Some good news on the audio front


#1

I bought the Hub hoping to play some hi-def audio (24/96) over the optical output to an external DAC.  I was really dissapointed when my external DAC indicated that the files were being down-converted to 48k.

More important to me was whether it was still outputing the 24 bit depth.  I couldn’t get any response clarifying this one way or the other, but I did notice that the 24/96 files did sound very good even when down-converted to 48k.

Tonight I tried to come up with a way of determining if the 24 bit depth was retained.  What I decided to do was create a test file at 24/96 containing a test tone recorded at -110dB.  If the audio were being truncated to 16 bits which has a noise floor of 96dB then the test tone would be lost in the noise.

My recording setup has a noise floor of about -80dB (noisy preamp) so in order to make the file, I recorded a 1kHz test tone at -60dB… 20dB above the noise floor and then used the amplify effect from Audacity to amplify the signal by -50dB.

This produced a test file with the test tone at app. -110dB and the appearant noise floor at -130dB.

I loaded the test file on to the Hub and took the fibre output to the SPIDF optical input on my Mac.  I recorded the output of the Hub using Audacity again at 24/48 (since the Mac indicated that the audio was being down-converted to 48k).

When I looked at the recorded file, there was my 1kHz test tone sitting at -110dB and the noise floor showing at -130.

I think this clearly indicates that the Hub will put out 24/48 which is great news since I am more interested in getting the best dynamic range and lowest quantization errors from my music.


#2

I’m glad its clear to you but you lost me at the numerous noise floors. :wink:


#3

Ya, I think I made it pretty confusing.  I was pumped as I was writing and ramling on a bit.

I basically applied a test tone to the analogue input of my M-Audio FastTrack Pro USB audio interface and recorded it at 24/96 using Audacity.

The test tone was at -60dB and noise floor was at -80dB (noise floor of the M-Audio interface).

I them applied a negative gain to the recording (-50dB) so that the levels were reduced by 50dB so the test tone went down from -60 (-50)  to -110dB and the noise floor went from -80 (-50) to -130dB.

These levels can’t exist in a 16bit file because the absolute minium signal 16 bit can have is -96dB.

Since I saw the test tone at -110 and the noise floor at -130 the playback from the Hub must have been 24bit.

I’m convinced enough that I am now saving my audio (vinyl transfers) at 24/48 to play on the Hub.