Calibration issue, or are my ears the problem?

[JP3D]

I've been building a set of active 2-way bookshelf speakers and have been going back and forth between quasi-anechoic measurements and listening to them. While not perfect, I'm happy with how they're measuring now. This is the latest spinorama (7.3ms gated, 87cm distance, nearfield spliced at 400Hz)

(Note I'm 90% sure the DI error around 3kHz and 4.5kHz is from diffraction. I found dropping the on-axis response there to smooth out the PIR easier to listen to. Tradeoffs)

The trouble is, I realized that I've been doing all this measuring with only the sound card calibration applied without the microphone calibration. The calibration file for my EMM-6 shows a +4-5dB shelf above about 6kHz, so in theory my speakers are 4dB too low in the high treble.

What I find odd though is in a previous listening test I did (without remembering exactly how the speakers were measuring) they sounded balanced to me on a lot of tracks (artists including The Hu, Taylor Swift, Diana Krall, plus some song recommendations from here like Lo Sono Metà and My Favorite Things). But after about 30-40 minutes my ears did feel a bit tired, so I figured that the treble was maybe a bit too hot. I put on some Led Zeppelin and Carrie Underwood and I found them ear-bleedingly bright. I was listening at about 78-82dB RMS 1m away. When I went to measure the speakers again I noticed a 2-4dB trough from about 1kHz-4kHz (aside from the diffraction compensation) and a high shelf about 3-5dB up above 5kHz. I did a bunch of tweaking on the DSP filters to get it to the spin above, but alas I didn't save that initial measurement.

Putting all that together, it means that for these to be measuring correctly with the mic calibration applied, I would have needed to make the high treble even louder by about 1-2dB, beyond what I found to be too bright.

So this leads me to the question: are my ears the problem? I don't have much by way of comparison - I do 95% of listening on AirPods Pro and have some Sonos speakers in my kitchen, but I've listened to tracks with KRK Rokit 5 G3s without them seeming too bright. Is the problem that my comparison devices all have rolled off treble and my ears aren't accustomed to "correct" treble balance? Or is there something else going on here?

FWIW, this is the harmonic distortion measurement.

 

[fpitas]

Pop recordings are often too bright for me, too. And maybe your ears are just naturally more sensitive to treble. Once upon a time we had tone controls for these very reasons...

 

[fpitas]

Certainly, late at night I like the treble dialed back a lot.

 

[JP3D]

Pop recordings are often too bright for me, too. And maybe your ears are just naturally more sensitive to treble. Once upon a time we had tone controls for these very reasons...

It might be worth noting I’m a 27y/o male, so my treble hearing hasn’t entirely deteriorated.

These speakers are for someone else, so my thought is to get them sounding good on a variety of music and measuring pretty well, then adding bass and treble balance knobs on them for the client to tune as they desire. But for the future I’d love to know what might be going on with my measuring setup, since my ears are telling me the opposite of what the calibrated measurements are saying.

 

[Frank2]

One meter is pretty close to the speakers. You'll typically get a bit more treble rolloff at larger distances (more indirect sound from the lower frequencies will cause the treble to be a bit lower in comparison).
That said, I do have similar experiences. It seems to vary a bit from day to day but as soon as my ears get a bit irritated I tend to focus on the treble and end up dialing it back. If I start a listening session again a few day later I may find the sound too dull and dial it a bit up again. It's hard to find the sweet spot, as it seems to be a moving target.. And it varies from recording to recording as well. Luckily I have a digital equalizer with presets
My gut feeling is that it is related to how tired you are and to the time of day as well. My ears tend to be more sensitive to treble in the morning, after not using them for the whole night.

All in all I tend to prefer a bit more treble rolloff than the "ideal room response curve".

 

[fpitas]

It might be worth noting I’m a 27y/o male, so my treble hearing hasn’t entirely deteriorated.

These speakers are for someone else, so my thought is to get them sounding good on a variety of music and measuring pretty well, then adding bass and treble balance knobs on them for the client to tune as they desire. But for the future I’d love to know what might be going on with my measuring setup, since my ears are telling me the opposite of what the calibrated measurements are saying.

It also occurred to me that if you're listening at relatively low levels, you might be compensating for the equal loudness curve at bass somewhat.

I'm 69, but my ears are still not fond of blasting treble.

 

[LTig]

What I can say is that the brain can adapt easily to any kind of deviation from flat frequency response and then take this as the reference to judge. Hence it is difficult to make objective statements what's correct and what's not.

When I built my first speakers in the 80's using Kef drivers (kind of a LS 3/5a clone) I found them dull but a friend congratulated me for their good sound. I adapted to them and later, when listening to my old speakers (cheap bright sounding "Taunus"-sound 3-way bookshelves) again I found them horrible.

