TANGZU BaJie EQ settings
Parametric EQ correction for the TANGZU BaJie, derived from measurements by Jaytiss. It moves the measured frequency response toward a neutral target.
How the BaJie sounds before EQ
Measured against a neutral target, the BaJie comes out warm. The low end sits ahead of the treble, which flatters bad recordings and rounds off detail.
The air region, openness at the very top, sits about 4.3 dB forward of neutral, so the very top sounds airy and open (measurements this high up vary between test rigs, so treat the exact number loosely). The presence region, vocal clarity and attack, sits about 1.9 dB behind neutral, so detail and clarity soften.
The largest single move in the correction is 5.5 dB around 20 kHz, on top of a broad tilt of up to 1.9 dB. That is not subtle. This model benefits clearly from correction.
These figures are read from the correction itself: whatever the EQ lifts, the headphone lacked, and whatever it cuts, the headphone had too much of. The target is the community's neutral reference, and how far above 10 kHz you can trust any measurement varies by rig. More warm models.
The correction curve
The solid line is the full measured correction (all 10 filters). The dashed line is what Atrium applies: the same shape at 65 percent strength, capped at 6 dB per band, so the headphone keeps its own character instead of being forced flat.
What the EQ fixes
Summed across every filter, these are the changes the correction actually makes, largest first:
- At the top of the range the correction cuts everything above roughly 12.4 kHz by up to 5.5 dB.
- Around 6.6 kHz the correction lifts the treble by 4.8 dB, over a narrow band about 0.3 octaves wide.
- Around 4.1 kHz the correction lifts the sibilance range by 4.3 dB, over a narrow band about 0.4 octaves wide.
The preamp of -4.8 dB offsets the boosts so nothing clips at full volume.
Parametric EQ settings
Preamp: -4.8 dB (headroom for the summed boost, not for any single row)
| # | Type | Frequency | Gain | Q |
|---|---|---|---|---|
| 1 | Low shelf | 105 Hz | -0.5 dB | 0.70 |
| 2 | Peaking | 237 Hz | -1.7 dB | 1.07 |
| 3 | Peaking | 4.13 kHz | +4.0 dB | 3.37 |
| 4 | Peaking | 6.57 kHz | +4.3 dB | 4.46 |
| 5 | Peaking | 1.32 kHz | -1.5 dB | 2.09 |
| 6 | High shelf | 10 kHz | -5.6 dB | 0.70 |
| 7 | Peaking | 10 kHz | +2.1 dB | 0.91 |
| 8 | Peaking | 73 Hz | +0.7 dB | 2.17 |
| 9 | Peaking | 35 Hz | -0.6 dB | 2.06 |
| 10 | Peaking | 5.31 kHz | -1.4 dB | 6.00 |
This is the published correction for the BaJie, unaltered, in AutoEq format. It pastes straight into Equalizer APO or Peace on Windows, and into most parametric EQ apps. Enter every row, including the ones that cut. These are the values to use: the numbers are not rounded off or simplified for the page, because changing them would change the result.
How to use these settings
Manually: enter the preamp and each filter into any parametric EQ. On Windows that is usually Equalizer APO with the Peace interface. On a Mac, eqMac or SoundSource. Set the type, frequency, gain and Q exactly as listed.
With Atrium: pick your BaJie in the app and it applies a tuned version of this correction: 65 percent strength with a 6 dB per-band cap, which fixes the rough spots without erasing what the headphone does well. Atrium also measures how you hear with a short listening test and layers your personal hearing profile on top. Two people with the same headphones do not hear the same sound. Headphone corrections are part of Signature Edition; the app itself is free.
BaJie vs other TANGZU models
Bass and treble are each model's average distance from the neutral target in that range. Positive means more than neutral, negative means less.
| Model | Signature | Bass vs neutral | Treble vs neutral |
|---|---|---|---|
| BaJie (this page) | warm | +0.7 dB | -1.4 dB |
| Fudu | bass-heavy | +2.6 dB | -2.1 dB |
| Princess Chang Le | bass-heavy | +2.9 dB | -1.9 dB |
| Shimin Li | dark | +0.5 dB | -2.6 dB |
| Wan'er 2 | mid-forward | -0.2 dB | -1.9 dB |
Headphones that sound similar
Closest matches across all 4,332 measured models, by distance between their deviation curves. Useful if you like how the BaJie is tuned and want alternatives.