Does harmonic distortion matter?

[Jiraya369]

Why do we use harmonic distortion when we don't really listen to single tones, even Dacs have multitone tests, why don't we perform any multitone test to find out how it is? Also, is there any way of gauging IMD in headphones? Would that not be more apt as that's what we find the most annoying?

 

[solderdude]

Here are some multitone measurements with headphones.

HIFIMAN Susvara Headphone Review

I have long been trying to find a Susvara review over here. (Only 2 months old) Then I gave up until a friend linked it to me. There are some decent things over here, but a negative Susvara review is the final nail in the coffin for ASR for me. From all DACs sound the same, to I can’t recommend...

www.audiosciencereview.com

Here is a reader's attempt to do this with speakers (similar issues arise)

Need help with multitone measurements

I am trying to figure out how bad my 25 year old speakers are and have some questions about multitone tests I did in REW, using UMIK. I measured on tweeter axis at 1m at several levels in 10 db steps, the reading from REW was 68, 78, 88 and 98. Even with earplugs and short measurements interval...

www.audiosciencereview.com

Potential issues are noise floor (both mic and surrounding sounds) at real world loudness and to mimic music signals the 'spectrum' should slope downwards (pink-noise spectrum).
To make it match with reality of how headphones are used in practice (not as loud as most people think) you would have to do the multitones at lower levels too... below 94dB peak.

One thing that is easily forgotten when evaluating plots is things like masking (which is also level dependent) as well as the dynamic behavior being key with evaluating sound.
The frequency response when looked at from the initial sound wave reaching the ears can differ from the measured frequency range so a headphone might sound a bit different from what the frequency response suggests even when all headphones are EQ'ed to the same target.
Then there is ringing at certain frequencies and cone break-up (nulls etc.) also resulting in distortion but only when 'excited' with those frequencies which are likely only harmonics.
The IM and distortion is also highly frequency dependent.

 

[Jiraya369]

Here are some multitone measurements with headphones.

HIFIMAN Susvara Headphone Review

I have long been trying to find a Susvara review over here. (Only 2 months old) Then I gave up until a friend linked it to me. There are some decent things over here, but a negative Susvara review is the final nail in the coffin for ASR for me. From all DACs sound the same, to I can’t recommend...

www.audiosciencereview.com

Here is a reader's attempt to do this with speakers (similar issues arise)

Need help with multitone measurements

I am trying to figure out how bad my 25 year old speakers are and have some questions about multitone tests I did in REW, using UMIK. I measured on tweeter axis at 1m at several levels in 10 db steps, the reading from REW was 68, 78, 88 and 98. Even with earplugs and short measurements interval...

www.audiosciencereview.com

Potential issues are noise floor (both mic and surrounding sounds) at real world loudness and to mimic music signals the 'spectrum' should slope downwards (pink-noise spectrum).
To make it match with reality of how headphones are used in practice (not as loud as most people think) you would have to do the multitones at lower levels too... below 94dB peak.

One thing that is easily forgotten when evaluating plots is things like masking (which is also level dependent) as well as the dynamic behavior being key with evaluating sound.
The frequency response when looked at from the initial sound wave reaching the ears can differ from the measured frequency range so a headphone might sound a bit different from what the frequency response suggests even when all headphones are EQ'ed to the same target.
Then there is ringing at certain frequencies and cone break-up (nulls etc.) also resulting in distortion but only when 'excited' with those frequencies which are likely only harmonics.
The IM and distortion is also highly frequency dependent.

thanks a lot for this

idk what it is but my headphones start distorting at a certain volume after loads of eq, and i find it quite uncomfortable at that volume i wont lie, i never listen that loud, but it's EXTREMELY easy to hear for me, prolly like 30% distortion or something.

My headphones have like 1% distortion at 94 db, i shouldnt be hearing all this distortion but i am. I did a test with two tones, one at 5-9khz (edit it was like 1-2khz lol) or something and another at 30-60 hz. After cranking up the volume to high levels i heard some really nasty distortion. Sounds like garbled mess

I'm using a Topping DX5 II at high gain with my Hifiman HE1000 V2 Stealth headphones

Harmonic distortion profile

My EQ :

I did klippel tests and after -15 db i genuinely cant tell the distortion so i can tell 20% and higher distortion and anything lower should be inaudible, but it's not. I for some reason heard it quite easily with songs and decided to make a test tone to see if i was hallucinating. It was reproducible.

If i turned the bass boost off the distortion was suddenly mostly gone. Is the fit unstable and the vibration from the bass is garbling the sounds idk? But that's why i wanted to know if multitone tests matter, cus by itself the bass or the tone dont create issues but together they do.

 

[solderdude]

Might have been the digital signal clipping or the amp.

