Review of behringer U-PHORIA UMC1820

I'll also refer you to read my review of the behringer U-PHORIA UMC404HD as a very similar product (less input and output channels, but with the ability to use some higher bit-rates—though that's mostly a pointless thing to do).  I bought it first, and upgraded to this model, later on.  Then lent the 404HD to a friend, and haven't asked for it back.

I bought this (on-line in December 2020 from Mannys for around $477AU) because I wanted a box with more input channels, as I'd been doing some music recording collaborations during the covid crisis with a few friends.  Though it arrived after most of the restrictions on the number-of-people congregating changed.  But I've been using it with recording my organ, which has been modified to provide several separate output channels for different sections (bass, reverb, synth, drumkit, etc).  So I don't have to do purely acoustic recordings, or simple stereo direct recordings, and can post-mix some individual sound levels.  I prefer playing the organ as an ensemble, simultaneously playing melody, accompaniment and bass.  I dislike playing one part at a time, doing overdubs (I find it hard to do, and it's not an enjoyable way to play).

When over-dubbing/multi-tracking, where you are making a new recording track while listening to playback of a previous one, there is a latency that needs to be compensated for (playback buffering & input delay).  Either you have to slip your new track back slightly to sync it up, or configure the latency into your recording software for it to automatically do it for you (and it could either pre-buffer the playback audio, or delay the recording to match the playback latency).  There may be software that can work out the delay by itself, but I don't know any, it should be possible.

But you can listen to your audio input sources latency-free while recording, via its analogue input monitor circuits.

I use this with Audacity on CentOS 7 (Linux), and it works as I expect.  I connect the box, and just use it, there's no futzing around with drivers.  All I have to do is select the right input and output devices to use (my computer has an on-board sound card, and the HDMI monitor has in-built speakers, they tend to be the default devices).

I can plug 8 things into the analogue inputs and record them individually, simultaneously.

I can also plug other things into the digital inputs, such as the electrical SPDIF output from my Lexicon MX200 reverb unit, or the optical SPDIF output from a CD/DVD player, and record up to 10 devices simultaneously (8 analogue and 2 SPDIF).  The device has up to 18 input channels when using analogue and optical inputs (that's in ADAT mode, rather than SPDIF mode), but I don't have any ADAT devices, and I've not managed to exceed 10 sources with Audacity, so far.

There's 10 analogue output channels, and 10 more digital output channels, but Audacity only plays back in two-channel stereo, so I can't use the rest of the outputs in any sensible fashion (it can render an output file as multi-channel audio, and you get to choose which input channels are routed to which output channels).  But you'll need to use something else as the player.

Also, my computer system decides any multi-channel output device is some kind of surround-sound device and always uses the output channels in its own fashion (I seem to recall that an older release would also let me configure things to use it as a basic two-channel stereo device, ignoring extra channels, but I don't have any option like that currently).

It's in a standard 1RU high 19″ form-factor box.  It has some small feet underneath it, but I found it doesn't have enough internal weight to stop it skidding around when it's just sitting on top of something.

Virtually everything is controlled by the switches on the box, making it a universal device that doesn't depend on specialist software to be useable.  The only thing I see that's computer-controlled is bitdepth and bitrate, with that being set by your audio software.

The UMC1820 has 8 analogue inputs, with individual channel switches for attenuation pads, high-impedance inputs (e.g. for guitar pickups), and gain controls (that's gain, not signal level—it's not a mixer).  Mike and line inputs are supposed to be able to handle a maximum of 11 dBu, with the high-impedance line input going up to 18 dBu.

It uses combination XLR & ¼″ mike/line input sockets.  I'm not a fan of these, they don't grip as well as dedicated connectors, and don't handle strain very well.  For the sake of reliability, and avoiding permanent damage, don't have heavy cables hanging off them, or leads stretched between things, they need some form of cable support.

XLR input impedance was around 13 kΩ (measured between pins 2 & 3), though the manual claimed 3 kΩ.  Normally the TRS inputs are around 150 kΩ, but when the Hi-Z instrument switch is pressed in they're 750 kΩ (measured tip to ring), so you can (supposedly) plug an electric guitar directly into it.  The manual claimed 1 MΩ for instrument inputs, and said nothing about normal line inputs.  The XLR inputs can supply +48 volts of phantom power, with two switches to switch phantom power on/off for inputs 1 to 4 as a group, and inputs 5 to 8 as another group.  It would have been nice to individually control power to each XLR input, but it'd need to be built into a bigger box to fit switches for all of that in.

