Topic: Level mismatch solutions - fixed and variable attenuators
Level mismatch is a continuing topic on our forum and support line. The most often reported problem is that active monitors are far too sensitive. The low volume click and pop when turning a unit on/off turns into a quite loud one. A volume control pot then has a very limited operation area, and the master fader in TotalMix has to be lowered to below -20 dB constantly. This also raises concerns about reduced audio resolution. With such a level mismatch hum and noise can also become audible.
What needs to be done in such cases is to make the connected amplifier or speakers/active monitors as insensitive as possible. Ignore whatever the current setting print on the monitor says, choose one that makes them as low in volume as possible.
Some customers encounter situations where this is not enough. We heard of professional and very powerful active monitors using a reference level of +4 dBu, which seems to equal full output power. Setting an RME unit to -10 dBV, which equals +4 dBu at full output level, then again requires to lower TotalMix FX to far more than -20 dB to be able to work at a typical studio listening level - plus having all the above disadvantages/problems.
The solution here is to use passive level attenuation between the RME interface and the amp/active monitor's input by a simple voltage divider, built of 2 (unbalanced) or three (balanced) resistors per channel. Do a Google search for 'line level audio attenuators' or just 'audio attenuators', and you will realize that this is not a RME problem, but a long time classic found everywhere - even in home HiFi. And this is also the reason that the web is not only full of examples on how to build such parts, including the part values and circuit diagrams, but why they are also available at reasonable prices in various models from various manufacturers.
If you don't want to waste your time with ordering parts, soldering, and drilling housings, here are two solutions that have been tested to work perfectly with RME devices:
For balanced outputs: JTS Mic attenuator MA-123, XLR to XLR, available for around 20 €, switchable attenuation of -10, -20 and -30 dB. Don't get confused with the term 'mic attenuator'. It works perfectly with line outputs. In fact it works even better as other 'line' solutions, because its impedance is high enough for the RME output stages to not cause distortion (1 kOhm), but very low for any following device (below 200 Ohm), which means there is no added low level noise and no high frequency rolloff caused by cable capacitance when plugged into the output of the RME device. Of course, cable influences can be ruled out completely even with higher impedance attenuators by plugging them directly into the amp/monitor's input jack - at the end of the cable.
For unbalanced outputs: IMG Line attenuator ILA-1020, RCA to RCA, switchable attenuation of -10, -15 and -20 dB. One pair is available for around 25 €. Note that these use much higher resistor values than the balanced ones, more than 10 kOhm input and around 1 kOhm output. Therefore they should be placed directly at the input of the receiving unit, not at the output of the sending one.
These are just two examples. There exist many other, like:
Pro Co Max20
Rothwell RCA attenuators
Harrison Labs RCA Line level attenuators
Another way to solve the level mismatch is by adding a variable resistor (aka potentiometer) instead of the fixed ones. Put the pot into a desktop housing and you have a variable volume control. These are also very popular, examples:
SM Nano Patch +, also available from JBL
They have a few advantages: volume controlled from your desktop, mute switch provided, also serves as format adapter as it provides XLR, TRS and small TRS stereo on input and output simultaneously. The current generation uses quad potentiometers, so work both balanced and in stereo, and are also automatically compatible to unbalanced operation. All passive!
Compared to the above fixed low impedance resistor solutions the variable ones have a technical disadvantage - they use pots with 10 kOhm resistance, 10 times as much as the 'mic' attenuators. Ideal they would use 1 kOhm pots, but maybe decided for higher values as some devices have much higher output impedances as RME units. With 10 kOhm and the pot in the middle position a series resistor of 5 kOhm is in the signal path, and a parallel one of 5 kOhm works as load. Depending on the used cables and input circuitry a slightly increased noise floor as well as a slight high frequency damping might occur. In real-world operation and usage user feedback is very positive when using these devices.
RME interfaces and converters provide several, user adjustable output levels, from -10 dBV up to +24 dBu. Most devices support +19 du, +13 dBu and +4 dBu (aka -10 dBV). Some devices like the Babyface and Babyface Pro do not support different ref levels at their outputs, for example the BF Pro XLR outputs are fixed at +19 dBu. As explained in the manuals a digital fullscale level in -10 dBV setting does not equal -10 dBV, but +2 dBV, or coarsely +4 dBu. The reason is that analog equipment, especially mixing desks working with this reference level, offer a very high headroom - not astonishing, as the internal circuitry ist still running with +-15V, the same voltage that is used for +24 dBu output level stages. We therefore defined our -10 dBV setting with 12 dB of headroom.
The attenuation required in mismatch cases as described above is typically -20 dB.