IPB

Welcome Guest ( Log In | Register )

5 Pages V   1 2 3 > »   
Reply to this topicStart new topic
Why’s FLAC ‘the’ lossless codec? Is it future-proof(able)? Comparisons, [TOS #6: was “Future proof lossless audio/codec” in General Audio]
eahm
post Jun 23 2012, 18:03
Post #1





Group: Members
Posts: 1063
Joined: 11-February 12
Member No.: 97076



I am trying to understand why is FLAC standardizing in the lossless world? Reading here: http://en.wikipedia.org/wiki/Comparison_of...chnical_details isn't True Audio more future proof? All these channels and huge range, isn't it the "best" option for the next "100 years"?

If one day everyone will start using FLAC, will there be a way to expand the number of channels available? For expample the "new new" 12.1 Surround System or whatever they will come up with.

Just some thoughts, don't want to get anyone upset, I use FLAC and I love it.

Thanks.

This post has been edited by eahm: Jun 23 2012, 18:03


--------------------
/lwAsIimz
Go to the top of the page
+Quote Post
saratoga
post Jun 23 2012, 19:04
Post #2





Group: Members
Posts: 4923
Joined: 2-September 02
Member No.: 3264



If people wanted more channels they'd just update the format. I don't think there is much interest in even multichannel music though.
Go to the top of the page
+Quote Post
Debby747
post Jun 23 2012, 19:26
Post #3





Group: Members
Posts: 9
Joined: 28-January 10
Member No.: 77624



I use and prefer WavPack for its features.

The reason Flac is still the most widely spread lossless compression format
is simply the fact that it is supported by many hardware devices.
I don't know why it is but it is, and it is most of the time the only compatible lossless format (except WAV).

So when you can't play TAKs or TAs or WavPacks on your stereo Hi-Fi or mobile player,
it's hard to justify their use. Not even all software media-libraries fully support those...

Greetings
D.
Go to the top of the page
+Quote Post
lvqcl
post Jun 23 2012, 19:51
Post #4





Group: Developer
Posts: 3362
Joined: 2-December 07
Member No.: 49183



Why care about future? FLAC is lossless and DRM-free so it can be converted (easily!) to any format of choice if necessary.
Go to the top of the page
+Quote Post
felix26591
post Jun 23 2012, 20:39
Post #5





Group: Members
Posts: 24
Joined: 17-May 09
From: Panama
Member No.: 69900



I don't know about the channels, but flac is in constant development. FLAC has become standard, one because it's open-source, two it's future proof, you will always be able to find a FLAC decoder and all future FLAC decoders are backwards compatible.

FLAC is excellent, it's versatile, it has very good compression and it encodes very quickly. You can convert it into anything without losing quality and this is the reason everybody is moving to FLAC. It was understandable before that people used mp3, and mp4, the sizes of their hard drives didn't allow for a lossless copy. I can carry my CD collection in lossless in a backup drive and have a high fidelity lossy copy of ogg vorbis or AAC in my phone or in my computer.

Also with FLAC you only have to rip once and you have all the codecs you need, if a new one comes along that you want to convert it too, you can without transcoding lossy to lossy.

Hope that helps.
Go to the top of the page
+Quote Post
Kohlrabi
post Jun 23 2012, 20:43
Post #6





Group: Super Moderator
Posts: 1024
Joined: 12-March 05
From: Kiel, Germany
Member No.: 20561



QUOTE (felix26591 @ Jun 23 2012, 21:39) *
I don't know about the channels, but flac is in constant development.

No, it's not, at least not the official reference implementation. From what I heard the lack of development is quite problematic, for example regarding large file support (>2GB, iirc), among some other missing features/bugs.

This post has been edited by Kohlrabi: Jun 23 2012, 20:47


--------------------
Ceterum censeo Masterdiskem esse delendam.
Go to the top of the page
+Quote Post
eahm
post Jun 23 2012, 20:48
Post #7





Group: Members
Posts: 1063
Joined: 11-February 12
Member No.: 97076



QUOTE (Debby747 @ Jun 23 2012, 11:26) *
I use and prefer WavPack for its features.

