Question about audio artifacts in WMA

[Radium[_4_]]

Hi:

Why do WMA artifacts sound different from MP3 artifacts?

What is the mechanism behind WMA compression that gives WMA its
characteristic artifacts?


Thanks,

Radium

P.S. Hydrogen-Audio has neglected me and I have lost my patience!!!!

http://www.hydrogenaudio.org/forums/index.php?s=8049b632e3397908e62dda977d8e01de&showtopic=56553

[[email protected]]

On Aug 2, 8:37 pm, Radium > wrote:
> Hi:
>
> Why do WMA artifacts sound different from MP3 artifacts?
>
> What is the mechanism behind WMA compression that gives WMA its
> characteristic artifacts?
>
> Thanks,
>
> Radium
>
> P.S. Hydrogen-Audio has neglected me and I have lost my patience!!!!
>
> http://www.hydrogenaudio.org/forums/index.php?s=8049b632e3397908e62dd...

I think its cuz it does multiple passes when compressing unlike MP3 so
the redundancy-detection threshold is probably higher... such as some
repeating verses in the song that make up 50% of the audio. I got a
couple 64kbps WMAs which retain the same quality as a 128kbps MP3.

Yeah, I think it does multiple passes since it takes like 4x longer to
encode/decode than MP3.

[Richard Crowley[_2_]]

On Aug 3, 5:59 am, wrote:
> On Aug 2, 8:37 pm, Radium > wrote:
>
> > Hi:
>
> > Why do WMA artifacts sound different from MP3 artifacts?
>
> > What is the mechanism behind WMA compression that gives WMA its
> > characteristic artifacts?
>
> > Thanks,
>
> > Radium
>
> > P.S. Hydrogen-Audio has neglected me and I have lost my patience!!!!
>
> >http://www.hydrogenaudio.org/forums/index.php?s=8049b632e3397908e62dd...
>
> I think its cuz it does multiple passes when compressing unlike MP3 so
> the redundancy-detection threshold is probably higher... such as some
> repeating verses in the song that make up 50% of the audio. I got a
> couple 64kbps WMAs which retain the same quality as a 128kbps MP3.
>
> Yeah, I think it does multiple passes since it takes like 4x longer to
> encode/decode than MP3.

"WMA" is actually one of four different codecs: Standard,
Professional;
Lossless or Voice. SOME of the variations and versions are 2-pass,
but most appear to be 1-pass. Some basic research may be in order...
http://en.wikipedia.org/wiki/Windows_Media_Audio

[Richard Crowley[_2_]]

On Aug 3, 5:59 am, wrote:
> Yeah, I think it does multiple passes since it takes like 4x longer to
> encode/decode than MP3.

It has never taken "4x" longer to encode in my experience.
I rarely notice the difference in encoding times.

It CLEARLY does NOT take "4x longer" to decode or you
wouldn't be able to play WMA in real-time. Think about it.

[Serge Auckland]

> wrote in message
ps.com...
> On Aug 2, 8:37 pm, Radium > wrote:
>> Hi:
>>
>> Why do WMA artifacts sound different from MP3 artifacts?
>>
>> What is the mechanism behind WMA compression that gives WMA its
>> characteristic artifacts?
>>
>> Thanks,
>>
>> Radium
>>
>> P.S. Hydrogen-Audio has neglected me and I have lost my patience!!!!
>>
>> http://www.hydrogenaudio.org/forums/index.php?s=8049b632e3397908e62dd...
>
> I think its cuz it does multiple passes when compressing unlike MP3 so
> the redundancy-detection threshold is probably higher... such as some
> repeating verses in the song that make up 50% of the audio. I got a
> couple 64kbps WMAs which retain the same quality as a 128kbps MP3.
>
> Yeah, I think it does multiple passes since it takes like 4x longer to
> encode/decode than MP3.
>

Whatever else it does, it *doesn't* do anything with repeating verses, as no
two repeating verses are *identical*, consequently none is redundant. The
codec doesn't know anything about what the music consists of, so *can't*
know if verses have been repeated unless the data (not the music) is
*identical*, which it never is.

S.

--
http://audiopages.googlepages.com

[isw]

In article . com>,
Richard Crowley > wrote:

> On Aug 3, 5:59 am, wrote:
> > Yeah, I think it does multiple passes since it takes like 4x longer to
> > encode/decode than MP3.
>
> It has never taken "4x" longer to encode in my experience.
> I rarely notice the difference in encoding times.
>
> It CLEARLY does NOT take "4x longer" to decode or you
> wouldn't be able to play WMA in real-time. Think about it.

