Yes, if you were using the digital output (if they had them), both CD players would have sounded identical.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Audiophile's Audio Format Showdown
- Thread starter Prodo123
- Start date
- Sort by reaction score
I agree with you guys. Psychological or not, AIFF does sound fuller than ALAC. I use a benchmark dac1 HDR for DAC and a pair of KRK speakers and the sennheiser HD800 as well. You can tell a difference that WAV/AIFF sounds fuller and more dimensional with jazz or classical music. However, all my music library is encoded in ALAC because it only take up half the space of WAV/AIFF. And that minimal difference is something I can live with.
Having a full Mac lineup, I don't use flac often enough to comment.
100% Psychological. Read the thread for the many reasons to support this, and why it can easily be proven that they sound exactly the same.
Wrong, transport specs and motors etc can introduce jitter into the digital signal. Some of the best CD transports are still the old and original Sony-Phillips which are used in some of the top CD Players (Sugden, Moon, Naim etc).
They sound nearly identical, but the compression method used in ALAC makes it unnoticeably different from AIFF. The data contained may be different, but the playback is affected. Only people with extremely careful or trained ears can hear this difference, and it does not detract from the overall quality of the recording.
That contributed to why I re-encoded my library back to ALAC. Now that it's open source, sounds 99.99999999% same and has more tagging options than AIFF, it's more practical, at least at the moment.
"Nearly identical" "Unnoticeably different" "99.99999999% same"
Do you suspect that the implementation is somehow faulty?
Last edited:
Yes, agreed, the data is the same but the device one use to deliver the sound can be affected. Which is why they have those expensive speaker cables, DACs, and power cables that tries to decrease the interference in the hook ups of the devices.
But again, everyone's ears and preferences are different. So something that may sound better to one may sound unnoticeably similar to another. And that other person will tell you they are exactly the same. One can train their ears to listen to detailed sound, but might have to stop listening to extremely heavy bass stuff because they do damage to the ears.
I have provided proof that they sound exactly the same. What have you to show apart from backtracking? Before you were convinced that ALAC was somehow "duller" than AIFF (makes zero sense in the digital realm), now the difference has shrunk to an unnoticeable 0.000000001%. Do the experiment I wrote about a few pages back and see for yourself.
Not faulty, but the design of the codec itself does affect the sound quality, since it takes more time for a computer to decompress an ALAC file and play it than to play the uncompressed AIFF file. It is not faulty in any way; you will almost never hear the difference, since the lag is only a couple microseconds, maybe less."Nearly identical" "Unnoticeably different" "99.99999999% same"
Do you suspect that the implementation is somehow faulty?
No, you have provided proof that all lossless formats have the same data within the file itself. Of course it would provide absolute silence when you invert the phase of the same audio data. Try recording the sound output of the AIFF and ALAC through a studio monitor in a recording studio. When you invert the phase of the recorded samples, you will NOT hear silence, that I can almost guarantee.
0.000000001% for most people. Not me; I can hear the difference. Years of music has changed my ears. Is the differentiation between "You" and "I" really that hard?
You're misunderstanding something. There will be no audio lag due to the decompression of the file. Modern processors can do more than enough operations per time period to write the decompressed data into a buffer way sooner than it's needed.
This proves nothing. There are so many factors to take into account, like the analog speakers and microphones used, the room and the air that's in it.
No, the DIGITAL data proves nothing. Are our ears digital, or are they "analog"? Do we hear music as bits and points that make up a curve, or do we hear pressure differences in the atmosphere? Comparing digital data to sound quality when dealing with lossless formats is like comparing apples to oranges.
This experiment has both the microphone and speaker, as well as their position, room condition, atmosphere, and most other things controlled. They are no longer variables, and will not affect the recording.
You can use any speaker you want. Any set, whether it's a $400 studio monitor or a $50,000 distortion-free set of speakers. The recorded data will show a difference.
