• 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!

Beta Test: Earful -- a hearing test App

hutt132

Member
Joined
Jul 25, 2019
Messages
49
Likes
84
I wasn't sure how to import Oratory's parametric AutoEQ settings into REW, so I just resized the graphs from Equalizer APO (Oratory) and REW (My equal loudness) and overlayed them in photoshop.
You can see how my ears shift the harman curve up more in the high end.

Headphones = HD600
Black = Oratory AutoEQ
Red = My Equal Loudness
74578e840d.jpg


My Equal Loudness
e0bb52d8c7.png


EDIT:
Spent a few hours flipping between my equal loudness curve and EqualizerAPO to create an EQ. I got them to line up pretty close and so far it sounds very good. I wasn't able to boost the bass as much or else I'd have to set a -15 gain vs -5.60.

Filter: ON LSC Fc 110 Hz Gain 5.4 dB Q 0.88
Filter: ON PK Fc 185 Hz Gain 0.6 dB Q 4.14
Filter: ON PK Fc 453 Hz Gain 0.6 dB Q 2.36
Filter: ON PK Fc 565 Hz Gain -1.3 dB Q 1.3
Filter: ON PK Fc 883 Hz Gain 1.3 dB Q 1.86
Filter: ON PK Fc 1500 Hz Gain 0.2 dB Q 2.2
Filter: ON PK Fc 2810 Hz Gain -0.3 dB Q 2.2
Filter: ON PK Fc 3810 Hz Gain -0.5 dB Q 1.04
Filter: ON PK Fc 6563 Hz Gain -2 dB Q 2.7
Filter: ON PK Fc 8202 Hz Gain 7 dB Q 3.1685
Filter: ON PK Fc 10249 Hz Gain -7.6 dB Q 2.2
Filter: ON PK Fc 16005 Hz Gain 3 dB Q 1.18

5e7959bc9a.png
 
Last edited:
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
I wasn't sure how to import Oratory's parametric AutoEQ settings into REW, so I just resized the graphs from Equalizer APO (Oratory) and REW (My equal loudness) and overlayed them in photoshop.
You can see how my ears shift the harman curve up more in the high end.

Headphones = HD600
Black = Oratory AutoEQ
Red = My Equal Loudness
74578e840d.jpg


My Equal Loudness
e0bb52d8c7.png


EDIT:
Spent a few hours flipping between my equal loudness curve and EqualizerAPO to create an EQ. I got them to line up pretty close and so far it sounds very good. I wasn't able to boost the bass as much or else I'd have to set a -15 gain vs -5.60.

Filter: ON LSC Fc 110 Hz Gain 5.4 dB Q 0.88
Filter: ON PK Fc 185 Hz Gain 0.6 dB Q 4.14
Filter: ON PK Fc 453 Hz Gain 0.6 dB Q 2.36
Filter: ON PK Fc 565 Hz Gain -1.3 dB Q 1.3
Filter: ON PK Fc 883 Hz Gain 1.3 dB Q 1.86
Filter: ON PK Fc 1500 Hz Gain 0.2 dB Q 2.2
Filter: ON PK Fc 2810 Hz Gain -0.3 dB Q 2.2
Filter: ON PK Fc 3810 Hz Gain -0.5 dB Q 1.0396
Filter: ON PK Fc 6563 Hz Gain -2 dB Q 2.7
Filter: ON PK Fc 8202 Hz Gain 7 dB Q 3.1685
Filter: ON PK Fc 10249 Hz Gain -7.6 dB Q 2.2
Filter: ON PK Fc 16005 Hz Gain 3 dB Q 1.18

Interestingly, the overall shape resembles my preferred target with headphones. A bump in the bass, below around 150Hz and a nearly flat line (or slightly sloping down) after that. Not sure what it means :)
 

hutt132

Member
Joined
Jul 25, 2019
Messages
49
Likes
84
Interestingly, the overall shape resembles my preferred target with headphones. A bump in the bass, below around 150Hz and a nearly flat line (or slightly sloping down) after that. Not sure what it means :)
That is quite interesting. I messed with the bass response a bit more to make it slope better towards 30Hz from 150Hz, which helped 'de-bloat' the bass from my initial EQ.