There's 10 analogue outputs, the first two being a stereo monitor output that's mixable between input-monitoring (switchable for either just channels 1 & 2 or 1 to 8, and switchable between stereo odd/even—left/right or mono monitoring) and playback-monitoring of channels 1 & 2, and an output level control with -18 dB dim and mute switches.  It would have been handy if that mono button monofied the monitor output no matter what it was monitoring, but you're dependent on either having an external mono monitor switch, or DAW software with a mono monitor function, if you want to mono-monitor playback.  The remaining 8 analogue outputs are dedicated for channels 3 to 10 playback-monitoring.  The line outputs are meant to be able to handle up to +16 dBu maximum

There's two stereo headphone outputs that (individually) can listen to the main monitor out (input and playback channels 1 & 2 mixer), or channels 3 & 4 playback, with their own switching and level controls.  They're meant to be able to handle up to +21.5 dBu maximum.

The manual lists various maximum output levels, but says nothing about nominal signal levels.  That may depend on your software, if it goofs in the way it handles internal bitdepths when driving external hardware with a different bitdepth.  I've seen audio software that internally clipped, without ever driving audio hardware to full strength.  And the opposite, of inputs driving hardware to clipping level, yet the audio recording software was only half-scale.

The mike and line inputs are balanced (both input poles are active).  But the line outputs are impedance-balanced (only the tip is driven, the ring is simply a dummy load to ground that's supposed to be the same value as the driven pole's output impedance).

There's digital inputs and outputs, supporting electrial SPDIF, and optical SPDIF or ADAT (apparently 10.6 and older MacOS releases can't support ADAT, and 10.5 or older is completely unsupported).

There's 5-pin MIDI in and out (though I have no instruments with 5-pin MIDI to try it out with).

And it's powered by a wall-wart plugpack, not the USB connection.

SPDIF/ADAT/SMUX input & output

S/PDIF (Sony/Philips Digital InterFace) is a two-channel optical (using plastic fibre with TOSLINK connections) or electrical (using RCA connections) uncompressed audio signal scheme based on AES3.  It can also be used for compressed digital audio for surround sound with DVD and Blu-ray players.  It can be 20- or 24-bit (20-bit being original spec, and 24-bit may not be supported on all equipment), and commonly uses 44.1 kHz or 48 kHz sample rates, though other rates are possible (e.g. 32 kHz).

ADAT (Alesis Digital Audio Tape) is a 24-bit four- or eight-channel optical audio signal scheme (using plastic fibre with TOSLINK connections).  Four channels at a high bitrate (up to 96 kHz), or up to eight channels at a lower bitrate (48 kHz).

S/MUX (sample multiplexing) is a two- or four-channel optical audio signal that's a variation on ADAT (also using plastic fibre with TOSLINK connections).  Two-channels at a high bitrate (up to 192 kHz) or four channels at a lower bitrate (up to 96 kHz).

The electrical and optical digital inputs and outputs are affected by a front-panel OPT I/O (the label suggests it only changes the optical connections).  The analogue inputs and output channels do not change any of their functions.  It's somewhat confusing to sort out, involves the hardware doing a quick reboot to change modes, and will upset any software making use of it (you'll have to restart it, or rescan its I/O interfaces).  This could simply be whatever DAW you're using, or reconfiguring your OS's audio I/O.

When the switch is released for SPDIF mode, the optical inputs are in SPDIF mode (channels 9 & 10), the electrical input is disabled, but both optical and electrical outputs are SPDIF channels 11 & 12.  The documentation didn't give any info about sampling rates.

And when the switch is depressed at 44.1 or 48 kHz sampling rates, it's in ADAT mode, the optical inputs are ADAT channels 11 to 18, optical outputs are ADAT channels 13 to 20, the electrical inputs are SPDIF channels 9 & 10, and electrical outputs SPDIF channels 11 & 12.

But when the switch is depressed in 88.2 or 96 kHz samping rates, it's in SMUX mode, the optical inputs are SMUX channels 11 to 14, optical outputs are SMUX channels 13 to 16, the electrical are SPDIF channels 9 & 10, and electrical outputs SPDIF channels 11 & 12.

