(c) DANGEROUS
Computer music... Music op inter-
rupt, DIGITAL STUDIO, COVOX, SB, GS,
GUS, TURTLE BITCH... MOD, STM, S3M, XM,
IT... Computer music compo... Computer
Computer engineers are in awe of computer sound.
ka...
Have you ever thought, respect-
dear gentlemen readers, why at all
the personal computer must retrieve
sounds? Why do many users
gu listen to her, collect musical talents
Dulya? Of course, computer effects
complement games and demos perfectly, but
Can we call them music? Like
"just a module from nowhere", storing hundreds
located on the disks of many users and
regularly listened to by them? Why,
for example, in the world of PC there is a constant
the phenomenon of new cards appearing in the Amiga world
There are new players that supposedly give better
neck sound, and in the world of Spectrum sound
evolved from beeper to GS? Looks like
that everyone in the computer world is crazy
on the technical characteristics of sound
cards, forgetting that the computer is what
no matter how powerful he is, he is not capable of giving birth
music. Having purchased a more advanced sound
a new card to replace the old one, we run home,
in anticipationenjoying a clean computer
yuter sound...
Is it worth it?
Based on an engineering approach and
based on technical specifications,
learned from the user manual,
we install our acquisition, trembling
out of impatience to hear the charming
sound and dive (or soar)
consciousness into the unknown expanses of Music-
cal Events. Of course, desire
compare the present with the past is great, and
we, having subconsciously absorbed the
reotype of the superiority of our acquisition
above the previous one, we begin to catch (salt-
knowingly, and sometimes on purpose) su-
existing and non-existent events. And
prove the appearance of these to yourself and those around you
yushnyh.
Having read “16” in the passport of your audio card
bit, 44 kilohertz" and comparing this with the characteristics
CD characteristics - 16 bit, 44
kilohertz, illiterate user
Renno declares to his illiterate friends
- my card has CD quality!
He subconsciously states the same thing
to himself and, having become “zombified” by his own
conclusion, listens with pleasure to the sound
sounds emitted by the audio card. When trying
demonstrate "super sound" nothing
parents who understand computers, mom
quickly retreats to the kitchen under
with the excuse that she was running out of soup, anddad
after a while it begins to threaten,
what if you don’t do something right now?
hey, then he's your computer... (at best
in case it turns off).
Why are those not burdened with knowledge about
bits and kilohertz people are not able to
listen to computer music, at best
Are you indifferent to her? At the same time
the user receives a certain satisfaction
the pleasure of listening to all sorts of things
mod-s, stm-s, hmm-s, etc. Paradox?
Let's deal with the notorious 16
bits and 44 kilohertz and then we will understand,
why not a single sound card, even the most
my dear, will not give birth to real music and
will not lose you a module with the quality of com-
CD But what is quality?
CD? It must be assumed that this is
the quality of the sound we hear when
playback Playing... again
on what? (It turns out there is a difference
where to put the CD).
First, let's discard the imperfection of the
computer sound and pay attention
that musical value is suppressed
of the majority of modules is aimed at
zero. The few remaining who, in
in principle, they can touch the soul, as if in
cage, fight, limited musical
mi and technical capabilities of audio
okart...
At the dawn of the computer boom, the fact
what a computer can producesounds,
similar to real ones, drove many people crazy
users, some even recorded this
"music" on cassette tapes and then listened. Not-
which users are simply obsessed with
on these electronic sounds, immersed
into the world of computer music, forgetting that
There is also other music - real music.
I foresee objections that, they say, all
modern music is created using
computers. The answer is quite simple:
first, it is created, not listened to.
Secondly, such low-field music
that, as a rule, accompanying (for
games, demos) or disco (so that your
do not interfere with intellect to “move your butt” and
think about people of the opposite sex).