Just listened (very shortly) to JT7 with 11V peaks on it and stayed clean (no EQ).
Well it became a bit 'shouty' because my ears don't appreciate me blasting my ears for a few seconds bit not distorted nor 'rough'.
That's 127dB peak and 6W power.
I'm sure the HE1000 won't do worse

 

[Jiraya369]

Might have been the digital signal clipping or the amp.

i start hearing issues at -8 db on the amp, and i have an appropriate negative pre-gain, at least based on what peace gui and squig link say.

 

[voodooless]

There have been many that asked this before, some can be found here:

Search results for query: multitone

www.audiosciencereview.com

 

[solderdude]

Both channels at the same time might be clipping.
The weird part is that it does not seem to happen without EQ.

The thing is that when one uses say +6dB in the bass the max. level in the mids will also drop 6dB so playing the HE1000SE loud on the DX5II it might be clipping.
When it suddenly starts above a certain level in both channels it most likely is clipping.
In headphones this is happening gradually with higher volume.
That is unless the membrane hits max. excursion (magnets or other limiter like what happens with some Focals).
Due to membrane tension with planars and thin membranes it is more likely to act more like a compressor/very soft limiter.

With 110dB/V you should be able to reach 130dB peak SPL with the DX5 before it clips.

 

[Jiraya369]

That's 127dB peak and 6W power

holy moly, i highly doubt i went above 96 db spl, i did a test with speakers and 88 db made me genuinely uncomfortable.

so playing the HE1000SE loud on the DX5II it might be clipping

on my dx5 ii, in high gain with -8 db i start hearing issues. And also i have a -8 db negative pre-gain on my eq app. That should mean that given 6 watts of power, i'm using like 16 db less. And since I'm cutting a lot of the mids, I'm actually hearing far less im guessing. At worst 100 db.

Idk how to explain it, it sounds like your voice when you talk through a fan, all rispy and stuff. The treble especially takes a nose dive and vocals seem weird. Super bassy songs suffer the most, I felt it in Future - Low Life, and Illenium - Free Fall.

I wish I could've measured the response to find out the flaw, but given the fact that the issue comes up when the bass shelf is on, I theorized that the headphone vibrating could be leading to the issue, it feels really nice the way it rumbles on my head but when it gets loud it vibrates super duper hard, maybe my fit is not stable enough to create a clean sound with loud bass?

I dont know man im just lost, I don't usually listen that loud but it feels bad that after spending so much and on something that measures well I'm getting this issue. Faulty unit perhaps?

 

[Jazigo]

thanks a lot for this

idk what it is but my headphones start distorting at a certain volume after loads of eq, and i find it quite uncomfortable at that volume i wont lie, i never listen that loud, but it's EXTREMELY easy to hear for me, prolly like 30% distortion or something.

My headphones have like 1% distortion at 94 db, i shouldnt be hearing all this distortion but i am. I did a test with two tones, one at 5-9khz (edit it was like 1-2khz lol) or something and another at 30-60 hz. After cranking up the volume to high levels i heard some really nasty distortion. Sounds like garbled mess

I'm using a Topping DX5 II at high gain with my Hifiman HE1000 V2 Stealth headphones

Harmonic distortion profile
View attachment 510339
My EQ :
View attachment 510340

I did klippel tests and after -15 db i genuinely cant tell the distortion so i can tell 20% and higher distortion and anything lower should be inaudible, but it's not. I for some reason heard it quite easily with songs and decided to make a test tone to see if i was hallucinating. It was reproducible.

If i turned the bass boost off the distortion was suddenly mostly gone. Is the fit unstable and the vibration from the bass is garbling the sounds idk? But that's why i wanted to know if multitone tests matter, cus by itself the bass or the tone dont create issues but together they do.

Could it be something with the EQ itself? I use a small dongle on my IEMs, which have EQ, and they can make some things sound very strange on certain settings. I have found that just tweaking bass, only lowering values (never boosting) and use low Q values seems to sound the best, but still there is something off sometimes with the bass. I never hear this on my other dongle without EQ, but then it is just to much bass for me.

 

[Jiraya369]

Could it be something with the EQ itself? I use a small dongle on my IEMs, which have EQ, and they can make some things sound very strange on certain settings. I have found that just tweaking bass, only lowering values (never boosting) and use low Q values seems to sound the best, but still there is something off sometimes with the bass. I never hear this on my other dongle without EQ, but then it is just to much bass for me.

could be, im using peace gui on windows cus i went over the bands allowed by dx5 ii's eq

 

[RexrothPigeon]

holy moly, i highly doubt i went above 96 db spl, i did a test with speakers and 88 db made me genuinely uncomfortable.

