New set of graphs for all speakers reviews

[sweetchaos]

More errata:

#1. IK multimedia has 2 models are they are

IK Multimedia iLoud MTM (on the left)
IK Multimedia iLoud Micro Monitor (on the right)

Spoiler

It's currently showing:

Spoiler

left picture is wrong
right picture is correct

#2. Please remove photos that have 2 speakers inside the one photo.
It's not consistent, as you most of the photos have a single speaker.

Spoiler

#3. What defines the 'quality' dropdown?
High, medium, low, unknown.

#4. When I selected 'others' in the dropdown:

Spoiler

I'm seeing results like:
Google Nest Audio => Napilopez (which shouldn't be visible, since Napilopez will come up with the napilopez dropdown).

Overall, the dropdown is not working perfectly.

#5.When I select 'audioholics' from dropdown:

Spoiler

Under the speaker photo, it shows 'misc'.
Since Audioholics did actually review the speaker, it shouldn't be 'misc', but 'audioholics' to be consistent.

 

[CJH]

Can we add EQs for LS50 Meta?
CJH

 

[pierre]

More errata:

#1. IK multimedia has 2 models are they are

IK Multimedia iLoud MTM (on the left)
IK Multimedia iLoud Micro Monitor (on the right)

fixed

#2. Please remove photos that have 2 speakers inside the one photo.
It's not consistent, as you most of the photos have a single speaker.

I agree but not a priority.

#3. What defines the 'quality' dropdown?
High, medium, low, unknown.

1. High: data from a klippel and not smoothed (asr+eac in practice)
2. Medium: spin from a big anechoic room and unsmoothed (most of Harman spin)
3. Low: Everything else (Princeton graphs, Audioholics, Pseudo anechoic ...)

#4. When I selected 'others' in the dropdown:
I'm seeing results like:
Google Nest Audio => Napilopez (which shouldn't be visible, since Napilopez will come up with the napilopez dropdown).
Overall, the dropdown is not working perfectly.

fixed

#5.When I select 'audioholics' from dropdown:
Under the speaker photo, it shows 'misc'.
Since Audioholics did actually review the speaker, it shouldn't be 'misc', but 'audioholics' to be consistent.

[/QUOTE]

done too.

 

[pierre]

Can we add EQs for LS50 Meta?
CJH

EQ for KEF LS50 Meta computed from ASR data
Generated from http://github.com/pierreaubert/spinorama/generate_peqs.py v0.7
Dated: 2021-05-07-13:43:21

Preamp: -1.5 dB

Filter  1: ON PK Fc  5159 Hz Gain -1.57 dB Q 1.27
Filter  2: ON PK Fc 11675 Hz Gain -1.16 dB Q 1.07
Filter  3: ON PK Fc  2275 Hz Gain +1.52 dB Q 1.84
Filter  4: ON PK Fc   825 Hz Gain +0.67 dB Q 12.00
Filter  5: ON PK Fc   969 Hz Gain -0.77 dB Q 7.98
Filter  6: ON PK Fc  4091 Hz Gain -0.88 dB Q 12.00
Filter  7: ON PK Fc 11026 Hz Gain +0.72 dB Q 12.00
Filter  8: ON PK Fc  4857 Hz Gain +0.50 dB Q 8.03
Filter  9: ON PK Fc   710 Hz Gain -0.59 dB Q 12.00
Filter 10: ON PK Fc  1375 Hz Gain -0.62 dB Q 11.67
Filter 11: ON PK Fc  8473 Hz Gain -0.40 dB Q 2.57
Filter 12: ON PK Fc   458 Hz Gain -0.66 dB Q 6.93
Filter 13: ON PK Fc   564 Hz Gain +0.58 dB Q 11.70
Filter 14: ON PK Fc   835 Hz Gain +0.41 dB Q 12.00
Filter 15: ON PK Fc  1702 Hz Gain +0.68 dB Q 8.04
Filter 16: ON PK Fc  5646 Hz Gain +0.26 dB Q 12.00
Filter 17: ON PK Fc   353 Hz Gain -0.45 dB Q 6.76
Filter 18: ON PK Fc  6952 Hz Gain -0.51 dB Q 12.00
Filter 19: ON PK Fc  1109 Hz Gain +0.31 dB Q 12.00
Filter 20: ON PK Fc  1467 Hz Gain -0.34 dB Q 12.00

I need to investigate why it doesn't show up on the website.

