• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Distortion from Software ?!

pwjazz

Addicted to Fun and Learning
Forum Donor
Joined
Sep 21, 2018
Messages
507
Likes
747
This will require some further digging, but I got a weird result just now that I need to share.

I was taking measurements to spot-check some EQ settings I've made for my LCD2C today. Measurements are taken on a MiniDSP E.A.R.S. with Room Eq Wizard, using the MiniDSP HEQ compensation (similar to Harman target). I'm playing the frequency sweep from Neutron Player on an LG V20, applying my EQ settings using its built-in DSP.

Lately I've been streaming from Neutron to gmediarenderer running on a Raspberry Pi (using pure Alsa as the audio subsystem). The Pi is connected via USB to a Topping NX4 DAC which is connected via fixed-level line-out to a JDS Labs Atom. I noticed a major rise in distortion in the treble which I'm not used to seeing with the LCD2C, and hadn't seen previously. Worried that it might be my new Atom, I repeated the measurement with the headphone out of the NX4 and got an almost identical result in the treble (bass and low mids vary somewhat, but they're pretty sensitive to room noise). On a whim, I then connected the LG V20 directly to the NX4 via USB (bypassing DLNA). To my surprise, on both the NX4 and Atom, the treble distortion rise disappeared! Distortion in bass and midrange looks similar across the board, they don't diverge until we get to over 3KHz, which is where things go off the rails.

I always assumed that DLNA to gmediarenderer would be bit perfect. Perhaps I'm wrong? But even if there's some resampling going on, that's a huge amount of distortion.

If anyone has any insights into this, they would be much appreciated.

amp_distortion.jpg
 
OP
pwjazz

pwjazz

Addicted to Fun and Learning
Forum Donor
Joined
Sep 21, 2018
Messages
507
Likes
747
I just measured Neutron -> Chromecast Audio -> Atom and it does NOT show the treble distortion rise, so I don't think it's a problem on Neutron's end anyway.
 
OP
pwjazz

pwjazz

Addicted to Fun and Learning
Forum Donor
Joined
Sep 21, 2018
Messages
507
Likes
747
Mystery solved (sort of). I set up another Pi with a Hifiberry DAC+ connected via I2S. I ran gmediarenderer on there as well, and it does NOT exhibit the treble distortion rise. So, there's something in the interaction of ALSA, USB and the NX4 that causes the issue. I have no clue what it would be though.
 
Top Bottom