Audio interface to outboard through headphone out or S/PDIF with DAC?

[mjw63]

I have an antelope audio interface with no analog line outs, but I want to use one or two pieces of outboard gear with it. I can physically do this - it has an independent second headphone out which could route out through a TRS->TS pair insert cable. It also has an S/PDIF out, which I could put into a DAC to get analog outputs.

My question is, is either of these options preferable to the other? Or I could upgrade the interface to get some line outs, but this seems like overkill if the HP/SPDIF options will do the job.

Background: I have a ZED R16 firewire desk, but am trying to slim down. The antelope interface really sounds incredible, and I'm happy to give up analog summing and mixing (I never use the EQ much since I prefer plugin options). The only thing I'd miss is the ability to use one or two outboard pieces, e.g. a Yamaha SPX unit and a stereo compressor. I'm happy to print these track by track, so doing it off the interface is a nice trick.

I'm going to try detailed experiments where I use the Zed R16 to print a reverb vs doing it via the headphone out on the antelope, but thought I'd check in case anyone does this routinely.

[thecontrolcentre]

I'd use the S/PDIF ... the headphone out is too hot for line level ins.

[mjw63]

Thanks. You can trim the headphone out in the software mixer, so you can get the output level to equal a line level. There is an impedance question I guess, but the answer isn't obvious to me. I'd rather not buy an extra DAC unless I really have to, but I'm very open to it.

What I might try whilst I still have the Zed R16 is an experiment. I can use the Firewire interface to route audio to the desk and use my SPX as an aux. Then I can compare the fidelity of the printed audio (which comes back through the Zed converters) to the antelope approach. If I can't tell the difference by ear it's not worth worrying (except a nagging suspicion that the S/PDIF might be better still!).

[mjw63]

Just checked out your music btw - love it! Blue is amazing.

[thecontrolcentre]

Just checked out your music btw - love it! Blue is amazing.

Thanks

[BertKoor]

You can trim the headphone out in the software mixer, so you can get the output level to equal a line level. There is an impedance question I guess, but the answer isn't obvious to me.

Impedance is a real issue when a high-impedance output is connected to a low-impedance input. Here you have the reverse, and that is not really a problem. Rule of thumb with connections not designed to draw all of the produced power (like with speakers) is that the input impedance should be at minimum 8 or 10x the output impedance of what you're connecting it with.

I used to think that headphone outputs should be used for headphones only, but in practice it's OK (or better said: not completely wrong) to connect them to line-level inputs. With small consumer gear like mp3 players it's often the only physical output you have, and the users guides say it's ok. So I've done the calculations, and the voltage is in the right ballpark! You got to watch the level indeed.

There is one rather big downside: noise. By trimming down the volume you're not using all of the available dynamic range and thus you get nearer to the noise floor. And the circuitry of the headphone amp might not be totally transparent. It might have a bit of harmonic distortion or emphasise some frequency ranges. You just don't know until you measure.

What I might try whilst I still have the Zed R16 is an experiment. I can use the Firewire interface to route audio to the desk and use my SPX as an aux. Then I can compare the fidelity of the printed audio (which comes back through the Zed converters) to the antelope approach. If I can't tell the difference by ear it's not worth worrying (except a nagging suspicion that the S/PDIF might be better still!).

Again: you just don't know until you measure. I would personally not go by ear alone. Objective measurements are what you want, not lunatic audiophile wordings like "warm, depth" describing what they hear as if they are tasting wine. And there are free tools to do so: RMAA - RightMark Audio Analyser https://audio.rightmark.org/products/rmaa.shtml

It sends test tones through the signal path you want to measure (eg audio interface outputs back to its inputs) and it will do just that: measure all aspects. Comparing results is easy. Flaws will stick out as a sore thumb. You can export the test tones to WAV, write them to CD and benchmark your cd players DAC. Or in your specific case: send the signal through the headphone outputs going into a line input, and record that with the best audio interface that has analog inputs you have available.

Good luck!

NB: I'm afraid a good DAC outperforms the headphone output. Simple S/PDif DACs don't need to be very expensive. Watch out for snake oil...

[mjw63]

That's amazing - thank you for the clear and excellent advice! I can bypass the SPX processing, so think it is straightforward - route the test tone through the system with the bypass on and see what comes back. I don't know when I'll get time to do this, but it's high on the list and I'll report back.

Do you have any opinion on a reasonable DAC? I've seen this recommended elsewhere:

https://addictedtoaudio.com.au/products ... -converter

That seems crazy cheap though.

The Antelope Zen has amazing converters. The sound quality is outstanding, and it would be great if this works out - I could have a permanent hardware FX send on it and retain all the benefits of the rest of the device. I want to get rid of the desk, and use that space to finally have my keyboard and couple of pieces of rack and tracking gear tidily arranged

[BertKoor]

Do you have any opinion on a reasonable DAC? I've seen this recommended elsewhere:

https://addictedtoaudio.com.au/products ... -converter

That seems crazy cheap though.

$35 is cheap indeed, but I did see them even cheaper like €20.

There are some tech specs quoted, and I'll comment on them.

It's a pity all my RMAA results are on a dead WinXP PC, otherwise I could compare these figures with various consumer CD / DVD / MP3 players I owned.

Power supply DC 5V
Size 61.6mm×49.1mm×21mm
Weight 50g

These figures are not so relevant I think.

Output amplitude 1.6V（0dB Fs）

This is a tiny little bit below the nominal level of +4 dBu (1.736 V) used in professional gear. Consumer gear has a nominal level of -10 dBV (0.447 V) so it's fairly 'hot' for that. Note that the difference between +4 dBu and -10 dBV is not 14 dB but closer to 12 dB. If you connect it to a mixer or something else that has a proper input level trim, then this is no issue at all. If you have to resort to lowering the master fader of your DAW or a volume setting in the mixer software of your audio interface, then you will lose about 2 bits of resolution.

Frequency Response 20Hz~20KHz

As expected. It's a pity they don't quote how near to 0dB the frequency response is. Would be nice if they had given any indication how flat it is, something alike +/- 0.1 dB or +/- 3 dB.

Signal to Noise Ratio ≥90dB

So this is effectively a 16-bit DAC, comparable to what you find in consumer gear like cd players. Not bad, but not stellar either.

Total harmonic distortion <0.0110mW

Ouch, this needs some calculations in order to make sense of. It's usually given as a percentage where above 1% is obviously audible and below 0.1% is where pretty good starts. If the peak voltage of 1.6 V has an impedance of 600 Ohm load (the reference for line level) then that's 4.3mW. So it's about 0.25% and that's the grey area of neither very bad or very good.

Crosstalk 75dB

That's not so good I'm afraid. Not that you can hear it on regular stereo material, but enough to clearly bleed through when one channel is silent and the other is not.

[mjw63]

Awesome, thanks. I've resolved to get it because it is super cheap, just to test that my routing idea works ok in principle. Then I can upgrade.

I'll be connecting it to the spx 990 and a stereo compressor, both of which have input trims. In fact the spx has a 20db pad on the input too. So might I be able to run the headphone amp fairly hot then slive 20db off at the input stage of the spx?

[BertKoor]

In fact the spx has a 20db pad on the input too. So might I be able to run the headphone amp fairly hot then slive 20db off at the input stage of the spx?

I think so, yes.

Good luck!