True, original music is impossible
create on a computer, since it never
may he not be able to experience too much in music?
to give away your soul (there is no soul). If
even take the most sophisticated musical
ny editor, then all the author’s efforts to
conveying your musical idea
will be broken by the imperfection of computer
th digital path. Even if in some
The computer has an expensive audio card.
In order not to seem boring when describing
all kinds of musical metaphors, debunking
I look forward to the possibilities of computers with purely technical
nic side. I'm sure this
quite enough. We'll have to start with sa-
I started. I’ll choose the style - “science-pop”,
therefore it should be easy to read. None
super higher education for absorption
the material presented below does not require -
Xia. It will be boring at first (maybe).
How is natural sound translated?
into digital code? It is known that, theo-
theoretically, a continuous signal with a limitation
a given frequency band can be single-
significantly restored from the values of the dis-
specific readings, if the condition is met
vie:F>2f, where F - sampling frequency
tions, and f is the frequency band of the encoded
mi signal (that is, the upper frequency
spectrum, which we plan to later restore
become). Sampling frequency is currently
calls how many times per second digital
zer will take a reading of the amplitude of our
signal, so that using an ADC (ana-
logo-to-digital converter) identification
validate each measured amplitude
binary code.
fig. 1.
All that remains is to write the codes into memory
- that's all. It's not hard to guess what
more often we will take readings from the signal
la, that is, increaseF,the more likely
the ability to record all sorts of small features
signal strength. There is one sub-
water stone. If the upper frequency of the digital
the signal we are detecting will exceed the half-
the fault of the sampling frequency, then when
listening we will hear essential
distortions in the form of extraneous sounds
kov (not noise, not creaking, but precisely
sounds that last synchronously with the os-
newsounds. Very noticeable on someone
Amiga computers, and on PC too). To
this didn’t happen, it should have been digitized
pre-cut the signal from above
low-pass filter (low-pass filter) with frequency
cutoff in our case is 22.05 kilohertz (at
actually a little lower).
What happens in practice? In your ru-
kah digitizer, you download the program,
which controls the digitizer. After her
download you can select the frequency of dis-
cretization (more if you need to digitize
better quality, or less, for
saving memory). Because for everyone
specific case, you need to have a specific
low-pass filter, it is not clear at what frequency
the filter needs to be adjusted if it were installed
in the digitizer. So it's just there
no. Further, practically no one (until now)
lately) did not digitize the signal
16 bit 44.1 kHz - too much memory
will be required. Let's count - multiply 2 bytes
at 44100. Roughly speaking, it will be 90 kilo-
bytes per second. Since the total length
the number of samples in the average parity module
duration can be 10 seconds, then
if you count again, it turns out that you can
a barrel of average lousiness should occupy
about 900 kilobytes - and this is only a sample
ly! Nobody actually does that.
On Amiga computers, for example, music
makers digitize samples with a frequency of about
8 kHz8-bit format (sound is
Amiga 8-bit). Looks modest
number - per second of sample sound -
about 8 kilobytes. Naturally the sound
suffers. But it’s okay, people listen, a lot
I even like it.
On PC, by the way, it’s not much better. Only in
16-bit ones have appeared recently
samples, but also digitally quite
low frequency. In addition, many
Thai SoundBlasterob analog part
so poorly executed that it is difficult to distinguish
8 bits and 16 are almost impossible.
Now comes the fun part. If you lose
put such samples on a sound card 16
bit 44 kHz, then, in all likelihood,
sample digitized at a frequency of 8 kilo-
hertz, should be lost 44/8=5.5 times
faster and at the same time become squeaky. What-
this would not have happened, the card driver
(or any other program, it doesn’t matter)
recalculates and forces you to throw
on DAC (digital-to-analog conversion
tel) the same byte until it fits
det time of arrival of the next byte if
the sample would be reproduced with its native
sampling frequency. So increase
high the sampling frequency, do not increase
- the sample was played as it was and will continue to be played -
xia, and there is no change in quality
there will be. Therefore toto the exclamations of your friend
telya "I play everything at a frequency of 44
kilohertz!" take it easy.