I strongly believe a lossless codec should be ONLY lossless, no other forms/situations.
In my opinion True Audio literally kills the competition.


QUOTE (felix26591 @ Jun 23 2012, 12:39) *
I don't know about the channels, but flac is in constant development. FLAC has become standard, one because it's open-source, two it's future proof, you will always be able to find a FLAC decoder and all future FLAC decoders are backwards compatible.

FLAC is excellent, it's versatile, it has very good compression and it encodes very quickly. You can convert it into anything without losing quality and this is the reason everybody is moving to FLAC. It was understandable before that people used mp3, and mp4, the sizes of their hard drives didn't allow for a lossless copy. I can carry my CD collection in lossless in a backup drive and have a high fidelity lossy copy of ogg vorbis or AAC in my phone or in my computer.

Also with FLAC you only have to rip once and you have all the codecs you need, if a new one comes along that you want to convert it too, you can without transcoding lossy to lossy.

Hope that helps.

Yes, I do the same as well and you didn't get my question. What you are trying to annoyingly explain is easily doable with other lossless codecs, my question was why FLAC since mathematically speaking there are better lossless codecs.

This post has been edited by eahm: Jun 23 2012, 20:51


--------------------
/lwAsIimz
Go to the top of the page
+Quote Post
skamp
post Jun 23 2012, 21:07
Post #8





Group: Developer
Posts: 1430
Joined: 4-May 04
From: France
Member No.: 13875



QUOTE (eahm @ Jun 23 2012, 21:48) *
In my opinion True Audio literally kills the competition.


Technically speaking, TAK rules all, IMO. It's (much) faster and compresses (significantly) better than pretty much anything else that's remotely comparable. But being closed source with very little support and no hope for progress on those fronts, it has no chance whatsoever of becoming a "standard" (a codec/format being very different from, say, music playback software, e.g. foobar2000).


--------------------
See my profile for measurements, tools and recommendations.
Go to the top of the page
+Quote Post
lvqcl
post Jun 23 2012, 21:08
Post #9





Group: Developer
Posts: 3362
Joined: 2-December 07
Member No.: 49183



IMO a codec that supports 4GHz samplerate is not better than a codec that support 'only' 655kHz.
Also, FLAC is very fast to decode.
Go to the top of the page
+Quote Post
db1989
post Jun 23 2012, 21:28
Post #10





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



QUOTE (eahm @ Jun 23 2012, 20:48) *
QUOTE (Debby747 @ Jun 23 2012, 11:26) *
I use and prefer WavPack for its features.
I strongly believe a lossless codec should be ONLY lossless, no other forms/situations.

Should we assume you’re referring to hybrid mode? Debby747 didn’t name any specific features, including that. Or do you just mean that compression time and so forth do not matter, as long as maximal compression ratio and the ability to support hugely inflated sampling rates, bit-depths, etc. are in place?

Either way, it’s probably good that you aren’t in a place to arbitrate over what codecs can and cannot do! Being lossless is of course the main factor, but why would you proscribe other options and benefits? That you don’t use them isn’t sufficient justification.

QUOTE
Yes, I do the same as well and you didn't get my question. What you are trying to annoyingly explain is easily doable with other lossless codecs, my question was why FLAC since mathematically speaking there are better lossless codecs.

What you are trying to annoyingly ask(!) is easily answered, since the (quasi-)mass-market niche for a lossless codec is incredibly small and was first entered by FLAC, making it the victor by default. Of course, this might change with the advent of ALAC, but it still seems that very few people care about lossless compression anyway. And those who do are probably just going to settle for whichever codec has the most support – which usually isn’t much in relative terms – anyway.

And again, the ability to support contrived parameters or combinations thereof is not the best metric of a codec’s performance. Look at features that are (much) more relevant in real use cases, or the level of active development. The latter is something with which FLAC could be more legitimately criticised, for instance.
Go to the top of the page
+Quote Post
Porcus
post Jun 23 2012, 21:50
Post #11





Group: Members
Posts: 1842
Joined: 30-November 06
Member No.: 38207



If I were dictator and could order specifications for The Next Universal Lossless Audio Codec, I can think of a quite a few improvements, but let me rather offer my $2e–02 on the 'why FLAC is standardizing' (which I am not sure will be true for forever, given the power of The New Microsoft with the fruity logo and name).

