[charonme]

Was the original question whether the DACs advertised as 24bit are really trying to take all the 24 bits of digital information and convert it to analog signal or they deliberately disregard a few bits? This would be just a question of circuit design, I sense it's been already settled that the difference wouldn't be detectable from the output analog signal because of noise.

 

[Nocturnal310]

Quote:

Originally Posted by metroidfox /img/forum/go_quote.gif Btw, I can hear above 22k - but pretty much anything (when I'm outside) in that range is just noise and hurts my head.

ya and i can see the God .


Most of us cant even hear above 18k by the time we reach 20. and u r speaking of 22k which is not biologically possible.


U must have heard of Teen ringtones? they work on this principle.


24/96 can only be enjoyed by new-born babies with highest quality of ears.

 

[gyrodec]

Nocturnal310

22k is biologically possible, it is simply rare. The human biology is quite varied person to person. If he has tested his hearing to 22k, then he can hear to 22k. Its not like he's claiming 44k.

 

[gregorio]

Quote:

Originally Posted by thomaspf /img/forum/go_quote.gif Here is what you can do. Create a 24 bit WAV file that cycles through all possible sample values of a 24 bit sample. Play that file through a digital output and record the result. You will find that 24 bit sound cards do indeed play back all samples with their correct values....

You've got yourself a bit twisted up here Tomaspf. As long as the audio stays in the digital domain then you are absolutely correct. However, the whole reason for a DAC to exist (the clue is in the name) is to convert the digital audio to analogue so it can be heard. It's at this point that no DAC on the planet can output anywhere near a 144dB dynamic range, which is the whole point of 24bit over 16bit.

24bit is useful but only in recording, not playback. Taking a 16bit recording and converting it to 24bit is a complete waste of time, unless you think there is a difference between a 16bit file and a 24bit file with the last 8 LSBs all set to zero.

The reality is that no one in their right mind would ever release a recording that uses the full dynamic range of 24bit because it would be classified as an extremely dangerous product, IE. It could potentially kill you! The usable dynamic range of commercial recordings is well within the 16bit format.

So us professionals are in a kind of weird vicious circle. Consumers erroneously believe 24bit is better than 16bit, so the demand for 24bit is growing. So we are releasing 24bit products which only utilise at best about 14bits of the 24bit format and the majority of commercial recordings utilise less than 6bits. In other words, there are no recordings on the market which take full advantage of even the 16bit format and even if there were you wouldn't be able to listen to them without causing hearing damage. So why on earth are the manufacturers convincing us to move to 24bit DACs when no content does or will ever exist in 24bit and their DACs couldn't resolve it anyway? If this all sounds ridiculous to you then join the club!!

 

[keal]

It's 2012, just doing a search on 24-bit audio as it seems life is getting to the point where consumers like me are starting to collect 24-bit audio files in FLAC format.
 
Currently, I have a bunch of music off the internet, and some of them I can tell they are indeed better than CD quality.  But some, like Pink Floyd's album Dark Side Of The Moon, I picked the 16-bit CD rip over the 24-bit version in a listening test.  However, recording friends playing acoustic guitar with a 24-bit recorder, and playing back the original recordings, as well as 16-bit wav file conversions, you can tell the difference.  3 of us could tell the differences every single time, in blind tests.  Doesn't mean much, because we didn't have conniptions or convulsions in full-blown ecstasy.  I'm not a militant 24-bit audiophile, yet
 
It boils down to the quality of the original recording, PLUS the option of whether that recording was mastered before finally creating the 24-bit version.  I don't want to waste time researching this stuff, so for the time being I'm settling on not getting anything in 24-bit if it's not a fresh recording.  Old remasters, they need to have full documentation on the process.  It's not happening yet.
 
If you have super awesome headphones, paired with a headphone amp, all I can say is, don't pay extra for the 24-bit versions until you know the quality of the material.  As of March 2012, this seems to be a transitional stage.  You could be taken advantage of, simply buying everything released as 24-bit audio files.  In a few years, it should all pan out.
 
Personally, I'm fine right now with 16-bit FLAC files.  I collect a handful of currently-recording jazz artists when their albums are 24-bit flac and it's not double the price of their 16-bit album, but I'm refraining from any old albums that are being reissued in the high-def formats.
 
This part is purely opinion, but I think that 16-bit audio CDs and MP3 downloads (at 320kbps) today are better than audio CDs that were being sold 20 years ago.  I was there for the very first few CDs like Dire Strait's 'Money For Nothing' along with a Sony CD boombox, and I can't say I was blown away then, but these days, I am always stunned by MP3 downloads.  Things have definitely improved!
 