Now pay attention again. We have not considered
Is there a question about the operation of the DAC? If you think-
those in CD players values
digital code is simply read from
disk, are thrown at the DAC, and then
ut on the way out, then you are completely mistaken. Yes
happens only in COVOXax, household computers
tern sound cards and computer
CD drives. Why? Because it's naked
digital-to-analog signal conversion
is not its restoration! In order to
to get what we (or not us)
once an ADC was supplied to the input, you need
make a restoration!
To restore the signal, it is necessary
dimo pass through a special thing
- V.F - Restoring Filter (filter
Kotelnikov). Then at the output we
we beam the desired signal. How to implement it
vat is a separate question that I don’t
I will cover it here (this is impossible to explain)
make it clear on your fingers), I’ll just say that
it can be placed in the passage path
signal either before the DAC or after it.
In the first case, the filter will be digital,
and in the second - analog. Analog
Kotelnikov filters do not make - very
difficult, but digital ones can be implemented
but... (In fact, idealF.K.realit is theoretically impossible to call, they do
only approximations with varying degrees of accuracy
ity - the more “ideal” the filter, the higher
its cost). Restore function
general filter - signal restoration in
operating frequency band, up to half an hour
sampling rates. And also suppression
spectral components beyond
operating band - above half the frequency
sampling.
Signal recovery is described non-
unknown formula Kotelnikova, about
which everyone who doesn’t now says
laziness, which is tricky and unprepared -
The reader will be horrified by:
S(T) - reconstructed signal.T - time
Si - the value of the i-th sample.
Part of this formula is the so-called
a constructed Nyquist function, which has
important in restoring ana-
of the digital signal, so
how is an impulse reaction of an idea
alny F.K.:
So, digital filter. His task is
accept the code sequence,
process it and throw it on the DAC PERFECT
ANOTHER SEQUENCE, which,
after passing through the DAC, it will turn into a real signal
cashIf you still penetrate a little
Kotelnikov’s formula, then we can arrive at
the following conclusion: the discretization operation
tions and recovery are fundamentally different
from each other. Sampling procedure
consistent, that is, sufficient
know only the value of the signal in ment
taking a count. A for reverse conversion
calling the sampled signal into
continuous need to know value
all counts at once! That is, ideal
F.K. must take in ALL the original
digital material, and then issue it
on the DAC with your own fabrications. If the theory is
technically accurate, then in fact the reaction
at the exit F.K. must appear before
how they give a signal to him - look,
the Nyquist function is also defined for the negative
Have a great time! In nature, this is
naturally impossible, but engineers
spin around and come up with all sorts of workarounds
maneuvers in the form of delay lines and elements
tov memory. In reality, the filter should
"remember" what happened before, as well as
"know" what will happen next. And the more
within wide limits he “remembers” and “knows”,
the better the filter.
What exactly does digital restoration do?
loading filter?
The essence of the process is as follows:
Yes, yescalledoversampling -
between existing samples with frequency
HHKHz are inserted additional, calculating
in accordance with the pulse ha-
characteristics of the Kotelnikov filter, which
which is reflected by the above function
Nyquist. (Impulse response -
this is a reaction to the so-called delta
impulse - a certain hypothetical impulse,
whose amplitude is infinite, and the length
ity is zero). In cheap losses
breakers use 8-fold re-
discretization (that is, between two
seven more are inserted by abacus), into the expensive
gikh - 256-fold. This means that in
on expensive players the values are thrown away
to a DAC with a frequency of 44.1*256=11289.6 kHz,
i.e. 11.2896 megahertz! By the way, here's one
from the explanations why two-speed CD-
ROM costs thirty dollars, but a good one
CD player - a thousand.