Well, no. (Hypothetically) if a computer can decode a second's worth of
MP3 in a tenth of a second, and it takes four times as long (four-tenths
of a second) for a WMV decoder to do its thing, both will play the one
second of audio in one second.

Isaac

[[email protected]]

On Aug 3, 9:12 am, Richard Crowley > wrote:
> On Aug 3, 5:59 am, wrote:
>
> > Yeah, I think it does multiple passes since it takes like 4x longer to
> > encode/decode than MP3.
>
> It has never taken "4x" longer to encode in my experience.
> I rarely notice the difference in encoding times.
>
> It CLEARLY does NOT take "4x longer" to decode or you
> wouldn't be able to play WMA in real-time. Think about it.

Bitch, you clearly don't know what the **** you're talking about. You
need to eat some ****.

[Phil Carmody]

Richard Crowley > writes:
> On Aug 3, 5:59 am, wrote:
> > Yeah, I think it does multiple passes since it takes like 4x longer to
> > encode/decode than MP3.
>
> It has never taken "4x" longer to encode in my experience.
> I rarely notice the difference in encoding times.
>
> It CLEARLY does NOT take "4x longer" to decode or you
> wouldn't be able to play WMA in real-time. Think about it.

Well, he may, or may not, be full of crap, but such
responses are clearly little better. MP3 decoding takes
less than 0.1% of my last-generation CPU, so "4x longer"
would still be trivial in real time.

Phil
--
Dear aunt, let's set so double the killer delete select all.
-- Microsoft voice recognition live demonstration

[Ron Capik]

wrote:

> < ...snip... >
>
> I think its cuz it does multiple passes when compressing unlike MP3 so
> the redundancy-detection threshold is probably higher... such as some
> repeating verses in the song that make up 50% of the audio. I got a
> couple 64kbps WMAs which retain the same quality as a 128kbps MP3.
>
> Yeah, I think it does multiple passes since it takes like 4x longer to
> encode/decode than MP3.

I would hope by "multiple passes" you don't mean it looks at the whole
song/file several times before encoding or decoding as that seems
rather absurd.

Thus I would hope that by " it does multiple passes" you mean it has a
larger chunk overlap. Maybe it works with larger chunks and has more
overlap or just a more complex algorithm than MP3. Those would all
have an impact on latency and processor load.


Later...

Ron Capik
--

[[email protected]]

On Aug 4, 8:10 am, Ron Capik > wrote:
> wrote:
> > < ...snip... >
>
> > I think its cuz it does multiple passes when compressing unlike MP3 so
> > the redundancy-detection threshold is probably higher... such as some
> > repeating verses in the song that make up 50% of the audio. I got a
> > couple 64kbps WMAs which retain the same quality as a 128kbps MP3.
>
> > Yeah, I think it does multiple passes since it takes like 4x longer to
> > encode/decode than MP3.
>
> I would hope by "multiple passes" you don't mean it looks at the whole
> song/file several times before encoding or decoding as that seems
> rather absurd.
>
> Thus I would hope that by " it does multiple passes" you mean it has a
> larger chunk overlap. Maybe it works with larger chunks and has more
> overlap or just a more complex algorithm than MP3. Those would all
> have an impact on latency and processor load.
>
> Later...
>
> Ron Capik
> --

Yes. However, that is what I *ASSUME,* cuz I don't see how else it
would outclass MP3 by 2x.

[Steven Sullivan]

In rec.audio.tech Radium > wrote:
> Hi:

> Why do WMA artifacts sound different from MP3 artifacts?

> What is the mechanism behind WMA compression that gives WMA its
> characteristic artifacts?


> Thanks,

> Radium

> P.S. Hydrogen-Audio has neglected me and I have lost my patience!!!!

> http://www.hydrogenaudio.org/forums/index.php?s=8049b632e3397908e62dda977d8e01de&showtopic=56553


actually you asked the same question twice, and it was answered on this thread

http://www.hydrogenaudio.org/forums/index.php?showtopic=54863

WMA is closed source. So for details of the codec you'll have to ask someone at Microsoft
who's willing to disclose what is probably proprietary information.

You can also search hydrogenaudio for threads about WMA, eg

http://www.hydrogenaudio.org/forums/index.php?s=&showtopic=54312

___
-S
"As human beings, we understand the world through simile, analogy,
metaphor, narrative and, sometimes, claymation." - B. Mason