A question about dBVU and dBFS and distortion

Support for JLM Audio Kits

Moderator: Joe Malone

Post Reply
tenchijin2
Posts: 12
Joined: Thu Mar 12, 2009 1:15 pm

A question about dBVU and dBFS and distortion

Post by tenchijin2 »

Hi Joe (et al),

There has been a lot of controversy about gain staging with regard to digital vs. analog gear. To summarize, here's what I've learned:

Most analog gear is designed to run at 0dBVU, with headroom above that generally +18dBVU to +22dBVU (allowing for some variation).

Most meters inside the sequencing software measure dBFS, and 0dBFS generally = +18dBVU to +22dBVU (again, allowing for variation).

Factually, I'm pretty sure I've got that right.

So here's where things get a bit troublesome. If my analog gear is designed to run at optimum signal 0dBVU, then when I am using my sequencer's meters (in the computer software) I should see about -18-(-22)dBFS depending upon calibration, right?

Wouldn't running hotter than that be simply increasing distortion in the analog device? In other words, if I'm reaching say -4dBFS on my digital track meter, I'd be running around +14dBVU. Isn't that about 14dBVU too hot for optimum performance?

People generally try to run the digital meters up near 0dBFS to get "the hottest signal possible" but again... doesn't that overdrive the optimum level of the preamp or other analog device?

I'm asking because I know that Joe (and probably some others among you) likely worked on analog studio gear before the age of digital. Was it typical to run your preamps or other analog gear at +whatever dBVU all the time, or did you guys shoot for 0dBVU?

Consider the can of worms open!

Aric Keith

chrisp
Posts: 184
Joined: Sun Mar 04, 2007 11:46 am
Location: Sydney, Australia

Re: A question about dBVU and dBFS and distortion

Post by chrisp »

Welcome to the forums, tenchljin

You've got the maths pretty much about right in terms of -20dbFS (or 2db either side thereabouts) = 0dbVU (an arbitrary number, its just a adjustment point on the VU) = +4dbu (=1.2V or so).

As to whether a really hot signal at +16dbu (with the VU meter nailed to the right) distorts the analog side depends entirely on the quality of the analog side. There's plenty of classic analog gear that can deal with incredibly hot signals, and even sound better when driven hard. Other stuff can wimp out even at +4dbu and sound bad. Its in the design and the quality of the gear.

One guide is the power supply. An op amplifier circuit, for example, cannot put out more voltage than its power supply rails. In the case of the JLM99V (original) op amp, for example, its power supply can be up to +/-34V, which is what, +33dbu !! Even at +/-24V like a simple BA build it can crank out something near to +30dbu. The more moderate +/-16V version could crank out +26db or something like that at maximum.

Of course, no circuit performs linearly through its possible operating range. Each has a sweet range of linearity, and outside this it distorts in either musically good or musically bad ways. Good equipment tends to have a wider, more useful sweet range, and distorts musically outside this (think tube sound or overdriven guitar as examples). A run of the mill opamp IC might look good on paper, but the question is how wide is its sweet range and what does it do outside that range? Again, this comes back to quality components and good circuit design (optimal IC operation, for example, depends on a really good power supply, and a really good chip like a Burr Brown can sound bad if not given a solid power supply - including looking at the current passing through the ground rail).

Most (good) "pro" analog gear is designed so that +4dbu lies in its sweet spot. Whether driving the gear to produce a hotter signal makes a better recording depends on the quality of your gear and its interconnections. Similarly, when playing back, there's no a lot of point in pegging the DAW needles to -1dbFS if the gear receiving the signal can't handle that. There are also plenty of audio reasons for not trying to max out the DAW, but that's the loudness war and its been done to death.
So here's where things get a bit troublesome. If my analog gear is designed to run at optimum signal 0dBVU, then when I am using my sequencer's meters (in the computer software) I should see about -18-(-22)dBFS depending upon calibration, right?
Yep. In some DAWs, that a just small glimmer at the bottom of the meter display. That's why better metering (either in software or hardware) is really useful.
Wouldn't running hotter than that be simply increasing distortion in the analog device? In other words, if I'm reaching say -4dBFS on my digital track meter, I'd be running around +14dBVU. Isn't that about 14dBVU too hot for optimum performance?
It will be a hotter signal than the device was designed to produce optimally. Whether that's good or bad depends on what you are trying to achieve musically, especially once the analog gear gets out of its zone of lineality.
People generally try to run the digital meters up near 0dBFS to get "the hottest signal possible" but again... doesn't that overdrive the optimum level of the preamp or other analog device?

There was a thought that not recording near to 0dbFS meant you were wasting digital "bits" and so losing resolution and worsening signal to noise. The mathematical argument might seem appealing but is flawed, especially in practical terms when dealing with 24 bit resolution. An yes, trying to get there using a preamp is likely to get you into troubles.
Was it typical to run your preamps or other analog gear at +whatever dBVU all the time, or did you guys shoot for 0dBVU?
Hey, I'm a child of the digital age, so others would be better able to speak. But basically, every recording setup was calibrated to some 0dbVU, and often that was at +4dbu, and yes, signals were aimed for that or for some things like snare drums, maybe a little hotter like an extra +2dbu. It depends whether the metering was peak or average, but they certainly did not aim for anything remotely close to +18dbu.
Chris P
---------------------------------------------------------
I do lots of things. I believe eclectic skills are best.

Post Reply