Review of a Soundcraft Spirit Folio 12/2 audio mixer

(picture of the mixer)
View of the mixer

At first glance, it looks a bit toyish.  It's small, it's light, it doesn't look robust, it's very plasticky.  More in-depth inspection doesn't give a better impression.  I had seen some people extol the virtues of Spirit Folio mixers, but I can't imagine why, unless they'd compared it to something much worse.

After a play around with it, the plasticky opinion springs foremost to mind.  Every rotary or slider control is wobbly, very wobbly.  That doesn't bode well for longevity, and feels horrible to use.  We rely on the pot shaft to stop external forces from breaking the internals of the pot, but a wobbly pot shaft isn't going to provide that protection.  And it seems like it'd be all too easy to snap the shaft off.

The controls are very tiny, and crammed together.  It's not designed to be operated with man hands.  And thanks to them being zig-zagged, it's not a quick and easy task to reset all the controls (turn auxillary sends to mute, and the equalisation to flat).

There's number of buttons that lock down (low pass filter, input gains, et cetera), that give little indication of whether they're pressed in, or not.  They're very low profile, there's no coloured edge around the mid-way point, nor any illuminations.

While it's nice to be able to select between pre or post for one of the auxilliary sends, unfortunately it's a master selector for all channels.  Likewise the phantom power switch is one master switch for all channels, a potentially destructive design.

The pre-fade/post-fade listen buttons do not lock down, which I find annoying.  It can be a pain to have to keep one finger down on a button while you're tweaking knobs, elsewhere.  Probably the one redeeming feature about the pre-/post-fade listen is that it functions in solo mode (main stereo monitoring is interrupted by the switch, rather than adding to it).

There's eight balanced microphone or line inputs, with microphones on XLR connectors, and lines in TRS jacks.  The XLRs have the locking latches removed, which is handy, and latches are pretty pointless on top panel mounted sockets.  Though I'm not too fond of top panel sockets, crap falls into them, and stuffs them up.  The mike and line inputs are wired in parallel, with only the attenuator, and phantom power isolation, circuitry between them.  It means that you can't plug one in to switch off the other, but on the other hand, it means that a worn out socket switch doesn't kill off the other input.  But, personally, I prefer mixers which let me use either of the XLR or TRS for line inputs.  I don't want to have to change leads, based on the strength of the signal at the other end of the cable.

There's two additional stereo balanced line inputs, using TRS connectors.  These only have a +4/−10 input gain switch, whereas the main channels have a continuously variable gain pot (but no pad switch).  And these stereo channels have a simpler equalisation stage, only providing bass and treble controls, whereas the main channels have treble, variable frequency mid-range, and bass controls, and a 100 hertz low pass filter switch.

And there's a stereo unbalanced line input, using TRS connectors, for monitoring the signal coming back from a recorder.  It can be routed to the mix output, or just to the monitor outputs.  The meter switches with the monitor, so the metering will show the signal coming back from the recorder.  There is an input gain control, so you can adjust the level correctly.

All the output sockets are TRS.  The main and auxilliary outputs are impedance balanced, the monitor line and headphone outputs are unbalanced.  The headphone socket cuts off the monitor line outputs, when headphones are plugged in.  That's not an option I'd choose, I tend to always use monitor speakers, but occasionally plug in headphones to listen carefully to something, while the speakers continue to play for everyone else in the room to listen.  And it's turned out to be a right pain when the switch contacts failed to reconnect the monitor line outputs after the headphones were unplugged (no, I'm not talking about theoretical faults, I've had to deal with this mixer doing that).

Impedance-balanced outputs are not the same as normal balanced outputs.  Normal balanced outputs, as traditionally used, have an in-phase and out-of-phase signal (positive and negative phase, or the so-called, and completely erroneously named, “hot” and “cold“ wires) carried across two wires.  Impedance balanced connection is the same as unbalanced connection, you have the signal and a ground, and the third connector is merely an impedance to ground that's the same impedance as the signal source.  It has the advantage that it's easy to connect the output to a balanced input, or an unbalanced input, without shorting any part of the output stage out.  But loses the advantages of a proper balanced signal, and that the audio signal isn't referenced to ground.  For this mixer, the output stage is apparently 75Ω, and the non-signal carrying negative phase just appears to have a 75Ω resistor to ground (that's been measured using a meter, rather than determined by having a look at the circuitry).

There's a built in 1 kHz tone generator, though it appears to be 10 dB too high in output level.  The set-up procedure for this desk indicates that the input channel faders' nominal holding position is the zero mark, which is 10 dB down from the top of the fader's travel; and the master faders' is also at the zero mark, but at the top of the fader's travel (where I prefer master faders to be, giving you a simple-to-use fade in/out control).  And playing with the pre-fade listen, while observing the metering, appears to confirm this.  But, if you turn on the oscillator, you need to turn the master faders down by 10 dB, so the tone is at reference level.  So I fixed this by inserting 109 kΩ of resistance between the tone switch and stereo bus.  Now it operates logically—with the master faders in the normal position, the tone produces reference level for the recorder.  If I really want to play around with above normal test tones, I plug in a separate test tone generator.

Initially I opened the mixer up to fix a fault in three of the input channels, they were hissy, and distorting.  The reason was immediately obvious, nearly every single solder joint was bad, there was probably only a dozen solder joints that weren't bad.  Those three faulty channels had dry joints where it was causing a noticeable fault to the signal, the rest of the board had obvious dry joints that hadn't yet caused an noticeable audio fault (but, later, did).  So I ended up resoldering the entire circuit board.  As well as the dry joints, many of the component leads didn't want to take solder, so it took a bit of heating and scraping to fix those connections.  I was taught to solder by an ex-navy guy, to military specifications (according to how he rated our soldering skills), so the mixer should probably well outlast its usefulness, now.  But dry joints continue to return to the mixer, and I'll blame the bad metal on the component's legs for that.