The figure shows a twofold re-
sampling (dashed lines) at
restoration of the signal taken from the figure
ka 1. Between existing readings there is a zone
new ones appeared - calculated overdisc-
retizer.
The drawing is quite idealistic, since
in fact, everything should be in here
digital "steps", I, of course,
having thrown them away, I show what I have turned into
the signal sounded.
Upon careful comparison of the drawings,
there is a noticeable difference in the signal shapes - the original
and restored. This is evidence
low initial sampling frequency, and
also low oversampling. (If
if there were no oversampling, then the form
the reconstructed signal would be different
from the original catastrophically.) Because
this is why information about the sites is lost
signalcharacterized by the most shi-
rocky spectrum. In the restored signal
the sharp peak of the original signal disappeared, and
the sharp decline also became smoother.
(By the way, it may turn out that taking
the countdown will still occur at the moment
"peak", but then after restoration
the signal level in the vicinity will rise
this "peak". Interesting fact - twice
we digitize the same thing, but the result is different
ny.)
It is important that this is not a drawback of the filter -
information about the signal was lost back in
process of conversion into digital code! Posse-
mu, no matter how good the restoration
current filter, it already processes this signal
will not restore it for sure. This is not at all
means that a digital grinding noise will appear.
So he just won’t appear. Filter
it will suppress him perfectly, but the musical
The signal strength will suffer greatly - primarily
First of all, dynamics and timbre. When using
using the signal as a special effect
defects in games - by and large, on-
I don’t care, but listen to normal music
It will be, to put it mildly, uninteresting. For example
measures, it will be impossible to distinguish, they are playing
whether the violins are live or whether they are imitated by synthetic
congestion, the sound of two different drum cymbals
installation will also be indistinguishable, rather
in total it will turn into a hiss
(with a good filter it’s very clean), that’s it
will sound dull, etc., in one word
-demonstration of dead sounds and complete
loss of means of expression.
People who are far from technology are looking for music
not primarily musical thought, but
when they are given computer packaging -
corvée, feel discomfort.
Why is it necessary to use
should I call it the Kotelnikov filter? Sweat-
mu that, guided by his work,
can be calculated as accurately as desired
TRUEintermediate values between
main readings.
If the sound card worked according to
in accordance with the IDEAL 256-fold conversion
discretization in accordance with the formula
Kotelnikov, then for a sample length of 1 kilo-
lobyte is the ideal algorithm for its operation.
kov: 44100*256=11289600 once per second
multiply 1024 numbers by the corresponding
coefficients, then sum everything up. A
for a sample 10 kilobytes long you will have to
multiply 10 times more. (Yes, in
the second channel will need to do approx.
essentially the same, but for the AWE64 card -
64 channels). In addition, in the process of calculating
numbers can occur with bit depth
up to 64.
If you continue like this, then by one sound
it will be possible to use a new cardmake it work
all the processors in the world. To simplify
the procedures do not take all the bytes, but, say,
50 nearest (real figure), then you-
numerical abilities of Pentium200
quite enough to restore the signal
huh, and then you can throw it on the DAC.
For example, in the General Sound map for pe-
Resampling costs Z80 at 12 MHz. Pe-
resampling - based on two neighboring samples
accounts, as it turns out, or rather, as successfully
Yes. Actually 3-4 times. Interpolation
there is not “according to Kotelnikov”, but the arithmetic average
metic but subjective purity of sound
improves very much. Filters
after DACs there aren’t any - they’re already there
there is simply nothing to restore.
