AIFF vs. FLAC file types

0 Members and 1 Guest are viewing this topic. Read 2336 times.

Pundamilia

  • Jr. Member
  • Posts: 141
AIFF vs. FLAC file types
« on: 9 Jan 2018, 04:37 am »
1. Since both are lossless, but FLAC files can be compressed, is the extra workload required of the BDP-2 to decompress the file of any consequence?
2. Would the BDP-2 processor be happier playing back uncompressed files (less workload)?
3. Are the two formats equivalent in terms of their capacity to transport metadata (e.g. Artist information) and can the BDP-2 pick up the metadata as easily from either format?
4. Are the two formats equivalent in SQ?

Bottom line: what format should I be ripping CDs to?

dB Cooper

Re: AIFF vs. FLAC file types
« Reply #1 on: 9 Jan 2018, 05:19 am »
Bottom
line: Use FLAC IMO.


Regards Question #1, FLAC files are compressed (you state "can be"), I can't think of any real 'advantage' to AIFF other than (as you mention) possibly slightly lower processor load, but no modern processor should be seriously challenged by this, witness the popularity of the Pi which has a fraction of the processor muscle of a full fledged PC, but sees wide use as a server/streamer.

2) Bryston would be the definitive source on this, but I doubt it.

3) Yes to the first part, defer to Bryston on the second.

4) Yes IMHO. Some claim to hear differences but I'm skeptical (to put it mildly) that these differences would survive blind testing.

I suggest FLAC.

charmerci

Re: AIFF vs. FLAC file types
« Reply #2 on: 9 Jan 2018, 05:35 am »
I'm fairly sure that guy will come on here and say that he thinks uncompressed WAV files sounds best. Though they take up much more space about twice as much.

Try both and see if you can hear the difference.

zoom25

  • Full Member
  • Posts: 632
Re: AIFF vs. FLAC file types
« Reply #3 on: 9 Jan 2018, 05:58 am »
I remember one data field was coming off incorrectly with AIFF previously with regards to file track number "01/11" vs. "1/11"...the lack of zero threw it off. FLAC showed up correctly. I don't know if there were other things, but FLAC definitely was the best suited for metadata. WAV was the worst with field entries.

As for SQ, you can also try FLAC uncompressed. Then you can have three uncompressed formats to choose from: FLAC uncompressed, AIFF, and WAV. All the same size as well. Pick a few tracks and convert them to as many formats as you want. You can do batch transcoding with XLD. Throw them on a drive and try it out for yourself.

timind

Re: AIFF vs. FLAC file types
« Reply #4 on: 9 Jan 2018, 12:29 pm »
Bottom
line: Use FLAC IMO.


Regards Question #1, FLAC files are compressed (you state "can be"), I can't think of any real 'advantage' to AIFF other than (as you mention) possibly slightly lower processor load, but no modern processor should be seriously challenged by this, witness the popularity of the Pi which has a fraction of the processor muscle of a full fledged PC, but sees wide use as a server/streamer.

2) Bryston would be the definitive source on this, but I doubt it.

3) Yes to the first part, defer to Bryston on the second.

4) Yes IMHO. Some claim to hear differences but I'm skeptical (to put it mildly) that these differences would survive blind testing.

I suggest FLAC.

You give a recommendation for FLAC without any reason, as far as I can read. What tips the scale to FLAC in your opinion?

CanadianMaestro

Re: AIFF vs. FLAC file types
« Reply #5 on: 9 Jan 2018, 01:28 pm »
A sonic preference for one lossless type over another lossless type can be readily explained by neurophysiology. Millions of hair cells are tuned to specific sound frequencies inside our ears. The sensitivity of these cells to becoming excited by their tuned freqs will vary with age: many actually die off with age (and with usage: louder music will kill off cells, leading to hearing loss). A perceptual difference between WAV/FLAC/AIFF is explained by the responses of the remaining hair cells to identical freqs of sound waves. To further complicate things, all cells are plastic -- they respond in a variable manner depending on prior experience (i.e. history of listening). Some hair cells also become stiffer with age or with chemical damage, making them harder to excite at specific freqs. (Taking certain antibiotics for a long period will kill inner ear cells, leading to partial hearing loss).

A few other variables also come into play, such as head position/angle and emotional mood when comparing file types on a system.

