Original Message                 Date: 28-Feb-02  @  09:02 PM   -   Why is digital less warm than....

Mindspawn

Posts: 659

Link?:  Link
File?:  No file

Hey folks, just some food for thought...

We all hear things like, digital isn't as warm as analog, and while I'm not here to argue that specifically, did you ever think about why that is? Outside of some of the basic physical properties, you'd think the two mediums wouldn't be that divergent.... and in some sense, maybe they're not. Now I'm gonna try a little heresy...

Maybe the reason mixes from analog sounds warmer, more musical, whatever, is: our techniques for recording, mixing, etc., are mostly built and modeled on analog experience. We've learned techniques for, say mic placement, that were establised in the analog realm... maybe we should be evaluating new ways of doing things....?

I mean think about one of the most basic differences between the two mediums, the level meter... Many of us that came from the analog world were sorely surprised to find out we couldn't push the LEDs "past the red" on a digital board... Now once I learned how to use digital LEDs, mt life, and my mixes, sounded better...

I'm not really trying to lay out new "rules" of digital recording/mixing, but just bouncing the idea off you all. If you have any experience with what I'm on about here, by all means share it. If you got a "warm sound" from all digital equipment, what was your methodology? Why do you think it worked that way? If you captured a digital take of a vocalist that just simply shimmers, did you do it the "traditional" (i.e., basically as it's always been done on analog equipment) way, or did you find a technique that is exclusive to digital?

Anyhoos, just some thoughts....

Peace All

[ back to forum ]                           [quote]

Message 131/157                 Date: 09-Apr-02  @  10:18 PM   -   RE: Why is digital less warm than....

influx

Posts: 7627

Link?:  Link

File?:  No file

hehe. you guys should see this one guy I know who is ALWAYS staring at his speakers

"look at the bass. its so stable"

[ back to forum ]               [quote]

Message 132/157                 Date: 10-Apr-02  @  12:30 AM   -   RE: Why is digital less warm than....

k

Posts: 12353

Link?:  Link

File?:  No file

ah but also hi quality converters on those old e-mu 12 bits

___________________________________

I had an idea for a script once. It's basically Jaws except when the guys in the boat are going after Jaws, they look around and there's an even bigger Jaws. The guys have to team up with Jaws to get Bigger Jaws.... I call it... Big Jaws!!!

[ back to forum ]               [quote]

Message 133/157                 Date: 10-Apr-02  @  12:56 AM   -   RE: Why is digital less warm than....

Brett

Posts:

Link?:  No link

File?:  No file

I have been seeing eIII's cheap and the 12 bit emax rack with the 8megs go for under $300.

[ back to forum ]               [quote]

Message 134/157                 Date: 10-Apr-02  @  01:14 AM   -   RE: Why is digital less warm than....

influx

Posts: 7627

Link?:  Link

File?:  No file

in rack shouldnt be more than $200 really.

[ back to forum ]               [quote]

Message 135/157                 Date: 10-Apr-02  @  01:53 AM   -   RE: Why is digital less warm than....

xoxos

Posts: 6231

Link?: Link

File?:  No file

simple.. higher resolution equals a higher rate of correction. if your speaker is at bit#1234 and it's next stop is 1334, if you insert a middle step at 1239 and say your speaker in reality is actually whanging at 1239.5 at that time, that middle step is counter productive. maybe more accurate, but we're talking about fluid movement here=speaker efficiency.

i reckon.

[ back to forum ]               [quote]

Message 136/157                 Date: 10-Apr-02  @  03:15 PM   -   RE: Why is digital less warm than....

realtrance

Posts: 184

Link?: Link

File?:  No file

Seems to make sense, but still asking a question out of ignorance here: wouldn't the D/A converter between the digital signal and the speaker complicate your description, Pongoid? Or, are you talking about a situation where you've got direct digital ins (although even there, wouldn't the internal D/A converters in the speaker do the same thing)?

rt

[ back to forum ]               [quote]

Message 137/157                 Date: 10-Apr-02  @  03:21 PM   -   RE: Why is digital less warm than....

realtrance

Posts: 184

Link?: Link

File?:  No file

Oh and sorry for not responding first -- knowa, yes, I'd forgotten that aspect. That's another important thing -- live performance, with the spontaneity and the kind of mastery to provide profound interaction between different musicians' personalities, souls, energies, you don't find a lot of that around. Real, improvisatory ensemble work is extremely important to music, and it risks being lost in the circuits.

