Moshe Werner wrote:
Hi to all,
I'm an audio engineer in my day (and night) work, but definitely not a
programmer.
From start of 1995 to mid 2000 I was a hardware design
engineer for
Fairlight ESP, manufacturers of the MFX2, MFX3 and MFX4 DAWs.
My questions would be:
1. Is it only me that can't here a difference between different DAWs mixing
algorithms?
Back in the late 1990s one the of big differences between the
Fairlight and Protools was that the Fairlight did is mixing on
Analog Devices Sharc 21k floating point DSP chips while Protools
was (I believe) built around the Motorola 24 bit fixed point
56k DSP chip.
The rounding and truncation behaviour of the floating point
chips sounded better than that of the fixed point chips (or
so I was told and actually makes some sense from an engineering
point of view).
All modern DAWs which do the audio processing on the host Intel/
AMD CPU should be using floating point or SSE instructions and
hence should have the good properties of floating point.
2. To the developers out there, what is your opinion?
Is there a
better/worse algorithm, or is the whole thing another "pay 600$ for this
software - it has great algorithms!!!" hype?
I really think there is only really one mixing algorithm, its
basically just multiply all the sources by their respective gain
values and sum it all together.
Other things that my be at play are things like different compressor,
limiter and filter algoriths all of which definitely do have multiple
ways of being implemnted.
However, the biggest differences are probably in the analog-to-digital
input or the digital-to-analog output. For instance, one thing that
made the Fairlight MFX really popular on the foley stages (recording
big bangs and crashes for sound effects) was the fact that the
Fairlight had a basically flat low end frequency response. The test
equipment we had at the time showed it was basically 1dB down at
10Hz which was the lowest the equipment measure. We suspected that
the -3dB point was actually around 5Hz.
Finally since A-D and D-A sections are highly susceptible to picking
up electrical/eloctronic noise from the rest of the system, a part
of the character of different machines is actually due to the noise
it picks up from the rest of the system its installed in.
4. Analog emulation plugins. How does one
"emulate" analog waveforms in a
digital world? That sounds like a paradox to me.
This is a very big field and can go from cloning the input/ouput
charavtyeristcs of a device to the extreme of programming digital
models of analog elctronics components and then hooking them
together.
HTH,
Erik
--
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/