AIFF vs. FLAC file types

0 Members and 1 Guest are viewing this topic. Read 7498 times.

skunark

  • Full Member
  • Posts: 1434
Re: AIFF vs. FLAC file types
« Reply #40 on: 24 Feb 2018, 01:00 am »
What's interesting about Roon is i've had some drop offs and playback speed issues when the source is Tidal.   Just pausing the song clears up the issue, but makes me feel that Roon isn't optimized and not nearly as mature as others media players.   

Jim

skunark

  • Full Member
  • Posts: 1434
Re: AIFF vs. FLAC file types
« Reply #41 on: 24 Feb 2018, 01:03 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?

Flac as it has a built-in file checksum that can be useful for detecting disk-rot and other problems.   Also if you rip with AcccuRIP/Exact Audio Copy then you know the rip is good and will always have a checksum to compare against.    Then you can batch convert from FLAC to any file format you need.

Krutsch

  • Full Member
  • Posts: 568
Re: AIFF vs. FLAC file types
« Reply #42 on: 24 Feb 2018, 01:56 pm »
On linux, you have to go through the ALSA drivers and everything needs to be converted to PCM.  So MPD/Roon will call a decoder for lossless uncompression for  flac (or ALAC, and a lossy uncompression for mp3, aac, etc). 

...

People hear what they hear and think what they think, but having a beer will be more impacting to the sound than the file format.

Just for clarification: Roon does any decoding on the Core device (your server or laptop or NAS box) and sends across as PCM wrapped within their proprietary protocol (RAAT). The BDP doesn't have to do this work as an endpoint.

As an FYI, I've loaded different encodings of the same audio track on my BDP-1 and monitored CPU utilization during playback (i.e. ssh into the box and look at the mpd process). WAV, AIFF and FLAC were pretty similar w/ 192/24 files, with FLAC requiring slightly more CPU, but mpd was still near single-digit % of CPU utilization, If I recall correctly.

The outlier was ALAC, which used a LOT more CPU for the same track (30+%? ). ALAC is more costly to decode because it allows meta data to be placed arbitrarily within the file. But, yes, in the end, they all produce the same PCM data. On a BDP-3, with its greater processing power, I would expect the differences to be minimal.

Finally, as I've stated before, FLAC is the best choice only because MPD and Manic Moose are far better at reading the meta data - I've had mixed results with both AIFF and WAV in this regard. If you think decompressing the FLAC file on playback is ruining your music, then you simply use FLAC without compression.

The beer part is probably true, but for me the audio enhancement tool is scotch.

Krutsch

  • Full Member
  • Posts: 568
Re: AIFF vs. FLAC file types
« Reply #43 on: 24 Feb 2018, 02:01 pm »
What's interesting about Roon is i've had some drop offs and playback speed issues when the source is Tidal.   Just pausing the song clears up the issue, but makes me feel that Roon isn't optimized and not nearly as mature as others media players.   

Jim

IMO, the speed of your Roon Core matters a lot more than Roon would like to admit. I used to run Roon Server on a Mac Mini (2012 model) and I would occasionally experience what you describe. When I moved to a MacBook Retina (quad i7 w/ 16 GB RAM, SSD), wired to the same switch as my BDP-1, all of that went away for me...

As always, YMMV.

R. Daneel

  • Full Member
  • Posts: 1086
Re: AIFF vs. FLAC file types
« Reply #44 on: 24 Feb 2018, 02:10 pm »
James,

Yet there are some who use the BDP (a great product that I highly endorse), and still claim to hear diffs between file types. With all else constant (cables, external drives, etc).

What gives?

cheers

Hi mate!

Like I said, providing (and only if) everything is fine with the encoder and decoder, there cannot be any differences. It is simply not possible because the bit checksum must be identical even if bit allocation is different. It is either this or the encoding process is not truly lossless. There is no third.

The bigger problem I see here is that threads like these are full of people mentioning type of ripping software and encoder they're using when I know for a fact they're flawed. As much as it grieves me to say it, there are far less ripping engines that do work than there are those that don't work. I always recommend Exact Audio Copy (EAC) in Secure Mode which is a bit complicated to set up but ultimately very worthwhile. Depending on condition of the CD and CD-ROM's capabilities, ripping a single disc may take anywhere from 15 minutes to an hour and a half. If you want to rip your CDs properly, this is what it takes.

