pitch is increased for
higher keys, the sound is duplicated, but at a higher pitch. The
more actual samples that are used, the more authentic the sound will
be, as each true root sample introduces the subtle variations that
exist between individual keys of, e.g., the piano. If too few samples
are used, as the pitch is increased for higher keys, the sound begins
to be inauthentic in the same way that "The Chipmunks" voices no
longer sound like human voices. Other anomalies can occur when the
sample, which must be cycled, or 'looped' or repeated when a note is
sustained, are poorly recorded, or are badly looped. The looping must
be done at what is called a 'null' point in the sample, where there
is minimal volume, and no transient sounds. If done poorly this can
result in 'clicks' or 'pops', or changes in the volume of the loop
which are clearly not part of the natural instrumental sound.
These can be found in less expensive keyboards, but in a quality
instrument like the Kurzweil, there are a sufficient number of
well-constructed samples that you will not hear such defects,
and the sound will not become 'stretched' so as to sound
inauthentic. The sound will still fall short of the original
instrument, in that the keys which use the same root sample
will have essentially the same dynamics, and will lack the
key-to-key variations of the actual instrument. Within the
context of a band or an orchestra, this will not be too
noticeable. It is more obvious when using the instrument in
playing solos. When played on a synth, instruments such as
guitar will lack the subtle sounds created when the left hand
changes chords and brushes the strings. And when playing the
flute on a synth, the 'tremolo' is programmed and relatively
the same regardless of the note, whereas a live flute player
will vary the tremolo on each note (I also play guitar and
flute, so I am acutely aware of these subtle differences).
There have been efforts to include, e.g., the natural sounds
of the guitar on separate, non-musical keys of the keyboard,
so that finger-scrapes can be included in a composition,
along with regular notes, however these are innately awkward
to incorporate, versus their natural presence when playing a
live guitar.
Piano - the sustain is noticeably shorter than the real
instrument. This is not especially noticeable in pieces played
at a faster rate, but is more obvious with very slow songs.
I was able to get around this by editing the sustain parameters
of the layers of the 'patch' - a patch being a selection on
the keyboard which may contain several layers, each with their
own sample(s) and programming parameters. Not all keyboards
allow for editing, so with some, you would be unable to change
the sustain in this way. The K2500 allows editing of almost
everything.
An mp3 recording of a piece, appropriately named 'Grand
Illusion', which uses the stereo piano samples from the K2500
Piano ROM block, can be found on the Kurzweil site, so you can
hear for yourself how authentic it sounds. It was actually
hearing this piece that convinced by to upgrade from a K2000
to a K2500:
http://www.kurzweilmusicsystems.com/html/piano_rom_demos.html
Flute - It is possible to work around the fixed tremolo built
into the standard patch. It is possible to create a more
randomized tremolo by using some of the complex or random
mathematical operators which can be assigned to vary the volume
and pitch of the sample.
Guitar - You can produce some modulation of the sound using
the modulation wheel. The built-in samples I have don't have
the added natural sounds of fingers scraping strings, or the
harmonics which can be created using a real guitar, but it is
possible to purchase sample sets which include these.
2 - As noted above, the inauthentic sound of "The Chipmunks"
can result when a sample is played too far above its natural
range. When played too low, you get a bass sound which is
muddied and indistinct, as opposed to the relatively precise
sound of a true bass guitar.
3 - Also discussed above, though, with sound effects, the lack
of authenticity often shows up more easily. A helicopter or dog
bark played at too high a pitch for the root sample sounds very
obviously inauthentic, since there is a very small range for
those authentic sounds. A helicopter sound, or a gunshot,
played too high, sounds like a fake cartoon sound, lacking
the bass which give them the authenticity of the real thing.
4 - I used the built-in sequencer in the K2500. I recorded a
round of the song on piano, using a built-in metronome function
to start and stop in the right places. I then copied the segment
to the same track (10) using Song Edit -> Track -> Copy 1:1 to
9:1, on track (10), to location 1:9. I repeated this, now copying
1:1 to 17:1 to location 17:1 -> Done -> Exit -> Save Song. This
gave me 4 rounds of the song.
Then, from track 10 in the song mode, I entered Edit ->Track ->
Copy from 1:1 to 31:1 (33:1 minus 2, to account for the later
entry of this track, so they will end at the same time) to track
9, location 3:1. Since this copied 'All' data to the new track,
I then had to go to Song Edit -> Event, and scroll through them,
finding all instances of Control Bank and PCHG (Program Change)
in order to assign a different instrument in another location.
'Da Clav', e.g. is patch #24, which would be Control Bank 0,
PCHG 24.
Then it's basically a matter of "Lather, Rinse, Repeat" to
assign identical changes to tracks 8 and 7, giving me four
instruments playing in a perfect round.
Had I not used the metronome, I could have afterward entered
Song Mode, and from each track, gone to Edit -> Track ->
Quantize, set the degree to 100% and hit 'Go', saving the
song after altering each track. This produces absolute
alignment with the beat.
As to contrasting the results with analog technology, I am
less familiar with professional analog mixing techniques,
though I have played with overdubbing on cassette and
reel-to-reel tapes. First of all, with tape, every overdub
introduces a measure of background and amplifier noise into
the mix, so by the time you get into your 4th overdub, it's
pretty poor quality. I also know of no comparable method to
produce the perfect alignment of the 'Quantize' function,
not to mention that this can also be used to introduce a
certain measure of artificial 'swing' into the music.
The digital sounds, on the other hand, particularly in
intimate solo sessions, just don't have the range of 'color'
which comes from a real piano using 3 strings for each mid
to upper range note, with each individual string lending
it's contribution to the overall sound, which is then
further sweetened by the wooden cabinet itself.
For solo piano concerts, I'd want a grand piano, but for
use in a band or for home recording, there's nothing like
a synthesizer!
If I left anything out, or can clarify anything, just let
me know!
Please do not rate this answer until you are satisfied that
the answer cannot be improved upon by means of a dialog
established through the "Request for Clarification" process.
sublime1-ga |
