Ronny wrote on Wed, 07 September 2005 01:50 |
What is the physical reason for not getting a good sound above -6dBFs, Bob? Not that I'm a stickler for squeezing every bit out, especially at 24 bit, but you'll find a trainload
|
Ronny, it's not the dBFS that Bob O was referring to, but the internal headroom of the analog electronics around and inside the converters. I've written extensively about this. Firstly, many lower cost (and even higher cost) pieces of gear are not designed with good internal analog headroom ABOVE the digital peak point.
Some integrated circuit opamps that I've encountered over the years seem to start to sound "ratty" as they approach their clipping points, their distortion characteristics shoot up as they approach clipping, especially at high frequencies. It is not fair to generalize, since the expertise of the designer, power supply design, bias, etc. count. But if I were to generalize, I'd say that caution is warranted as you approach the clipping level of an opamp, especially an integrated circuit opamp. I've encountered MANY opamp-based circuits that sound a lot better when driven to a max of 6 dB below clipping, rather than within 1 dB of clipping, which is a common practice in many designs to try and get high output levels.
Of course this is not universally true. The increase in distortion as clipping is appraoched is dependent on the implementation, brand and type of opamp, power supply voltages, how the opamp is biased. Regardless, I suggest caution in any case, especially in mixed signal (digital and analog) environments if the clipping point of the analog circuitry is being approached along with 0 dBFS (and measured 0 dBFS+ overs as well) at the same time, it's a recipe for ratty sound.
I've built unbalanced IC-based mixers with + and - 18 volt power supplies, using Burr Brown and other high quality opamps and despite the pedigree, I found they just sound better running "0 VU" at lower than 0 dBu, even as low as -6 dBu (0.0775 volts). Clipping point of these, running unbalanced, is nominally +20 to +22 dBu.
Thus, we are finding balanced preamps of high quality now which clip at far above +24 dBu, as high as +37 in some cases. The term I've been using for this necessary "analog headroom above 0 dBFS" is a "cushion". It's not a very scientific term, but it's the best I've got. To summarize: you should never operate certain solid state gear close to clipping; align the clipping point of the analog gear to AT LEAST 6 dB above the 0 dBFS point of the associated digital circuitry.
As for "people on the net saying you get a better sound when you are as close to 0 dBFS as possible", I tend to disagree. Given the evidence above, I'd say that is a generality as close to "he said, she said" as I've heard. Please let's hear some specific quotes from individuals who have tested and listened and reached these conclusions. If I were to generalize, I'd disagree; I'd say that "in general", you're a lot safer peaking somewhat lower than to full scale, for some of the reasons I mentioned above, and in addition, inaccuracies of digital meters, 0 dBFS+ signals, etc.
BK