Audient ID48 - Initial Transient Distortion

[jumper981]

Hi to everyone!
I recently bought an Audient ID48.
I found that in some conditions, the outputs will distort/clip on initial transient.
DAC outputs should have CS43198 despite Audient says “Sabre” on his site.

I’ve conducted some measurements and attached there.


I reproduced this issue in this way:

Ableton Live
Open a MIDI Channel with a good low frequency sine wave with VST (Serum 2 used here)
Attack Parameter relatively fast and sine wave phase starting from triggering.
Filter it with a steep low pass filter to clearly leave only the sine wave.
Do some fast triggering and you will find the distortion on triggering only.

Other method:
I’ve downloaded a high quality 36Hz sine wave with phase starting at 0deg.
Do some fade at both starting and ending.
Filter it same as before.
Distortion only on fade portion.

 

[Curvature]

In your attachment you have:

• Line OUT 1+2 into 1+2 Line Input (Sine Wave Download, Fade): Noise dominates over an extremely quiet 36Hz sine. See below
• Sine Wave Fade: There is literally no signal. Only 0s.

I suspect what's happening is that you are causing clipping by playing files with an abrupt start or by triggering the files on/off. There is nothing wrong with your interface.

DAC outputs should have CS43198 despite Audient says “Sabre” on his site.

This doesn't matter.

 

[jumper981]

Hi, thanks for your response.
I checked the files personally and left near 10mS start if I remember correctly.

All my other interfaces and DACs doesn’t exhibit this behavior with the same files.

I quoted the Cirrus Logic chips cause the distortion vs output level of this chip could be strange as posted here in a research from a user in AudioScienceReview.

I suspect something from the CS chip or the audio asio driver of the interface.

 

[Curvature]

You are using contrived, impossible signals to test your interface. They have nothing to with regular operation or music. It doesn't make sense to spend time this way.

 

[jumper981]

I noticed it while making music.
if I’m synthesizing a kick drum it make absolutely sense to use a sine wave in this way.
Despite this, how I use it it’s not an argument of this post.

 

[Curvature]

if I’m synthesizing a kick drum it make absolutely sense to use a sine wave in this way.

Even sines for kick drums need fade-ins/outs. This is a common issue and concerns all digital signals that start abruptly.

 

[jumper981]

As I said before, original audio playback have the right fades.
Also, those were starting at phase 0, so it’s not even necessary.

 

[Curvature]

I played back the first file very loudly and heard clipping. When looking at the waveform in the spectrogram I see a lot of spectral noise exceeding a very low-level sine, which means: clipping.

The second file contains literally nothing. No audio data.

File Summary
File Name : Sine_Wave_Fade.wav
sox WARN wav: wave header missing extended part of fmt chunk
Channels : 2
Sample Rate : 44100
Precision : 24-bit
Duration : 00:00:00.68 = 29975 samples = 50.9779 CDDA sectors
File Size : 240k
Bit Rate : 2.82M
Sample Encoding: 32-bit Floating Point PCM

File Detail
sox WARN wav: wave header missing extended part of fmt chunk
Samples read: 59950
Length (seconds): 0.679705
Scaled by: 2147483647.0
Maximum amplitude: 0.000000
Minimum amplitude: 0.000000
Midline amplitude: 0.000000
Mean norm: 0.000000
Mean amplitude: 0.000000
RMS amplitude: 0.000000
Maximum delta: 0.000000
Minimum delta: 0.000000
Mean delta: 0.000000
RMS delta: 0.000000

It is a problem to do with the digital domain, not your gear. How you are creating and processing the signal.

 

[jumper981]

Those file should be the recorded in/out of the interface. I think that I didn’t include the original sine wave.

 

[Curvature]

Post the original digital signal using a much longer sine wave (like 10 seconds) as well as the recorded version.