Not sure of the future of 24-bit FLAC files, but I want it to be the end-all for music nuts.  I simply DO NOT want to keep wearing white gloves to look at a jazz vinyl collection!  Shoot me now or give me 24-bit FLAC
 
 

 

[Roseval]

First of all 16 or 24 is a numerical description, it simply states the word length the DAC (the chip doing the actual DA conversion) expects at the input.
[size=10pt] [/size]
[size=10pt]If you send a 24 bit file to a 16 bit DAC it won’t play. In practice it does as your media player will truncated it to 16 bit and dither.[/size]
[size=10pt]If you play a 16 bit file on a 24 bit DAC it won’t play either. In practice your media player will append 8 zero bits to get the desired 24 bit word length.[/size]
 
[size=10pt]Dynamic range[/size]
[size=10pt]A bit= 6 dB so 16 bits allow for 96 dB and 24 for 144 dB.[/size]
[size=10pt]Some do think using the full 144 dB is dangerous, it will produce 144 dB over the stereo.[/size]
[size=10pt]In practice it is the reverse, 24 bits allows you to play softer![/size]
 
[size=10pt]There is the maximum output of the DAC, over RCA this is in general 2V.[/size]
[size=10pt]This is called 0 dBFS.[/size]
[size=10pt]A 16 bit recording can go down to -96 dBFS, a 24 to -144 dBFS.[/size]
[size=10pt]A 24 can resolve micro details below -96 dBFS and 96 is pretty soft already.[/size]
[size=10pt]Some gear do have a noise floor around -95 but even the best gear does have a noise floor well above -144, say -110 is a pretty good value.[/size]
 
[size=10pt]16 or 24 is a numerical indicator, not a performance indicator.[/size]
[size=10pt]The performance is the linearity of a DAC, the ability to translate the numerical value in to an equivalent voltage.[/size]
[size=10pt]Cheap DACs are linear to 12 bits, good ones goes up to say 18 and the very good ones with the whisper quit circuits up to 21/22 bits.[/size]
 
[size=10pt]This is no surprise.[/size]
[size=10pt]16 bits = 65536 discrete steps[/size]
[size=10pt]One step=4/65536=0.00006104V
24 bits = 16777216 discrete steps
One step=4/16777216 = 0.00000024 V[/size]
[size=10pt]This is such a small value no DAC at the present can resolve this properly.[/size]
 

 

[UtzY]

Quote:

First of all 16 or 24 is a numerical description, it simply states the word length the DAC (the chip doing the actual DA conversion) expects at the input.
[size=10pt] [/size]
[size=10pt]If you send a 24 bit file to a 16 bit DAC it won’t play. In practice it does as your media player will truncated it to 16 bit and dither.[/size]
[size=10pt]If you play a 16 bit file on a 24 bit DAC it won’t play either. In practice your media player will append 8 zero bits to get the desired 24 bit word length.[/size]
 
[size=10pt]Dynamic range[/size]
[size=10pt]A bit= 6 dB so 16 bits allow for 96 dB and 24 for 144 dB.[/size]
[size=10pt]Some do think using the full 144 dB is dangerous, it will produce 144 dB over the stereo.[/size]
[size=10pt]In practice it is the reverse, 24 bits allows you to play softer![/size]
 
[size=10pt]There is the maximum output of the DAC, over RCA this is in general 2V.[/size]
[size=10pt]This is called 0 dBFS.[/size]
[size=10pt]A 16 bit recording can go down to -96 dBFS, a 24 to -144 dBFS.[/size]
[size=10pt]A 24 can resolve micro details below -96 dBFS and 96 is pretty soft already.[/size]
[size=10pt]Some gear do have a noise floor around -95 but even the best gear does have a noise floor well above -144, say -110 is a pretty good value.[/size]
 
[size=10pt]16 or 24 is a numerical indicator, not a performance indicator.[/size]
[size=10pt]The performance is the linearity of a DAC, the ability to translate the numerical value in to an equivalent voltage.[/size]
[size=10pt]Cheap DACs are linear to 12 bits, good ones goes up to say 18 and the very good ones with the whisper quit circuits up to 21/22 bits.[/size]
 
[size=10pt]This is no surprise.[/size]
[size=10pt]16 bits = 65536 discrete steps[/size]
[size=10pt]One step=4/65536=0.00006104V
24 bits = 16777216 discrete steps
One step=4/16777216 = 0.00000024 V[/size]
[size=10pt]This is such a small value no DAC at the present can resolve this properly.[/size]
 

Do you have some examples of very good DACs with 21/22 ENOB? I know that in the good category are dacs like Benchmark DAC1, Anedio D2...know "better" ones? (Measured)

 

[astrallite]

Quote:

u r speaking of 22k which is not biologically possible.
 