In the end, a lot of things can happen before the sound waves reach my brain to allow me to perceive a music file as sounding the "same" or "different".

cheers
« Last Edit: 9 Jan 2018, 03:05 pm by CanadianMaestro »

dB Cooper

Re: AIFF vs. FLAC file types
« Reply #6 on: 9 Jan 2018, 03:27 pm »
You give a recommendation for FLAC without any reason, as far as I can read. What tips the scale to FLAC in your opinion?
I just don't see a compelling case for taking up twice the storage space (and bandwidth, if one is using a NAS setup), so I think a case for uncompressed has to justify that. Since the FLAC file (unlike lossy formats like mp3 and Ogg) decompresses into a bit-perfect replica of the original, I don't think the case is made. HD space is cheap though, so just pick whichever you want.

dB Cooper

Re: AIFF vs. FLAC file types
« Reply #7 on: 9 Jan 2018, 03:58 pm »
A sonic preference for one lossless type over another lossless type can be readily explained by neurophysiology. Millions of hair cells are tuned to specific sound frequencies inside our ears. The sensitivity of these cells to becoming excited by their tuned freqs will vary with age: many actually die off with age (and with usage: louder music will kill off cells, leading to hearing loss). A perceptual difference between WAV/FLAC/AIFF is explained by the responses of the remaining hair cells to identical freqs of sound waves. To further complicate things, all cells are plastic -- they respond in a variable manner depending on prior experience (i.e. history of listening). Some hair cells also become stiffer with age or with chemical damage, making them harder to excite at specific freqs. (Taking certain antibiotics for a long period will kill inner ear cells, leading to partial hearing loss).

A few other variables also come into play, such as head position/angle and emotional mood when comparing file types on a system.

In the end, a lot of things can happen before the sound waves reach my brain to allow me to perceive a music file as sounding the "same" or "different".

cheers

None of the above, as far as I understand your argument, has anything to do with the file type per se- your hair cells, whether healthy or not, do not change characteristics when you are playing a FLAC file (which, to repeat myself, unpacks into a bit-perfect replica of the original file) vs, say, an AIFF or WAV file. Neither do the file characteristics change when the angle of your head (or your state of mind, or attentiveness), change- although any of these things can affect our experience of a musical event, they have nothing to do with the characteristics of the file itself. I might be more relaxed on one occasion and my system 'sounds better' and bothered by something another time- when I was a younger person, my stereo used to sound better after some herbal 'enhancement' even though I know that the sound of my system didn't actually change, just my perception thereof- but those are not inherent SQ issues, which as I understand it is the original topic.

CanadianMaestro

Re: AIFF vs. FLAC file types
« Reply #8 on: 9 Jan 2018, 04:19 pm »
From pundamilla:
4. Are the two formats equivalent in SQ?

dBCooper:
"Inherent SQ" is impossible to define or measure for lossless files. SQ is purely subjective, regardless of file type. If you really meant frequency spectra, than yes, those are measurable, but are not equal to "Inherent SQ".

All the zeros and ones are the same with lossless files -- that fact does not explain WHY PERCEPTION of the files differs between SOME individuals. Brain circuitry can explain that.

In the end, the brain determines auditory perception; it is the final "instrument" regardless of file type. [I don't give a rat's ass about blind A/B comps or whatever. Your brain changes from moment to moment, across all sensory modalities].


cheers

zoom25

  • Full Member
  • Posts: 632
Re: AIFF vs. FLAC file types
« Reply #9 on: 9 Jan 2018, 04:20 pm »
There's a easy way to test this blindly on MPD. Pick one song, have a compressed version and an uncompressed version added say 5 or 10 times each. Then hit the shuffle button. See if you can figure it. Get another person to note down whether it was the FLAC or WAV version.

Before somebody comes in and says for the billionth time "they are all lossless and bit-identical"...just know that this isn't about that but rather the noise generated due to the CPU and/or I/O of compressed vs. uncompressed. (EDIT: CanadianMaestro beat me by a minute).

http://thewelltemperedcomputer.com/KB/WAV-FLAC.htm

Or you can stick with maximum compressed FLAC to save the most space and have perfect metadata, and have it sent as PCM to the BDP-1. Roon does this. I think other DLNA programs also sends PCM?

This conversation has been done many more times on Naim forums. Those guys would also find WAV to sound better against FLAC when the decoding took place in the box, but when they started transcoding to PCM before being sent to their Naim boxes, the differences went away.

With this approach, you get maximum compression (storage space WIN), best metadata of your choosing, and best (same) SQ. You'll never have to worry about file formats again as far as SQ goes.