As for your comment on how "pointless" it is to convert between lossless formats, there's many different methods by which data and metadata are handled. For example, WavPack's hybrid mode allows for a user to keep a lossy version of a song which can be upgraded to a lossless copy in the presence of the master file. This is an excellent concept, but has failed to gain traction. FLAC does not handle audio tagging the same way as ALAC, etc.
From my understanding, the buffer (or the file, dependent on codec) is read from continuously by the DAC. Otherwise it would create an even greater jitter than the amount reduced from using a buffer, since data is sent in packets instead of a continuous stream. Since it is read from constantly, and the decoded data is written at a slower pace than it is read, a jitter of a couple microseconds or less is inevitable.
Last edited:
What you are now talking about falls outside the scope of this conversation. All the analog elements including the DAC are irrelevant. Other problems all together. This disscution covers the digital decoding performed by the CPU.
No, this discussion is about the sound quality differences between ALAC and AIFF, which can be a byproduct of the digital decoding process required by ALAC. Sound quality is not measured in bits; it is all analog. It falls within the scope of the conversation, and you have been presented with nearly definitive proof that digital data is not an accurate way of measuring sound quality.
Until the signal becomes analog there is no "sound quality". It's all meaningless numbers when it gets to DAC and at this point in the signal chain they are all exactly the same whether they're coming from AIFF, ALAC, FLAC...
Edit :
There's no point in me going any further. This thread has all the input that's needed from me for people who may be reading to come to their own conclusions. See ya!
Last edited:
Which one sounds better is sound quality, is it not?
Again, with your misinformation!
It's all meaningless numbers, but those meaningless numbers that each codec tries to get across are the exact same. What actually comes out of your speakers will be different for certain formats, contrary to their purpose. When the compressed data is uncompressed and reaches the DAC, it will have unnoticeable lags for many people. Uncompressed formats do not have this phenomenon. I don't say problem because it does not affect many people, and it usually is not a problem.
So FLAC, ALAC and especially WavPack (since the audio data is split to 2 files in hybrid mode) will have this phenomenon. Uncompressed WAV and AIFF do not.
It seems an infographic is necessary for any viewpoint to be argued nowadays...
Are you saying that with ALAC the clock is less accurate than with AIFF? If so you're going to have to provide evidence (not just made-up statistics) in order for it to be anything more than polemic.
Have a look at these two links:
http://developer.apple.com/library/mac/documentation/MusicAudio/Conceptual/CoreAudioOverview/ARoadmaptoCommonTasks/ARoadmaptoCommonTasks.html#//apple_ref/doc/uid/TP40003577-CH6-SW1
http://developer.apple.com/library/mac/technotes/tn2097/_index.html#//apple_ref/doc/uid/DTS10003287-CH1-SECTION4
What is delivered to the output unit must be Linear PCM even if you're handling an ALAC file. Modern processors can easily decode ALAC to fill the audio buffer in time. It simply makes no difference whether the data originally came from an uncompressed file or had to be decompressed first.
Last edited:
Has there ever been a controlled double-blind study conducted where someone was able to consistently tell the difference between high quality audio formats? It would seem to me that if there were real differences that someone would be able to demonstrate that they can consistently tell them apart.
If someone has demonstrated this ability, can you point me to the experiment?
EDIT: Never mind. A little googling reveals that this is well traveled ground, with supporters on one side of the issue and supporters on the other side of the issue well dug-in to their positions and never likely to budge.
The comments section of the following thread has a lively and well-informed debate.
http://www.avguide.com/forums/blind-listening-tests-are-flawed-editorial
If someone has demonstrated this ability, can you point me to the experiment?
EDIT: Never mind. A little googling reveals that this is well traveled ground, with supporters on one side of the issue and supporters on the other side of the issue well dug-in to their positions and never likely to budge.