I've even seen people comparing two rips of the same CD claiming there is a difference when in fact one of the two was ripped with the channels reversed! No doubt, the CD-ROM used was to blame since some of the drives based on LG platform have this particular "feature".

Perhaps that an extreme example but is speaks volumes about the necessary preparation these people have taken (or rather, didn't take) before they started comparing.

The point is, people will compare anything not knowing what they're really doing or not even having enough patience to do a bit of reading before they start doing it.

Some ripping engines and encoders have the ability to change the overall SPL of the rip not to mention all sorts of filtering that can be applied, and usually is applied by default unless you make a change in the settings menu. Yet, nobody is talking about that, only that they can "hear a difference".

I have a few discs ripped in WAV that sound terrible, simply terrible. The original disc from which they were ripped sounds much better. I also have a number of files ripped to 320kbps MP3 that sound very close to the original CD and generally, much better than the improperly-ripped WAV files.

There are some really very expensive HDD/network players out there. Naim in particular makes a couple of them and these include CD rippers. Naim claims their engine is better than anything else on the market (don't they all?) but that's not really true. It can only be as good as something like EAC, not better, because EAC already incorporates everything. Naim has success with these devices PRECISELY because it is easier to spend money than to spend time! Of course Naim is better! It's better than most of the stuff out there. It just isn't better than something good, something that requires a bit of effort.

Wow, this has somehow turned into a long post!

Cheers!
Antun

maty

Re: AIFF vs. FLAC file types
« Reply #45 on: 24 Feb 2018, 02:34 pm »
2018 or 2008 year?

The musical information is the same because they are lossless but...

The problem is the compression. More decompression work -> more CPU -> more jitter!

Yes, the CPU generates jitter than we can not eliminate but yes minimize.

In my system, I can differentiate between FLAC 8 and FLAC 0. And more easy between 16/44, 16/96, 24/96 too.

With AIFF, WAV or FLAC 0 not.

By the way, it is important to update FLAC 1.3.2

And to optimize the S.O. to play multimedia. The soft players too. And...

Krutsch

  • Full Member
  • Posts: 568
Re: AIFF vs. FLAC file types
« Reply #46 on: 24 Feb 2018, 02:52 pm »
2018 or 2008 year?

...

The problem is the compression. More decompression work -> more CPU -> more jitter!

...

Nope. Jitter is not created or added-to in the digital processing domain (i.e. while decompressing a file).

maty

Re: AIFF vs. FLAC file types
« Reply #47 on: 24 Feb 2018, 02:57 pm »
The CPU is not an ideal machine.

And if we forced the CPU to work with avx2 instructions with sound files -> less jitter -> better sound.

Of course, with good records and system with low noise. Bad records and systems with high noise mask the jitter. That is why many are unable to appreciate the difference and not because they are deaf and others golden ears.

One of my sayings in audio: noise (of all kinds) is our enemy.

maty

Re: AIFF vs. FLAC file types
« Reply #48 on: 24 Feb 2018, 03:27 pm »

skunark

  • Full Member
  • Posts: 1434
Re: AIFF vs. FLAC file types
« Reply #49 on: 24 Feb 2018, 05:14 pm »
IMO, the speed of your Roon Core matters a lot more than Roon would like to admit. I used to run Roon Server on a Mac Mini (2012 model) and I would occasionally experience what you describe. When I moved to a MacBook Retina (quad i7 w/ 16 GB RAM, SSD), wired to the same switch as my BDP-1, all of that went away for me...

As always, YMMV.

I think this makes Roon a poor choice, it's using .net/mono so it's not at all optimized for linux or mac, maybe optimized for windows.   So its bloatware to begin with, but it does simplify the client.    The RAAT protocol can't seem to synchronize multiple devices very well either. Airplay had this down easily, but then its forces 24/48 playback so the task is easier.   

I've been finding the Roon iOS app buggy and the Tidal integration is very poor.

Jim

skunark

  • Full Member
  • Posts: 1434
Re: AIFF vs. FLAC file types
« Reply #50 on: 24 Feb 2018, 05:33 pm »
2018 or 2008 year?