First: there need not be any 'good' reason why one product dominates over another market-wise. And especially not here, where compatibility is more or less everything as soon as features are 'good enough'. If N equally good codecs share the market evenly, that is most likely an unstable equilibrium. Whichever gets the upper hand (even if just by coincidence) will have better chances of growing quicker. And when concerns 'good enough' ... that TTA can sample at 4 GHz is certainly nothing compared to compatibility.

Some time ago, I tried to start a thread asking why these formats came about – which problems did they set out to solve?
http://www.hydrogenaudio.org/forums/index....showtopic=79491 Didn't really find the answers.


I would suppose that the main reason for FLAC's success, is being at the right place doing the right thing at the right time. And I can only guess that parts of this 'right thing', were
- optimized for decoding speed. Ten years ago, I think that was fairly important (and maybe even today for portable devices): it didn't require much CPU load. (AFAIK, no codec improved over FLAC on both compression and speed simultaneously until TAK, am I right?)
- licensing. Free, could be used in any device.
- quite future-proof feature-wise, at least for quite a few years' horizon. (Compare to Shorten and WAV ... and their tagging ...)
- good enough compression ratio. I mean, most don't even use the –8 switch.
And the right place, would be under the wings of Xiph, who had gained a reputation after doing Vorbis at the time when manufacturers were worried about licensing costs for MP3.


Compare to the competition at (approximately) that time: WavPack (1998) and Monkey's (2000?) decoded slower, and not to mention the closed-sourcers LA (2002) and OptimFrog (2003). LPAC was also closed source, and not to mention WMA Lossless, which was release at a time where Microsoft was getting knocked big time for their attempts at locking-in, and at the same time everything DRM-enabled was getting the middle finger big time. Meridian Lossless has some market though, on physical media, but has never target the 'decode yourself' market.



I have to say though, that I don't really understand why WavPack didn't catch on, being so early available. Though I don't know what features it had back in last century.

This post has been edited by Porcus: Jun 23 2012, 21:56


--------------------
One day in the Year of the Fox came a time remembered well
Go to the top of the page
+Quote Post
2012
post Jun 23 2012, 21:55
Post #12





Group: Members
Posts: 68
Joined: 7-February 12
Member No.: 96993



QUOTE (Kohlrabi @ Jun 23 2012, 20:43) *
QUOTE (felix26591 @ Jun 23 2012, 21:39) *
I don't know about the channels, but flac is in constant development.

No, it's not, at least not the official reference implementation. From what I heard the lack of development is quite problematic, for example regarding large file support (>2GB, iirc), among some other missing features/bugs.


Wrong link, the right one wink.gif
http://git.xiph.org/?p=flac.git;a=summary

Also:
http://git.xiph.org/?p=flac.git&a=sear...mit&s=large
Go to the top of the page
+Quote Post
db1989
post Jun 23 2012, 22:26
Post #13





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



QUOTE (Porcus @ Jun 23 2012, 21:50) *
First: there need not be any 'good' reason why one product dominates over another market-wise. And especially not here, where compatibility is more or less everything as soon as features are 'good enough'. If N equally good codecs share the market evenly, that is most likely an unstable equilibrium. Whichever gets the upper hand (even if just by coincidence) will have better chances of growing quicker. And when concerns 'good enough' ... that TTA can sample at 4 GHz is certainly nothing compared to compatibility.
[…]
I would suppose that the main reason for FLAC's success, is being at the right place doing the right thing at the right time. And I can only guess that parts of this 'right thing', were
[…]

This was pretty much exactly what I meant in the second part of my above post, which was probably too brief/vague, so thanks for substantiating. smile.gif
Go to the top of the page
+Quote Post
greynol
post Jun 23 2012, 22:49
Post #14





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



QUOTE (Porcus @ Jun 23 2012, 13:50) *
(AFAIK, no codec improved over FLAC on both compression and speed simultaneously until TAK, am I right?)

If you mean encoding speed vs. compression, then no, not really; though this is a moving target.

