Almost always a ground loop, and in Toluca Lake the reason is usually the age of the wiring rather than the gear. The neighbourhood was subdivided in 1923 and most of the original houses went up before the war, so a large share of the outlets are two-prong with no equipment grounding conductor behind them, and the ones that were retrofitted were retrofitted at different times by different people. Two devices then sit at slightly different ground potentials, current flows down the shield of the cable between them, and you hear 60 Hz on everything.
Why does this neighbourhood in particular have the problem?
Because of when it was built and how it was rewired. Toluca Lake was developed as Toluca Lake Park in 1923 and filled in through the 1920s, 1930s and 1940s in Spanish Colonial, Tudor Revival and American Colonial, with post-war ranch houses filling the gaps (Toluca Lake, Los Angeles). A house from that period was wired without an equipment grounding conductor. Eighty years of additions, kitchen remodels and garage conversions then left a building where the kitchen circuit is modern, the living room is not, and the garage is on something a previous owner ran himself.
That is the condition that produces loops. One circuit has a real ground back to the panel, another has none, a third has a ground bonded at a different point, and any cable you run between rooms connects two of them.
The administrative half matters too, because Toluca Lake is a neighbourhood of the city of Los Angeles that spills across the line into the separate city of Burbank. The Toluca Lake Chamber of Commerce serves the community on both sides, while the city of Los Angeles treats its Toluca Lake as lying entirely inside its own boundary. Which building department issues a permit for a dedicated circuit depends on which side of that line the house sits on.
What is a ground loop, actually?
Ordinary house wiring behaves as a set of loops carrying alternating current, and the magnetic fields from that current induce small voltages in everything running alongside it. The result is that the safety ground connections in different rooms are not at the same potential. The differences are small, often a few tens of millivolts, and they are enough, because the shield of a balanced cable is a fraction of an ohm (Bill Whitlock, Jensen Transformers).
A few millivolts across a fraction of an ohm is a real current, and it flows in the shield of the interconnect between two grounded devices. That current is the hum. Nothing is broken, no component has failed, and no amount of expensive cable changes the arithmetic.
Why is it called a pin 1 problem instead of a cable problem?
Because the fault is inside the box, at the point where the shield lands. Neil Muncy named it in the *Journal of the Audio Engineering Society* in June 1995: on badly built equipment the shield connection at pin 1 of the XLR is routed to the internal audio reference instead of straight to the chassis, so the shield current runs through the circuit ground on its way out. Whitlock puts that mechanism behind the overwhelming majority of system hum (Grounding and System Interfacing).
The practical consequence is worth stating plainly: swapping cables is the most popular fix and the least likely one to work. If the shield lands in the wrong place inside a device, every cable in the world lands in the same wrong place.
How do you find it in twenty minutes?
Five steps, in order, listening after each one.
1. Unplug everything from the interface except the monitors, and leave both on the same outlet. Note what noise remains. That is the floor. 2. Add one device at a time, listening after each. The device that raises the noise is one half of the loop. 3. Move that device onto the same power strip as the interface. If the hum drops, the loop is in the power path, not the signal path. 4. Swap the interconnect for a known-good one of the same length. If nothing changes, the cable was never the problem and you can stop buying cables. 5. Break the loop at the receiving end with a transformer, not at the sending end and not at the wall.
Never lift the safety ground to test a theory. It removes the fault path that keeps a chassis from becoming live, and it is the one experiment in this list that can kill somebody.
Is it 60 Hz, 120 Hz, or hash?
The spectrum names the fault, and any analyser with a 1 Hz resolution will show it.
A strong 60 Hz fundamental with weak harmonics is magnetic induction into a loop, and it responds to moving cables away from power runs. A 120 Hz component with a long series of harmonics above it is rectifier current or a phase-control dimmer chopping the waveform, and it responds to putting every dimmer in the house at full or off. Broadband hash that sits above the audio band and folds down into it comes from a switch-mode supply, usually a phone charger or a laptop brick sharing the strip.
In an old house the dimmer is the most common recent addition, because LED retrofits went into fixtures whose dimmers were designed for filament lamps.
What does the code actually allow?
If a house has no equipment grounding conductor at an outlet, the National Electrical Code permits replacing the two-prong receptacle with a ground-fault circuit interrupter, and requires that the receptacle or its cover plate be marked "GFCI Protected" and "No Equipment Ground", visible after installation. An equipment grounding conductor may not be run between the replacement receptacles (NEC 406.4(D)(2), Up Codes).
Read that carefully before spending money on it. The GFCI protects people. It does not create a ground reference, so it does not change the hum by a decibel. Cutting the third pin off a plug or fitting a cheater adapter is not a fix either, and it defeats the only protective path in the system.
The change that helps both problems is a dedicated circuit with a proper grounding conductor run back to the panel, with every piece of audio gear on that one circuit and nothing else on it.
What actually fixes it in a rented pre-war house?
In order of cost.
Put everything on one power strip on one circuit, which costs nothing and solves a surprising share of cases. Use balanced connections wherever both ends support them, since a balanced input rejects the voltage difference that drives the loop rather than passing it (Rod Elliott on balanced interfaces). Insert an audio isolation transformer in the offending line, which breaks the shield continuity while passing signal. Replace switch-mode wall warts with linear supplies where the device tolerates it. Fit an isolator on the cable television or internet coax, which is a ground path most people never think of and which is bonded outside the house. Only after all of that, call an electrician about a dedicated circuit.
When is it cheaper to stop and drive?
When the fix requires a permit and the session is next week. Toluca Lake to my room at 2226 N Catalina St in Burbank is about four miles, ten to fifteen minutes along Riverside Drive outside rush hour, and a vocal that has to be clean is worth the drive rather than an afternoon spent chasing a loop in a house you do not own. What to bring and how to prepare the files either way is in what should be in the folder you send to a mix engineer. Credits with sources are at /credits.
If you have hum and a session booked
Send a thirty-second recording of the noise with nothing playing, at a known gain, and a list of what is plugged into what. The spectrum of that file usually names the fault before anyone drives anywhere. Contact at /about.