Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: 24-bit accuracy Decoders (Read 6550 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

24-bit accuracy Decoders

What are the recomended decoders out there w 24 bit accuracy? And, is it advatageous to use Cool Edit along w these decoders? What is obtained by using one of these encoders as opposed to just directly putting them on a cd? Thanks Jeff

24-bit accuracy Decoders

Reply #1 –
1 - Please, pretty please, read David Robinson's decoder test. Most of the questions you asked here regarding MP3 decoders is already answered there.

2 - CoolEdit uses the FhG decoder, I.E: it's the same as Winamp (2.666 and on) decoder.
Unless you use PP's MP3 filter, then it uses Lame decoder (mpglib from mpg123).

Quote
What is obtained by using one of these encoders as opposed to just directly putting them on a cd?


3 - These what encoders? And them what you want to put on a CD?

Regards;

Roberto.

24-bit accuracy Decoders

Reply #2 –
Quote
Originally posted by jjarmak
What are the recomended decoders out there w 24 bit accuracy? And, is it advatageous to use Cool Edit along w these decoders? What is obtained by using one of these encoders as opposed to just directly putting them on a cd? Thanks Jeff



dunno if that could help but MAD is a 24bit decoder plug-in for Winamp. http://www.mars.org/home/rob/proj/mpeg/

24-bit accuracy Decoders

Reply #3 –
Does anybody use the 24 or 32 bit mppdecoders? They work nicely via the batch encoder, and I think I hear a difference, but it could all be psychological-it is referred to as HQ, though but everybody seems to use winamp exclusively, and the 24 bit decoders are not updated with the other, main decoder.  I wonder why..?

24-bit accuracy Decoders

Reply #4 –
Quote
Originally posted by rjamorim
1 - Please, pretty please, read David Robinson's decoder test. Most of the questions you asked here regarding MP3 decoders is already answered there.

Roberto.


Anyone planning on doing updated tests with current versions?

It's a nice block of work,  but it would be nice if some of the decoders that have been developed further get re-reviewed..

24-bit accuracy Decoders

Reply #5 –
Quote
Originally posted by jjarmak
What is obtained by using one of these encoders as opposed to just directly putting them on a cd

As rjamorim said look for answers on Davis Robinsons site.

I understand (but can't check:)) that 24 decoders are mainly of interest for those with 24 bits sound cards.
I've have been preparing wavs to make audio CDs by going the cumbersome way of decoding with l3dec (has a problem with long file names) to ascii H24 format, converting to 24 bit (packed) wav, resample to 16 bits with CoolEdit using dither and noiseshaping. All this is pretty slow also because of the large intermediate files.
The result was OK but I found fast and easy decoding with MAD (with Speeks handy frontend) sounding at least just as good to me. Maybe the fact that I use the MAD plugin when playing MP3's directly has conditioned me?

So , it not worth much trouble to have that last 20+ bit accurate when resampling to 16 bit afterwards. Having some form of dither while going to 16bit might help, but it's a matter of taste.
--
Ge Someone
In theory, there is no difference between theory and practice. In practice there is.

24-bit accuracy Decoders

Reply #6 –
Quote
Originally posted by JonPike
It's a nice block of work,  but it would be nice if some of the decoders that have been developed further get re-reviewed..
Which decoders are you thinking of?
Seems pretty up to date to me, although some of the players have newer versions I'm not aware of real decoder changes. If I missed something I'd like to know.
--
Ge Someone
In theory, there is no difference between theory and practice. In practice there is.

24-bit accuracy Decoders

Reply #7 –
I completely agree with GeSomeone:   

  IMHO the best way to create CD from MP3 is :
1) decode the file using L3dec in 24 bits or better with Foobar in 32 bits (or 64 bits if you own Soundforge 6)
2) remove DC Offset and dither (noise shaping in better) rectangular 1 bits noise shaping ultra/high pass contour
3) then you get a better (theorically) ready-to-burn wav files

But If want only play mp3/aac/mpc... I'll stick to Foobar in 32 bits (and 48kHz resampler for multimedia cards which badly resample).

After many test i've tried some decoders (old CEP filter based on Maplay,L3dec and Foobar)
and my impression is Foobar produces the best sound of all.

24-bit accuracy Decoders

Reply #8 –
@Nick Jr III

Can you actually hear any difference or is this the placebo talking?
"You have the right to remain silent. Anything you say will be misquoted, then used against you."

24-bit accuracy Decoders

Reply #9 –
>>  /\/ephaestous

I think you're ok w/ me to say better > good:
I mean why not to use the best technique (and not just a god one).
And I truly feel 24/32 or even 16 dithered is better on some samples:
        i feel traditional  truncation makes the music lifeless on some songs.

So i can't tell why A is better than B but my hears feels so...

24-bit accuracy Decoders

Reply #10 –
People here will tell you that your ears don't feel, but instead, hear.
It is proven that feeling conditions can make you hear things differently when they are not. That's what here call as "placebo effect".

Also, if you can hear a difference with decoding mp3's to 24bits instead of 16bits, mp3 might not be good for you all in all.

So, concluding. It is more probable that you are imaginating that, than the other way around.

24-bit accuracy Decoders

Reply #11 –
Jaz >

do you really think that hearing is just objective ?
Do you hear that a sample is lifeless or do you feel it ?

I sincerely choose the second option

there is a point where the feeling begin and the hearing stops. (IMHO)
have you never heard of impressions like " song is better than song b but i can't tell why"... ?