MAC is pretty swift at its lowest compression which is generally stronger than flac at its most aggressive settings. WavPack also fares better than flac by virtue of it being stronger at compression as well, though not as fast as MAC for compression levels where it can compete. TAK generally does a faster job than MAC at encoding at the weaker end of where they share common compression levels. Once TAK is configured to encode aggressively, its speed drops substantially, more so than MAC the last time I looked at data, but things may have changed since then. WavPack's encoding speed drops even more sharply.

When you take it into perspective, however, we're only talking about a 2% improvement in compression over flac at -6 at which point flac begins to lose efficiency.

This post has been edited by greynol: Jun 23 2012, 23:43


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
IgorC
post Jun 23 2012, 23:22
Post #15





Group: Members
Posts: 1556
Joined: 3-January 05
From: ARG/RUS
Member No.: 18803



QUOTE (skamp @ Jun 23 2012, 17:07) *
Technically speaking, TAK rules all, IMO. It's (much) faster and compresses (significantly) better than pretty much anything else that's remotely comparable.

Is 3% compression gain significantly better? ( TAK vs FLAC)
Less than 3% for profiles with the same decoding speed.
An encoding speed isn't issue with today hardware. I don't see anybody talking about speed issues and FLAC. Not even closer.

People prefer TAK less and less
2012 ripping/encoding general poll
2011 ripping/encoding general poll

LA is better than FLAC by laughable 5.3%. Big unbelievable deal?
http://www.synthetic-soul.co.uk/comparison...sion&Desc=0

This post has been edited by IgorC: Jun 23 2012, 23:24
Go to the top of the page
+Quote Post
Porcus
post Jun 23 2012, 23:28
Post #16





Group: Members
Posts: 1842
Joined: 30-November 06
Member No.: 38207



QUOTE (greynol @ Jun 23 2012, 23:49) *
QUOTE (Porcus @ Jun 23 2012, 13:50) *
(AFAIK, no codec improved over FLAC on both compression and speed simultaneously until TAK, am I right?)

If you mean encoding speed vs. compression well it's a moving target, so no, not really.


Decoding speed and compression. And by 'both [...] simultaneously', I meant offering not both higher decoding rate in one setting and better compression in another, but a setting where your files both (I) smaller and (II) decode faster. A Pareto improvement, that is: one that is better for any trade-off over those two parameters.

As an example (now I dug to find one that isn't TAK): in Josef Pohm's test, http://synthetic-soul.co.uk/comparison/jos...Rate&Desc=1 , there is – apart from YALAC née TAK – one and only one way to improve over one and only one FLAC setting: MP4 ALS - Garf -a -o4 Pareto-improves over flac -3.


And of course there is the test at flac.sourceforge.net, which is hardly neutral ground, but shows performance on slow hardware (... how fast a CPU does a portable player have these days?)


--------------------
One day in the Year of the Fox came a time remembered well
Go to the top of the page
+Quote Post
greynol
post Jun 23 2012, 23:49
Post #17





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



MAC at it's weakest setting compresses better than flac at its strongest setting, so there is no way to make a comparison. The same is true for other compressors such as TrueAudio which itself is very fast.

Furthermore, flac may still decode faster than TAK at any setting. If this hasn't changed then there is no arbitrary line to be drawn for any comparison.

This post has been edited by greynol: Jun 23 2012, 23:51


--------------------
I should publish a list of forum idiots.
Go to the top of the page
+Quote Post
yourlord
post Jun 24 2012, 01:01
Post #18





Group: Members
Posts: 210
Joined: 1-March 11
Member No.: 88621



FLAC was early on the scene. Performance is good enough. Royalty free. BSD licensed libraries.
FREE is part of the name.

Go to the top of the page
+Quote Post
Takla
post Jun 24 2012, 01:13
Post #19





Group: Members
Posts: 169
Joined: 14-November 09
Member No.: 74931



Some other factors which imo made/make for widespread adoption:

FLAC is not only open source but is available with simple installers and/or really easy to follow instructions for lots of different OS. This is hugely important for less technical users and of course for distributors and vendors. Anyone who wants to offer FLAC to consumers/purchasers doesn't want to be spending much time with support queries.