The musical information is the same because they are lossless but...

The problem is the compression. More decompression work -> more CPU -> more jitter!

Yes, the CPU generates jitter than we can not eliminate but yes minimize.

In my system, I can differentiate between FLAC 8 and FLAC 0. And more easy between 16/44, 16/96, 24/96 too.

With AIFF, WAV or FLAC 0 not.

By the way, it is important to update FLAC 1.3.2

And to optimize the S.O. to play multimedia. The soft players too. And...

CPUs fetch 64 bytes at time, more than enough decode and keep a sound card buffer full on a dedicated music player.    Jitter doesn't get introduced until the PCM single-bit stream anyway.


skunark

  • Full Member
  • Posts: 1434
Re: AIFF vs. FLAC file types
« Reply #51 on: 24 Feb 2018, 05:36 pm »
The CPU is not an ideal machine.

And if we forced the CPU to work with avx2 instructions with sound files -> less jitter -> better sound.

Of course, with good records and system with low noise. Bad records and systems with high noise mask the jitter. That is why many are unable to appreciate the difference and not because they are deaf and others golden ears.

One of my sayings in audio: noise (of all kinds) is our enemy.

vector instructions just helps efficiency, you won't see that in a BDP or RPI for a very very long time.

maty

Re: AIFF vs. FLAC file types
« Reply #52 on: 24 Feb 2018, 05:41 pm »
I have been for months, almost a year, doing it that way and the sound is much better, more pleasant, with more bass, more believable, more exciting.

maty

Re: AIFF vs. FLAC file types
« Reply #53 on: 24 Feb 2018, 06:10 pm »
Just now with a new 24/192 FLAC vinyl rip (TELARC). The music fills the room. From good sound -> better sound +  EMOTION.

Boito*, Verdi*, Robert Shaw, The Atlanta Symphony Orchestra* And Chorus*, John Cheek, The Young Singers Of Callanwolde, The Morehouse-Spelman Chorus ‎– Prologue To Mefistofele / Te Deum (1980)

https://www.discogs.com/Boito-Verdi-Robert-ShawAtlanta-Symphony-Orchestra-And-Chorus-John-CheekYoung-Singers-Of-CallanwoldeM/release/4534535

---------------------------------------------------------------------------------------------
 DR      Peak      RMS      Filename
----------------------------------------------------------------------------------------------

 DR14      -1.05 dB    -22.10 dB    A Prologue To Mefistofele - Composed By - Arrigo Boito - I Prelude And Chrous, II Instrumental Scherzo And Dramatic Intermezzo. III Vocal Scherzo.aif
 DR13      -0.47 dB    -19.95 dB    B1 Prologue To Mefistofele - Composed By - Arrigo Boito - IV Final Psalmody.aif
 DR15      -0.91 dB    -22.12 dB    B2 Te Deum (No. 4 Of Quattro Pezzi Sacri) - Composed By - Giuseppe Verdi.aif
----------------------------------------------------------------------------------------------

 Number of files:   3
 Official DR value:   DR14

==========================================

* avx2 also with multimedia players and not only sound files.

Krutsch

  • Full Member
  • Posts: 568
Re: AIFF vs. FLAC file types
« Reply #54 on: 24 Feb 2018, 09:32 pm »
...

I've been finding the Roon iOS app buggy and the Tidal integration is very poor.


What? Where do you think Tidal integration is done better?

skunark

  • Full Member
  • Posts: 1434
Re: AIFF vs. FLAC file types
« Reply #55 on: 25 Feb 2018, 12:19 am »
From Roon, you can't update or modify a playlist but for some reason you can delete it?
From Tidal, when you update or modify a playlist those changes don't appear on Roon after syncing.   Seems like you can only duplicate the playlist on the desktop app and add to it before you sync it with Roon.   

Rather basic integration features missing from two premium products.   

As in better, well beets does a good job for regular expression playlists and can sync across mpd, plex and other services and even users. 



R. Daneel

  • Full Member
  • Posts: 1086
Re: AIFF vs. FLAC file types
« Reply #56 on: 25 Feb 2018, 09:37 am »
2018 or 2008 year?

The musical information is the same because they are lossless but...

The problem is the compression. More decompression work -> more CPU -> more jitter!

