Hot Take (or not), if you EQ your headphones to your ears, you can easily beat ANY unEQ'd headphone.

[pfsoft]

Only thing I will never get on headphones is the bass, will never extend as low

That's not true. Many headphones have a more extended frequency response toward low frequencies and are more linear than traditional speakers.
You can never compare the two systems (always at low frequencies) because you don't physically perceive low frequencies through headphones.

 

[JoelDollie]

That's not true. Many headphones have a more extended frequency response toward low frequencies and are more linear than traditional speakers.
You can never compare the two systems (always at low frequencies) because you don't physically perceive low frequencies through headphones.

Yeah that's true, well I meant that it doesn't feel quite the same, not physical, but planars can definitely produce super low frequencies, closed backs as well.

 

[StephanB]

Coming late to the party.

• I got nice results with David Griesinger's DGSonicFocus application.
• I am developing a simplified and more portable version of his application percep-head-eq, which runs in the browser. My version outputs a csv file that you can import into Room EQ Wizard (REW) to produce an EQ in the format of your choice.
• In both cases I use a convolution filter with an HRTF before EQ. The HRTF is in the public domain and produced with an artificial head, it's not tuned to my head, but I still find it effective.
• I am struggling to find a good bass response curve. I lack the knowledge in psychoacoustics. If someone is interested to collaborate on the app, please contact me.

 

[HissingFree]

Coming late to the party.

• I got nice results with David Griesinger's DGSonicFocus application.
• I am developing a simplified and more portable version of his application percep-head-eq, which runs in the browser. My version outputs a csv file that you can import into Room EQ Wizard (REW) to produce an EQ in the format of your choice.
• In both cases I use a convolution filter with an HRTF before EQ. The HRTF is in the public domain and produced with an artificial head, it's not tuned to my head, but I still find it effective.
• I am struggling to find a good bass response curve. I lack the knowledge in psychoacoustics. If someone is interested to collaborate on the app, please contact me.

So as I understand it, we would equalize by comparing a given frequency to a 500Hz reference tone instead of sine sweeping, right? so this instead of moving the frequency slider up and down to quickly try to detect a much higher or lower sounding frequency.

 

[StephanB]

So as I understand it, we would equalize by comparing a given frequency to a 500Hz reference tone instead of sine sweeping, right? so this instead of moving the frequency slider up and down to quickly try to detect a much higher or lower sounding frequency.

Yes, I believe Dr Griesinger has done the homework on this. He recommends to: 1. compare to a reference tone, 2. the duration should be around 1 sec per tone, 3. use 1/3 octave bandpass noise rather than pure tones.

Point 3 engages with the whole discussion on ERB (Equivalent Rectangular Bandwidth) in psychoacoustics. Should we use 1/3 octave across the whole spectrum, or vary the filter width with the frequency? Btw. you can listen to the filtered noise with my app in the browser, DGSonicFocus is a Windows app. You can also produce 1/3 octave filtered noise in REW, but it's slower.

The last question is: what reference curve do we EQ to? Dr Griesinger lets you determine the curve by listening to the tones on a loudspeaker first, so the reference is tailored to your hearing. I use a shortcut: I set the RMS of the filtered noise to 0dB, then I apply the ISO equal loudness contour at 80phons plus a room curve. I would love to have this reviewed by someone with better academic knowledge than me.

 

[IsaacOscar]

The last question is: what reference curve do we EQ to? Dr Griesinger lets you determine the curve by listening to the tones on a loudspeaker first, so the reference is tailored to your hearing. I use a shortcut: I set the RMS of the filtered noise to 0dB, then I apply the ISO equal loudness contour at 80phons plus a room curve. I would love to have this reviewed by someone with better academic knowledge than me.

This sounds similar to what Peace's hearing test feature does. I've been meaning to try it again (last time I tried it I gave it the minimum volume that I could hear each tone at, but that was too difficult; so next time I'm going to try it based on equal loudness to a 500 Hz tone)

 

[stalepie2]

This amp seems to offer something new? AI-assisted EQing:

Luxsin Audio X8 | Desktop DAC and Headphone Amp

Luxsin X8 features 8 CS43198 DACs, dual-mono parallel design, 4800mW output power, a proprietary Digital Audio Core, and the world’s first AI-assisted EQ.

bloomaudio.com

"Siri, drop the bass..."