 

[JP3D]

It also occurred to me that if you're listening at relatively low levels, you might be compensating for the equal loudness curve at bass somewhat.

I'm 69, but my ears are still not fond of blasting treble.

The loudness curves occurred to me as well. I think that’s definitely happening in the bass (I’m measuring at about the same volume I’m listening), but from my understanding I would need to make the high treble even louder at these lower volumes, beyond what I would need to compensate for the mic calibration.

 

[JP3D]

What I can say is that the brain can adapt easily to any kind of deviation from flat frequency response and then take this as the reference to judge. Hence it is difficult to make objective statements what's correct and what not.

When I built my first speakers in the 80's using Kef drivers (kind of a LS 3/5a clone) I found them dull but a friend congratulated me for their good sound. I adapted to them and later, when listening to my old speakers (chesp bright sounding "Taunus"-sound 3-way bookshelves) again I found them horrible.

I feel like this really might be it. If the tuning of the AirPods Max is anything like the tuning on the AirPods Pro, then I’m used to hearing the treble rolled off ~4dB.

Apple AirPods Max Review (Noise Cancelling Headphone)

This is a review and detailed measurements of the Apple AirPods Max wireless Bluetooth noise cancelling headphone. It is on kind loan from a member. The retail cost is US $549 but I see it on Amazon for $479 including Prime shipping. I am used to shocking good looks in Apple products but this...

www.audiosciencereview.com

It makes me wonder if I listened to tonally flat speakers for a while if I would go back to these ones I’ve made and find them dull.

 

[fineMen]

... If the tuning of the AirPods Max is anything like the tuning on the AirPods Pro, then ...

Best served in the DIY department?

 

[JP3D]

Best served in the DIY department?

Thanks for mentioning that, I should’ve taken a better look at the different forums before posting. I figured it would make sense here since it isn’t strictly about DIY audio but more measurements and listening preferences.

I did find this thread about a target FR curve in the DIY forum, and it looks like there are others that also roll off the treble in the same way I inadvertently did.

Preferred Target FR Curve

Hi Everyone! TLDR; what's your preferred FR Curve and EQ goals when using DSP? I've built my own pair of speakers of my own design, you can find my write up of it here: Still Haven't Named Them. I stopped messing with them for a bit and am very happy with its sound signature except I found...

audiosciencereview.com

 

[alex-z]

You need proper free-field mic calibration loaded for accurate high frequency response. Many measurements mics, including the EMM6, have diffuse field calibration by default, for measuring room acoustics.

Acoustic Methods of Microphone Calibration

An article detailing the frequency response of a measurement microphone, general acoustic calibration, free-field calibration, pressure calibration, etc.

audioxpress.com

It is also important that your measurement setup have no reflections of its own. It is possible for the microphone or speaker stand to add reflections.

You should measure your KRK Rokit 5 G3 to give yourself a reference point. You may indeed prefer rolled off treble, or perhaps the R5 G3 has much different directivity than your DIY speaker.

 

[JP3D]

You need proper free-field mic calibration loaded for accurate high frequency response. Many measurements mics, including the EMM6, have diffuse field calibration by default, for measuring room acoustics.

Acoustic Methods of Microphone Calibration

An article detailing the frequency response of a measurement microphone, general acoustic calibration, free-field calibration, pressure calibration, etc.

audioxpress.com

It is also important that your measurement setup have no reflections of its own. It is possible for the microphone or speaker stand to add reflections.

You should measure your KRK Rokit 5 G3 to give yourself a reference point. You may indeed prefer rolled off treble, or perhaps the R5 G3 has much different directivity than your DIY speaker.

I think at this point the answer to my question is that yes, my ears are the problem and I’ve rolled off the treble on these speakers by forgetting to apply the calibration. For future measurements I’ll be sure to have it applied with the knowledge that in practice I may prefer something other than flat.

Unfortunately i don’t actually own or have the Rokit 5s, otherwise I’d do your suggestion to compare the two speakers to see what’s different.

 

[fineMen]

Thanks for mentioning that, I should’ve taken a better look at the different forums before posting.

Most of the people in this section don't fumble around with measurements for a variety of reasons. That's why I suggested the DIY section.

Anyway, believe it or not, I've calibrated my mike against a well known tweeter or better to say two of different, but premium manufacturing. An SEAS metal dome and a Tymphany aka Peerless soft dome. The results were cross-checked and found quite consistent one versus the other and with independend measurements from others. Additionally I calibrated my absolute level to that tweeters with quite reproducible results.