Tone1 vs Tone2 Pro

I have both the Tone1 and the Tone2 Pro, so I pulled out REW and connected my Focusrite Scarlett 2i2 gen2 audio interface to see if I could measure any differences between the two. The 2i2 is far from the level of equipment over at ASR (and by Khadas) to measure performance, but I wanted to see what I could learn with what I have at hand.

First up is a frequency plot of the two DACs with an input sweep from 0 to 48 kHz and sampling rate of 44.1 kHz (sorry for the duplication if you’ve seen this in one of my other posts):

Note that the vertical scale is very tight. But it is interesting that we can see some differences, even if they are very small.

Next, let’s take a look at the step response:

Similar but, obviously, the Tone2 Pro inverts phase. Not something that you’re likely to notice, but I wonder if this is intentional. Can be easily solved using Roon’s DSP if anyone cares.

Next, we’ll look at distortion. I like to compare THD to the noise floor, which is admittedly high in my little 2i2 audio interface. But, as long as distortion stays near or below the noise floor, I consider that pretty good. Unfortunately, that’s not the case for the Tone2 Pro.

I thought this might just be an anomaly, so I checked one of my other measurements (48 kHz instead of 44.1), and I see the same thing.

Thankfully, once we get to 96 kHz and above, things look good.

Edit: The distortion I see for DSD64 is the same for all of the DACs that I have, so what you see below must be an issue with my format conversion for the DSD64 sweep. I may try something different tomorrow:

Based on these results, for folks who like using something like HQPlayer for upsampling, I’d recommend PCM 192 kHz for best results with the Tone2 Pro and DSD128 for the Tone1.

For more details, feel free to download my REW *.mdat files from here.

I’m very curious to know what other Tone1 / Tone2 Pro users are seeing and if Khadas is able to reproduce these results.

2 Likes

Another plot, here’s the Tone2 Pro at 44.1 kHz showing each component of distortion.

I’m curious to know what’s causing this and how audible it is. The more significant red line is 2nd order.

Hi.
●The 1st diagram shows that the cut-off frequency of Tone2 Pro is lower then Tone1,because the integrating capacitor value of Tone2Pro’ LPF is higher than Tone1.Thus it can weaken the high-frequency signal.
●The 2nd digram I want to know which signal did you test,RCA balanced?RCA Single end?4.4 JACK or 3.5 JACK? Two products may have 180 degree phase difference.
●The 3th digram I doubt there probably are some problems with the test way. I can show you the digram about the THD+N Radio Vs Frequency at 44.1KHz/48KHz/96KHz sample rate by the APx555.

3 Likes

For this diagram, I used the RCA single-ended output. I’m Looking forward to getting those balanced RCA to XLR cables in so that I can re-test. :slight_smile:

Yes, your APx555 is the proper way to test, I’m sure. I will re-run those tests today with a cleaner setup and see if I can get closer to your results. Thanks so much for having a look!

2 Likes

I made some improvements to my simple test rig today that seem to be giving me results that are much closer to yours (within the limits of my $149 audio interface). Thanks for having a look!

1 Like

Amir also observed poor THD at ‘low’ sample rates in his measurements and bizarre filtering behaviour:

1 Like

Yes. The Tone2 Pro does seem to benefit from upsampling for rates below 96 kHz. I’ve configured the following custom upsampling rules in Roon’s DSP:

44.1 kHz -> 176.4 kHz
48   kHz -> 192.  kHz
88.2 kHz -> 176.4 kHz
96   kHz -> 192   kHz

Everything else is left alone. Seems to result in the best sound.