Help equalizing bluetooth speaker and headphones separately

[Volutrik]

Hi, I want to follow this equalization method https://sourceforge.net/p/mesh2hrtf-tools/wiki/Earful_tutorial/

The bluetooth speaker and headphones will alternate sound, so I'd like to know how's it possible to apply equalization for the bluetooth speaker only, making it sound flat, but not for the headphones, as they will be equalized during the session. I want to use Equalizer APO or any other application that works for this. I don't have an audio interface and the headphones will be connected through Jack input. Thanks!

 

[RayDunzl]

Does the EQ software have the ability to save the settings and pick a setting you want to use?

I'll assume you are not listening to both at the same time.

 

[kemmler3D]

Hi, I want to follow this equalization method https://sourceforge.net/p/mesh2hrtf-tools/wiki/Earful_tutorial/

The bluetooth speaker and headphones will alternate sound, so I'd like to know how's it possible to apply equalization for the bluetooth speaker only, making it sound flat, but not for the headphones, as they will be equalized during the session. I want to use Equalizer APO or any other application that works for this. I don't have an audio interface and the headphones will be connected through Jack input. Thanks!

To use different EQ settings with EQAPO at the same time, you need a different output device for each EQ setting.

In this case, that might be the bluetooth adapter vs. the analog output.

If they're using the same audio device, you will need to set up different files for each device and enable / disable them when you switch.

My comment: This method makes some sense, but the slow-switching method of matching tone volume by taking the headphones on and off seems difficult. IME it's really hard to judge the volume of tones (or anything) if you aren't doing fast switching (< 1s at the slowest) or listening simultaneously.

 

[AwesomeSauce2015]

In order to do that process you posted, you need a measurement mic to calibrate the bluetooth speaker, and then you need to EQ the speaker.
Personally, I would use two source devices, one with EQ for the bluetooth speaker, and one which handles the headphones. -- Although with my audio setup I can drive multiple, separately-equalized devices from any number of separately-equalized sources.
What you could also do is set EqualizerAPO to EQ specifically the bluetooth speaker device, and not touch the headphone output. This is probably what the writer of the software you are trying to use intended. You would then be able to correct the frequency response of the bluetooth speaker, and use that to do whatever it is you are trying to do with your headphones.
(Source: I used to use EQAPO to handle my system EQ, using multiple outputs on my Focusrite interface. It was a bit unreliable, and the equalizer implementation seemed off sometimes, but it did sort of work. I also uninstalled it immediately after setting up my new hardware DSP solution...)

One thing to keep in mind though: Many active / bluetooth speakers, and some headphones have loudness compensation and dynamic limiting built in. This means that the frequency response changes with volume level changes. This is just something to keep in mind as it could impact your test. In my opinion, this really would only work with passive headphones and a passive, coaxial, full(ish)-range speaker like KEF's Q150. - Though other studio monitor speakers like the JBL 305 or Genelec 8010 could also work.

 

[kemmler3D]

One thing to keep in mind though: Many active / bluetooth speakers, and some headphones have loudness compensation and dynamic limiting built in. This means that the frequency response changes with volume level changes. This is just something to keep in mind as it could impact your test. In my opinion, this really would only work with passive headphones and a passive, coaxial, full(ish)-range speaker like KEF's Q150. - Though other studio monitor speakers like the JBL 305 or Genelec 8010 could also work.

Good point, dynamic compressors are very common on most small bluetooth speakers.

I wouldn't say it's a deal breaker, but you need to run some tests to check for this before starting.

Do several sweeps at different volumes through the bluetooth speaker. If there is dynamic compression (likely) you will see the relative level of bass (probably 60-100hz and lower depending on the speaker) change depending on the volume setting on the speaker.

At some point the shape of the curve in the bass region should stop changing. This is the level you can't exceed during your calibration testing. Set the gain lower than this, after any corrections you apply.

Or just forget all that and use a normal monitor if you can.

 

[Volutrik]

Personally, I would use two source devices, one with EQ for the bluetooth speaker, and one which handles the headphones.

Yeah, I could use two laptops for this! Even though using a single one is better because Earful switches the two outputs automatically.

One thing to keep in mind though: Many active / bluetooth speakers, and some headphones have loudness compensation and dynamic limiting built in. This means that the frequency response changes with volume level changes. This is just something to keep in mind as it could impact your test. In my opinion, this really would only work with passive headphones and a passive, coaxial, full(ish)-range speaker like KEF's Q150. - Though other studio monitor speakers like the JBL 305 or Genelec 8010 could also work.