 

[sweetchaos]

I'm loving the new improvements and bug fixes...

I'm having a difficult time finding further things to complaint about...haha

Thanks for the effort!

 

[SDX-LV]

I don't mean to be pedantic, but I want to remind everyone that Klippel NFS data is anechoic. What's being referred to as 'anechoic' in this thread are anechoic chambers, which are not fully anechoic in the entire bass range.

I suggest that the quality rankings look like this:
Klippel NFS > Anechoic chamber > Quasi-anechoic > Unknown > Vendor

May I suggest:

Klippel NFS > Anechoic chamber (virtually no spins like this) > Vendor data (Anechoic or NFS) > Free field (Audioholics) > Quasi-anechoic > Unknown (I wonder where you can find that?)

 

[sweetchaos]

Few more...
1.

All 4 of the "EQ links" on the right-hand side, go to the same link:
https://raw.githubusercontent.com/pierreaubert/spinorama/develop/datas/eq/Neumann KH 80/iir-autoeq.txt
Which is the "asr-v3-20200711".

Which means the EQ links for the other 3 should either be:
- removed, since they don't correspond to the correct review
- OR if you're going to run the EQ algorithm against the other 3, then those EQ links would need to be updated.

2.
https://raw.githubusercontent.com/pierreaubert/spinorama/develop/datas/eq/JBL 305P Mark ii/iir-autoeq.txt
should say "computed from Erin data", not "ASR data".

3. NHT Pro M-00
Review by Amir
Shows a blank page
https://pierreaubert.github.io/spinorama/NHT Pro M-00/ASR/index_asr.html
Is it because Amir never released the spin data?

4.
I noticed some of your older graphs scale doesnt go down to 20hz.
For consistency, I think it's a good idea to regenerate those.
Example, goes down to 500hz.
https://pierreaubert.github.io/spinorama/JBL PRX635/Princeton/index_princeton.html

5.
The following speakers are active (not passive, as are currently classified):
Adam T7V
Devialet Phantom Reactor 900'
Fluance Ai60
Google Nest Audio
JBL LSR305
JBL LSR308
JBL One Series 104 (it's both passive and active, but the active measurements are a lot better).
Presonus Eris E8 XT
Sonos Move
Vanatoo Transparent Zero

6.
The following speakers are passive (not active, as are currently classified):
NHT SB2
Neumi BS5
OSD Audio AP650

7.
GGNTKT M1 is missing the source link: https://ggntkt.de/model-m1/technische-daten/

8.
Missing Erin's reviews on ASR:
Revel PerformaBe F226Be https://www.audiosciencereview.com/...be-f226be-floorstanding-speaker-review.16702/
Klipsch The Fives https://www.audiosciencereview.com/...fives-powered-bookshelf-speaker-review.22892/
Selah Audio Purezza https://www.audiosciencereview.com/forum/index.php?threads/selah-audio-purezza-review.15850/
Philharmonic BMR https://www.audiosciencereview.com/forum/index.php?threads/philharmonic-bmr-speaker-review.14781/
Neumi BS5 https://www.audiosciencereview.com/forum/index.php?threads/neumi-bs5-bookshelf-speaker-review.14404/
JBL HDI-4500 https://www.audiosciencereview.com/...hdi-4500-center-channel-speaker-review.22025/
Jamo S807 https://www.audiosciencereview.com/...amo-s807-floorstanding-speakers-review.14314/
Micca MB42X Mark III https://www.audiosciencereview.com/...b42x-mark-iii-bookshelf-speaker-review.15307/
Bose 901 Series V https://www.audiosciencereview.com/forum/index.php?threads/bose-901-series-v-speaker-review.14865/
Klipsch Heresy IV https://www.audiosciencereview.com/forum/index.php?threads/klipsch-heresy-iv-speaker-review.17853/

 

[pierre]

Few more...
1.

View attachment 129054

All 4 of the "EQ links" on the right-hand side, go to the same link:
https://raw.githubusercontent.com/pierreaubert/spinorama/develop/datas/eq/Neumann KH 80/iir-autoeq.txt
Which is the "asr-v3-20200711".