The comments section of the following thread has a lively and well-informed debate.
http://www.avguide.com/forums/blind-listening-tests-are-flawed-editorial
Last edited:
Hydrogen Audio has posted tests of many double blind tests. However, last time I checked..and I admit it has been a while...no one was interested in testing lossless versions against each other. The main rational in the threads I have read is that the data that it sends to the DAC is exactly the same, so there is no reason to test them.
If for some reason your computer can't handle decompressing a file, that would say more about your computer than it would about the two formats.
If the system worked as you think it does then, yes it would have the defects you describe. But it doesn't work like that.
It does not matter how long it takes the computer to decompress the data. In NO CASE is the data ever decompressed synchronously. It is always done it spurts where many samples are processed and then for a few milliseconds no samples are. This is even true for uncompressed audio. the bytes are read off the disk if "chunks". None of this matters because after reading from the disk and decompressing or not the data are then placed in a FIFO buffer and clocked out at a fixed rate. The process that unbuffers the data can't know if the data were ever compressed. Next there is another re-clocking that happens inside the audio interface hardware. The DAC is controlled by a hardware crystal oscillator.
Even a physical CD player uses buffer. The disc motor typically only has a few fixed speeds and almost never would the mechanical speed of the CD match the 44.1K sample rate. So the CD player has a FIFO buffer where the samples are place into the buffer at some odd ball non constant rate but are removed by a crystal controlled hardware clock at a fixed rate.
On a computer typically the data lives in the buffer for maybe 500 sample periods before it goes to the DAC. At can be less, some time just over 100.
Ok an analogy.... The samples are people. They arrive at a subway station in a train where hundreds get off all at once. But at the exit is a turnstile that rotates at exactly one person per second. As you watch people leave you see them come out the door at a fixed rate even if a train is a few seconds late or early it does not matter. In fact S/PDIFF fiber cable works this way. First it delivers all the bits for one speaker then for the next and so on. Each audio channel has it's own "turn style"
Last edited:
As a wanna-be audiophile, I'm excited and pleased that this thread exists. With that said, I do have questions for the brain trust:
1) Will there be an audible difference between AAC and ALAC/FLAC if I'm only playing on my 4th gen iPod Touch with headphones, and occasionally via Airplay to my AppleTV?
2) We're running Windows XP on a Dell Dimension pc. What is the best free program with which to convert files downloaded in FLAC from say HDtracks, Linn Records, iTrax, etc? Because those services very obviously exclude ALAC support, but I want to give their music a try.
3) Is there an iOS app that plays FLAC files AND which can Airplay?
1) Will there be an audible difference between AAC and ALAC/FLAC if I'm only playing on my 4th gen iPod Touch with headphones, and occasionally via Airplay to my AppleTV?
2) We're running Windows XP on a Dell Dimension pc. What is the best free program with which to convert files downloaded in FLAC from say HDtracks, Linn Records, iTrax, etc? Because those services very obviously exclude ALAC support, but I want to give their music a try.
3) Is there an iOS app that plays FLAC files AND which can Airplay?
I prefer FLAC for the simple reason that the Cowon J3 plays it (with my Shure SE535's and custom EQ it sounds amazing), even with the device been highly unstable when connected to my Mac via USB it is worth the hassle!
Fortunately Disk Utility can rescue the Cowon when you accidentally unplug it without ejecting it first.
If you have an iPod or iPhone then AIFF is the best way to go as I have never been able to tell the difference between it and a FLAC file and iTunes seems to work fine around AIFF files, even so far as to having a a conversion tool in iTunes (iTunes > Preferences > Import Settings > AIFF encoder).
Fortunately Disk Utility can rescue the Cowon when you accidentally unplug it without ejecting it first.
If you have an iPod or iPhone then AIFF is the best way to go as I have never been able to tell the difference between it and a FLAC file and iTunes seems to work fine around AIFF files, even so far as to having a a conversion tool in iTunes (iTunes > Preferences > Import Settings > AIFF encoder).