Tone2 filters - objective and subjective difference

Hey everyone,

I have played around with the Tone2 for about two months now.
My first question is, how those filters differ from an objective viewpoint. What do they do with the audio signals? I found some information for the brick wall filter, but for the rest, no luck. Would love to hear some explanations or to see some references to explanations.

Having tried the different filters, I have a hard time to figure out which filter I like the most. I have tried 2 different headphones and 4 sets of amp/speakers combos. First impression: It is very difficult to hear any difference at all, and often I am not sure if I only imagine hearing a difference. Would love to read about your experience.

As for now I can say, that with one speaker set, the difference was tremendous. Switching to linear slow made this speakers “open up”, I felt more integrated into the experience. Have to ask the friend who owns them which pair it was. I think it was a pair of Edifier R1700BT. I think he also had some Edifier R1280DB, which also sounded better with a different filter, but the difference was not as apparent as with the R1700BT.

Then I have a set of custom made speakers with about 20 year old visaton drivers and an harman kardon HK 660 amplifier. I cannot hear the difference between the filters, but I felt like mini slow sounded the nicest on them. It sounded a bit more well rounded.

At last I tried the filters with a pair of Criterion TA T-160 with an Aura VA-50 amplifier (I think), which btw. sound marvelous together. Could not perceive any difference by using different filters.

I own 2 crappy headphones, one 50€ in-ear pair and a HyperX Cloud gaming headset. Could not make out any difference using different filters.

Generally I feel like the slow filters feel more “natural” and “warm”, but I think they are a little less precise. Brick Wall seems to be very precise, but I am not sure yet what the drawbacks are.

Looking forward to learn from your knowledge and experience.

1 Like

Personally I cannot comment any subjective feel here, and just share the information below FYI:

  • LINEAR FAST: Linear-phase fast roll-off filter
  • LINEAR SLOW: Linear-phase slow roll-off filter
  • MINI FAST: Minimum-phase fast roll-off filter
  • MINI SLOW: Minimum phase slow roll-off filter
  • APODIZING: Apodizing fast roll-off filter
  • HYBRID FAST: Hybrid fast roll-off filter
  • BRICK WALL: Brick-wall filter

Check the document below from ESS official website for further information:

Hope somebody from our community can comment here :blush:

Have fun!

1 Like

Link doesn’t work…

I think the filters are not working. See Tone 2 pro.: don't worlk the filters

1 Like

Nope, it works.

And actually, the software issue @Eric68 mentioned is setup is not memorable and once reboot the T2P will go back to the default Filter mode.

The ESS Filters page not available, check similar one here:

Good day!

I connected the Tone2 Pro to an audio interface (Focusrite Scarlett 2i2 gen2) and used Room EQ Wizard (REW) to measure the frequency response after cycling through each filter setting. Here’s what I found at 44.1 kHz (rate where the filters will show the greatest difference):

As you can see, the results are identical. I separated the curves a bit so that you could see them. If I did not, you would only see a single line!

To make sure that I was not doing something wrong, I used the same process to take similar measurements of the different filters offered by my S.M.S.L. Sanskrit 10th V2. Here’s what I saw there:

I was expecting to see something similar for the Tone2 Pro. Hopefully, after the next firmware update, we’ll get different results.

1 Like

I didn’t test filters for now, but what’s strange is that the test on ASR was showing all different filters working, with the exception that the curve was going too much high after the cutoff (-60dB) (may need a fix)
And why yours would be cut at 17kHz ?

Hi everyone, the firmware for the first batch of Tone2 Pro devices had a small problem. We have fixed the bug and have tested the white noise with different filters once again.

You can see the result below - it is better than before! This firmware will be released soon.


Really nice to know, question, can we be able to update using pure USB, or do we need an ST-LINK v2?

In case to use the ST-LINK is this one ok:


Hi for the first time, you’ll require an ST-Link v2 hardware tool to flash the MCU bootloader to the MCU. After you have applied the v1.2 MCU and XMOS update, only USB-C will be required to flash both the MCU and XMOS in the future.

Please refer to: Tone2 Pro - Firmware v1.2 + MCU Update


I’m curious to know how Amir got this to work when it does not on mine. Could be that my particular Tone2 Pro is defective? I’m 100% certain that I’m cycling through the filters properly (based on light colors, etc.).

As for the cut off, Amir’s dB scale is way more relaxed than mine. He’s showing you a dB range of 110 dB on the vertical axis while I’m showing only 1 dB. I’m zoomed in over 100x what he is because I’m interested in what’s happing within the audible band while he’s focused on what’s happening above 22.05 kHz.

Maybe the amirm version is different than yours?

1 Like

Mine is from the second batch (serial number starts with “11”). Perhaps the one he reviewed was from the first batch (serial number starting with “0C”). He reviewed it well before the third batch was released, so unlikely that it was one of those, unless he received a pre-release sample.

Not easy to identify the trace colours for the new filters …

I think the point was that this graph with the new firmware looks better than before. Previous firmware had more high frequency noise (around -60 dB):

With the new firmware, there is less high frequency noise (around -80 dB):


Perhaps it varied from unit to unit, but as I posted before, with v1.0 firmware, there was zero objective/subjective difference. After upgrading MCU and XMOS firmware, filter selection works as advertised.

~For now, Linear Fast is my favorite.~ :slight_smile:

Update: I’ve done more listening, and subjectively speaking, I actually prefer Hybrid Fast. According to my measurements, high frequencies are a little rolled off compared to the others, but I don’t think that explains what I’m hearing.

Perhaps it’s a time-domain thing? When I zoom in to the impulse response, I do see less pre-ringing for Hybrid Fast than the other six options.

It’s difficult to believe that I can hear a half ms or so of pre-ringing, but this sort of thing does not happen nature, so perhaps this explains my preference. Surprising since, looking at the frequency response, I was sure that I’d prefer the sound of Linear Fast.

For those who have the latest MCU and Firmware v1.2 (or later), I’m interested in any filter preferences you may have discovered.