While it was open, I decided to have a look at what's in the mixer, although this required removing every single knob, screw, bolt, and connector fastener, to get the circuit board out of the box.  While some may say that the all-one-board approach is better for reliability, I prefer the modular approach for serviceability.  And the board's many dry-joint faults back up my preferences, with faults everywhere, rather than just one or two channel boards which could, otherwise, be removed.

The pots are all ALPS brand, the first eight channels head-amp stages use four transistors, but most of the rest of the amplifiers are TL072 ICs (not the quietest of op-amps), although there's some NE5532s in the main output stage (that's the dual version of the NE5534, that isn't quite as good as the NE5534, and noticeably worse than the NE5534AN).  The peak level meter is driven by a collection of LM2901 quad comparator ICs, rather than a special meter driver IC.

I haven't checked whether the meter uses any standard ballistics, but it has a near to instantaneous rise-time as you can tell by eye, and takes about three seconds to decay from full scale illumination to none.  There is no peak-hold, nor any average level indication.  Like most simple meters, it's good enough to avoid clipping, and to keep your signal within the optimum range of the equipment, but not good enough to mix by (you couldn't easily set audio to a specific signal level).  There are no clip indicators anywhere, not on the input channels, nor the output.  You'll have to set up properly, and not over-drive the meter—which goes up to +12 above reference, which would be +16 dBu out, and the mixer output is rated to +22 dBu.  So if your meter readings are kept below full-scale, you should be at least 6 dB below the clipping level.  Though, with the absence of input indicators, it's easily possible to overdrive an input stage, while under-driving the output stage.

There is no power switch, so if you want to kill the mixer while you patch, you have to pull the power plug out.  It uses an external transformer, supplying low voltage AC to the desk through a three-pin connector.  For those wanting to replace their power supply, you need a centre-tapped transformer, 20.7 volts – 0 volts – 20.7 volts, 400 mA.  The power supply is not earthed, so you may need to be connected to something else that is earthed to reduce noise pick-up in some environments.  That's not too hard to do, and easier than trying to deal with an earth loop between your mixer and other equipment that it connects to.

All the TRS jacks are domestic A-gauge 6.35 mm phone plugs, the type used by older headphones and guitar leads.  B-gauge is the type used by ancient telephone switchboards and professional audio patch panels (the tip of B-guage plugs ends in a ball, not a point).

The XLR sockets are plastic Neutrik connectors, which do not ground the shell of the plugged in connector.  So you better not be in an electrically hostile environment, nor touch the shells, or you may get hum into microphone signals when the gain is set very high.  I've never been fond of Neutrik connectors, they're brittle plastic affairs, and the cable-mounted ones don't survive being slapped into the floor while cables are wound up, stepped on, squashed by closing doors, or being crushed by heavy objects, like the older metal Cannon connectors could withstand.  Pin one of the XLRs goes to signal ground, rather than chassis frame, so that can lead to some noise problems, though it may be the simplest solution for allowing unbalanced microphones to be used (but who uses unbalanced microphones with XLR leads?).

So, what does it sound like?  When working properly (after fixing the dry joints), it sounds reasonably good.  Not too noisy, tone controls sound good, but the low-cut switch barely makes a difference (about all it can do is slightly limit sub-sonic rumbles, it's certainly no good for reducing wind noise picked up by a microphone.  It doesn't have a great deal of microphone gain, enough for recording a band where they play and sing loud, but rather inadequate for recording something that's quiet.

And what's it like to use?  The level meters appear to be peak reading meters of some variety, so they're useful for avoiding over-driving the signal levels, but not for mixing by.  Which is probably good enough for a lot of people, where your concern is about putting reasonable sound levels through the desk, and using your ears to mix by.  And, at least, they don't madly flutter about like the near-useless meters in the Yamaha MG12/4 audio mixer.  The wobbly and crammed-together controls are annoying, and the linear sliding faders have a nasty grinding feel to them, like your scraping over a rough surface, through a thick coating of vaseline.

Unusually, all the line outputs (main, auxilliary, and monitors) seem to be 0 dBu, rather than much more common +4 dBu.  Which may be helpful with people using amateur equipment, so you don't blast their input stages with a lot of signal, but it's too low for pro and semi-pro gear.  There are no insert points on the channels, only a pair of insert points just before the main stereo bus master faders, and at another odd level of around −10 dBu.  So your options are quite limited if you want to use any external effects devices.  It's a nuisance that the line inputs only have +4/−10 dB input gain choices, with no variable gain, as experience has shown that line sources are all over the place, especially if you have to connect non-pro equipment, and you end up having to use their main faders at inconvenient positions.  It's normal to adjust input gains so fader pots are operated around 10 to 15 dB down from the top.  This gives you most of the pot travel to do smooth fades, or adjust mix levels.  With a bit of extra space should you need to boost a signal a bit, and avoiding having the pot right at the top where it might go off the top of the track.  But when you have to use faders nearly all the way up or down, you lose the best part of the pot's operating range.

Having impedance balanced outputs is a bit of a let down.  While it helps users with unbalanced gear, so that they aren't shorting out one leg of a balanced output, there are other ways of making a balanced output that doesn't care about that (the so-called servo-balanced outputs, or just having sufficient output impedances on each leg that you can't cause a dead-short).  And should someone connect a phase-reversed lead between an impedance balanced output and input, they'll get no sound, at all.  But the biggest issue with not having floating outputs is having to use out-board devices to get rid of any ground loops.


Written by Tim Seifert on 29 September 2022, and last updated on 10 October 2022.