By the way, if anyone directly compares
Nival sound of GS and Amiga, could sometimes
notice that on GS some game modules
bark suspiciously dully, and some
- completely normal. This is a direct re-
Result of resampling. In samples,
digital with a frequency of 8 kilohertz, higher
4 kilohertz there is no musical in-
formation - therefore everything that is higher is
is pressurized by the resampler. Here's the poetry
This is where “deafness” appears. Amiga
delivers, along with samples, a decent
the share of high-frequency noise - hence the
buzzing sonority. If the sample is digital
set it to 22 kilohertz, then there will be a difference
already almost invisible.It is known that Pentiu sound cards
they don’t install mov200. They don’t put them in pro-
CD players. In CD players the role
digital filter-regenerator performs
uses a special signal processor,
who only knows how to do this. A in
PC and Amiga now write special
players, which with grief half-and-half produced
They do some kind of interpolation. About her ver-
no one knows the truth except the authors of these
programs. One thing I'm sure of is the filter.
Kotelnikov doesn't even smell there, but the sound
they improve. (Or rather, they don’t steal, but kill
they express outright digital dirt, and at the same time
but also a considerable part of the true musical
information, introducing your own, false). Here
now, the higher the oversampling, the
cleaner sound. It is cleaner, not more correct!
So forget about live music.
Another disappointment is that the computer does not
can perform its direct function by
purpose - its processor is loaded
you to the limit. About interpolated mu-
language in dynamic games is still out of the question
no either.
Now I’ll debunk another “digital
myth" generated mainly by advertising
campaign. This is the dynamic range in
9bdB. For those who don’t know, this is for now
maker of the relationship between maximum and mini-
small levelssystem volume.
Described by the expression D=20*lg(max/min).
If we calculate the dynamic
ical range of a 16-bit system, then
we just get the number 96. And what in re-
ness? Any digital system has
one fundamental drawback - with ma-
Disturbances in the volume level occur, so
called quantization distortions. After all,
the lower the volume, the lower the bit
used for conversion! When ma-
scope volume level high bits digital
they sit in silence and cannot hear a word
hear about the signal. It turns out that with
Quiet sounds are a problem. Therefore, how
Practice has shown that sounds are digitized from
level -CHODB, are no longer very similar to
real, when digitalized -bODBfrom sound
only horns and legs remain... And what music -
what about without real quiet sounds? Really
usable dynamic range - 30-
CHODB. Therefore, the destiny of computer music is
only techno and pop music, since it has the
gives an exceptionally narrow D. (In losing
CD makers are again struggling with this
using other methods).
I suggest Amiga fans do the math for themselves.
dynamic range of 8-bit system
(by the way, this is quite enough for effective
tov andaccompanying music in games).
So maybe let's not be so serious...
be careful about computer music,
audio cards, music compo, because everything is mu-
level -CHODB, are no longer very similar to
real, when digitalized -bODBfrom sound
only horns and legs remain... And what music -
what about without real quiet sounds? Really
usable dynamic range - 30-
CHODB. Therefore, the destiny of computer music is
only techno and pop music, since it has the
gives an exceptionally narrow D. (In losing
CD makers are again struggling with this
using other methods).
I suggest Amiga fans do the math for themselves.
dynamic range of 8-bit system
(by the way, this is quite enough for effective
com and accompanying music in games).
So maybe let's not be so serious...
be careful about computer music,
audio cards, music compo, because everything is mu-
vocal gadgets are made in order to
to make games more interesting to play. Or
Am I wrong?