I have to say, though, the HD600 sound thin, tinny, and not as good now with my EQ off. You've ruined them! ;)
 

jae

Major Contributor
Joined
Dec 2, 2019
Messages
1,208
Likes
1,508
That's strange. The warning should only come up one time, the first time you launch Earful. It should remember that you've agreed to it the next time. Do you see the warning every time you start? Also, once you get past the warning, you should be able to pin the icon to the taskbar. Does that not work?

The warning is coming up for me every time the software is opened.
 

Spyart

Member
Joined
Dec 25, 2020
Messages
54
Likes
22
Thank you so much for this app! Finally I know that my equal loudness is far from Harman or any preference curve around the web. If you curious I could describe my work flow for headphone corrections to make it like near field monitor setup here. But for now just a few requests.
1. Could you make at least Warble tone and Pink noise at the same volume (perceived? Don't know but I don't think so.) The issue is when I use it together if struggling at some point there is a significant drop in Warble loudness. Maybe good for testing spl sensitivity but... It would be great if we can adjust a volume of each test tone by volume sliders for instance.
2. There is also a bag while switching test tones in separate channel test, sometimes new tone clashes with old one and appears in non tested channel.
3. Subtraction two filters right in Earful. It's already possible to upload two graphs and if would be great to specify target from them and subtract it from another and then save (yeah, Greisenger method). It would prevent for messing with Rew filters and we still use one app Earful for both tests and it seems to me perfectly right to use it for subtraction purposes. I also love the curves of Earful.
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
Thank you so much for this app! Finally I know that my equal loudness is far from Harman or any preference curve around the web. If you curious I could describe my work flow for headphone corrections to make it like near field monitor setup here. But for now just a few requests.
1. Could you make at least Warble tone and Pink noise at the same volume (perceived? Don't know but I don't think so.) The issue is when I use it together if struggling at some point there is a significant drop in Warble loudness. Maybe good for testing spl sensitivity but... It would be great if we can adjust a volume of each test tone by volume sliders for instance.
2. There is also a bag while switching test tones in separate channel test, sometimes new tone clashes with old one and appears in non tested channel.
3. Subtraction two filters right in Earful. It's already possible to upload two graphs and if would be great to specify target from them and subtract it from another and then save (yeah, Greisenger method). It would prevent for messing with Rew filters and we still use one app Earful for both tests and it seems to me perfectly right to use it for subtraction purposes. I also love the curves of Earful.

Hi, and you're welcome, glad you like it!

1. These were set to the same level of output energy (power), but subjectively, it's much harder to make them sound the same. One tone compared to a noise signal of many noise-shaped frequencies combined within limited frequency range is not the simplest thing to make sound reliably at the same level. I can certainly adjust it more, for example using partial loudness or other loudness measures. Let me see what I can dig up.

2. This is a known bug, and I'll fix it in the next update.. When switching tone types, it's better to stop the playback and then start again, for now.

3. Good idea. Let me see if there's a simple way to add this.
 

Spyart

Member
Joined
Dec 25, 2020
Messages
54
Likes
22
Hi again! I've just tested two device equal loudness and I think it sounds great! However due to blocking ear canal during test by wearing headphones (in my case it's Superlux HD681, semi-open) I reduced 10khz or so so much and I feel I need some compensation kind of high shelf filter but I don't understand from which frequence I should start, Q and amount of dB. Maybe you have some thoughts on this compensation? Cheers! (Also interested in same thing for in ears)
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
Hi again! I've just tested two device equal loudness and I think it sounds great! However due to blocking ear canal during test by wearing headphones (in my case it's Superlux HD681, semi-open) I reduced 10khz or so so much and I feel I need some compensation kind of high shelf filter but I don't understand from which frequence I should start, Q and amount of dB. Maybe you have some thoughts on this compensation? Cheers! (Also interested in same thing for in ears)

Try to create an equal loudness curve with the headphones on and then off to see what effect the headphones have on the frequency response. Do it at your normal listening level, and adjust the level between tests to approximately the same level, by ear. By comparing the two results you should be able to see where the frequency response starts to diverge.
 