Which means the EQ links for the other 3 should either be:
- removed, since they don't correspond to the correct review
- OR if you're going to run the EQ algorithm against the other 3, then those EQ links would need to be updated.

fix: only print eq/score for best data

2.
https://raw.githubusercontent.com/pierreaubert/spinorama/develop/datas/eq/JBL 305P Mark ii/iir-autoeq.txt
should say "computed from Erin data", not "ASR data".

fix: put correct origin in eq documentation

3. NHT Pro M-00
Review by Amir
Shows a blank page
https://pierreaubert.github.io/spinorama/NHT Pro M-00/ASR/index_asr.html
Is it because Amir never released the spin data?

yes

4.
I noticed some of your older graphs scale doesnt go down to 20hz.
For consistency, I think it's a good idea to regenerate those.
Example, goes down to 500hz.
https://pierreaubert.github.io/spinorama/JBL PRX635/Princeton/index_princeton.html

data from Princeton starts at 500hz.

5.
The following speakers are active (not passive, as are currently classified):
Adam T7V
Devialet Phantom Reactor 900'
Fluance Ai60
Google Nest Audio
JBL LSR305
JBL LSR308
JBL One Series 104 (it's both passive and active, but the active measurements are a lot better).
Presonus Eris E8 XT
Sonos Move
Vanatoo Transparent Zero

6.
The following speakers are passive (not active, as are currently classified):
NHT SB2
Neumi BS5
OSD Audio AP650

7.
GGNTKT M1 is missing the source link: https://ggntkt.de/model-m1/technische-daten/

8.
Missing Erin's reviews on ASR:
Revel PerformaBe F226Be https://www.audiosciencereview.com/...be-f226be-floorstanding-speaker-review.16702/
Klipsch The Fives https://www.audiosciencereview.com/...fives-powered-bookshelf-speaker-review.22892/
Selah Audio Purezza https://www.audiosciencereview.com/forum/index.php?threads/selah-audio-purezza-review.15850/
Philharmonic BMR https://www.audiosciencereview.com/forum/index.php?threads/philharmonic-bmr-speaker-review.14781/
Neumi BS5 https://www.audiosciencereview.com/forum/index.php?threads/neumi-bs5-bookshelf-speaker-review.14404/
JBL HDI-4500 https://www.audiosciencereview.com/...hdi-4500-center-channel-speaker-review.22025/
Jamo S807 https://www.audiosciencereview.com/...amo-s807-floorstanding-speakers-review.14314/
Micca MB42X Mark III https://www.audiosciencereview.com/...b42x-mark-iii-bookshelf-speaker-review.15307/
Bose 901 Series V https://www.audiosciencereview.com/forum/index.php?threads/bose-901-series-v-speaker-review.14865/
Klipsch Heresy IV https://www.audiosciencereview.com/forum/index.php?threads/klipsch-heresy-iv-speaker-review.17853/

fix: numerous metadata errors

Note: you can easily correct the metadata directly if you know a tiny bit of git and a text editor. Pull requests are welcome.
Edit https://github.com/pierreaubert/spinorama/blob/develop/datas/metadata.py directly in your browser or in your favorite editor, save and submit a pull request.
There is a check_meta.py python file that checks if your changes are valid but that's not required.

Thanks for the feedback.
Pierre

 

[pierre]

Thanks to the hard work of speaker reviewers and graphs manually digitalised by motivated contributors,
we now have 300 speakers (355 measurements, 286 EQs) in the database!

Thanks you @amirm, @hardisj, @sweetchaos!

 

[sweetchaos]

Found the NHT Pro M-100 spinorama
https://www.audiosciencereview.com/...equivalent-sinad-discussion.10818/post-421708

 

[TLEDDY]

I have attempted to look through this thread. The contributors operate at an intellectual level I find amazing!

Thank you for your hard, time-consuming work... Reminds me of peer review in my clinical laboratory field.

Tillman

 

[MZKM]

JBL LRS25P Spinorama in vector quality.
https://jblpro.com/en/site_elements/lsr25p-owners-manual

MSRP was $960/pair (so well over $1000 with inflation) and it shows.