The case of the missing 6 dB - PubMed

In 1933 data were reported which indicated that thresholds of hearing for frequencies apparently depended upon whether the source was an earphone (MAP) or a loudspeaker (MAF). A decade or so later the same type of discrepancy appeared when loudness balances were made at 100 Hz between an...

pubmed.ncbi.nlm.nih.gov

 

[solderdude]

holy moly, i highly doubt i went above 96 db spl, i did a test with speakers and 88 db made me genuinely uncomfortable.




on my dx5 ii, in high gain with -8 db i start hearing issues. And also i have a -8 db negative pre-gain on my eq app. That should mean that given 6 watts of power, i'm using like 16 db less. And since I'm cutting a lot of the mids, I'm actually hearing far less im guessing. At worst 100 db.

Idk how to explain it, it sounds like your voice when you talk through a fan, all rispy and stuff. The treble especially takes a nose dive and vocals seem weird. Super bassy songs suffer the most, I felt it in Future - Low Life, and Illenium - Free Fall.

I wish I could've measured the response to find out the flaw, but given the fact that the issue comes up when the bass shelf is on, I theorized that the headphone vibrating could be leading to the issue, it feels really nice the way it rumbles on my head but when it gets loud it vibrates super duper hard, maybe my fit is not stable enough to create a clean sound with loud bass?

I dont know man im just lost, I don't usually listen that loud but it feels bad that after spending so much and on something that measures well I'm getting this issue. Faulty unit perhaps?

Perhaps take a phone and record the effect ?
When it happens on both channels at the same time it isn't a faulty headphone.
The fact that it does not happen without EQ is a sign there could be something wrong with the EQ how it is handled in the internal DSP.
It can't be seal I reckon as planars are pretty seal insensitive.
A bad seal produces even more bass.
What you can do is test (with no glasses) and when you hear it compress the pads a little (the odd shape is bound to give seal issues) and also lift both pads slightly (so the seal breakage is similar and guaranteed.

 

[Jiraya369]

Perhaps take a phone

i have a umik-1, would that be better than the phone?

When it happens on both channels at the same time it isn't a faulty unit.

technically speaking, it doesn't happen at the exact same time but that's not because one channel is faulty or anything. I hear less with my left ear, i have a few db lowered on my left channel. When turning it up after some point it starts on the left ear and then suddenly it seems sort of equally fuzzy distortion in the middle, if that makes sense.

Does it happen without EQ too ?

Not at the volumes I tried out. When I turned off the bass shelf the distortion was not there

 

[solderdude]

i have a umik-1, would that be better than the phone?

Best to put it on your head and 'record' the sound from the outside of the headphone about 10cm away or so.

You can also do this when it is lying on a desk.
If the bass distorts you'll hear it (but softly)

Also you can put it on a stack of books to mimic the seal the headphone would be getting on your head and record/listen from the back of the headphone.

In any case it won't be 'harmonic distortion' but clipping or another side effect the DSP produces the 'effect' you are hearing.

 

[Jiraya369]

Best to put it on your head

im sorry i cant put the umik-1 in my head it's a one long mic

You can also do this when it is lying on a desk.

problem is, it has some weird phase things, it sounds garbled and messy when away from my head, but when i put it on it sounds clean. It sounds EXTREMELY distorted at the same levels but it sounds "only" distorted when on my head. If that makes sense.

Also you can but it on a stack of books to mimic the seal the headphone would be getting on your head and record/listen from the back of the headphone.

i can try that but idk man if i cover the back of the open back headphone, it gets all weird and phase-ey and FR is horrendous

I'll do all the tests i can and send the files

 

[solderdude]

Leave the backside open.
It radiates the same sound on both sides.
So when on a stack of books you can hear the sound from outside but much attenuated (bass will be rolled off)

Have you tried the same using EQ from a PC and use the amp as is ?
Have you tried different programs for the bass boost ?

 

[Jiraya369]

Leave the backside open.
It radiates the same sound on both sides.
So when on a stack of books you can hear the sound from outside but much attenuated (bass will be rolled off)

Have you tried the same using EQ from a PC and use the amp as is ?
Have you tried different programs for the bass boost ?

funnily enough, i made a post a while ago saying playing sounds without exclusive mode sounds worse. Someone said i was most likely tweaking. But now that you mention it, i never noticed this issue when i used a similar eq on my topping dac's internal EQ alongside an output to the topping asio driver OR exclusive mode output. I was starting to think i was hearing things cus of placebo so i didnt mention it. I'll perform some tests when i get some time.

 

[Blorg]

Exclusive mode eliminates all possible sources of Windows dicking with the sound.

Non-exclusive mode can be made to sound the same, but there is just more opportunity for there to be something in there that is changing the sound, you have to look around and make sure there isn't. I've had that same experience, where exclusive mode sounded right and not didn't, and it was on investigation because Windows had enabled spatial sound after an update, or there was some other APO applying an effect to the audio.