Another important factor is easy and flexible metadata handling. FLAC is as easy to use as any codec can be, and so is metaflac with and its simple, adaptable key pair tag format. If you want to integrate a lossless codec into rippers, library managers, taggers, players, scripts, devices etc. then free, easy and convenient is a pretty good combination.

Looking back at early large capacity portable/personal players it's hard to see what other lossless format would have been considered by manufacturers who were significant at the time, such as Rio or iRiver. Portable lossless playback capability was probably never important for more than a tiny minority but it was a feature that always gained very positive reviews and generated interest.

I also find it very hard to imagine record labels being keen to pay patent licensing fees and maybe even per CPU or per server software licenses when there is a high quality, gratis, liberally licensed equivalent that their customers know about and want to use.

This post has been edited by Takla: Jun 24 2012, 01:13
Go to the top of the page
+Quote Post
eahm
post Jun 24 2012, 01:54
Post #20





Group: Members
Posts: 1063
Joined: 11-February 12
Member No.: 97076



I am not talking about speed, compression, hardware support, software support, just features. Can everything in the audio world (DTS, Dolby, Studio production, DJ/Mixing, Sharing) be used in the FLAC codec? No. What about True Audio since the range of its features is so big? Yes? This I don't know.

I completely understand these are all extremely important factors to determine which codec will become the standard but if we have to change again in 10 years why don't everyone picks the "most complete one"? This was my thought.

People get angry, please try to understand English is my third language, sometimes I invert words and you may get something else other than what I really wanted to say, sorry about this.

This post has been edited by eahm: Jun 24 2012, 02:09


--------------------
/lwAsIimz
Go to the top of the page
+Quote Post
Takla
post Jun 24 2012, 02:52
Post #21





Group: Members
Posts: 169
Joined: 14-November 09
Member No.: 74931



That's fine but most people, even music lovers and audio equipment hobbyists, just don't care about surround sound, are not involved in music production and aren't DJs. Multi channel audio has been heavily promoted to domestic buyers/listeners several times since the 1970s and it's still the case that hardly anyone cares enough about it to buy SACD discs or players, just like they didn't buy Quadraphonic equipment or DVD-A. But they did buy many millions of radio cassette players and Walkmans, then radio CD players and CD Walkmans, and then iPods and similar. And how exactly do you promote the benefits of high fidelity multichannel playback to people who listen with earbuds?

Even a huge proportion of people who do care a lot about sound quality are content with stereo reproduction, and have also seen numerous formats and expensive next-big-things arrive and then expire over the years.

I think this idea of One Complete Codec has much to do with the human affinity for the mythical complete and universal answer, aka religious type feeling, and not much to do with what actually happens now or is likely to happen in the future. Probably what will continue to happen is that people involved in audio production will use various formats and codecs that suit their purpose, and people who consume/listen will use other, different, formats and codecs which suit theirs. The particular formats and codecs will inevitably change but the differences between production and consumption will remain. A tiny minority of consumers will want the format with all the extra features and the other 99.99% of the world will want something that sounds good (stereo without obvious artefacts is good enough), and works on their iPod and in the car and at home, and which they can easily share with their friends and family.

This post has been edited by Takla: Jun 24 2012, 02:54
Go to the top of the page
+Quote Post
skamp
post Jun 24 2012, 08:53
Post #22





Group: Developer
Posts: 1430
Joined: 4-May 04
From: France
Member No.: 13875



QUOTE (IgorC @ Jun 24 2012, 00:22) *
Is 3% compression gain significantly better? ( TAK vs FLAC)

If compression factor is a priority, then yes.

QUOTE (IgorC @ Jun 24 2012, 00:22) *
An encoding speed isn't issue with today hardware. I don't see anybody talking about speed issues and FLAC. Not even closer.

If your point is that FLAC is fast enough for most people, then I agree.

QUOTE (IgorC @ Jun 24 2012, 00:22) *
People prefer TAK less and less

Like I said. The way things are, TAK can't possibly grow its user base to the point of dethroning FLAC.