Thanks for your comments!

rt

[ back to forum ]               [quote]

Message 138/157                 Date: 10-Apr-02  @  06:14 PM   -   RE: Why is digital less warm than....

knowa

Posts:

Link?:  Link

File?:  No file

j-type: "distortion = greater perceived loudness" this makes sense.

ok so what do analog sound generators--with no 'bits' to speak of--feed to speakers?

or A/D converters as rt mentioned.

this is related to why samples of synths sound different than the synths themselves playing the same parts?

16-bit samples of SP-1200 hits don't sound as cool as the SP itself because they spit too many bits at the speakers?

so this lo-fi magic is happening *right* at the moment that electrical energy is converted into acoustic energy?

how 'bout on a Mac ;)

[ back to forum ]               [quote]

Message 139/157                 Date: 10-Apr-02  @  07:10 PM   -   RE: Why is digital less warm than....

j-type

Posts: 40

Link?:  No link

File?:  No file

The output of a DAC is an analogue signal, but its the DAC's 'guess' at what analogue signal the digitally encoded signal represents. The more 'steps' in the digital signal, the better the guess is going to be as the DAC has more info to go on. A 12 bit signal (4096 steps) is going to sound grainy next to a 16 bit signal (65536 steps) no matter how good the DAC is, because there's only so much a DAC can do with the information its given.

Analogue synths present a smooth, continuous signal to a speaker - but so does a DAC on a digital synth, only the DAC signal is less smooth because of the way it was constructed.

Let me use an analogy - imagine a smooth, well defined surface that represents an analogue signal from an analogue synth. Now instead of the smooth surface, replace it with a closely-packed array of spikes that describe the same surface (think Pinhead in Hellraiser). This array of spikes represents the digitally encoded signal. Now lay a sheet of rubber over the spikes (analogous to the action of a DAC) - you get a smooth surface once again, but its not perfectly smooth since between the spikes there's no information about the original surface. The rubber sheet fills in the gaps, just as a DAC does. The more spikes you have, the closer to the original surface you will get. Same with digital audio, but replace 'spikes' with 'bits'.

This analogy could explain why lower bit depth gives greater perceived volume - imagine having two representations of the original surface in front of you, one with loads of very densely packed spikes under the sheet, giving a smooth contour, and the other with much fewer spikes, resulting in a more choppy, irregular surface. Which one would you look at first? Which grabs your attention? Which is visually 'louder'?

[ back to forum ]               [quote]

Message 140/157                 Date: 17-Apr-02  @  05:29 AM   -   RE: Why is digital less warm than....

tortoise

Posts: 107

Link?:  No link

File?:  No file

Actually, a DAC can perfectly reconstruct a smooth waveform; it is basic to the mathematics that go into their design. The idea that a DAC puts out a waveform with polygonal segments is false. The stair-step waveform is sent to an analog smoothing filter in the DAC that removes everything above the Nyquist limit, giving a perfectly smooth waveform. The stair-step shaped components of the signal only exist the Nyquist limit (NOTE: massive simplification), so all you have to do is run your stair-step waveform into an LPF with the cutoff freq set just below the Nyquist limit, with a little fudge factor thrown in to account for filter slope. The 44.1kHz sample rate for CDs is not an accident; it allows for the DAC smoothing filter cutoff to be set at 20kHz, which is nominally the limit of human hearing (which in practice only goes to about 15-16kHz).

What bit depth DOES give you is extra headroom and dynamic range, which allows things to sound less "compressed" (a non-issue at 24-bits for most purposes). When you get crappy quality out of a DAC, it is usually caused by poor clocking in the stairstep generator (which adds nasty harmonics that can get through the filter) or a poor quality filter design for the Nyquist smoothing filter.

[ back to forum ]               [quote]