Spyart

Member
Joined
Dec 25, 2020
Messages
54
Likes
22
Try to create an equal loudness curve with the headphones on and then off to see what effect the headphones have on the frequency response. Do it at your normal listening level, and adjust the level between tests to approximately the same level, by ear. By comparing the two results you should be able to see where the frequency response starts to diverge.
Thanks! I have already tried a different way. Just put only left side of cans and turn the left monitor off. Then I've set in Earful left channel to prevent right side of cans messing around with brain and here we go - just imagine a warble tone that comes from the right ear to left, wow, I could hear 1db difference easy. So in the conclusion difference from 2khz to 10khz is about 6 dB like a large peak. Then I added 112db redline monitor, turn crossfeed to 90 degrees and phantom center to minus 1.5 and here we go, I can mix in superlux right now. Thank you for this app!
 
Joined
Dec 22, 2020
Messages
38
Likes
6
1.Is the work still in progress? I don't want to urge it, but the only mainthing preventing me from doing the two devices equal loudness is the small overlap in transit. The short loudness raise here is still too confusing, a (very) short pause would cause much less distraction for the brain.

2.Do you eventually plan to integrate a two channel eq with both (different) channel eqs applicable at the same time?

3. The smyth realiser a8 / a16 provides a "manLOUD eq" where the test signal frequency bands can be played in one run, which really helps to detect aberrations of loudness. Hope you can add that useful method, too.

Thanks! I have already tried a different way. Just put only left side of cans and turn the left monitor off. Then I've set in Earful left channel to prevent right side of cans messing around with brain and here we go - just imagine a warble tone that comes from the right ear to left, wow, I could hear 1db difference easy. ...!
I don't think it's feasible proceeding like that, because
Even with left speaker off you will hear the right speaker also on the left ear and the perceived loudness is always composed of the left AND right ear, so seal the unnecessary ear is still advisable.

I quote myself:
For anybody, who wants to proceed like me for matching different loudnesses of the same noise band from headphone and speaker, bear the following in mind:
The perceived total loudness is composed of the loudness of the left and the right ear, so listening to a single speaker regardless of mid or left/right position will always affect both ears.
For that reason comparing the loudness of one headphone channel (left or right) with the one from a single speaker using BOTH ears doesn't make sense.
I infer, it's indicated to seal the unnecessary ear during the test, but even then the signal will be heard (at lower loudness) by that ear.
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
1.Is the work still in progress? I don't want to urge it, but the only mainthing preventing me from doing the two devices equal loudness is the small overlap in transit. The short loudness raise here is still too confusing, a (very) short pause would cause much less distraction for the brain.

All my software is always "work in progress" :) I looked at the change you asked for, and found it wasn't as simple as I thought. I should be able to get back to it in a few days and try to get a version out that doesn't have the tone overlap.
 

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,209
Likes
2,674
Location
Salvador-Bahia-Brasil
For anybody, who wants to proceed like me for matching different loudnesses of the same noise band from headphone and speaker, bear the following in mind:
The perceived total loudness is composed of the loudness of the left and the right ear, so listening to a single speaker regardless of mid or left/right position will always affect both ears.
For that reason comparing the loudness of one headphone channel (left or right) with the one from a single speaker using BOTH ears doesn't make sense.
I infer, it's indicated to seal the unnecessary ear during the test, but even then the signal will be heard (at lower loudness) by that ear.

how true is that in the practice?

when you hear music you also hear with both ears the signal which is played on one channel
 
Joined
Dec 22, 2020
Messages
38
Likes
6
how true is that in the practice?

when you hear music you also hear with both ears the signal which is played on one channel
I don't know it exactly but the problem here is the difference between the perceived loudness curves of the left and right ear.
The aim is to determine the loudness curves for the left and right ear separately, because they often differ by nature. But this is not achievable with both ears unsealed, (except for a method using sth. like crosschannel cancellation)

Spyart compares the loudness of one ear (for instance left, muting the right channel on headphones) with the one of both ears (right speaker on - sonic reaches both ears, right one is more affected).
When your hear music over headphones with one channel off, you don't perceive the sound with both ears.
I conclude, his measurements lack comparability this way.

I also plan to determine the dampening factor of my (open-back) over-ear headphones with in-ear mic measurements to increase the speaker volume appropretly.
 

Spyart

Member
Joined
Dec 25, 2020
Messages
54
Likes
22
I don't know it exactly but the problem here is the difference between the perceived loudness curves of the left and right ear.
The aim is to determine the loudness curves for the left and right ear separately, because they often differ by nature. But this is not achievable with both ears unsealed, (except for a method using sth. like crosschannel cancellation)

Spyart compares the loudness of one ear (for instance left, muting the right channel on headphones) with the one of both ears (right speaker on - sonic reaches both ears, right one is more affected).
When your hear music over headphones with one channel off, you don't perceive the sound with both ears.
I conclude, his measurements lack comparability this way.