Best Regards, 

24-bit accuracy Decoders

Reply #12 –
oh just another thing:
let me justify myself,

we're not talking about artefacts (audible) but we're talking about bit resolution , the difference is this way much more subtle, it's nearly a feeling.

24-bit accuracy Decoders

Reply #13 –
I try to maintain the technical aspect of these boards, which don't "believe", but rather prove.

If you talk about low frequency sounds, I can accept that you feel them, instead of hearing them, but bit resolutions are not felt, but rather heard (Basically, a difference at -96db, which is the case of 16bits, can't move anything, it just gives a different noise).

About the objectiveness of hearing, I only can accept that mood can alter it, which is against a trully difference in the origin, but in the person itself.

Note that I have replied to you this way, because (I'm quoting you) you manifested that:

A) After many test i've tried some decoders (old CEP filter based on Maplay,L3dec and Foobar)
and my impression is Foobar produces the best sound of all.

B) And I truly feel 24/32 or even 16 dithered is better on some samples:
i feel traditional truncation makes the music lifeless on some songs.

And then,

C) So i can't tell why A is better than B but my hears feels so...


All in all, you think there's a difference, but you haven't found it. This generally corresponds to placebo. It's your work to discover if it is, or it is not. I can't help there.

24-bit accuracy Decoders

Reply #14 –
May be you're right 
but how can you prove something you can't describe ( in this case a sample which is "lifeless" than an other)  ?

24-bit accuracy Decoders

Reply #15 –
Quote
May be you're right  
but how can you prove something you can't describe ( in this case a sample which is "lifeless" than an other)  ?

Well why don't you have a friend play the same song using different decodings blindly to you. You can then decide which song "feels" the best and then you could see which decoder is "the best". In order to do this correctly your friend should randomly switch the decoder around and you should listen to each decoder about 5-10 times. If you consistently pick the same decoder as "feeling" better then that is definitely the best decoder for you. If you pick each decoder as best an equal number of times then that would probably prove that you were just suffering a placebo effect.
Sorry, I have nothing witty to say here.

24-bit accuracy Decoders

Reply #16 –
flloyd >

i agree with your methology !

thanx to all for your replies !

24-bit accuracy Decoders

Reply #17 –
methology > methodology 

24-bit accuracy Decoders

Reply #18 –
Of course it's placebo. You can never achieve SNR better than the original recording. And we know the encoding already decreases that SNR.

24-bit accuracy Decoders

Reply #19 –
tangent > you're right,absolutely right but I don't think bit depth only makes change the SNR.
I'll ABX some samples for Monday, april 20th , at this time I'm On Vacation!
I'll test Foobar 32/64 diskwriter,Winamp 2.81 output and problably L3dec.

Regards,

24-bit accuracy Decoders

Reply #20 –
Nick Jr III <=

Hi,
Just a remark about mp3 and bit resolution:  remember that, even at ~320kbps, there are many bits of resolution lost during encoding, at various frequencies. The time resolution is lower and pre-echo is always present too.

So, whether you choose to decode the 2-bit..12-bit numbers to 16 or 24 bits usually won't matter at all...

See:  16 bit (cd) ->  2..12 bit (mp3)  -> 24 bits ? 

24-bit accuracy Decoders

Reply #21 –
http://ff123.net/madchallenge.html

FWIW the value of each sample in a decoded mp3 will be dramatically different from the value of that sample on the original CD. However, when the value has been calculated, you can round it to 16bits, dither it to 16bits, or round it to 24 bits (or... etc etc).

The noise and distortion added by mp3 is one type of effect; the noise and distortion added by rounding or truncation is another type of effect altogether.

I don't claim to be able to hear the effect of truncating to 16-bits on normal music at normal listening levels. However, there are those who do. If they really do hear something, it is not unthinkable that they would be able to separate the distortion introduced by mp3 encoding from the distortion introduced by truncating the output to 16-bits.


Whatever - I'm sure 95% of the reports of the "better sound" of 24-bit decoding are placaebo, especially on 16-bit equipment. I wouldn't bet my life on it all being placeabo though - some 24-bit DACs are measurably different if fed with a correctly dithered 16-bit signal + 8 bits of noise compared to being fed with a correctly dithered 16-bit signal + 8 zeros.

And, should you own a recording with a 96dB dynamic range (very unlikely!), and should you be able to stand to listen to the loudest part at 100dB SPL, then the difference between dithered 16-bit and 24-bit in the quietest section may be clearly audible in a near-silent listening room. To put this in context, most pop music has much less than 48dB dynamic range!

Cheers,
David.

P.S. because of the different measures of dB SPL for pure tones and noise; and because of the different (and usually confused) meanings of dynamic range and signal-to-noise ratio, it's possible to have what seem like impossible combinations. e.g. a 30dB tone clearly audible in the presence of 40dB white noise - it sounds like nonsense until you look at the spectrum - then you find that the tone is way above the noise!
 

24-bit accuracy Decoders

Reply #22 –
NumLOCK >

hi,

you wrote:
there are many bits of resolution lost during encoding, at various frequencies.
what do you exactly mean ?

2Bdecided >
hi,
you wrote :
most pop music has much less than 48dB dynamic range!
1/you're right but don't ya think it can improve sound quality when it comes to classical music ?
2/IMHO 24 bits decoding is useless for modern/electro song but for guitar songs (I want to talk about formants which are problematic in terms in fidelity).

ThnX you very much !