Krutsch

  • Full Member
  • Posts: 502
Re: AIFF vs. FLAC file types
« Reply #10 on: 9 Jan 2018, 04:25 pm »
You give a recommendation for FLAC without any reason, as far as I can read. What tips the scale to FLAC in your opinion?

Manic Moose and MPD are far better at reading tags (meta data) from FLAC files than from AIFF or WAV. There is no other reason to consider one over the other - they will sound exactly the same.

I've actually measured CPU utilization on a BDP-1 with different codecs, using high-res files. The only file type that really presented a significantly different result was ALAC. Apple Lossless allows for extremely flexible tagging (i.e. tags can appear just about anywhere in the media file), which causes a bump in CPU utilization when decoding and playing back the file. I can't recall the numbers, but it was a large difference. FLAC, WAV and AIFF were very similar, with respect to the CPU during playback.

Pundamilia

  • Jr. Member
  • Posts: 141
Re: AIFF vs. FLAC file types
« Reply #11 on: 9 Jan 2018, 04:36 pm »
Thank you folks for your opinions and sage advice. Subject to my own personal listening tests, it looks like FLAC is the winner.

I have to agree with the comments regarding the subjectivity of *perception* of SQ. It is certainly personal and I suspect (as others have suggested) that is also the function of many other factors (mood, alcohol or other substances consumed, situation, etc.).

I appreciate the discourse and hope that others have learned as much from it as I have. :D

Syrah

  • Full Member
  • Posts: 564
Re: AIFF vs. FLAC file types
« Reply #12 on: 9 Jan 2018, 04:47 pm »
I have a dedicated CAPS server using JRiver and Audiophile Optimizer.  I did blind A/B comparisons.  AIFF beat FLAC on my system to my ears.  Am I going to re-rip my entire CD collection and buy an additional NAS for the extra storage.  Uh, no.  Not anytime soon.  But I am re-ripping the stuff I really like, in AIFF.

My experience.  Potential technical reasons for this - I don't know.

charmerci

Re: AIFF vs. FLAC file types
« Reply #13 on: 9 Jan 2018, 05:28 pm »
Since you have JRiver and if sound is that important - this thread talks about better sounding programs -

http://www.audiocircle.com/index.php?topic=154436.0

though to use them, you have to pay yearly fees.

Syrah

  • Full Member
  • Posts: 564
Re: AIFF vs. FLAC file types
« Reply #14 on: 9 Jan 2018, 06:33 pm »
Cool.  Thanks.  I have been thinking about Roon.  I read a post somewhere that it plays very nicely with old Squeezebox devices too.  So it might be a good whole house solution, in addition to potentially having better sounds, and working well with Tidal, etc.

TJ-Sully

  • Jr. Member
  • Posts: 233
Re: AIFF vs. FLAC file types
« Reply #15 on: 9 Jan 2018, 10:58 pm »
Have you guys tried A/B tests of tracks ripped from dBpoweramp vs. iTunes?
I've ripped about 100 CD's using iTunes....but thinking of getting specific software to complete this task more accurately.
any thought on this?
T

CanadianMaestro

Re: AIFF vs. FLAC file types
« Reply #16 on: 9 Jan 2018, 11:02 pm »
Have you guys tried A/B tests of tracks ripped from dBpoweramp vs. iTunes?
I've ripped about 100 CD's using iTunes....but thinking of getting specific software to complete this task more accurately.
any thought on this?
T

I highly recommend a separate rip software. Peace of mind, mostly. No idea what Apple puts into their software...(if their iPhone stuff is any indication... :o ).

cheers

Edit: You answered your own query: ACCURACY

TJ-Sully

  • Jr. Member
  • Posts: 233
Re: AIFF vs. FLAC file types
« Reply #17 on: 9 Jan 2018, 11:41 pm »
what software do you use, CM?

CanadianMaestro

Re: AIFF vs. FLAC file types
« Reply #18 on: 9 Jan 2018, 11:44 pm »
what software do you use, CM?

Chief, I use XLD because I am a Mac fanboy.  :lol:
But dBPoweramp is very good, I'm told.
Deadly accurate. 

OTM

  • Jr. Member
  • Posts: 74
Re: AIFF vs. FLAC file types
« Reply #19 on: 10 Jan 2018, 03:15 am »
Chief, I use XLD because I am a Mac fanboy.  :lol:
But dBPoweramp is very good, I'm told.
Deadly accurate.

I have been using dBPoweramp for a couple of years now, pleased with results, and easy to use.