| Author |
Message |
mr_alex
Guest
|
 Posted:
Wed Nov 09, 2005 3:47 am Post subject:
Re: dac clock rates |
|
|
Ok, yes I definitely understand that changing the DAC clock speed will
alter tempo & pitch as they are irrevocably related; your example with
the 440Hz note cleared that up (thank you for that).
So, if I go about changing the DAC clock speed can I get the same
results as a cd mixer/mp3 mixing device (changing tempo & pitch) or
will I encounter audio problems? Commercial units must work by
buffering a sizeable amount of audio data and performing the sample
rate change either through DSP or DAC clock rate alterations...
But, if you can tell me what will happen if the DAC clock speed is
altered in terms of the quality of the audio and any other issues that
may arise?
Cheers, Alex
|
|
| Back to top |
|
 |
Guest
|
| Posted:
Wed Nov 09, 2005 4:09 am Post subject:
Re: dac clock rates |
|
|
mr_alex wrote:
| Quote: | Ok, yes I definitely understand that changing the DAC clock speed will
alter tempo & pitch as they are irrevocably related; your example with
the 440Hz note cleared that up (thank you for that).
So, if I go about changing the DAC clock speed can I get the same
results as a cd mixer/mp3 mixing device (changing tempo & pitch) or
will I encounter audio problems? Commercial units must work by
buffering a sizeable amount of audio data and performing the sample
rate change either through DSP or DAC clock rate alterations...
But, if you can tell me what will happen if the DAC clock speed is
altered in terms of the quality of the audio and any other issues that
may arise?
|
Unfortunately, now you get into the implementational details of
one system or another. One manufacturer may do it right, another
my screw it up horribly out of sheer incompetence.
You have another problem Simply "changing the DAT rate" all by
itself doesn't necessarily do anything to change the rate at which
data is being fed to the DAC> In that sense, you are wholely unclear
as to how you prpose to accomplish this.
For example, if by "DAC" you mean an external converter that's
fed by a S/P-DIF link from a CD player, you CAN'T change the
clock rate on the DAC, because it is largely derived from the
incoming data stream. The stream itself has a data header that
says "I'm 44.1 kHz" AND the data is coming from the CD player
at 44,100 samples per second.
If, instead, you're talking about opening up a CD player and fiddling
the clock signals that go to the DAC< you're likely to break things.
These CD players with the varispeed wheel do it by altering the
manipulating how the data is being buffered from the CD, and the
drive servo runs only approximately at the DAC speed: it actually
varies all over the place as the buffer needs to keep itself reasonably
full to work right. In other words,, the player mechanism and
electronics are specifically designed with this task in mind.
If you're talking about writing a program to do this from, say,
the equivalent of data files, then you're going to have to find a
DAC that suppoert any arbitrary clock speed over the desired
range, and as far as such existing that will interface readily to
a computer, they're pretty few and far between. Even having found
one, you're now going to have to invent all this stuff from scratch,
stuff that people with a lot more experience and expertise have
already done and moved on.
And if you want TRUE varispeed, i.e. the ability to change tempo
over a continuous range and in real time, that, as I have said, is
a real tough task to get right.
Why not go out and buy a box that does what you want. It'll range
anywhere from a handful of dollars for some piece of software
(like GoldWave) that won't do it in real time, and has audble
imperfections
to tens of thousands of dollar for professional workstations that will
do an absolutely fantastic job over a range of -25+30% with almost
NO audible artifacts of any kind, and within +-10% maintain CD quality
audio. |
|
| Back to top |
|
|
Dave H.
Guest
|
| Posted:
Wed Nov 09, 2005 5:37 am Post subject:
Re: dac clock rates |
|
|
"mr_alex" <dancedynamix@hotmail.com> wrote in message
news:1131486443.111069.246090@g14g2000cwa.googlegroups.com...
| Quote: | Ok, yes I definitely understand that changing the DAC clock speed will
alter tempo & pitch as they are irrevocably related; your example with
the 440Hz note cleared that up (thank you for that).
So, if I go about changing the DAC clock speed can I get the same
results as a cd mixer/mp3 mixing device (changing tempo & pitch) or
will I encounter audio problems? Commercial units must work by
buffering a sizeable amount of audio data and performing the sample
rate change either through DSP or DAC clock rate alterations...
But, if you can tell me what will happen if the DAC clock speed is
altered in terms of the quality of the audio and any other issues that
may arise?
Cheers, Alex
|
Hi Alex,
I'll probably be shot down in flames for this, but...
Depending on the depth of your electronics background [1], you might want to
pick up a thrift-store CD deck to experiment with - once you get the case
off, take a look to see whether it runs from a single crystal oscillator
(probably around 45MHz) or a bunch of different crystals - if a single
oscillator functions as the master clock, it may be possible to replace it
with a variable-frequency oscillator to obatain varispeed (with a matching
variation in pitch). Knowing the way everything's pared to the bone in the
interest of "cost effectiveness", it's likely that the lower-end players
will be more suitable for this approach - and using a digital out is going
to be near-impossible.
The majority of CD players use a buffer to match the data from the laser
pickup to the DAC data rate (after de-interleave, error correction etc.),
the buffer state is used to regulate the speed of the disc motor, and the
feedback loop so created *should* ensure that the disc rotation is adjusted
(within limits) to match the varispeed - a pair of SL-1210's is a lot more
elegant though!
Hope this helps,
Dave H.
(The engineer formerly known as Homeless)
[1] Ham radio experience is probably appropriate here - the variable
frequency oscillator will need to be both wide-ranging and stable, not an
easy combination, requiring temperature compensation, non-microphonic
construction, well-regulated power supplies etc. |
|
| Back to top |
|
|
Joe Kesselman
Guest
|
| Posted:
Wed Nov 09, 2005 5:42 am Post subject:
Re: dac clock rates |
|
|
| Quote: | what will happen if the DAC clock speed is
altered in terms of the quality of the audio
|
Same things as speeding up or slowing down tape, more or less. Or same
things as adjusting the pitch by repeating or discarding samples every
so often, which is the more common professional solution.
| Quote: | Commercial units must work by
buffering a sizeable amount of audio data and performing the sample
rate change either through DSP or DAC clock rate alterations...
|
Commercial units (and commonly available software) manipulate tempo
independently of pitch by breaking up the audio stream into short blocks
(long enough to exhibit the right overtones, not long enough that humans
notice them) and either repeating or dropping those to stretch or shrink
the sound's duration. To change pitch without changing duration, they do
the resampling mentioned above, then use this block-splicing trick to
convert it back to the original time duration.
A bit of additional work is needed to keep all those splices from
introducing noticable amounts of noise, but it's reasonably
straighforward signal processing. |
|
| Back to top |
|
|
|
|
|
|
FAQ
Memberlist
Register
Profile
Log in to check your private messages
Log in