Argthtjtr

From the wikipedia page about redundancy:

[...] redundancy is the duplication of critical components or functions of a system with the intention of increasing reliability of the system, usually in the form of a backup or fail-safe [...]

This practice is observed every time one deals with something critical: it is not uncommon for flight instruments to have even triple redundancy -- there are three instruments measuring the same value, and if one of them disagrees with the other two then it is considered broken.

I was listening to Snarky Puppy's Sylva, a massive non-stop 1-hour-long modern jazz performance executed in front of a live audience. I thought: what if the PC hosting the DAW crashed? What if a cable broke while they where recording? What if there was whatever hardware problem?

It is unusual for such a live concert to be non-stop and to be recorded, so on big stages if something breaks there's often some PA guy promptly running to the rescue. The incriminated cable is quickly swapped, the guitar player is up to speed again. Twenty seconds of guitar solo have been lost, no big deal. And if it's a studio recording -- well -- the sound engineer might just solve the problem and ask kindly for another take.

So here comes the question: what is standard practice for such "expensive" live performances? There are multiple cameras recording videos, but is there such a thing as multiple microphones/cables/DAWs recording the same audio?

So, I work in live production, specifically in audio. With very, very few exceptions the mics themselves and the mic cables have no immediate redundancy other than someone being ready to deal with a problem should one occur (though if a lead vocalist is using a wireless mic, there may be a wired mic tucked away someplace on stage the talent can get to quickly if such an arrangement is practical.) The reality is that the mics aren't terribly prone to failure while in use (except in musical theatre, but that's a different lecture). Cables? Most of them, once plugged in and laid out aren't being touched or moved; they're also not likely to fail. There are exceptions of course, but most of those scenarios have been replaced with wireless anyway. For all sorts of reasons, anything critical on wireless will typically have some form of redundancy available.

In the pure analog days, all the mic inputs went to analog transformer isolated splits. One and only one console would provide phantom power for those mics and DI boxes (which convert the signal from instruments like electric guitars to a level and type useful for mixing consoles) that require an external power supply; typically this would be the monitor console. The remaining splits (however many required - would go to FOH audio (the PA), broadcast, multitrack recording etc...) so that each would get its own clean perfect copy of all signals that they could manipulate, process, and mix as required.

So, at that point, a failure of something with the PA for example would leave the broadcast or recording unaffected. Now, in terms of multitrack recording, there will often be yet another split and multiple recorders, DAW or other multitrack recording devices.

Things are only sometimes different today. Most shows/concerts (though not all) use digital consoles. Virtually all these consoles are designed with the idea that your inputs go to one set of analog to digital converters and the signal then gets split and distributed to the various systems digitally. This does happen sometimes, but at this point, I'm still seeing analog splits and the various systems using their own converters after the analog splits. This choice gets made for a variety of reasons including, but not exclusively, to maintain an increased level of redundancy (other reasons include negating the need for the various audio engineers to agree and decide which engineer gets control over the analog gain applied to each signal prior to the A/D converters and what that gain should be set to.) We tend to be a prickly, opinionated lot and wars have been started over less. It's kind of a miracle you got us all to agree to the mic selection.

So basically so long as the producers/promoters will pay for it, we incorporate as much redundancy as is technically practical and financially feasible in any place that is particularly prone to failure. Many top end consoles for example have two mix engines (the computer part of a digital console); if one fails, you can switch over near seamlessly to the other from the control surface (which at this point is pretty much a glorified USB keyboard and less likely to experience a failure than the mix engine).

There are several youtube videos showing live setups for large bands, and you'll see that there is a LOT of redundancy in there. Thinking of Billy Joel's keyboard player David Rosenthal - he has two separate identical racks of equipment, set up so that pressing a single button can send the midi data to the B rack while the A rack is restarted, and he has a basic Hammond keyboard on stage in case all MIDI dies.

https://www.keyboardmag.com/artists/billy-joel-keyboardist-and-music-director-david-rosenthal

For guitar players - Brian May has his favorite guitars, plus several replicas, in case he breaks a string. Every thing else is also duplicated. https://www.premierguitar.com/articles/21168-rig-rundown-queens-brian-may

For recording, we duplicate everything. Even in our small church on Sunday, I'll feed the video camera with a shotgun mic and a bus feed from the board, plus a USB recording of the main outs, just in case something goes wrong. When working on larger events, we'll run similar systems (we set up bus mixes both pre and post eq, and record multitrack from each). We have a 40 track recorder (it can record all 40 at once), so we have several channels to record to (and could add more recorders if needs be, syncing them to an external clock); and we're a tiny company, just bodging things together in our own way.

I think you need to look at this historically.

Back in the days when the only real mobile multi-track recording facility was owned by the Rolling Stones, then you would be lucky to even be recording to multi-track at all live, let alone have any redundancy in the system.

The only 'redundancy' I can think of was that they had to run 2 totally separate sound systems, one for live & the other for broadcast. Remember images of such as Roger Daltrey singing into two mics, taped together?

Even more recently, I'm sure there was many a recording done on little more than a wing & a prayer... plus some rugged, expensive gear & some well-practised crews looking after it - & a car park full of diesel generators for mains power redundancy.

These days, when redundancy is simply hanging another SSD off the board, then it's easy.

Not a product recommendation, but there is gear like this:

https://www.scan.co.uk/products/tascam-da-6400dp-compact-64-channel-digital-multitrack-recorder

Feature headlines include:

• The DA-6400dp includes a redundant power supply
  In addition to the DA-6400 standard model, the DA-6400dp is available with two power supply circuits to allow redundant power from another AC circuit. This provides failsafe operation in case AC power is interrupted on one of the circuits.


• This (optional interface) card includes a redundant coaxial connection so that MADI is passed through even if power is lost, making it an ideal backup recorder for DAW sessions

In other words, you have some MADI stage boxes, and stick this device upstream of your main recording setup. It then records up to 64 tracks of anything passing through it. It has its own redundant capabilities, but even if it dies completely, it still passes through the signal to the next part of the chain.

So you could do a live mix down, a 32/64 track recording into whatever software you usually use, and still have a backup recording of the raw inputs. Combine that with spare hardware down on the stage, and you'll have access to pretty much everything that happened on the night.

there such a thing as multiple microphones/cables/DAWs recording the same audio

Multiple tape machines or hard disk recorders is pretty standard and has been for decades.

Multiple microphones and cables, not really that I know of: if a microphone gets trashed a stage person will go up there and replace it.
Of course that particular track for that particular song will be lost.

The particular instrument will be overdubbed in the studio or the whole song will be removed or replaced with the dress rehearsal.

Digital stage boxes often have the ability to share their inputs with mixers and software workstations connected to the network, allowing separate redundant machines to perform the same recording task. The sound engineer at the FOH desk may then record a mix directly from the mixer, using an external recorder, while the sound engineer's laptop records individual tracks as a backup.

In the past, it was common for that purpose to use (analog) stage boxes that had a parallel output for each input, isolated by an audio transformer. This stagebox was commonly located right next to the monitors mixer location (often to the side of the stage), allowing the monitors mixer to use the outputs directly, while the FOH mixer would use the multicore cable connected to the stage box. This way, both the monitors and the FOH positions may record the show, a mix or even individual tracks, provided they had the required material.