S(T) - reconstructed signal.T - time Si - the value of the i-th sample. Part of this formula is the so-called a constructed Nyquist function, which has important in restoring ana- of the digital signal, so how is an impulse reaction of an idea alny F.K.:
So, digital filter. His task is accept the code sequence, process it and throw it on the DAC PERFECT ANOTHER SEQUENCE, which, after passing through the DAC, it will turn into a real signal cashIf you still penetrate a little Kotelnikov’s formula, then we can arrive at the following conclusion: the discretization operation tions and recovery are fundamentally different from each other. Sampling procedure consistent, that is, sufficient know only the value of the signal in ment taking a count. A for reverse conversion calling the sampled signal into continuous need to know value all counts at once! That is, ideal F.K. must take in ALL the original digital material, and then issue it on the DAC with your own fabrications. If the theory is technically accurate, then in fact the reaction at the exit F.K. must appear before how they give a signal to him - look, the Nyquist function is also defined for the negative Have a great time! In nature, this is naturally impossible, but engineers spin around and come up with all sorts of workarounds maneuvers in the form of delay lines and elements tov memory. In reality, the filter should "remember" what happened before, as well as "know" what will happen next. And the more within wide limits he “remembers” and “knows”, the better the filter. What exactly does digital restoration do? loading filter? The essence of the process is as follows: Yes, yescalledoversampling - between existing samples with frequency HHKHz are inserted additional, calculating in accordance with the pulse ha- characteristics of the Kotelnikov filter, which which is reflected by the above function Nyquist. (Impulse response - this is a reaction to the so-called delta impulse - a certain hypothetical impulse, whose amplitude is infinite, and the length ity is zero). In cheap losses breakers use 8-fold re- discretization (that is, between two seven more are inserted by abacus), into the expensive gikh - 256-fold. This means that in on expensive players the values are thrown away to a DAC with a frequency of 44.1*256=11289.6 kHz, i.e. 11.2896 megahertz! By the way, here's one from the explanations why two-speed CD- ROM costs thirty dollars, but a good one CD player - a thousand.
The figure shows a twofold re- sampling (dashed lines) at restoration of the signal taken from the figure ka 1. Between existing readings there is a zone new ones appeared - calculated overdisc- retizer. The drawing is quite idealistic, since in fact, everything should be in here digital "steps", I, of course, having thrown them away, I show what I have turned into the signal sounded. Upon careful comparison of the drawings, there is a noticeable difference in the signal shapes - the original and restored. This is evidence low initial sampling frequency, and also low oversampling. (If if there were no oversampling, then the form the reconstructed signal would be different from the original catastrophically.) Because this is why information about the sites is lost signalcharacterized by the most shi- rocky spectrum. In the restored signal the sharp peak of the original signal disappeared, and the sharp decline also became smoother. (By the way, it may turn out that taking the countdown will still occur at the moment "peak", but then after restoration the signal level in the vicinity will rise this "peak". Interesting fact - twice we digitize the same thing, but the result is different ny.) It is important that this is not a drawback of the filter - information about the signal was lost back in process of conversion into digital code! Posse- mu, no matter how good the restoration current filter, it already processes this signal will not restore it for sure. This is not at all means that a digital grinding noise will appear. So he just won’t appear. Filter it will suppress him perfectly, but the musical The signal strength will suffer greatly - primarily First of all, dynamics and timbre. When using using the signal as a special effect defects in games - by and large, on- I don’t care, but listen to normal music It will be, to put it mildly, uninteresting. For example measures, it will be impossible to distinguish, they are playing whether the violins are live or whether they are imitated by synthetic congestion, the sound of two different drum cymbals installation will also be indistinguishable, rather in total it will turn into a hiss (with a good filter it’s very clean), that’s it will sound dull, etc., in one word -demonstration of dead sounds and complete loss of means of expression. People who are far from technology are looking for music not primarily musical thought, but when they are given computer packaging - corvée, feel discomfort. Why is it necessary to use should I call it the Kotelnikov filter? Sweat- mu that, guided by his work, can be calculated as accurately as desired TRUEintermediate values between main readings. If the sound card worked according to in accordance with the IDEAL 256-fold conversion discretization in accordance with the formula Kotelnikov, then for a