I also plan to determine the dampening factor of my (open-back) over-ear headphones with in-ear mic measurements to increase the speaker volume appropretly.
Funny enough but I tested today Dr. Greisenger update app for achieving equal loudness via comparing frequencies to the reference one and found that I got a huge gap in low frequencies an overall disbalance, I mean non equalized that way HD681 sounds much better. But the way I found described above is a huge improvement into my headphone mixing\mastering work. Actually both for HD681 and HD650 (HD650 still gets it's own vibe even with needed deep for my ears -10db at around 200hz according my tests). After doing impulse responce via rePhase based on Earful measurements I started experimenting with localization in headphones. Three ways got my attention - 112db Redline Monitor (somewhere in between defaul settings), dSoniq realphones binaural ambience only and Shannon Pearce collection of binaural room responces from his github (actually room 01 fits better to mine at 60 degrees). Currently I prefer the last one, sounds so natural. So as you can see without that Earful method I couldn't give any step in proper monitoring with headphones.
 

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,209
Likes
2,674
Location
Salvador-Bahia-Brasil
with the threshold method we can isolate one ear because the pluged ear will never hear sound since the open ear reaches the threshold before.
we can then make a threshold meassurement with the headphone on the same side, import both curves to REW and calculate the diference.
the question is if the percieved FR of loudpeakers is dominated by the "direct ear" only, or if what we hear has the mixed FR of both ears.
this can actualy easy be tested by listening to pink noises and comparing speakers to the corrected phones.

when we aply binaural DSP afterwards, it is obvious that we need the pluged ear. actualy we should be able to create our own binaural correction if we take 4 meassurements; by plugin one ear and then the other - per channel (and a little delay to the crosschannel)
 
Joined
Dec 22, 2020
Messages
38
Likes
6
I am not interested in doing the threshold method, that's why I suggested the two device loudness comparison.
The equal loudness is not my goal, it's to make the headphones sound like the linear (measured) playing speakers, e.g. what Griesinger persues.
1. Comparing the loudness of the same frequency band on hp/speaker is much easier than comparing the loudness of two different frequency bands.
2. The equal loudness threshold (0 phon) curve differs from the 40/80 phon curve suitable for listening to music. (according to iso226-2003)
I know, the base range is mostly affected by this, but it is left for correction and I want two avoid those uncertainities.
3. With one center speaker only left/ right HP loudness needs to be compared directly with the speaker loudness, two steps less than with griesingers method.

But even with a plugged ear the sound is perceptible by it, although it adds much less perceived loudness.
 

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,209
Likes
2,674
Location
Salvador-Bahia-Brasil
The equal loudness threshold (0 phon) curve differs from the 40/80 phon curve suitable for listening to music. (according to iso226-2003)

if we have a perfect reference, and then create our personal equal loudness threshold curve for it,
and then do the same for a second, imperfect source,
we can then aply the diference of both those curves to the second, imperfect source and have a perfect copy (in theory; we imply here that we can make perfect threshold curves) of the reference FR.
the ISO is not relevant here at all, since we are using our own threshold
 

dasdoing

Major Contributor
Joined
May 20, 2020
Messages
4,209
Likes
2,674
Location
Salvador-Bahia-Brasil
Comparing the loudness of the same frequency band on hp/speaker is much easier than comparing the loudness of two different frequency bands

without a doubt. it is without a doubt better then finding the threshold, too....I mentioned the threshold as the best option to measure ear isolated respose of a speaker
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
Why would you pack such a simple and small software with an installer? Just so it bloats my PC after i use it once or twice?

Sorry for the tone but this is really uncommon in my experience, and I'm getting very aware of bloatware recently.
 
OP
pkane

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,632
Likes
10,205
Location
North-East
Why would you pack such a simple and small software with an installer? Just so it bloats my PC after i use it once or twice?

Sorry for the tone but this is really uncommon in my experience, and I'm getting very aware of bloatware recently.

What bloatware? The installer is only 2.5MB, and includes all the software that it needs, and all of these are internally used libraries. There's no other software installed. Here's what's installed, in total, 3.15MB:

1613049964503.png
 
Last edited:
Top Bottom