Apple TV versus Airport Express - which sounds best?

[Charlie Jefferson]

I'm taking delivery of an ATV tomorrow and will be able to do my own comparison, but what are the opinions of those of you with both devices.

Both my ATV and AE will be linked optically to my DAC, so will they produce identical sounds when fed the same lossless files from my Mac or are there differences between the two devices when relaying audio?

 

[Charlie Jefferson]

Btw, I'm buying the ATV mainly for it's video and photo streaming capabilities but presumed it will be at least the equal of the AE for audio too.

 

[Paul.]

In theory, the express should be better as it leaves the audio at 44k unmolested, whereas the atv2 upsamples to 48k. I am unable notice any difference tho.

 

[Anonymous]

The interpolation is very good but it's still interpolated. Get the Airport Express if you want bit-perfect (at full volume - all other levels will interpolate).

The only problem with full volume with (16bit) digital is that the DAC can (does on modern pop) clip quite audibly as the signal overloads the DAC outputs, often sounding like a loose connection. So sometimes better quality is found at lower volume levels where the digital signal is still clipped but now fails to overload the DAC. So most of the time you may be interpolating anyway.

This effect can be particularly bad on pro-audio DACs that are not designed to handle the clipping normal for all domestic applications. Of more importance than (good) interpolation is a) clipping and b) DAC up-sampling, linearity, filtering followed by c) jitter and d) adequate analog circuitry after the DAC output.

Effectively interpolation in the digital and/or analog domain happens quite a bit before the 16bit/44.1kHz signal gets to your amp, the 16/44 is a compressed audio format and needs careful handling to turn back into real analog music.

 

[altruistic.lemon]

Hey, mate, I have them both, too. Can't tell the difference between them, but as they're going into the same DAC much as expected.

Globsy, I've always run both at max vol on the mac and used the amp to adjust. Never heard any clicks, pops, drop outs, degradation at all on the ATV. Maybe the interpolation thing - and I doubt doubt you're 100% spot on - is to small to be measurable, or fixed by some error correction.

 

[Anonymous]

I too am interested in this issue. I have been using an AE optical out to an external DAC to stream iTunes and Spotify (via Airfoil) to my preamp. I just got an ATV 2, mostly for Netflix and sports. That seemed to make my AE redundant, as the ATV also has an optical audio output. So I took my AE out of the loop. I thought I heard a slight degradation of audio quality with the ATV 2 via Spotify (which replaces iTunes for music already stored on the drive) but am not at all convinced that this is anything more than a palcebo effect. It's impossible to do a real A/B test. So my question is: Is there any technical difference between the AE optical output (going to a DAC) and the ATV 2 optical output also going to the DAC? Has anyone out there with Golden Ears detected a difference?

Thanks, GH

 

[John Duncan]

I too am interested in this issue. I have been using an AE optical out to an external DAC to stream iTunes and Spotify (via Airfoil) to my preamp. I just got an ATV 2, mostly for Netflix and sports. That seemed to make my AE redundant, as the ATV also has an optical audio output. So I took my AE out of the loop. I thought I heard a slight degradation of audio quality with the ATV 2 via Spotify (which replaces iTunes for music already stored on the drive) but am not at all convinced that this is anything more than a palcebo effect. It's impossible to do a real A/B test. So my question is: Is there any technical difference between the AE optical output (going to a DAC) and the ATV 2 optical output also going to the DAC? Has anyone out there with Golden Ears detected a difference?

Thanks, GH

There is a technical difference, yes - the ATV2 upsamples to 16/48, compared to the AEX which leaves it unadulterated at 16/44. I'd be surprised, however, if it makes a blind bit of difference to the sound quality. One of these days I'll finally get an ATV2 (I had the original which was brilliant) and find out

 

[Anonymous]

I have ATV2 and AE both, mainly for listening to CDs on my setups. So 16bits/44.1K is the only thing I care.

This is something I tested.

Test environment:

MacMini ->wireless-> AE(or ATV2) ->optical-> DAC -> Integrated Amp -> Speakers

1) Don't use Apple lossless (ALAC) format, AIFF is a better choice than WAV. Here is the chart I got (I was using an optical capturing to compare with the original file)

Format; MacMiniOpticalOut; AEOpticalOut

ALAC; Filtered; Filtered (differently from MacMini)

WAV; Bit-accurate; Sometime filtered, sometime not filtered with extra repeat samples (Timing skew?)

AIFF; Bit-accurate; Bit-accurate

2) ATV2 upsampling to 48K, make the quality worse.

What I noticed so far is, any changes (filter or duplicate samples) on the digital side make the sound worse. Even the changes is relatively small, but still noticable.