if you're featuring these numbers have a reality of sound to the degree that the numbers suggest per se, you're barking up the wrong tree.

what's crucial for a samples library developer is not necessary for the end user of the samples. one may even find some information to suggest that the nyquist limit isn't abolutely the limit of hearing, but certainly 96khz is far far above what anyone hears. that sort of criteria is a matter of headroom and ceiling room for the process of recording and editing in development.

but those of us that actually use the samples need the performance to be high, and 96k is rather too taxing for the CPU for most people's workflow I think.

"not even half the quality" is an absurd conclusion. And I thought you were asking. Now you're pronouncing.

 Very few studios work at 96KHz, because there simply is no gain in quality by trying to record frequencies that your microphone, mixer, speakers and your ear cannot process. 

Most studios work at 24-bit/44.1KHz or maybe at 48KHz for film work.  96KHz is generally only used for the sake of saying that it was used.

The actual quality of any recording nowadays has more to do with the final D/A used to listen to the music, not the sampling frequency.

@tfkbucncv said:

Hello I work only in 24 BIT 96 KHZ and wanted to know what format your packages work that I saw they were working on 44 KHZ and it does not format my work to what you do not put the high quality raw material that is very important. ?

What sample libraries do you currently use then?

DG

@tfkbucncv said:

I wanted to know if for example I buy a package SPEICAL EDITION STRING Can I get the raw material inside the package at a frequency 96 KHZ frequency course if you tape this what the original frequency of the sampling before conversion to 44 ? /

Simple answer. No you can't.

DG

@tfkbucncv said:

no true does hear the difference while you have a powerful computer I year ago I moved to 96 KHZ than 44 and the difference is very significant sound more deep, sharp, clear, I would never go back to 44 the difference is very significant, even after you finish the mix and gives the song for distribution in 44 KHZ clarity and sharpness continues to be done two experiments recorded once on 44 KHZ and I mix and once on 96 KHZ and converted to 44 KHZ after mixing the difference is very significant, no doubt, at 44 KHZ still indulge in that sort of clarity and depth recording of the 96 KHZ! .

It's not the 44.1KHz vs 96KHz difference you are hearing,  it is what must be the poor quality conversion between the two frequencies.

What software did you use to convert the 96KHz to 44.1KHz ?    And what D/A converter are you using to hear the sound?

Welcome tfkbucncv,

@tfkbucncv said:

[...] I saw they were working on 44 KHZ [...]

Vienna Instruments will work in any typical sampling frequency, and of course 96 kHz is amongst the supported values. The built-in real-time sample rate conversion will take care for that.

The instruments' samples themselves are recorded in 96 kHz, 24 bit. They are stored and edited in 96 kHz, 32 bit FP. -> http://www.vsl.co.at/en/65/72/103/20.vsl

They are converted to the internal delivery format of 44.1 kHz, 24 bit in a final step before mapping, using a proptietary hi-end tool that was developed specifically for VSL. This format is widely regarded as the best compromise between sound-quality, file-size and CPU/RAM-requirements.

The samples are not accessible individually, but delivered in a data-compressed, monolithical format. Data gets un-compressed in real-time by the Vienna Instrument Player software.

HTH,

The argument for high SR recording is that it pushes the brick wall filter's ringing well out of the audible range.

That doesn't mean it's worth losing half your performance when working with samples - as Dietz says - but you don't have to be delusional to want to work at 96k.

Dan Lavry (who might have to say a thing or two about digital audio in general and converters in particular) offers an even more differentiated POV. In his famous White Paper from 2004 he discusses the benefits of high sampling rates and comes to the conclusion:

"Pro:

Easier filter
Overcome Sinc problem

Con:

Reduced accuracy [!]
Significant increase in data files size
Significant increase in processing power required"

Assuming I understood Mr. Lavry correctly, he thinks that a properly built converter would ideally be operated around 65 kHz sampling rate.

Kind regards,

A high quality converter is key at 44.1