24bit 96khz encoding

is it currently possible to encode to that yet, or do we have to have te new sound cards?

There are soundcards available from Terratek which can record 24/96, but they don’t work properly with some VIA chipsets :frowning:

that blows.

No, it’s actually quite funny. The lower end boards use the ICEnsemble Envy24 controller which is now owned by VIA…

Originally posted by kwkard
is it currently possible to encode to that yet, or do we have to have te new sound cards?

What would this 24bit 96khz encoding do for you? and what are the current standards? Sorry for the newb question :slight_smile:
JR

higher fidelity audio. If you’re an audiophile, then you’ll love this. It sounds much better because of a higher sampling rate.

BTW as long as sound cards work with analogue signals, 20 bit precision is the most what is possible. 24 bit is fake, the last 4 bits are merely “guessed”, but never measured.

The Terratec EWX24/96 works just fine on my Abit KT7 (VIA KT133/686A). However, if you don’t have the PCI and AGP timings set properly, problems will show up. (The SBLive! specific settings do not work properly, for instance.)

Once you get a 24/96 file, you can compress it nicely using Monkey’s Audio.

Alexnoe is also correct about the resolution… but having a card that can actually operate accurately at just 16+ bits (instead of the full 24) will still give you the full 96dB CD dynamic range and then some. :slight_smile:

Originally posted by kwkard
higher fidelity audio. If you’re an audiophile, then you’ll love this. It sounds much better because of a higher sampling rate.

No one ever proved he could hear the difference between 48 kHz 24 bits and 96 kHz 24 bits in a blind ABX test. Even between 48/16 and 96/24 there is quite no difference.

In theory higher sample rate can’t bring anything to the sound, since only inaudible frequencies are added. It might even introduce distortion, since audio equipment doesn’t support such frequencies and can intermodulate.

In practice, it remains an open question until we get more ABX results.

I tried recording an old full analog vinyl (Sandra 10/10) in 44.1/16 and 96/24 with the Marian Marc 2 soundcard.
The 24/96 recording seemed immediately to get all the analog definition of the vinyl, while the 44.1/16 one was sounding “digital”.
Once the samples fed into the abx, all the differences vanished. I again listened to them in CoolEdit, where they sounded different, and had to admit that I was just hearing what I wanted to hear : there was actually not the slightest audible difference between the sound of each file. I was just paying more attention to the texture of the sound listening to the 96/24 files, and to its harshness listening to the 44.1/16 one.

So for me, I started convinced that 96/24 sounded better, I could hear it, then the ABX opened my ears, and proved me there was no difference between my recordings.

EDIT : ABX is a way of performing blind tests. A and B are the tested devices, and are visible. X is randomly chosen between the two and is hidden. The listener can listen to a, b, and x at will, and must guess if x is a or if x is b.
If he succeeds 8 times out of 8, there are 255 chances out of 256 that he actually hears the difference, and 1 chance out of 256 that he just guessed right.
With 16 sessions, the probability of guessing is 1/65536