A year ago, I've done a quick and not-so-polished test to my two listening setup using lossy and lossless file formats:
Apple TV:
First generation apple tv playing AAC 256k VBR files
CD player (original):
Cambridge Audio 540D DVD player playing the original CD
CD player (decompressed):
Cambridge Audio 540D DVD player playing a CD-burned version of the AAC files
Both setup were routed to:
Analog connection to Cambridge Audio Azur 540A (version 1) amplifier
B&W speakers
My wife and I took turns in operating the switch between various source with the listener indicating which source was preferred. For all types of tracks we listened too, the CD player was systematically preferred. We weren't able to distinguish between the original CD and its decompressed version. I must add that both of us embarked on that test thinking that lossy compression would lose hands down.
Subjectively, we described the sound coming from the CD player as being fuller (much more spatially distributed). We are not audiophiles at all but we both gave a clear preference for the CD player over the apple TV. But at the same time, we are lazy and love the convenience of the Apple TV listening...
My conclusions at that time was this:
- I will see no problem in buying my music online as AAC files.
- All CDs bought will be ripped to AAC 256k in iTunes.
- Apple TV DAC (or line out amplifier?) is inferior to what is inside my CD player, but I can't use my CD player's DAC with the apple TV... So when funds are available I should shop for an external DAC for the apple TV that would hopefully match the performance of the CD player's DAC (I believe the Cambridge Audio DAC magic is the same DAC).
In retrospect, two variables have been ignored in the test: 1) the cables between the playing device and the amplifier were not switched (they are both cheap RCA [red/white] cables); 2) the input used on the amplifier were different. Would it be worth reconnecting the test while controlling for these variables?
I would really like if someone could redirect me to similar or contradictory results? I'm sure similar test have been performed by others but I couldn't locate them using the search feature.
You have to ensure that the levels between devices are matched and that the results were truly double blind. At this point I don't think you've met the burden of proof set out by our Terms of Service (http://www.hydrogenaudio.org/forums/index.php?showtopic=3974) to claim that your CD player sounds better than your first generation Apple TV.
Yes, level-matching is critical in these kinds of comparisons, and while I can't find any actual measurements of the output of the Cambridge or Apple TV, Stereophile did run their full suite of measurements on a first-generation Airport Express (http://www.stereophile.com/digitalprocessors/505apple/index.html), and found its analog output to be 2.3 dB lower than the usual 2V full-scale output of most CD/DVD players.
Whether or not the same holds true for the Apple TV is unknown, but given that they are rather similar devices from an audio standpoint, produced by the same company around the same time frame, I'd say there's a good chance the output of your Apple TV is lower than that of your DVD player, and could certainly account for the differences you describe.
Welcome!
It's really unlikely that two components from different manufacturers would be level matched to the degree necessary to do an accurate comparison. If you did not measure levels with an accurate meter and make adjustments, your test results are not to be trusted. You description of a "fuller" sound from the CD is consistent with it being slightly louder. Loudness can affect preference without being perceived as obviously louder. Personally, I would switch to Apple TV, turn up the volume a click and be happy.
Your results on compressed vs. uncompressed are not surprising to us here at HA. You do not have tin ears. It often takes extreme material ("killer samples") to hear compression artifacts. At these higher bitrates, even that doesn't always work.
Thanks for the info. Very interesting review of the Airport Express.
Greynol: I fully acknowledge that the test I have conducted is not objective at all. Unfortunately, from what I get, doing full ABX test with output level matched would require non-trivial hardware investments. Let me know if there is a relatively easy way to set this up. My question was more to see if someone else had conducted a similar test (basically testing the quality of the DAC in the Apple TV) with a more appropriate methodology. I'd be very interested in those results.
Notat: You are right that the CD player sounds a bit louder (that further diminishes the objectivity of my test!), but even by turning the volume up on the amp I can't obtain the "full" sound I was getting with the CD player. Actually, with the Apple TV I rapidly hit a point where the music sounds "unpleasant" (as in too loud, probably some sort of distortion kicking in), in contrast I have the feeling that I can play the CD player louder without feeling annoyed. In the absence of any measures, I wouldn't put much weight on that though.
http://listening-tests.freetzi.com/ (http://listening-tests.freetzi.com/)
http://listening-tests.freetzi.com/mf-128-1/results.htm (http://listening-tests.freetzi.com/mf-128-1/results.htm)
Given high quality of Apple AAC encoder at 128 kbps it will have a high grade of transparency at 256 kbps.
Given high quality of Apple AAC encoder at 128 kbps it will have a high grade of transparency at 256 kbps.
Good news! :-)
According to another thread (here (https://discussions.apple.com/thread/2359127?start=0&tstart=0)), the DAC in the Apple TV is a Realtek ALC885. And my DVD player uses a Crystal CS4360. Would either qualify as a low quality DAC that likely needs to be replaced, or are they both decent DAC?
I understand that following the DAC, a line amplifier must be present in both component. Is it possible that the difference I perceive comes from this amplifier?
Finally, in order to redo my test in a proper way, is it correct to volume match by reducing the output of the DVD player?
Thanks again for your help.
Would either qualify as a low quality DAC that likely needs to be replaced, or are they both decent DAC?
I don't know, but it's cheap & easy* to build a DAC that's better than human hearing.
And, it's not just the chip... It's the whole circuit & design. Where I work (non audio) we measure the accuracy of the DACs & ADCs in the final circuit, and then every unit we ship gets it's own "correction curve" in firmware/flash to get better performance (better accuracy) than the chip is designed for!
If you're not hearing noise, the quality is probably adequate. Noise can sometimes be an issue with soundcards/soundchips inside computers where audio/video performance is secondary to "computing power", but it's rarely a problem in audio/video equipment. Frequency response and distortion are probably not an issue unless it's a very poor design.
I understand that following the DAC, a line amplifier must be present in both component. Is it possible that the difference I perceive comes from this amplifier?
Yes. If it's not built into the DAC chip, there may be an analog filter and/or "buffer" amplifier. Again, these are cheap & easy to build. (It's not like you are building a high-gain phono preamp or microphone preamp, which is slightly more difficult.)
Unless there's something wrong with your hardware,
your 1st assumption was right.... The DAC and associated analog circuitry should do far-less "damage" than AAC or MP3 encoding. But, I'm not saying you can always
hear the "damage" from lossy encoding.
If you have a (good) digital source, your main concern should be speakers (and having enough clean-power to drive the speakers.) If you change speakers you
will hear a difference! Simply re-positioning your speakers should make a bigger difference than switching between your Apple TV and yoru CD/DVD player! (Again assuming nothing's defective.) And, I'm not really saying
you should be concerned with
your speakers or that there's anything wrong with your speakers... I'm just saying that any change of speakers will make a BIG difference and everything else (on the hardware side) will usually make no difference or a very-small difference.
Finally, in order to redo my test in a proper way, is it correct to volume match by reducing the output of the DVD player?
Sure, if there a volume control on the DVD player. (There may be some issues if the volume is adjusted/controlled
digitally, before hitting the DAC.) You'd need to make some test-tones and use a voltmeter (or some kind of meter) to match the levels. ProperlABX testing is not quite as easy as comparing sound files.
*24-bit DACs cost more than 16-bit DACs, and a stand-alone "audiophile" DAC built in small quantities costs more than the one built into a DVD player. But still, you are unlikely to
hear a difference.