Help finding USB audio interface with digital out controlled by monitor volume knob

[banjobiotic]

Hello. I am setting up a windows 10 PC based home recording system. My PC has USB 3, but no thunderbolt. I am going to be using a pair of KEF LS50W speakers as monitors.

I have read Phil Ward's article in SOS from January 2018 and would like to know if anyone has any tips on a audio interface that would negate the need for something like the TC Electronic BMC-2 DAC/monitor controller. Basically I would need a USB based audio interface with spdif/digital out that could be controlled by the monitor out volume knob on the interface. I read that some of the Universal Audio Apollo interfaces that have digital outputs can do this, but they only have two USB interfaces and neither has a digital output. I do not require lots of inputs, mostly just recording single instruments one at a time.

I am able to find loads of interfaces at sweetwater with digital outputs, but there is no indication of being able to use the monitor volume control with the digital output.

There are more and more active monitors these days with built in DSP, so I would think that this would start to be more common among interfaces.

[Drew Stephenson]

Can I ask why you're so keen to avoid using the analogue output from an interface? You're massively restricting your options so I'm curious what's driving that?

[banjobiotic]

Can I ask why you're so keen to avoid using the analogue output from an interface? You're massively restricting your options so I'm curious what's driving that?

It seems like having the interface convert an analogue signal to digital than back to analogue then the speakers convert the signal back to digital and back to analogue is a lot of conversion that would potentially introduce latency and or sound quality issues. I'm really just going off of Phil Ward's 2018 SOS article on the LS50W.

[Drew Stephenson]

DACs and ADCs are so good these days you'd need to be looking at upwards of half a dozen conversions before it even thought about becoming a problem. I think Hugh Robjohns did some testing a while back and found that fewer than 10 conversions was undetectable, and that was about a decade ago. I may be misremembering the exact numbers but I think it was of that order.
Latency might be more of an issue though, how critical is that for your use?

[Sam Spoons]

Latency of the processors and AD/DA conversions in the speakers should be pretty low, certainly no more than the total latency through my X32 desk which is 1ms, that's AD>mixing engine/eq/dynamics/etc*>DA. FWIW 1 ms latency is equivalent to sitting 12" further from the speakers. I found I could detect 6ms latency when listening on IEMs but I doubt you'd notice it on speakers.

* I suspect adding fx processing will add extra latency but don't know how much.

[banjobiotic]

I guess I am possibly trying to solve a problem that doesn't exist. The article I read got it into my head that it would be best to keep the signal digital as long as possible, but I guess I didn't really think about why. I've got these speakers set up in my jam/hangout space primarily for listening enjoyment. I recently decided that I'd like to mess around with doing some recording in the same space (mostly just vocals and string instruments) and I'm trying to just use what I already have as much as possible. I suppose for multi tracking one would just be wearing headphones and the speakers wouldn't apply. I can potentially see using a midi controller for live jamming, but other than that I guess latency with the speakers is a non issue. I'm new to home recording. Are monitors used for anything besides mixing?

[RichardT]

I have a pair of LS50Ws. You can simply use the remote control app for the speakers, or the volume button on the master speaker, to control the volume and leave the digital signal alone.

If you're worried about accidental speaker overloads (it would take longer to respond to that with an app rather than a knob), you can put a limiter on the master bus.

[Nazard]

You can reduce the digital output level from 0 dBFs on my Lynx Hilo and it acts by reducing the level feeding the D/A converter to attenuate a 'hot' signal, (as the Hilo can accept many different input sources simultaneously), and reduce distortion. However, it is not to be used as a volume control, (the manual says NOT!).

[CS70]

The issues that come to mind with feeding less than a full signal out of the digital outs are that you cannot just truncate the words (i.e. reduce the significant bits) but you need to dither, or do tricks like using a lager word length, scaling up the significant bits, then reduce/dither ; and the fact that if you do that before conversion, you will reduce the amount of signal, while the DA converter noise will stay the same - effectively decreasing the signal/noise ratio. As opposite, if you use an analogue attenuator after conversion, you will reduce both signal and noise to the same degree, so the S/N noise stays the same (at least at the outputs).

What certain interfaces have (for example my Apollo Twin) is "digitally controlled" analogue outputs - meaning that the attenuator is set by a controller chip (fed with data by some software). That means that the attentuation level is very precise and "recallable", and when you move the knob manually, the chip can tell you exactly where the level is.

As for the conversion, you find online quite a few well made tests that show that with decent converters there's no appreciable signal degradation until you go thru many, many generations. It is good to keep the signal digital when you can, simply because you will (most often) avoid introducing noise, crosstalk and other typical analogue problems.. but the speakers are you last ring in the chain anyways, and I would guess they're typically way more noisy than any converter..

[Hugh Robjohns]

I guess I am possibly trying to solve a problem that doesn't exist. The article I read got it into my head that it would be best to keep the signal digital as long as possible, but I guess I didn't really think about why.

Yes, it is quite logical to feed a DSP-based monitor speaker with a digital signal and thus avoid unnecessary additional format conversions. However, the practicalities sometimes outweigh the theoretical benefits.

There certainly are some interfaces that allow the user to assign a rotary control to the digital output. I think the Prism Orpheus does, for example. Alternatively, some DAWs have a dedicated monitor control section which you can direct to any physical output, using the virtual controls on the DAW to adjust the volume -- or co trol them remotely from a MIDI controller, etc.

However, in both of these cases, a computer crash would remove all control while quite possibly also generating a very nasty, very loud noise at the same time.

For that reason, and because I'm old enough to be proper old school, I much prefer an independent analogue monitor controller. yes, it requires an extra D-A/A-D conversion, but that is inaudible -- as Drew says, the tests I did n the mid 90s proved you'd need something like 10 conversions to hear artefacts and converters are even better these days. And although the conversions would also add a small amount of latency too, around 1ms perhaps... the DSP processing latency in the speaker will dwarf that, so its not going to create any more problems than you've already got.

[Aled Hughes]

RME Babyface Pro allows you to control the optical (ADAT or SPDIF) output with the volume control (as does the MADIFace Pro with MADI outputs 1+2).

[RichardT]

I run my LS50Ws directly from the computer using USB without going through an audio interface. I use the controls on the speakers and KEF Control app to control the volume. I’m not sure you need anything else....