Testing audiophile claims and myths

[analogsurviver]

Although not directly applicable to the FFT noise spike, you can test your PC  if it has the performance required for conversions on the fly etc by downlading jRiver
http://www.jriver.com/  free trial. There is a benchmark test in its Help menu and in jRiver's blog you can figure out how many "points" are required for certain operation(s). 
 
It can be simple playback of MP3 - up to converting on the fly while performing correction for the acoustics of the room or EQing headphones. And accordingly, the load on the computer can vary wildly as well.  Although jRiver does not achieve the same quality as native DSD, converting MP3 on the fly up to DSD256 ( using DSD capable DAC of required performance ) can significantly improve the SQ of those MP3s. Or CD redbook... - the reason is the ability to use MUCH simpler filtering not affecting the soundstage nearly as much as CD redbook has to - by default.
 
For Mac, an "equivalent" is Amarra or Audirvana - being Windows user, no direct experience with these. 

 

[Phishin Phool]

...converting MP3 on the fly up to DSD256 ( using DSD capable DAC of required performance ) can significantly improve the SQ of those MP3s. Or CD redbook... -

That is quite a claim - got anything to back that statement up?

 

[lamode]

  With 2 different analyzers.. the same recording as before 96/24 from ECM...
 
any suggestions of where this large spike come from?
 
this begins with the ending of file then to the beginning... both very silent passages...
 
could it be a mosquito repellent? as someone suggested?

 
It's just one possibility, but electronic ballasts for fluorescent lights operate in that band. Without being at the venue, it is just guesswork.  What kind of lighting was installed there?

 

[lamode]

Originally Posted by analogsurviver /img/forum/go_quote.gif
 
converting MP3 on the fly up to DSD256 ( using DSD capable DAC of required performance ) can significantly improve the SQ of those MP3s.
 
 

 
Oh good grief... more ridiculous claims. How can adding noise and distortion "improve" the sound?

 

[stv014]

Originally Posted by analogsurviver /img/forum/go_quote.gif
 
Although jRiver does not achieve the same quality as native DSD, converting MP3 on the fly up to DSD256 ( using DSD capable DAC of required performance ) can significantly improve the SQ of those MP3s. Or CD redbook... - the reason is the ability to use MUCH simpler filtering not affecting the soundstage nearly as much as CD redbook has to - by default.

 
Upsampling CD quality audio does not restore the lost information above 22.05 kHz (not that it should normally be needed anyway, as the listener is most likely not a bat), with the "simple" filter at most it produces high frequency images of the content in the audio band. The filter used in the upsampler has to deal with the same trade-offs as any other (hardware or software) reconstruction filter for 44.1 kHz PCM.

 

[OddE]

could it be a mosquito repellent? as someone suggested?

-Could be, but my money would be on the alarm system; motion detection sensors typically rely on ultrasound and doppler shift on the echoes.

Though it would make sense for the sensors to be disabled when the alarm system was disabled - which it presumably would be when recording was in progress... Hm.

 

[maverickronin]

  converting MP3 on the fly up to DSD256 ( using DSD capable DAC of required performance ) can significantly improve the SQ of those MP3s. Or CD redbook...

 
That's like saying you took a picture of an apple with a cheap 2MP point and shoot camera, printed it out, took a picture of that picture with a TOTL DSLR and quality glass, and then claiming the second picture contains more detail of the original apple.
 

  It's just one possibility, but electronic ballasts for fluorescent lights operate in that band. Without being at the venue, it is just guesswork.  What kind of lighting was installed there?

 
That's a good one too.  I forgot about EMI.

 

[StanD]

   
Oh good grief... more ridiculous claims. How can adding noise and distortion "improve" the sound?

I see everyone is still busy feeding the troll.

 

[MacacoDoSom]

...another strange one for you guys... HDtracks... 192/24....
 

 

[MacacoDoSom]

Please try to keep it serious, people...

 

[analogsurviver]

-Could be, but my money would be on the alarm system; motion detection sensors typically rely on ultrasound and doppler shift on the echoes.

Though it would make sense for the sensors to be disabled when the alarm system was disabled - which it presumably would be when recording was in progress... Hm.

Alarm sensors and emergency lights (which can also emit ultrasonics ) are prescribed to be ON by the law - at least whenever people are in the building. Else, the police, fire department or security is  obliged to come for a check if everything is in order - and those visits are not free. It varies in regulation across the globe, but for these reasons , people responsible are VERY reluctant to turn off anything that might result in a "visit" and a resulting fee - even during recordings.
 
It is hard enough for them to be persuaded to kill the ventilation or heating devices -  but these are at least audible to them and they can understand why they have to be shut down for the recording. But they usually do it reluctantly, because powering down the "climate" in a large hall might take some time to really stop - and at least the same amount of time to get it up again. 
 
It may sound trivial - but once such noises land on the recording, it is only recording engineer who gets the blame. Sometimes, more energy must be spent into assuring the decent noise free environment is achieved - than in recording itself.

 

[lamode]

  ...another strange one for you guys... HDtracks... 192/24....
 

 
There's not much point in posting every example. The same list of suspects applies to each one. My money is on the lighting.

 

[MacacoDoSom]

   
There's not much point in posting every example. The same list of suspects applies to each one. My money is on the lighting.

...could be... thanks...

 

[analogsurviver]

   
There's not much point in posting every example. The same list of suspects applies to each one. My money is on the lighting.

Something else we may agree on. 
 
No matter how we look at it, it is the real thing. We as audience do not get "quieter treatment" than during the recordings - that is certain to be true.

 

[analogsurviver]

   
Oh good grief... more ridiculous claims. How can adding noise and distortion "improve" the sound?

The proof of the pudding is in eating it - isn't it ? Or in this case, listening.
 
This is not something only me will attest to be true - you can go to the Best classical recording ever thread here on head-fi, download the entire Bach organ works in AAC256 I posted link there - for free - and listen to them straight and as converted DSD128 and higher - and THEN tell me what SOUNDS more realistic to you.
Stepping foot in a church and listening to some real organ may provide reality check after the CD redbook diet.
 
The worst legacy CD did inflict on public at large is the sterile black nothingness in sound that is NEVER present in live music. No performance heard live or recording in the studio is so "quiet" as it normally lands on the CD. That "noise and distortion" are actually filtered out by the steep filter , inherent to the CD redbook - there is nothing in nature filtering the content above 20 kHz ( 22050 Hz, to be precise ) - and these natural noises are responsible for the correct spatial localization - no matter how small in amplitude, removing them is detrimental to the realism obtained. You can add to that more than 16 bits required to capture the "noise" that is actually the part of the real music - more accurately so.
 
I am aware of the out of "audible" band increase of noise with DSD. For this reason, I believe letting out DSD64(SACD) was a BIG mistake. It is (too) poor - period.To get the "silence" above "audible band" required to be able to silence even the likes of hard core CD redbook supporter, DSD512 is required. DSD128 is the minimum I consider to be fit for musical reproduction - not totally quiet, not totally indistinguishable from the source (live mic feed ) - but much better than CD redbook.