HOW TO ADJUST AD LEVELS IN THE STUDIO? What is a reference meter bridge? What is headroom? What is Clipping?
Introduction:
When drawing a picture on a page, one must keep the pen inside the page. The page itself sets the limits of the drawing. Similarly, there are two “hard limits” set by any digital audio format. We simply can not “go above” the maximum binary code (all 1’s) or bellow the minimum binary code (all 0’s).
When feeding an AD with a signal (time varying analog voltage), we expect the result to be a (time varying) sequence of digital code with a tight correspondence between the analog and digital. That correspondence will take place as long as we stay within digital “hard limits” of the conversion.
Clearly, going outside the allowed (available) digital range will be manifested as a hard limit, often referred to as clipping, generally a bad form of distortion coupled with a lot of unwanted aliasing (non musical sonic energy).
Note that the allowed range is set by the digital capability of the converter. It is about a binary codes (all 1’s or all 0’s). “Converter A” may reach the peak code (all 1’s) at say 5V, “converter B” may reach the same peak code at 18V. The relationship between the limit and the input voltage may differ from setup to setup, and from between converters.
The problem:
It is easy matter to apply a constant signal to an AD and to make level adjustment by use of meters (using UV, LED, software or other meters). The question is – adjust to what?
Clearly, one wishes to adjust the recording level so that the loudest peak will be shy of the clipping point. The problem is, of course, that the loudest peak may not be available at the time of the adjustment. For example, one may be setting up for a performance where the peak loudness will not be present until 30 minutes after the music began. In other words, we need a way to pre adjust the levels ahead of time!
Clearly, one can keep a large margin away from clipping by adjusting to very low levels. However, adjusting too low is also undesirable, because the signal gets closer to the noise floor, which will amount to reduction of dynamic range. Any attempt to correct it later – (amplifying the signal) will also amplify the noise.
So we are looking for a good compromise point – loud enough for best utilization of the dynamic range, soft enough to make sure we will not clip…
Where is the best setting?
Clearly, the relationship between the peak and the “average” loudness depend on the music itself. Classical orchestral music tends to have a lot of loudness variations, Hip Hop tends to have nearly the same loudness throughout the performance, and jazz falls in between… So setting the peaks levels for different styles of music at the same point (very near clipping) means setting the “average level” at different points. For example, a classical music recorder is likely to set the average level lower than say the rap music recorder, because classical music calls for more headroom for the loud music passages.
What is a good estimate for headroom? The specific adjustments are based on real life experience.
The standard method to adjust AD level:
The method requires a reference analog test tone, typically 1KHz sine wave with an amplitude of 4dBu (1.228 volts). That tone level is somewhat analogues to an “average” recording level. The tone is fed to an AD through a circuit with variable gain (volume pot). That volume pot is adjusted to make the AD converter generate a digital level with the desired headroom.
Experience shows that the desired headroom adjustment (with the specified 4dBu tone) for all types of music falls somewhere between -20dBFS and -10dBFS. The term dBFS means dB below full scale, 0dBFS is the clipping point. So in principle, one can just look at the AD level meter and make the adjustment to the desired level. Many of the classical music recording adjust to about -16dBFS or -14dBFS. Others may opt towards -12dBFS (it is not my business to supply or dictate specific guidelines)…
Reference meter bridge:
A front panel meter adjustment is often not optimal to make precise adjustments. For example, a meter made out of lamps spaced by say 3dB (0, -3, -6, -9…) is just too crude, because it can be off by as much as 3dB. The meters I use on my AD are 1 db steps (0, -1, -2, -3, -4….) which can introduce a 1dB error. A “reference meter bridge” is a way to expand the meter adjustment to a higher resolution. The user first sets the reference point for the bridge. The reference point is the desired headroom point we already discussed above (somewhere between -10dbFS and -20dBFS). The meter show the DIFFERENCE between the incoming signal analog level and the set reference point. The user can now adjust the volume pot for zero difference (between the adjusted analog and the digital reference set point).
It may sound complicated at first, but it is easy to learn and execute. The advantage of the reference meter bridge method is resolution. A meter aimed at 1dB steps in normal monitoring mode can be transformed to say .1dB steps in reference meter bridge mode. A .2dB step display will require 100 LED’s for a 10dB range! A reference bridge can yield the same precision with say 10 lamps, spaced by .2dB with the reference set in the middle. The question being asked here is: is the signal greater, smaller or equal then the reference. One adjusts the signal lamp to coincide with the reference lamp.
Note:
The precision of reference meter bridge is also helpful to ensure that all channels are set to the same level.
Best Regards
Dan Lavry
Property of Lavry Engineering Inc.
Look for Lavry Handbook in Late Spring 2005
Author
Topic: AD level adjustment and the Reference Meter Bridge (Read 3942 times)