sample length of 1 kilo- lobyte is the ideal algorithm for its operation. kov: 44100*256=11289600 once per second multiply 1024 numbers by the corresponding coefficients, then sum everything up. A for a sample 10 kilobytes long you will have to multiply 10 times more. (Yes, in the second channel will need to do approx. essentially the same, but for the AWE64 card - 64 channels). In addition, in the process of calculating numbers can occur with bit depth up to 64. If you continue like this, then by one sound it will be possible to use a new cardmake it work all the processors in the world. To simplify the procedures do not take all the bytes, but, say, 50 nearest (real figure), then you- numerical abilities of Pentium200 quite enough to restore the signal huh, and then you can throw it on the DAC. For example, in the General Sound map for pe- Resampling costs Z80 at 12 MHz. Pe- resampling - based on two neighboring samples accounts, as it turns out, or rather, as successfully Yes. Actually 3-4 times. Interpolation there is not “according to Kotelnikov”, but the arithmetic average metic but subjective purity of sound improves very much. Filters after DACs there aren’t any - they’re already there there is simply nothing to restore. By the way, if anyone directly compares Nival sound of GS and Amiga, could sometimes notice that on GS some game modules bark suspiciously dully, and some - completely normal. This is a direct re- Result of resampling. In samples, digital with a frequency of 8 kilohertz, higher 4 kilohertz there is no musical in- formation - therefore everything that is higher is is pressurized by the resampler. Here's the poetry This is where “deafness” appears. Amiga delivers, along with samples, a decent the share of high-frequency noise - hence the buzzing sonority. If the sample is digital set it to 22 kilohertz, then there will be a difference already almost invisible.It is known that Pentiu sound cards they don’t install mov200. They don’t put them in pro- CD players. In CD players the role digital filter-regenerator performs uses a special signal processor, who only knows how to do this. A in PC and Amiga now write special players, which with grief half-and-half produced They do some kind of interpolation. About her ver- no one knows the truth except the authors of these programs. One thing I'm sure of is the filter. Kotelnikov doesn't even smell there, but the sound they improve. (Or rather, they don’t steal, but kill they express outright digital dirt, and at the same time but also a considerable part of the true musical information, introducing your own, false). Here now, the higher the oversampling, the cleaner sound. It is cleaner, not more correct! So forget about live music. Another disappointment is that the computer does not can perform its direct function by purpose - its processor is loaded you to the limit. About interpolated mu- language in dynamic games is still out of the question no either. Now I’ll debunk another “digital myth" generated mainly by advertising campaign. This is the dynamic range in 9bdB. For those who don’t know, this is for now maker of the relationship between maximum and mini- small levelssystem volume. Described by the expression D=20*lg(max/min). If we calculate the dynamic ical range of a 16-bit system, then we just get the number 96. And what in re- ness? Any digital system has one fundamental drawback - with ma- Disturbances in the volume level occur, so called quantization distortions. After all, the lower the volume, the lower the bit used for conversion! When ma- scope volume level high bits digital they sit in silence and cannot hear a word hear about the signal. It turns out that with Quiet sounds are a problem. Therefore, how Practice has shown that sounds are digitized from level -CHODB, are no longer very similar to real, when digitalized -bODBfrom sound only horns and legs remain... And what music - what about without real quiet sounds? Really usable dynamic range - 30- CHODB. Therefore, the destiny of computer music is only techno and pop music, since it has the gives an exceptionally narrow D. (In losing CD makers are again struggling with this using other methods). I suggest Amiga fans do the math for themselves. dynamic range of 8-bit system (by the way, this is quite enough for effective tov andaccompanying music in games). So maybe let's not be so serious... be careful about computer music, audio cards, music compo, because everything is mu- level -CHODB, are no longer very similar to real, when digitalized -bODBfrom sound only horns and legs remain... And what music - what about without real quiet sounds? Really usable dynamic range - 30- CHODB. Therefore, the destiny of computer music is only techno and pop music, since it has the gives an exceptionally narrow D. (In losing CD makers are again struggling with this using other methods). I suggest Amiga fans do the math for themselves. dynamic range of 8-bit system (by the way, this is quite enough for effective com and accompanying music in games). So maybe let's not be so serious... be careful about computer music, audio cards, music compo, because everything is mu- vocal gadgets are made in order to to make games more interesting to play. Or Am I wrong?
Share your thoughts about the article