Airport Express/Streaming Audibility Tests

Thanks chaps, a nice range of opinions there.

This first test was intended to set the baseline for future experimenting, so with that in mind:

A - CD rip. Nothing done to it other than trimming of the track length.

B - Same file as 'A' played on iTunes and sent wireless from a Macbook Pro to the Airport Express. Analogue output of the AE then captured via ADC and recorded digitally onto another Macbook Pro.

So, ADC aside, you have compared the standard analogue output of the AE with a straight rip.

Why is this stage important?
Well it gives you some idea of what output from £80 worth of combined psu/router/wireless/dac/analogue stage in something the size of a cigarette pack actually sounds like when played blind.

My opinion (carrying a health warning as I arranged the test) was 2A in both instances.
Truth be told that surprised me as I'd expected it so sound pretty bad.

The next stage is capturing the digital out though I take on board SQs reservations.
I can only see that reservation relating to jitter so it doesn't itself make such a test pointless IMO. I say that because there is surely more to good digital than jitter performance, and on the AE it measures comparably to a decent CD transport if tests to date are correct. Then of course we have the question of many modern dacs re-clocking the stream, so does it then matter if a recording process does the same?

Lots of question and so its open to the floor for suggestion on how to proceed?

 

I thought Lush was too noisy to be a suitable test track and I thought New was so full of synth sounds it was almost impossible to pick out what you normally look for. Neither of these tracks would ever get played in my house.

Well recorded rock or acoustic next time Rob, a Chesky will do.

 

JC,

Technically better? Not sure what you mean by that, but I can't say that either track really appealed to me. (My tastes are rather eclectic — Beethoven and Bach one day, Bauhaus and Beatles the next.)

I tend to prefer music with a strong rhythmic component, so the more bouncy New clip should have appealed more so than Lush, but neither had pressed my buttons to be honest.

In any case, regardless of my musical preferences, I listened to both blind and preferred the CD rip to the analogue out on the AEX through a ADC.

I think this has settled whether I'd want to use the audio out on the AEX. It's fine for casual listening, but when I'm in full-Spock mode using my finely tuned Vulcan ears it will not be quite good enough.

Joe

 

If I have any certainly.

But it should still be possible to get a good feel on the tracks used IMO.
I hear technically far worse frequently from TTs and those get compared without too many issues.

What's the general view - reapeat with different music or move on?

 

I think when the digital input is recorded to the computer, it won't show jitter at that stage, as the computer will just wait for the data and write it in to the next waiting block in the file.

However, jitter can still appear from the stage where the DAC receives the input from the Airport Express, and also from the ADC sampling it back to digital. This is still twice as many instances for jitter to become audible as were you playing the DAC directly in to your hi-fi and listening.

As such I think it is a fair enough test to play the Airport Express digi out to a DAC (non-reclocking) and then record that analog signal back to digital with an ADC. If a difference can be reliably detected, then perhaps more thinking is required, but if not it's safe to assume it won't be audible in a normal system when listening.

 

Rob,

Your test answered whether an AEX is for me, but I'm happy to try again with different cuts.

What is interesting is that despite being hardcore audio nerds we aren't unanimous, which proves to me that we have different levels of system resolution and hearing acuity, so what may be unacceptable for you may not be for me and vice versa.

It also puts some of the heated exchanges into perspective -- why some get really animated when talking about seemingly small differences and others just shrug and say meh.

Joe

 

Joe, were you planning to use it via the analogue output or into a dac?

Very true, though sometimes supposed differences are just so miniscule (or entirely absent) that hearing ability in advance of that available to any human would be required (Vulcan perhaps ). In this case and using the analogue output we can see the difference exists (from the file exam in a file editor) and everyone heard 'something' different with these files.

These things only ever act to steer thought and opinion - they very seldom give difinitive answers. This test proved quite educational for me, as I espected posts such as 'Uh thats bloomin horrible' or 'sounds like its been through a mangle' but they didn't materialise. For me that is very useful as I'm deeply sceptical of much accepted internet wisdom when it comes to audio.
The only way to climb up and out from this sludge it to challenge, challenge challenge and carry on testing. Truth be told I find much in audiophile land to be utterly bonkers, and the odd blind test proves quite cathartic

Anyway enought of that - if you are thinking of buying an AE and want to hear what a modestly priced dac brings to the AE I can repeat these tests for you but via a Cambridge Dacmagic.

 

Rob,

The audio out. I don't have a stand-alone DAC.

Probably just as well that I don't drop a few bucks on an AEX, as cheap as they are, because I just spent a pile on a new computer (my old G5 died) and I'm due for a new cartridge, the funds for which got raided to buy a new computer.

Joe

 

Rob,
Using poor quality, high noise tracks or completely synthesised music will not give you representative results of the limits of peoples ability to discern between certain aspects of the the replay chain.