ADAT not supported on some older MacOS versions (10.6).  ADAT/SPDIF mode switching requires restarting the OS's sound system on some older MacOS versions (10.8, 10.7).

It should be noted that electrical SPDIF connections on RCA sockets are meant to use a proper 75 Ω coaxial cable, not ordinary shielded audio cable.  Though shielded audio cable may work fine with short leads, it's against all recommendations.

TOSLINK cables use plastic lightpipes, and have a length limit of around 5 to 10 metres.  Glass fibre optic cable can be used, but it's not the norm.  Longer lengths may work depending on the equipment involved, and the type of cables.  Because they're optical, no ground/hum loops can be created by them, and they're not susceptible to electrical noise.

The digital outputs can be used as a word-clock to other devices (that'll depend on them being able to sync to it).  Otherwise, the UMC1820 will synchronise to the external SPDIF source.  I don't have any ADAT equipment to know how it handles multiple external sources, they probably have to be synchronous with each other.

I've only been able to test inputting electrical SPDIF signals from my Lexicon MX200 reverb unit, and optical SPDIF from a DVD/CD player.  The DVDs were my own recordings, I haven't tried anything that might be copyguarded.  I haven't, yet, tried connecting a Bluray player to it.  These are the only digital audio sources I have to test against.

Audacity only outputs to channels 1 & 2, so I can't test driving the other analogue or digital outputs (they're all on higher channels).

Sampling rates

The device uses 24 bits per sample, at 44.1, 48, 88.2, or 96 kHz sample rates.

The worth of high bitrates and bitdepths, or not

To the human ear, and using ordinary audio equipment, most of the high bitrates (bits per second) are a waste.  Likewise with large bitdepths (how many bits in each sample).  You can't hear any alleged benefit.

But when you are modifying (editing) your data instead of just recording and playing it back, there can be tangible benefits.  You can slow down (re-pitching or re-timing) high bitrate samples (somewhat) without making them sound jangly.  And greater bitdepths should allow for a larger dynamic range, or potentially allow wider gain adjustments without making quantising steps noticeable.

Analogue input and output levels

This section not written yet...  The manual promises a higher maximum output levels than the 404HD, mentions nothing about nominal levels, and I haven't got around to measuring anything, so far.

Thoughts

I've been using this for a few years now, and a few things spring to mind…

I feel that it's better than the 404HD, and not simply because it has more inputs and outputs.

Like it's smaller relative, the 404HD, I find that line/instrument impedance switch behaves in an odd manner.  I don't really have any high-impedance equipment to test with, but when I use the transformer balanced outputs from my organ, I get some strange distortions in high-impedance mode and unpredictable signal levels.  As with the 404HD, it only affects the ¼″ input jack, the XLR is unaffected by it.  It might have been useful to have a low/medium impedance switch for the XLR, it's fixed at 13 kΩ (which I thought a bit high, there are times when you do want a low impedance input).

And the pad switch is similarly painful.  A 20 dB attenuation switch is rather radical change, and I've had times where you either end up with the input gain turned way-down with the pad off (and potential headroom problems), or way-up with the pad in (which makes for a fair amount of unwanted white noise).  My old Yamaha MC802 audio mixer had 10 dB pad switches, which I found quite practical.  And my ancient Astor mixer had multi-step attenuator switches for the signal levels it was expected to be used with in its day (+8 dB or -20 dB line inputs, and -40 dB, -60 dB or -75 dB mike inputs).

In a lot of ways I wished that all the input connectors were on the back panel, rather than having two on the front, but there isn't really the space for them.

It would be nice if the mono button for the monitor output affected everything it outputted, rather than just the input source monitoring.  A mono-monitoring option is a very valuable feature that's overlooked on an awful lot of equipment.

The SPDIF/ADAT optical I/O mode switch operates in a very confusing manner.

There's very little physical difference between a pressed-in or released button, there's no marker on the button to help you tell, and no lights (except for one phantom power indicator nowhere near the phantom power switches).

The manual leaves a lot to be desired, it's very little more than a listing of the knobs and sockets with brief descriptions.  I wouldn't call it an instruction manual.