This is indeed something I haven't thought about. Thanks for talking about it! I don't have the option to use studio monitors right now as I'm trying to get the best out of headphones using binauralization. I'll have to try it out with a bluetooth speaker, which I also don't have as of yet I'll buy everything specifically for this. I'm also looking to buy the Dayton Audio iMM-6 to perform the measurements.

Do several sweeps at different volumes through the bluetooth speaker. If there is dynamic compression (likely) you will see the relative level of bass (probably 60-100hz and lower depending on the speaker) change depending on the volume setting on the speaker.

The JBL Go 3 seems to be a good option as it has 1.36dB of Dynamic Range Compression if compared being played at 76dB SPL and it's max volume, which is 79.6dB SPL. Using it at 65-70dB SPL seems to be fine - https://www.rtings.com/speaker/reviews/jbl/go-3

 

[staticV3]

@Volutrik Equalizer APO must be installed for each output device separately.

Install it only for the Bluetooth speaker and EQ will be applied just to it.

 

[AwesomeSauce2015]

Yeah, I could use two laptops for this! Even though using a single one is better because Earful switches the two outputs automatically.

Nice. I figured that the software would probably do that, I just didn't want to look into it last night as I was quite tired lol.
You would then just only install EqualizerAPO to the bluetooth speaker output, and it would only touch your bluetooth speaker output and not the headphones. (as @staticV3 said above)

This is indeed something I haven't thought about. Thanks for talking about it! I don't have the option to use studio monitors right now as I'm trying to get the best out of headphones using binauralization. I'll have to try it out with a bluetooth speaker, which I also don't have as of yet I'll buy everything specifically for this. I'm also looking to buy the Dayton Audio iMM-6 to perform the measurements.

Personally I would use something like the miniDSP UMIK-1, if you don't already have a decent mic interface (say a Focusrite Scarlett or similar), or the Dayton EMM-6 if you do. Device headphone jack mic preamps can have some big frequency response issues, which could impact your measurements. A USB mic or one with a good interface solves this problem. I also just don't like the headphone jack form factor.
It sounds like you may be on a budget, but I would still recommend saving up for the UMIK-1 / EMM-6+good preamp, it is just way more versatile and will probably last longer.

I can't comment on the quality of the JBL Go 3, but so long as you stay at the same SPL level throughout your testing, it should work fine. The issue comes when you try to run at different SPL levels. I mean, unless you already have it, there are other speakers for $50 that will not have these dynamic compression issues, and you may also be able to find better deals on ebay/craigslist.

 

[kemmler3D]

Yeah, I could use two laptops for this! Even though using a single one is better because Earful switches the two outputs automatically.



This is indeed something I haven't thought about. Thanks for talking about it! I don't have the option to use studio monitors right now as I'm trying to get the best out of headphones using binauralization. I'll have to try it out with a bluetooth speaker, which I also don't have as of yet I'll buy everything specifically for this. I'm also looking to buy the Dayton Audio iMM-6 to perform the measurements.



The JBL Go 3 seems to be a good option as it has 1.36dB of Dynamic Range Compression if compared being played at 76dB SPL and it's max volume, which is 79.6dB SPL. Using it at 65-70dB SPL seems to be fine - https://www.rtings.com/speaker/reviews/jbl/go-3

View attachment 380138

The JBL Go is good for what it is (I own one) but do make sure you've set the volume correctly so that compression isn't activated. Also, you should still do the test I mentioned... this chart doesn't actually show a normal compression test result. Obviously the JBL Go 3 cannot do 76dB at 20hz...

 

[staticV3]

this chart doesn't actually show a normal compression test result. Obviously the JBL Go 3 cannot do 76dB at 20hz...

The JBL Go 3 definitely cannot do 76dB at 20hz, and yet, the chart shows a normal compression test result.

When measuring loudspeaker compression, we pick a "comfortable" loudness for the speaker, say 70dB at 400Hz, then gradually increase the volume and plot the difference in frequency response, compared to that reference.

Normalizing the results to that 70dB response means that it is now plotted as a straight line, which however does not mean that the speaker has a ruler flat, 20Hz to 20kHz response at 70Hz.

An example:


If one is unfamiliar with the methodology and it or the graph is not clearly explained by the reviewer, then yes, it could be easily misinterpreted.

Btw, rtings explain their compression measurement methodology here:

Our Speaker Sound Tests: Dynamics

When you're looking at a speaker's sound profile, it's important to look at its dynamics performance.

www.rtings.com