search

pierreaubert / spinorama

A library to display and compare spinorama (speakers measurements) graphs.
GNU General Public License v3.0
119 stars 12 forks source link

Neumi BS5 with/without EQ "tilt" is incorrect #244

Open Plompudu opened 1 year ago

Plompudu commented 1 year ago

Describe the bug When looking at the Neumi BS5 in-room-Target "tilt" it's -10db@20kHz instead of the normal tilt

To Reproduce Steps to reproduce the behavior:

  1. Go to https://www.spinorama.org/speakers/Neumi%20BS5/ErinsAudioCorner/index_eac-v2-20211212-ported.html
  2. Click on Ref vs EQ
  3. Scroll down to Estimated in room
  4. See error

Expected behavior The Target Should be identical for with and without EQ

Desktop (please complete the following information):

pierreaubert commented 1 year ago

Thanks. One of the autoEQ version didn't control the slope enough.

I am not sure the target should be the same. This one is likely bright in room. I ran the current optimiser and I got this:

filters_eq

Sounds better to me. WDYT?

If you are listening near field, then flat on axis is likely better.

filters_eq

Which one do you want?

pierreaubert commented 1 year ago

EQ for flat LW:

EQ for Neumi BS5 computed from ErinsAudioCorner data
Preference Score 2.93 with EQ 4.54
Generated from http://github.com/pierreaubert/spinorama/generate_peqs.py v0.23
Dated: 2023-07-07-12:06:04

Preamp: -2.8 dB

Filter  1: ON PK Fc    68 Hz Gain +2.65 dB Q 5.25
Filter  2: ON PK Fc   289 Hz Gain +2.50 dB Q 1.05
Filter  3: ON PK Fc   873 Hz Gain +2.94 dB Q 5.92
Filter  4: ON PK Fc  1180 Hz Gain -2.96 dB Q 2.78
Filter  5: ON PK Fc  1704 Hz Gain +2.30 dB Q 5.59
Filter  6: ON PK Fc  8766 Hz Gain +0.31 dB Q 3.19
Filter  7: ON PK Fc 12662 Hz Gain -1.76 dB Q 2.54

EQ for higher score (and flatter PIR)

EQ for Neumi BS5 computed from ErinsAudioCorner data
Preference Score 2.93 with EQ 5.39
Generated from http://github.com/pierreaubert/spinorama/generate_peqs.py v0.23
Dated: 2023-07-07-12:10:06

Preamp: -2.9 dB

Filter  1: ON LS Fc    68 Hz Gain +2.72 dB Q 1.05
Filter  2: ON PK Fc   237 Hz Gain +2.10 dB Q 1.51
Filter  3: ON PK Fc   447 Hz Gain +1.53 dB Q 0.73
Filter  4: ON PK Fc  1141 Hz Gain -2.70 dB Q 3.43
Filter  5: ON PK Fc  1594 Hz Gain +1.37 dB Q 4.58
Filter  6: ON PK Fc  5490 Hz Gain -1.43 dB Q 1.18
Filter  7: ON PK Fc 13092 Hz Gain -2.04 dB Q 1.16
Plompudu commented 1 year ago

Thanks!

I would prefer the flat on-axis. Could you do the same for the Sealed Measurement?