U.S. Air Force has a sound test program and some service members have managed to test out to 23 KHz...I don't think 22K is out of the question.

 

[khaos974]

 
Quote:

Do you have some examples of very good DACs with 21/22 ENOB? I know that in the good category are dacs like Benchmark DAC1, Anedio D2...know "better" ones? (Measured)

Weiss DAC, most DACs from Weiss reach 21/22 ENOB, possibly more.
Same with the Ressonessence Invicta, or the high end from Moon Audio.
 
In more reasonable prices, the Anedio D2 is probably the on with the best measurements.
 
 

 

[joe_cool]

Quote:

U.S. Air Force has a sound test program and some service members have managed to test out to 23 KHz...I don't think 22K is out of the question.

References, please. This is not he Sound Science forum!

 

[rahzim]

Quote:

[size=10pt] [/size]
[size=10pt]If you send a 24 bit file to a 16 bit DAC it won’t play. In practice it does as your media player will truncated it to 16 bit and dither.[/size]
[size=10pt]If you play a 16 bit file on a 24 bit DAC it won’t play either. In practice your media player will append 8 zero bits to get the desired 24 bit word length.[/size]

    This brings up a question I have regarding 24bit recordings.  I downloaded the album Hotel California on HDtracks in the 24/96hz bitrate.  Is there a way I can tell what the bitrate is leaving my laptop?  I'm waiting on my dac to arrive in the mail, and I'm curious as to whether the signal is being down converted from 24/96 to 16 bit/44.1.  The only comparison I can do right now without an external DAC is between the HDtracks version and the 128VBR version I already had.  The HDtracks version sounds slightly better, but definitely not night and day.
 
Fubar2000 says the file playing is 24/96, but I'm not sure if that's before or after conversion, or if in fact any conversion is being done at all.

 

[proton007]

Quote:

    This brings up a question I have regarding 24bit recordings.  I downloaded the album Hotel California on HDtracks in the 24/96hz bitrate.  Is there a way I can tell what the bitrate is leaving my laptop?  I'm waiting on my dac to arrive in the mail, and I'm curious as to whether the signal is being down converted from 24/96 to 16 bit/44.1.  The only comparison I can do right now without an external DAC is between the HDtracks version and the 128VBR version I already had.  The HDtracks version sounds slightly better, but definitely not night and day.
 
Fubar2000 says the file playing is 24/96, but I'm not sure if that's before or after conversion, or if in fact any conversion is being done at all.

 
Foobar tells you the bit-rate of the song its playing. The actual truncation happens at the decoder/DAC, so if you know which sound chip/framework you have, you can know if it supports 24/96 playback.
For example, if your laptop implements intel HD audio, chances are it supports 24/96.

 

[rahzim]

Quote:



24/96 can only be enjoyed by new-born babies with highest quality of ears.

Thanks for the tip proton.

This is why trying to learn about high quality audio can be so confusing. Up to this point everything I've been reading has been talking about how cd's are deficient for top quality playback because they only do 16 bit /44hz . It appears as if the op is saying there's no point to higher Rez playback than cd quality.

 

[proton007]

Quote:

Thanks for the tip proton.
This is why trying to learn about high quality audio can be so confusing. Up to this point everything I've been reading has been talking about how cd's are deficient for top quality playback because they only do 16 bit /44hz . It appears as if the op is saying there's no point to higher Rez playback than cd quality.

 
That's up for debate...the audio resolution is guided by the requirements. From my own understanding of signals, using a higher sampling rate will not help much, because it is redundant, it does not give you any more information than 44.1kHz. Going by the 16 bit or 24 bit argument, it depends on whether it makes any difference. Ofcourse, 24 bit is closer to the analog signal because it increases the number of discrete steps, but its more useful for recording and editing rather than playback. Its possible our ears cannot discern a step change of 16 vs 24 bit at that high a frequency.
Ethan Winer has put up links on his website, you can actually listen to audio encodings at different bit rates, from 24 all the way to 9 bits.
http://www.ethanwiner.com/BitsTest.html
 
Try to discern them in a blind test. I can hear differences only at around 11 or 9 bits.