________
JBL LRS32 & LRS28P Spinoramas in scanned quality:
https://jblpro.com/en/site_elements/lsr-series-owner-s-manual

 

[flipflop]

Now that @sweetchaos has run out of things to complain about, I'm going to list a few myself

• The pictures on the home page are too large, which makes them look intimidating, and the quality is also being compromised by the upscaling. I realize that the responsive design of the website makes the page look different depending on the resolution of the device its being viewed on. I've attached a screenshot of how it looks on my 1440p monitor.
• When selecting a speaker, I would appreciate if the spinorama pages would show 'With EQ' spinoramas for both EQs.
  
  
• On the EQs page, I'd like to see 'autoEQ' be renamed 'Automatic' (so that it doesn't get confused with jaakkopasanen's AutoEQ project) and 'flipflopEQ' be renamed 'Manual' (so that other users don't get the impression that I'm the only one who can contribute with EQs to the site).
• The EQs page should also have a 'information box' at the top, instructing people on which EQ to pick. It could read something along the lines of:

  Manual EQ provides the highest fidelity but requires the full set of 20 filters. The filters are sorted by frequency, from lowest to highest.
  Automatic EQ allows you to use any number of filters. The filters are sorted by improvements to fidelity, from highest to lowest.
  Both adjust frequencies from 300 Hz and up only. Additional room EQ is advised.

  
  
  
• The 'Score w/EQ' column (on the Scores page) shows both manual and automated EQ scores, but the '(eq)' only links to the automated ones.
• I would like to see a 'Score w/EQ+w/Sub' column added to this page.
  
  
• On the Statistics page, the 'Score per speaker.' and 'Score per speaker with an extra perfect subwoofer.' graphs need to have their y-axes labelled. It took me a while to figure out the speakers are sorted alphabetically.
• Having the scores being represented by bars rather than dots could make the graphs easier to read.

 

[SDX-LV]

JBL LRS25P Spinorama in vector quality.
https://jblpro.com/en/site_elements/lsr25p-owners-manual

Nice find on the vector quality. I only saw the scanned quality previously.

For these 20 year old models that are 3+ generations obsolete it is certainly not the highest priority to get all the data published So this is hardcore completionist territory.

 

[SDX-LV]

@pierre & @sweetchaos - you are both killing it I was not expecting that someone will digitize all Spinorama plots known to the internet!

One improvement suggestion that jumps at me: For quite a few speakers where the data is taken from speakerdata2034.blogspot.com the source of the data is entered quite wrong: for example, there is no such source as "Speakerdata2034" and "Vendors-Bowers & Wilkins" is also very wrong.

The correct source of data can be seen in the Spinorama Index table in column H: https://docs.google.com/spreadsheet...t-ckjJ9BQf9N9dGSUYZKAptY4_Jzxh/pub?output=ods

Great work!!!

 

[sweetchaos]

Digitized latest Audioholics\Perlisten S7t Tower.

 

[pierre]

Found the NHT Pro M-100 spinorama
https://www.audiosciencereview.com/...equivalent-sinad-discussion.10818/post-421708

Done. Partial data but better that what we had before.
- allow symmetry for partial measurements in klippel format
- add more complete data for NHT Pro M-00

 

[pierre]

Digitized latest Audioholics\Perlisten S7t Tower.

Hello I merged the data with the ground plane measurements. Score is 6.3 which is surprisingly low.
Confidence in the data is also very low.

 

[sweetchaos]

Now that we have all 3 Revel FxxxBe series measurements...
Amir made a good point here about rear-ported speakers (like the F328Be) that will have more output at the back of the speaker, when compared to other series like F228Be (that are front-ported).


F226Be:

F228Be:

F328Be:

As you can see from the above, F328Be's results for -3db and -6db point are worse than F226Be....which of course we know is not true.

Not sure what can be done about this, but thought I should mention it.
Hopefully I'm not misunderstanding it.

 

[richard12511]

Now that we have all 3 Revel FxxxBe series measurements...
Amir made a good point here about rear-ported speakers (like the F328Be) that will have more output at the back of the speaker, when compared to other series like F228Be (that are front-ported).


F226Be:
View attachment 131792

F228Be:
View attachment 131793

F328Be:
View attachment 131794

As you can see from the above, F328Be's results for -3db and -6db point are worse than F226Be....which of course we know is not true.

Not sure what can be done about this, but thought I should mention it.
Hopefully I'm not misunderstanding it.

Is that using the sound power curve to determine the -3 and -6 points? The SP curve extension should be closer to correct for those big rear ported speakers(from what I understand).