• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

autoeq.app Is a Web Application for Equalizing Headphones the Easy Way

Do you plan of gathering analytics? Or ideally providing an API to gather statistics about the EQ settings people select for specific headphones?

Where I think this would be amazing is as raw data into an understanding a preference curve.
 
I’ve used SoundSource for years and this thread just led me to its built in Headphone EQ that supplies all the parameters from AutoEQ - a very good place to be led! Thanks.
 
Do you plan of gathering analytics? Or ideally providing an API to gather statistics about the EQ settings people select for specific headphones?

Where I think this would be amazing is as raw data into an understanding a preference curve.
I hadn't think of that but now that you mention, it could be interesting
 
@jaakkopasanen I know a bit unrelated to the thread, but FYI I contacted rogueamoeba and they fixed the parsing problem for the filter names generated by AutoEQ in their latest build of SoundSource, so in their next version of their software it should not be necessary to use the old ones. Maybe both will work(?), perhaps it could be worth to keep it as it is for compatibility with old versions, I don't know.
 
Any chance to get a target just like the one that Crinacle uses for over-ear headphones? It's pretty much like "Harman Over-ear 2018" but completely flat below 900 Hz.
 
Any chance to get a target just like the one that Crinacle uses for over-ear headphones? It's pretty much like "Harman Over-ear 2018" but completely flat below 900 Hz.
Just turn bass boost to zero
 
Just turn bass boost to zero
If you had checked that before posting a reply you would have seen that the result has nothing to do with "flat under 900 Hz". The result of turning the bass boost to zero is more like "-2dB under 900 Hz" and has a couple of negative extra bumps. It's just not the same thing.
 
Have the ACS profiles been removed recently? They were available in AutoEQ but can't find them now.
 
They make custom IEM moulded ears (acscustom.com). The EQ profiles for their emotion & a couple of other are no longer in the list in the AutoEQ app.
If those were measured by Reference Audio Analyzer then they are gone. I dropped RAA entirely because the measurements were very questionable and they changed the site structure so the crawler didn't work anymore and I couldn't justify putting in the work for such in accurate measurements.
 
What do you mean by ACS profiles?
They make custom IEM moulded ears (acscustom.com). The EQ profiles for their emotion & a couple of other are no longer in the list in the AutoEQ app. I was using the auto
If those were measured by Reference Audio Analyzer then they are gone. I dropped RAA entirely because the measurements were very questionable and they changed the site structure so the crawler didn't work anymore and I couldn't justify putting in the work for such in accurate measurements.
Ah, fair enough. Yes they were RAA. That is such a shame as the autoeq made the IEMs sound fantastic with the qudelix q5. Just wish I'd taken a screen shot of the EQ settings before they disappeared.
 
I think this could be interesting to EQ speakers, too! I was thinking of getting a graph from Spinorama, converting it to a CSV file and running it through Auto EQ with a flat target curve, but you could do this with room measurements too! You could even apply a house curve, the same way you can do in REW, too. Maybe this is a better alternative to REW's auto EQ. I've found REW's auto EQ function to give bad results, if compared to EQs done with Crinacle/Squig tools using the same target curves, for example, even using impulse responses. For headphones, that is.

I tried doing this by using a Spinorama graph and that's the result I got:
 

Attachments

  • Screenshot.png
    Screenshot.png
    126.6 KB · Views: 57
Last edited:
I think this could be interesting to EQ speakers, too! I was thinking of getting a graph from Spinorama, converting it to a CSV file and running it through Auto EQ with a flat target curve, but you could do this with room measurements too! You could even apply a house curve, the same way you can do in REW, too. Maybe this is a better alternative to REW's auto EQ. I've found REW's auto EQ function to give bad results, if compared to EQs done with Crinacle/Squig tools using the same target curves, for example, even using impulse responses. For headphones, that is.

I tried doing this by using a Spinorama graph and that's the result I got:
You totally can equalize speakers with AutoEq but the workflow of doing everything else in REW and always exporting to AutoEq for eq optimization might be a bit clumsy. You probably need to drop the treble transition frequencies to around 500 Hz and limit the treble gain though. Also keep in mind that the equalized curve isn't going to be predictable in higher frequencies with speakers and a room as it is with headphones
 
You totally can equalize speakers with AutoEq but the workflow of doing everything else in REW and always exporting to AutoEq for eq optimization might be a bit clumsy. You probably need to drop the treble transition frequencies to around 500 Hz and limit the treble gain though. Also keep in mind that the equalized curve isn't going to be predictable in higher frequencies with speakers and a room as it is with headphones
Hi Jakko. Could you tell please how To push AutoEq follow the target 1:1? I always get some variation and low passing in high end region (I am gonna use my own target btw)
 
Hi Jakko. Could you tell please how To push AutoEq follow the target 1:1? I always get some variation and low passing in high end region (I am gonna use my own target btw)
There's 99.9% chance that this is not what you you actually want. You won't get more accurate results by following the target more precisely because the measurements themselves aren't accurate enough, nor is it the target. And because of this it's not possible to completely disable smoothing.
 
Last edited:
Hey Jaakko. Is headphone FR matching as simple as setting the other headphone's FR as the target in this version?
 
Hey Jaakko. Is headphone FR matching as simple as setting the other headphone's FR as the target in this version?
 
Thanks a lot! Sorry if I'm asking too much, but what is the 711/5128 Delta target useful for?
 
Back
Top Bottom