Headphone output vs. Master output sound quality

[junebug]

I am assuming that somehow the "Master Output" from a synth would be a higher quality audio signal than the "Headphone Output" from the same synth?

For individual recording purposes I mean. Not just for playing.

All of my Synths "Master Outputs" are plugged into my mixer.
I do not want to record the synths from the output of the mixer - it adds more noise than I am happy with.

If they are typically equal quality audio signals, that would make it a lot easier to just grab a recording from a synth's headphone output jack with a Y-cable, without having to unplug/re-route my Synth's Master Output cables every time.

Any thoughts on the master vs. headphone output?

[BertKoor]

Master or line out is what you need indeed when going to a mixer or whatever. The headphone output has an extra amplification stage, in theory this makes it a bit more noisier.

Some gear I have has headphone output only and the manual states it can be used as line outputs as well. I always was a bit skeptical about it, but after doing the math: it is possible.
Common problem though is that headphone outputs are louder than line level needs to be. So the signal/noise ratio goes slightly into favour of the noise.

If you care about quality, you do some tests and measurements. Each synth is different.

Maybe you'd be interested in either a better mixer or a patch bay. Also it's possible you can improve the way you use the mixer. Feed it a high level signal, trim back if necessary, tap off for recording as soon as possible. Either with a Direct output or a separate group or even an Aux Send can be used for that.

[junebug]

Thanks for the reply, Bert!

[shmuelyosef]

I agree on the 'better mixer or patch bay'. I'm a keyboard player, and have for years used a small format Mackie (1402VLZ) for keyboard patching and effects routing before going into a patch board that feeds my converter rack (Focusrite).

The flexibility lets me get a lot of work done faster, as I never handle cables except occasionally to add something outboard through the patchbay...all my main gear is always online.

[junebug]

Thanks for the reply, shmuelyosef !

The thing is, i'm using a brand new Soundcraft Signature MTK 22 mixer that I just got a few months ago.

and yet -

It's sends more noise than I am comfortable with out through the USB port and waaay too much noise than I want to record with from *any* of the analog outputs - even with every volume and gain turned to zero and nothing plugged in.

When I plug a synth or an instrument directly into, say - a focusrite scarlett 2i2, there is very little noisefloor in comparison...

~sigh....

[deastman]

It has been my experience that headphone outputs on synths are often much more noisy than main outputs.

[BertKoor]

i'm using a brand new Soundcraft Signature MTK 22 mixer that I just got a few months ago. and yet - It's sends more noise than I am comfortable with out through the USB port and waaay too much noise than I want to record with from *any* of the analog outputs - even with every volume and gain turned to zero and nothing plugged in.

On paper it looks alright though:

Noise - Mono Mic Input E.I.N. (max gain): <-124dBu (150Ω source)

So it could be a level mismatch or gain staging problem. Let's talk a bit about Level mismatch on analog connections.

It's possible the mixer sends out at +4dBv "pro" line level and your audio interface expects -10dBu "consumer" level. So the mixer is about 10dB too loud (also account for difference between dBv and dBu) or in other words: the noise floor goes up by 10dB. If you combine that with the possibility the mixer has 25dB headroom but your interface allows only 15dB headroom, then another 10dB of S/N ratio goes down the drain.

Still, you'd have a signal/noise ratio of at least 100dB, which for most purposes is fine. Analog tape had only 70dB S/N ratio. But: if you listen to that silent noise with headphones cranked up, it seems quite loud indeed. But as soon as you play full-blast audio on the same level, you'll rip off the cans very quickly. So this is not a realistic way of asserting the noise level.


What I'd advice you to do is go through a calibration session. Take a test tone (1000 Hz sine for example) of a known level: usually -18dBfs which is 18 dB below digital clipping level. Most audio editors can generate that for you. Use that audio editor to send it out of your interface. It's a rule of thumb that this level is near 0dB VU, so you get 18dB headroom for peaks. Usually that's plenty.

First you need to assert that the levels you're reading on the inputs of the audio interface are correct. There are all sorts of volume controls on these things, on the outputs & inputs, they all should be set to gain factor zero. So feed the output direct to the input (watch out for feedback loops: disable input monitoring on the interface) and assert it comes in at -18dBfs as well. If not, adjust levels until it does.

With the levels on the interface sorted out, you can insert the mixer: output of the interface into line inputs of the mixer, then an AUX or Group output goes to input of the interface. Make sure your test signal shows up as 0dB VU on the metering of the mixer. Also monitor the Aux or Group level: it should be the same 0dB VU. Then check that what comes into your interface again is -18dBfs, and if not: adjust it and note down the level difference for future reference.

Chances are that what comes into your interface is too loud. As explained before: this can be due to the difference between consumer and pro line levels. If this happens: too bad. You now know what level to send out for the interface to be happy.

With the levels now set correctly: record some silence. Measure the noise level with the analysis tools of your audio level. Don't just look at the peak level: that's 10dB louder than the averaged RMS level of the noise. Or do look at the peak level and know it's 10dB louder than what all papers will say. Now I'm curious what the noise level is you get if you follow this procedure. Please report it back.


The mixer also has a digital connection over USB. If you use this you won't have the difference between pro and consumer line levels. But still you need to go through calibration, so you know what level is zero dB on the VU meter of the mixer.

Feed it the test signal, make sure it shows as 0 dB on the master. Then watch what the level is you get into the computer. Now you know the rest is headroom.
Again record some silence and analyse it's level. Please report it back.

[junebug]

Thanks for the help, Bert!

I am confused. lol

My audio editor is Sound Forge on PC.

I generated a 1000 hz sine wav at -18db (not sure about the "fs" part?), but I can't figure out how to get Soundforge to play it, and also monitor the input level.