[Shike]

Quote:

http://www.aes.org/e-lib/browse.cfm?elib=11981
 

 
The abstract doesn't seem to say much in regards to what you said.
 
 
Quote:

A dynamic range of 118 dB is determined necessary for subjective noise-free reproduction of music in a dithered digital audio recorder.

 
See what I mean?
 
Moreover, just to be sure - we're talking about SNR correct?  I want to confirm whether your claim is a 3dB difference in SNR is audible.  In relation to THD, 3dB may or may not be audible.  If you're saying +3dB of noise onto whatever signal is already playing - then it's not necessarily impossible that it be audible depending on what level the signal is playing at.  Then again, the SNR would be terrible . . .

 

[tomb]

I'm not sure this is fruitful going down this road.  I try to interject when possible whenever the masses are being misled about DACs.  You can tell differences, whether one wants to believe that or not.  You challenged me for a citation.  This is a typical tactic with the same crowd trying to justify their cheap* soundcards and DACs by trying to tell everyone there's no appreciable difference.  I'm not saying that you were doing that, but it wouldn't be surprising if a double-blind test was mentioned at some point, too.

 
In any event, the citation is meant to illustrate that human hearing (educated hearing) can indeed detect noise floors up to a range of 118 dB.  You challenged the idea of whether someone could tell the difference between noise at -87db vs. noise at -90db.  I submit that they can under the right conditions and circumstances.  That citation at least proves that human hearing can detect noise beyond that.  If you want to focus on the mention of a "recorder," SN ratio vs Dynamic Range, etc., then I can't help there.
 
 
* There's nothing wrong with "cheap."  I build, sell, and have helped design cheap DACs - some of the most inexpensive on the market.  The difference is that I don't go around trying to tell everyone that there are no differences among them and that someone should settle for one of them, or a soundcard, or someone's inexpensive but seemingly good-test-results amplifier - instead of continuing a journey through the audiophile experience.

 

[stv014]

Quote:

I'm not saying that you were doing that, but it wouldn't be surprising if a double-blind test was mentioned at some point, too.

 
No subjective comparison of DACs can be taken seriously if it is sighted, unless one of the devices being compared is very bad. The differences are simply too small, and are entirely outweighed by psychological bias. In sighted tests, people can "hear" differences even between binary identical lossless audio files.
 
Quote:

Originally Posted by tomb /img/forum/go_quote.gif
 
In any event, the citation is meant to illustrate that human hearing (educated hearing) can indeed detect noise floors up to a range of 118 dB.

 
The article you cited is for professional audio recording, as Shike already pointed out, but you conveniently ignored. For studio recording, the extra headroom is needed for the processing, mixing, and dynamic compression that will be applied to the signal. On the other hand, for simple playback of 16 bit CD audio, 118 dB is more than enough as the data format itself has much higher noise level anyway. By the way, some currently available "cheap soundcards" can actually reach 118 dB A-weighted signal to noise ratio.

 

 

[Shike]

The recording chain and reproduction chain are different.  It's one of the main arguments of when to use 24 bit vs. 16 bit for example.  I don't have access behind the AES paywall, but I did find another related PDF from Julian Dunn that did a later AES convention that cited it and found findings we're interested in (at least in regards to noise).  It appears in a domestic situation, 101dB-108dB could be necessary under the most quiet of circumstances for domestic listening, but it also mentioned the use of noise shaping to gain the necessary subjective improvement (psychoacoustics) to reach the 101-108dB requirement.  This is cited as important due to CD's limited 93dB dynamic range.
 
So a 101-108dBA SNR or better is necessary to eliminate noise effectively.  So, there we go - we now know the law of diminishing returns are.  No, a UCA202 won't be perfect in this sense.  However, it seems a DAC Magic for example would be - etc etc.

 

[Currawong]

stv014: The OP asked why amps are so expensive, not about whether you don't respect subjective impressions of sound. Please stay on topic.

 

[sphinxvc]

Quote:

I'm not sure this is fruitful going down this road.  I try to interject when possible whenever the masses are being misled about DACs.  You can tell differences, whether one wants to believe that or not.  You challenged me for a citation.  This is a typical tactic with the same crowd trying to justify their cheap* soundcards and DACs by trying to tell everyone there's no appreciable difference.  I'm not saying that you were doing that, but it wouldn't be surprising if a double-blind test was mentioned at some point, too.

 
In any event, the citation is meant to illustrate that human hearing (educated hearing) can indeed detect noise floors up to a range of 118 dB.  You challenged the idea of whether someone could tell the difference between noise at -87db vs. noise at -90db.  I submit that they can under the right conditions and circumstances.  That citation at least proves that human hearing can detect noise beyond that.  If you want to focus on the mention of a "recorder," SN ratio vs Dynamic Range, etc., then I can't help there.
 
 
* There's nothing wrong with "cheap."  I build, sell, and have helped design cheap DACs - some of the most inexpensive on the market.  The difference is that I don't go around trying to tell everyone that there are no differences among them and that someone should settle for one of them, or a soundcard, or someone's inexpensive but seemingly good-test-results amplifier - instead of continuing a journey through the audiophile experience.

Nice post.
 

 

[leeperry]

Anyway, the definitive answer to the OP is: design and build your own amp, and you'll find the commercial ones cheap as hell =)
 
It's all about the details, and just like most/many DAC manufacturers like to boast about the DAC chip model number and use cheapo output stages, Many amp manufacturers will skimp in the areas that matter the most: the PSU and the attenuator. A crappy pot and you're on a sinking boat....stereo misbalance, random distortion 

 

[Spareribs]

Putting together a good quality amp by hand is work and not like making a ham sandwich. If you ever built a quality amp with meticulous attention, you can see that it would be an insult to charge $149. I mean c'mon, stop being so cheap and realize that quality products and workmanship deserves it's credit and craftsmen need to make money to survive.