spinorama.org

[pierre]

FYI, Theory Audio has Klippel NFS measurements posted on their site.

Theory Audio Design

SURFACE MOUNT LOUDSPEAKERS

www.theoryaudiodesign.com

thanks that's nice. Do you see a file with data I could download? Or maybe you can ask them to share the data?

 

[pierre]

Hello,

you may have notice that I changed the URL for the site. It used to be on github, hosted for free but I hit the limit in terms of size, size of files, size of git pack etc too many times and it became really complicated to deal with the limits.

I chose the easy solution: self hosting.

I bought the spinorama.org domain that was not used for the next 10 years.

So now:

A collection of loudspeakers measurements

www.spinorama.org

is the main site
https://dev.spinorama.org is the dev site (which may or may not work)

The site is hosted on 1 machine at the moment but soon 2 in 2 different places and I have a lot more bandwidth at the second one.
Load balancing will be done by the DNS so you may experience some outages when IPs change or when one server is rebooting.
Uptime will not be as good at github but should be good enough (99% is likely).

I was surprised to see that the audience is much higher that what I thought: 10k visitors per day from all over the world (after removing the bloody crawlers). So thank you all!

 

[beaRA]

thanks that's nice. Do you see a file with data I could download? Or maybe you can ask them to share the data?

Ah sorry, I'm not aware of an available data file. I wasn't sure if you were doing some chart digitizing from images to gather manufactacturer data sources.

 

[pierre]

Ah sorry, I'm not aware of an available data file. I wasn't sure if you were doing some chart digitizing from images to gather manufactacturer data sources.

i sometimes digitalized the spin (takes 10 minutes) but I am too lazy to do the 72 measurements when they are available.

 

[jtk9412]

I appreciate your effort and thanks for the work overall. Whenever I check Edifier MR4's EQ, I see an updated version constantly and the sound is getting better each update.

 

[jtk9412]

@pierre After applying your Edifier MR4 EQ settings, should I use the speakers on monitor mode or the music mode? Which option would give a better and flatter result with your EQ settings?

 

[firestart3r]

Hello ! This tool is absolutly amazing, great work!

I have a UX suggestion:
- Could it be possible to select the line or lines we want to see in the graph ?

For example, when I compare two speakers, I would like to be able to see the "on axis of speaker A" and "on axis of speaker B".
Or "on axis speaker A" + "on axis speaker B" + "sound power di speaker A" + "sound power di speaker B"
Or "on axis speaker A" + "sound power di speaker B" (even if doesn't make sense).

The reason is that I have difficulty comprehending the big picture of the multiples lines displayed at the same time...
It could be a nice way to "unclutter" the graph for newbies like me. What do you think ?

Thank you

 

[CJH]

Pierre:
Just wanted to thank you for the continued improvements you have made in your EQ algorithm along with updating your website. My KEF LS50 Metas have never sounded better. Thanks.
CJH

 

[jtk9412]

Which one should I use? @pierre

 

[sweetchaos]

AutomaticEQ (which are PEQs) are the most accurate.

But since not every hardware supports PEQs, you’ll likely use GEQs in those cases, which are less accurate but better than nothing.

For example:
-VLC media player doesn’t support PEQs, but it supports 31 band GEQs.
-Equalizer APO (for Windows) supports PEQs.
-MiniDSP hardware supports PEQs.

Most people should be using PEQs since most modern hardware supports it.

 

[pierre]

Hello ! This tool is absolutly amazing, great work!

I have a UX suggestion:
- Could it be possible to select the line or lines we want to see in the graph ?

For example, when I compare two speakers, I would like to be able to see the "on axis of speaker A" and "on axis of speaker B".
Or "on axis speaker A" + "on axis speaker B" + "sound power di speaker A" + "sound power di speaker B"
Or "on axis speaker A" + "sound power di speaker B" (even if doesn't make sense).

The reason is that I have difficulty comprehending the big picture of the multiples lines displayed at the same time...
It could be a nice way to "unclutter" the graph for newbies like me. What do you think ?

Thank you

You can add/remove a graph by clicking in the legend. Each click toggle between visible/hidden states.

 

[pierre]

AutomaticEQ (which are PEQs) are the most accurate.

But since not every hardware supports PEQs, you’ll likely use GEQs in those cases, which are less accurate but better than nothing.

For example:
-VLC media player doesn’t support PEQs, but it supports 31 band GEQs.
-Equalizer APO (for Windows) supports PEQs.
-MiniDSP hardware supports PEQs.

Most people should be using PEQs since most modern hardware supports it.

True. At the same time the differences are not so large. Here is an example with score being 7.23 with PEQ and 7.13 with GEQ, the difference 0.1 is not statistically significant. Honestly a basic GEQ like the DBX 231s for ~ 150$ does a good enough job.

 

[firestart3r]

You can add/remove a graph by clicking in the legend. Each click toggle between visible/hidden states.

Amazing! Thank you Pierre

 

[tiago.trindade]

Hello,

Can we export the measurements data to use it as a “target/house curve” in REW?

Thanks

 

[beaRA]

@pierre heads up that your price ranges are not inclusive on either side, so anything right on the limits will get filtered out. Example: the Polk XT15 is filtered out if you select either "<100$" or "100$-200$".

The price ranges don't seem to consistently filter on price per speaker either. Example: the same 4 Polk models are displayed whether you choose "100$-200$" or "200$-300$"

Any chance of implementing a slider for price filtering?

 

[kernelpanic]

@pierre heads up that your price ranges are not inclusive on either side, so anything right on the limits will get filtered out. Example: the Polk XT15 is filtered out if you select either "<100$" or "100$-200$".

The price ranges don't seem to consistently filter on price per speaker either. Example: the same 4 Polk models are displayed whether you choose "100$-200$" or "200$-300$"

Any chance of implementing a slider for price filtering?

I was just thinking about this the other day. A slider for pricing filtering would be amazing. The current dropdown selection has very limited filtering ranges. It would be nice to see all <$500 or <$1000, etc.

 

[pierre]

I was just thinking about this the other day. A slider for pricing filtering would be amazing. The current dropdown selection has very limited filtering ranges. It would be nice to see all <$500 or <$1000, etc.

ok sliders it will be. I also have new parameters (size, weigth, …, sensitivity, max spl etc) most would be better with a slider.

 

[bennybbbx]

I had some time to kill in a plane and I played with all the data’s that Amir gives us.
you can see them on www.spinorama.org The code to generate them is On github Too.

your display site is great, only small change, can you please change the low freq limit (currently it is 100 hz )in the SPL contour graph from 100 to maybe 50 or at last 80 hz ?. the directivity of subwoofer dipol or not is report here. https://www.linkwitzlab.com/AES'98/aes-98.htm

there are diffrences . also a rear bass port can act as a little dipol at some frequencies. so it is usefull to see below 100 hz because below 100 hz also the bass port do putput more

 

[pierre]

your display site is great, only small change, can you please change the low freq limit (currently it is 100 hz )in the SPL contour graph from 100 to maybe 50 or at last 80 hz ?. the directivity of subwoofer dipol or not is report here. https://www.linkwitzlab.com/AES'98/aes-98.htm

there are diffrences . also a rear bass port can act as a little dipol at some frequencies. so it is usefull to see below 100 hz because below 100 hz also the bass port do putput more

yes, that’s easy to do.

 

[beaRA]

@pierre another thought: the EQ graphs are rescaled to align the response with 0dB. It might be useful to skip rescaling to illustrate how much sensitivity is given up for EQ.