--------------------
See my profile for measurements, tools and recommendations.
Go to the top of the page
+Quote Post
Porcus
post Jun 24 2012, 10:57
Post #23





Group: Members
Posts: 1842
Joined: 30-November 06
Member No.: 38207



@ Greynol: you don't seem to get the point. I didn't claim – nor did I want to imply that it would be crucial – that FLAC would be unbeatable on each parameter (of the two I would guess were the key ones). Rather, I was saying that there was nothing that would beat it on one without having to sacrifice on another (suppose it were, there would be harder to explain FLAC's relative success). It's a partial ordering) issue.


@ eahm: I do agree with you that 8 channels might not be enough in [fill in number] years. For all that I know, someone might at some point want to deliver the multitrack recordings, 'one track per instrument' style, in a format which specifies the mix but allows for the user to add effects to each. Just to take an example that none of the 'consumer-oriented' lossless formats (to my knowledge) offer. I would guess that the 'next' such format would be a container format which also supports e.g. video, and which would be able to store multiple discrete streams of a certain codec (want 24 channel ALAC? Store three 8-channel streams.).


QUOTE (IgorC @ Jun 24 2012, 00:22) *
I don't see anybody talking about speed issues and FLAC. Not even closer.


Not even on portable devices? Even when the CPU can decode in realtime (a must!), then higher effort means lower battery time.


--------------------
One day in the Year of the Fox came a time remembered well
Go to the top of the page
+Quote Post
dumdidum
post Jun 24 2012, 16:35
Post #24





Group: Members
Posts: 57
Joined: 21-January 12
From: Germany
Member No.: 96595



QUOTE (Porcus @ Jun 23 2012, 22:50) *
First: there need not be any 'good' reason why one product dominates over another market-wise. And especially not here, where compatibility is more or less everything as soon as features are 'good enough'.

tbh, as an economist, i am a bit offended by your statement. of course, there are good reasons why a product dominates the market despite the presence of competitors. individuals do not roll dice to determine which product to adopt. individual decision-making is not arbitrary, and therefore what is happening in the aggregate isn't arbitrary.

i do agree that once most competing products are "good enough", network effects and other externalities become a bigger and bigger part of the story, importantly affecting both consumers' and firms' decisions. nevertheless, we can still explain how a product became number one in the first place while competitors faded into obscurity. for example, various game-theoretic models illustrate how a monopoly can arise out of oligopolistic competition where one firm has a first-mover advantage.

QUOTE
If N equally good codecs share the market evenly, that is most likely an unstable equilibrium. Whichever gets the upper hand (even if just by coincidence) will have better chances of growing quicker.

homogeneous goods (and services) are rare. i think codecs vary significantly and competition between them is imperfect. so, i can very well imagine situations were various codecs all have non-trivial market shares. of course, it's all a question of how narrowly you define the market. but, i think, for reasonable definitions of the relevant market, there can be equilibria in which several codecs own non-trivial shares of the market. look at youtube. they encode videos using no less than three lossy codecs. look at beatport, they offer four different audio formats. bandcamp offers even more. look at the variety of codecs--even lossless ones--employed by HA readers. etc.
Go to the top of the page
+Quote Post
Dario
post Jun 24 2012, 16:36
Post #25





Group: Members
Posts: 158
Joined: 20-September 11
Member No.: 93842



QUOTE (skamp @ Jun 23 2012, 22:07) *
Technically speaking, TAK rules all, IMO.

I second this. Simply said, TAK can’t be beaten by any other lossless codec that is currently present. It's a shame that there are no cross-platform binaries. Oh, and the lack of Unicode support it annoying, too.

QUOTE (IgorC @ Jun 24 2012, 00:22) *
QUOTE (skamp @ Jun 23 2012, 17:07) *
Technically speaking, TAK rules all, IMO. It's (much) faster and compresses (significantly) better than pretty much anything else that's remotely comparable.

Is 3% compression gain significantly better? ( TAK vs FLAC)

The compression gain of TAK over the reference FLAC implementation is something like 6%, not 3%. And 6% is a lot (even 3% is a lot).

This post has been edited by Dario: Jun 24 2012, 16:37
Go to the top of the page
+Quote Post

5 Pages V   1 2 3 > » 
Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 30th August 2014 - 04:40