Yes, the CPU generates jitter than we can not eliminate but yes minimize.

In my system, I can differentiate between FLAC 8 and FLAC 0. And more easy between 16/44, 16/96, 24/96 too.

With AIFF, WAV or FLAC 0 not.

By the way, it is important to update FLAC 1.3.2

And to optimize the S.O. to play multimedia. The soft players too. And...

Jitter is generated in clocking mechanisms and input receiver chips. Internet is full of those "who just know". Don't buy into their stories.

maty

Re: AIFF vs. FLAC file types
« Reply #57 on: 25 Feb 2018, 10:46 am »
Software digital processing do not generate jitter.

But the CPU hardware yes. In modern Intel/AMD CPU you can minimize the jitter generation with avx2 instructions. Maybe noise and RF/EMI interferences too.

I also doubted in his day but I decided to do the test and in my system the difference is very much noticeable.

And the improvement I notice is consistent with what others say when they reduce jitter. The key word for me is EMOTION.

http://thewelltemperedcomputer.com/KB/BitPerfectJitter.htm

At the end of page:

Quote
The sound of jitter

In general I find jitter to cause a loss in "inner detail" which usually relates to a "flatter" sound. It looses the "liveness", it becomes boring. Sometimes I find jitter also effects bass significantly. This is strange because I would expect high frequencies to be more susceptible, but I frequently hear a significant improvement in bass articulation when decreasing jitter.

Its hard to listen to a piece of gear and say "thats got high jitter" just by listening, because many other things can cause similar sonic effects. By building my own gear I've been able to do quite a few tests where I can hold everything equal except change jitter and can definately hear major improvements in sound by lowering jitter.

But frequently similar changes can be had (for example) by upgrading the power supply of the preamp.

Another issue is all jitter is not the same. The spectrum has a lot to do with it. I have one receiver with 200 ps of jitter that sounds significantly better than another with 50ps, BUT the spectrum is radically different between them. Thus just picking the lowest published "jitter number" will not guarantee the best sound.

John Swenson
« Last Edit: 25 Feb 2018, 12:20 pm by maty »

maty

Re: AIFF vs. FLAC file types
« Reply #58 on: 25 Feb 2018, 12:14 pm »
Software jitter, Paul McGowan

http://www.psaudio.com/pauls-posts/software-jitter/

Article and comments. Jitter, noise, EMC interferences, power...

And I am sorry by the long off topic.

maty

Re: AIFF vs. FLAC file types
« Reply #59 on: 26 Feb 2018, 07:58 pm »
Again, with a new 24/192 FLAC vinyl rip. The sound is much better now and not only better.

Nicolo Paganini* - Salvatore Accardo, Charles Dutoit, The London Philharmonic Orchestra ‎– Violin Concerto No.2 La Campanella (1977)

https://www.discogs.com/Nicolo-Paganini-Salvatore-Accardo-Charles-DutoitLondon-Philharmonic-Orchestra-Violin-Concerto-No2-La/release/7738883


----------------------------------------------------------------------------------------------
 DR      Peak      RMS      Filename
----------------------------------------------------------------------------------------------

 DR15      -0.68 dB    -21.89 dB    A1 Konzert Fuer Violine Und Orchester Nr. 2 H-Moll Op. 7 'La Campanella' - 1. Allegro Maestoso.aif
 DR15      -0.64 dB    -22.32 dB    A2 Konzert Fuer Violine Und Orchester Nr. 2 H-Moll Op. 7 'La Campanella' - 2. Adagio - 3. Rondo.aif
 DR18      -0.84 dB    -26.68 dB    B1 La Primavera Andante Sostenuto - Larghetto - Variazione - Allegro Moderato - Tema - Variazioni I-IV - Finale.aif
 DR19      -1.03 dB    -26.36 dB    B2 Introduzione E Variazioni Sul Tema 'Non Piu' Mesta' Da 'La Cenerentola' Di Rossini Introduzione - Adagio Cantabile - Tema. Moderato - Variazioni I-IV - Finale. Allegro - Tempo Del Tema - Piu' Presto.aif
----------------------------------------------------------------------------------------------

 Number of files:   4
 Official DR value:   DR17

=======================


- The End -