Sound Quality iPad + external audio interface + Macbook

KB1234

Hi there,

Does it make a difference in term of sound quality if i hook up my iPad to the input of my external audio interface and out the signal to my mac and record inside the Daw VS goes directly from my iPad usb to usb macbook with the Inter device audio protocol and record inside the daw.

Is the external audio interface will improve sound vs the Inter device audio when recording the app sound ( like any music app ) through the Daw ?

Thanks

Telefunky

If you hook up the iPad via protocol, it transfers the plain unchanged sound data to the Mac.

In the other case the iPads‘s output is influenced by it‘s output stage, the input stage of the interface, the interfaces clock and converter quality.
(input impedance and (if present) transformer balancing may have an influence, too)

Long story short: it‘s very unlikely, that the cable path will „improve“ the signal, but sometimes this may happen. That‘s why I mentioned (quality) transformers.
They are rarely used in interfaces (expensive), but their presence is often perceived positive.
Without transformers... almost impossible (imho).

uncledave

Any A/D conversion adds noise to an audio signal. When the audio is from a continuous source (mic, instrument pickup), the sampled result should be a pretty good approximation of the source. But when the source is already digital (digital recording, computer synth), converting it to analog and back to digital adds noise, basically because the samplers are not synchronized. The D/A output will be slightly smoothed, and the A/D samples that signal at unsynchronized times, so the recorded signal will not match the original. That difference is noise that you will carry through any further processing. It is always preferable to keep digital sources in the digital domain, where they can remain "perfect", until they are reproduced as sound.

KB1234

Hi Telefunky,

Thank you very much to take the time to answer my question.

So in this situation regarding what you just said and considering my interface is an Rme babyface fs, will you suggest to use the interface instead of directly plug iPad to mac through the protocol ?

Ps : Usb cable from Rme to Mac is an UDG gold plate for info.

Regards

KB1234

Hi Uncledave,

Thank you for your answer, crystal clear. I learned something.

All the best..

Telefunky

The Babyface is known as a very(!) clean interface... it should exactly reproduce what‘s fed to it. Of course it will add noise, because that‘s the nature of things, but this will be hard to detect (if at all).
The unknown part of the equation is the performance of the iPad‘s output stage when it‘s „loaded“ by the interface. In case the output circuit is optimized for a low impedance inductive load (ear/headphones), then there may be differences.
It might not sound identical, but that‘s no more than a possibility.

I have one of these „inductive load“ cases which took me hours to solve and I still have no clue what‘s going on exactly (in technical details).
It‘s an old BOSS ME-5 pedal, that delivers a tremendous sound via headphones, but using the regular line out (as supposed for recordings), it‘s rather pale in performance... nothing to write home about. I tracked it with every interface, different converters, same result with both line out and headphone out into interface.

As a last resort I built a pair of fake headphones from 2 quality transformers (input impedance about 20 Ohm) and that did the trick. Recording result exactly as monitored.
The strange thing is that both line and headphone output are identical, feeding the same opamps in the exact same circuit design.

KB1234

Thank you Telefunky for your well détail answer.

I appreciate the respond of the community.

Have a great day.

BroCoast

@Telefunky said:
The Babyface is known as a very(!) clean interface... it should exactly reproduce what‘s fed to it. Of course it will add noise, because that‘s the nature of things, but this will be hard to detect (if at all).
The unknown part of the equation is the performance of the iPad‘s output stage when it‘s „loaded“ by the interface. In case the output circuit is optimized for a low impedance inductive load (ear/headphones), then there may be differences.
It might not sound identical, but that‘s no more than a possibility.

I have one of these „inductive load“ cases which took me hours to solve and I still have no clue what‘s going on exactly (in technical details).
It‘s an old BOSS ME-5 pedal, that delivers a tremendous sound via headphones, but using the regular line out (as supposed for recordings), it‘s rather pale in performance... nothing to write home about. I tracked it with every interface, different converters, same result with both line out and headphone out into interface.

As a last resort I built a pair of fake headphones from 2 quality transformers (input impedance about 20 Ohm) and that did the trick. Recording result exactly as monitored.
The strange thing is that both line and headphone output are identical, feeding the same opamps in the exact same circuit design.

That’s kind of weird.

I have no such issues with the outputs of my ME-5.

Telefunky

It‘s in fact weird, I‘ve asked myself a hundred times if what I heard was true or fiction.

I got the pedal for 50bucks for it’s 5 position midi switcher, with the bonus of a bunch of original BOSS analog pedals on a single circuit board, considering the only digital part (delay/reverb) a flaw initially... until I auditioned it.
This mid/late 80s technology simply created a strange (and very special) space and soon became my favorite feature of the ME-5.

Capturing that spacial impression in a recording was the problem - there may be (almost ?) no difference if you use the ME-5 as a straight mono guitar input.
(f.e. with compression, distortion, eq in front of guitar amp)

That‘s my Telecaster, d‘Addario ECG flats straight into the reverb of the ME-5, no other section of the pedal engaged, no post processing.

Btw the reverb is implemented in a gate array (similar to the more famous Lexichip), and part of probably all the early BOSS RV pedals, but these use a simple resistor network as their DA conversion stage, while the ME-5 has a dedicated DA converter.

BroCoast

@Telefunky said:
It‘s in fact weird, I‘ve asked myself a hundred times if what I heard was true or fiction.

I got the pedal for 50bucks for it’s 5 position midi switcher, with the bonus of a bunch of original BOSS analog pedals on a single circuit board, considering the only digital part (delay/reverb) a flaw initially... until I auditioned it.
This mid/late 80s technology simply created a strange (and very special) space and soon became my favorite feature of the ME-5.

Capturing that spacial impression in a recording was the problem - there may be (almost ?) no difference if you use the ME-5 as a straight mono guitar input.
(f.e. with compression, distortion, eq in front of guitar amp)

That‘s my Telecaster, d‘Addario ECG flats straight into the reverb of the ME-5, no other section of the pedal engaged, no post processing.

Btw the reverb is implemented in a gate array (similar to the more famous Lexichip), and part of probably all the early BOSS RV pedals, but these use a simple resistor network as their DA conversion stage, while the ME-5 has a dedicated DA converter.

Lovely flatwound tone.

It’s such a good unit. I also paid next to nothing for all my units.

I had all my patch switching automated by Ableton at one point but I am yet to set that up on iOS.

What I really like is the ability to dial in almost any distortion type with the comp>drive>eq combinations.

Another sleeper I think is the Digitech RP20. Can be had for $50-$100 and has a tube preamp running at proper voltages. Very powerful midi implementation too.

Telefunky

Yes, a lot of „old“ gear is much underrated
Back to @KB1234 ‘s Babyface... it‘s technically state of the art and a solid base for a(ny)mixing setup.
Choose whatever is more convenient from either cable or software channel, there will be no dramatic difference in quality, if any at all.
We‘re often fooled by signal level: if a source happens to be just 0.5 louder in monitoring, it‘s almost guaranteed that it will be perceived as „better“, but few people are able to detect that it‘s just loudness level.

I fooled myself with the BIAS amp sim for a long time, even considering an earlier release had a different sound engine... while it was just 1.5dB of virtual preamp gain that made the older version sound better.