Frankly you couldn't be queering the pitch any more if you tried.

 

If we stuck to the same tracks as those in the AAC test it might help people to become accustom to what they are listening for when looking for differences. I thought the Rage track and 'Even After All' were good ones to use.

 

Haven't read through the whole thread yet, so apologies if someone's already posted on this, but I thought I'd share my experience with my Audio Synthesis DAX Decade (a high-end DAC from over a decade ago, so excellent quality but no longer state-of-the-art tech).

Prologue: if you plug a decent CD transport into the DAX, it will lock onto the incoming signal with some sort of fancy crystal lock (called X-lock). If the signal has high jitter, X-lock won't be possible (though the DAX will still produce sound, of a lower quality).

I've got my AirPort Express plugged into the DAC, fed over wifi from my Mac Mini. At first I tried it straight from iTunes. This caused the DAX Decade to acquire X-lock initially, but then drop it a few seconds later, reacquire it, drop it, reacquire it, etc. Each time it would produce a popping sound, so this was unacceptable.

So I had the idea to pump it through AirFoil instead. I don't know exactly what sort of modification AirFoil does to the signal, but I think it's supposed to be bit-perfect. However, it presumably does some sort of mixing or processing on it. This is borne out by my findings - namely that the signal from the AE to the DAX when fed by AirFoil never even attempts to acquire X-lock! It just sits at "regular" lock perpetually. This is great for me, as it means I don't get the popping sounds. Sound quality is presumably reduced, but as I mainly just use the AE for background music, this isn't really an issue - as soon as I want to do some proper listening, I stick on a CD (or a record! )

But this demonstrates two things: 1) the AE on its own generates more jitter than a decent CD transport, but perhaps not lot more; and 2) AirFoil does something to the signal that increases jitter significantly.

Thoughts?

Dunc

 

Read my first post in this thread.. I think it's on page 1.

AirFoil seems to produce a wondering sample frequency between 44.0kHz and 44.2kHz. As such it could be seen as jitter, but the frequency of the variation is so slow I don't think it produces an audible issue, it tends to take about 1minute to very from 44.1kHz to 44.2kHz. It is a strange effect though, and I'd like to know why it does it.

 

I don't agree that they are poor quality.

Or rather, they are easily good enough to reveal differences.
You surely can't be arguing that we have to use some Chesky audiophile disc to show differences between equipment.

 

Only issue with using stuff that listeners might own (or can download easily) is that they have access to the WAV and can view it in software. Therefore, if one of the test files is a CD rip it can be identified very quickly. The other file is then automatically seen as inferior, at least potentially.

I've a stack of demo and cover discs here so perhaps if i presented say 10 short MP3 samples we could agree on some references.

I know that sounds quite cynical, but it happens.

I've some very boring work to do this weekend so will need need lots of breaks. I'll see what can be found. In the meantime, consider where to go next with the testing. Going nowhere is an option if you've all had enough

 

I agree it's unacceptable, but it's a fault of the dac, not of the Aex. The Aex stops the digital stream when there is no information to transmit, and then resumes when necessary. This caused the dac to drop lock and then resume when the stream re-appears. This results in a pop when lock is achieved. Other dacs mute their output until lock is achieved so you don't get pops. I bought a very expensive Apogee dac which had this fault and returned it for a refund.

Airfoil transmits continuously but slowly alters the stream to keep it at the right rate. It doesn't stop and start the stream like the Aex does and therefor your dac doesn't pop.

I don't think it's anything to do with jitter, and a good dac will cope with jitter anyway without any audible effects.

JC.

 

Hi JC,

Appreciate the input, but you misunderstand - the x-lock pops on and off while streaming music, constantly. Not just when you start playback. Hence, completely unacceptable.

Dunc

 

Ok dunk, but it's the same problem. The newer Aex's adjust their rate very quickly and the dac thinks it has stopped and started again, hence the pops. Some dacs do seem to do this and others can track the signal without problems.

I would change the dac, (as I did with the apogee) but the older (g not n) Aex's don't switch so rapidly and the dacs can track the signal more easily, so one of them might suit you better. It still wont cure the pops on actual start and stop though. The only answer to that is a dac that mutes until lock happens.

JC.

 

p.s. It's just occurred to me that you could try a New Apple Tv, they're only £9 more money than an Aex. They do the same job as an Aex as far as audio is concerned but will receive video as well if you wish. But I haven't tried one myself yet.

Apple give a 14 day return if not satisfied facility.

JC.

 

Back to the test of New A and B.

I'm not sure this test was a test of the Aex. It could just as easily have been a test of the adc used to produce the digital file from the Aex's analogue output.

JC.

 

Why are they adjusting their rates while playing?

Paul