• 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!

Some measurements of IDT 92HD93 headset out in Dell Latitude E6330

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,397
Likes
3,351
Location
.de
I've been meaning to do this forever now.

While the combo of Focusrite Forte and RMAA (both a bit flaky) was subjectively causing grey hair to sprout at an alarming rate at first, things ultimately went reasonably smoothly. Signal chain:
Battery-powered Latitude E6330 (w/ Win10 22H2, Win 7/8.1 driver as supplied by Dell, MME) headset out --> TRS --> 2x XLRm (Sommer Cable Basic HBA-3SM2-0150) --> 2x XLRf --> TRS (Cordial CFM 0.6 FV) --> Focusrite Forte Line-In (+3 dB in input gain gives a calibration level of -1.0 to -1.4 dB at full output depending on sample rate, or +9 dB at output 67 = -6 dB; this points to a maximum output level of about +5 dBu or 1.4ish V)

E6330-rmaa-summary.png

Full set of results

The most shocking part of all this is that I found precious little to be glaringly wrong. There is almost as much dynamic range as you could ever want for a line-level source, even if with the usual rise in noise towards the low end as typical for consumer sound chips, and the noise floor rises by a moderate ~3 dB in THD vs. DR tests. Distortion is low at 1 kHz even if IMD does degrade towards the treble, again nothing unusual. (Estimated THD+N about -90 dB at 1 kHz.) Playback speed appears to be accurate to within 25 ppm (0.0025%). Periodic filter ripple is nothing much to write home about at about +/- 0.05 dB with roughly a 6 kHz period (Julian Dunn would not be proud), but even that is nothing I would call terrible. Is that really enough for my ears to complain that the treble ain't right?

44.1 kHz plain vanilla:
E6330-44k-fr.png


Same if upsampled to 192 kHz via Windows audio stack (Windows audio stack resampling has come a long way...):
E6330-44k@192k-fr.png

She's a beaut! :D

The difference:
E6330-44k-filter-ripple-subtract44k@192k.png


48 kHz plain vanilla:
E6330-48k-fr.png


48k upsampled to 192k:
E6330-48k@192k-fr.png


The difference at 48 kHz:
E6330-48k-filter-ripple-subtract48k@192k.png


(And yes, I subtracted both 44.1 and 48 kHz graphs in the same order, so yes, the difference is real!)

The designers clearly did not give a hoot about hi-res audio for bats, with filters becoming increasingly relaxed until 192 kHz looks like this:
E6330-192k-fr.png


So this is not the one to use with software stereo MPX for your pirate FM radio station - not like anyone would, mind you...
 
Last edited:
Top Bottom