[CraftyClown]

So I'm really making an effort to manage my listening levels and with some help from other forum members I now have a rudimentary SPL measurement setup for my IEMs, consisting of a Vibro Veritas coupler (calibrated with a multimeter and some Etymotic IEMs) and the SPL meter in REW.
 
I can now take a rough measurement of music playing in my DAP through my IEMs into the coupler, however I'm a little unsure of what results I should be taking notice of.
 
I'm aware that if I want to listen to music for a prolonged amount of time, let's say up to 8 hours, then I shouldn't excede 85db, however I have no idea whether it is the max level or the peak ;evel that should be monitored.
 
I've spent several hours trying to google this, but unfortunately I'm still none the wiser. Could someone help me out here please? I don't think my hearing is in the best shape after decades of unintentional abuse and I really don't want to make it worse.
 
Cheers
 

 

[RRod]

For listening to music, some averaged value will be most pertinent as you probably won't be hitting peak levels that would be instantly damaging. There are of course standards, but obviously companies care less about your ears than you do and fight tight regulations, so it pays to be conservative. If your meter is reading a pretty constant 85dB (A-weighted, slow reponse), then I'd say that is pretty loud and you might consider only having one "concert" of a couple of hours at such a level each day. A lower level would mean you can jam out longer, and higher level shorter. I tend to prefer the latter and crank it when I listen but don't have music on constantly. See what sounds bordering on "too loud" to your ear for a favorite album and see where that ends up through the coupler into the meter. That might be a good place to start for a ceiling.

 

[Joe Bloggs]

The main problem is, how do you calibrate your mic's loudness metering in the first place?

 

[CraftyClown]

The main problem is, how do you calibrate your mic's loudness metering in the first place?

My ER4-XRs are 104db @ 0.2v so I play a 1khz tone and attenuate my volume until my multimeter reads 0.2v

I then calibrate REW with my 104db tone played into the Veritas coupler.

This is the rough calibration I then use to measure my IEMs. I'll usually play a random selection of music for about 20 minutes and monitor the averaged A weighted SPL.

I know it's flawed, but it should give me a ballpark attenuation of my volume that is within safe levels.

 

[GRUMPYOLDGUY]

My ER4-XRs are 104db @ 0.2v so I play a 1khz tone and attenuate my volume until my multimeter reads 0.2v

I then calibrate REW with my 104db tone played into the Veritas coupler.

This is the rough calibration I then use to measure my IEMs. I'll usually play a random selection of music for about 20 minutes and monitor the averaged A weighted SPL.

I know it's flawed, but it should give me a ballpark attenuation of my volume that is within safe levels.

 
Assuming flat gain is usually not a good assumption. 

 

[CraftyClown]

   
Assuming flat gain is usually not a good assumption. 

 
Sorry, could you elaborate a little on that?

 

[GRUMPYOLDGUY]

   
Sorry, could you elaborate a little on that?

 
Gain at 1KHz will not necessarily be the same at some other frequency. On the transmit side or the receive side. 

 

[CraftyClown]

   
Gain at 1KHz will not necessarily be the same at some other frequency. On the transmit side or the receive side. 

 
Oh I see, yes of course.
 
For sure it's a flawed measurement system, but I figured it should at least give me ball park figures and guarantee I'm not going way too loud with my listening.
 
Could you think of a more accurate way of doing this? At least one that doesn't involve me buying expensive measurement equipment 

 

[GRUMPYOLDGUY]

   
Oh I see, yes of course.
 
For sure it's a flawed measurement system, but I figured it should at least give me ball park figures and guarantee I'm not going way too loud with my listening.
 
Could you think of a more accurate way of doing this? At least one that doesn't involve me buying expensive measurement equipment 

 
You'd have to sweep a tone at a constant level across the band to characterize the response of the mic and ADC front end. To generate that swept tone you could use the FR data for your Ety's (which I assume is available, somewhere) and subtract it from a boxcar in the frequency domain to get a filter response. Then take the IFFT of that to generate coefficients for a FIR filter to run your LFM through. Not ideal, but cheap (aka free), and if your filter length (i.e. number of points in your FFT/IFFT) is long enough it should be okay. Now you can use the post-filter swept tone to generate a wav file. In theory, this wav file should modulate the tone as it goes through the frequencies so that the output level (SPL) is constant throughout. The assumption is you source is flat and very clean (not likely, but again we're going for cheap). What you measure on the recording side will give you the response of the mic and front-end, which you can now cal out in your measurement. You'll have to crank through some math to figure out what the expected SPL is for the test file, but it's not very involved if you have good specs for your hardware. With all of that, you could test all the music your little heart desires, instead of just a tone.   
 
It seems like more work than it is... GNU octave is an open source tool that is pretty easy to use which will have a lot of these functions. The hardest part will be translating the FR chart for your IEMs into a vector to use in octave. 
 
I wouldn't fault you for doing it your way though. The process above is pretty imperfect. My post was just meant to be a caution about the pitfalls...
 
And for those still keeping score at home:
This would be a case where using (A) weighting would be useful. But here we're trying to actually determine what the sound will be like at the ear. When trying to determine what the sound will be like at the output of a DAC upstream - not so useful. Sorry, had to sneak one in there. I'm done now. 

 

[CraftyClown]

   
You'd have to sweep a tone at a constant level across the band to characterize the response of the mic and ADC front end. To generate that swept tone you could use the FR data for your Ety's (which I assume is available, somewhere) and subtract it from a boxcar in the frequency domain to get a filter response. Then take the IFFT of that to generate coefficients for a FIR filter to run your LFM through. Not ideal, but cheap (aka free), and if your filter length (i.e. number of points in your FFT/IFFT) is long enough it should be okay. Now you can use the post-filter swept tone to generate a wav file. In theory, this wav file should modulate the tone as it goes through the frequencies so that the output level (SPL) is constant throughout. The assumption is you source is flat and very clean (not likely, but again we're going for cheap). What you measure on the recording side will give you the response of the mic and front-end, which you can now cal out in your measurement. You'll have to crank through some math to figure out what the expected SPL is for the test file, but it's not very involved if you have good specs for your hardware. With all of that, you could test all the music your little heart desires, instead of just a tone.   
 
It seems like more work than it is... GNU octave is an open source tool that is pretty easy to use which will have a lot of these functions. The hardest part will be translating the FR chart for your IEMs into a vector to use in octave. 
 
I wouldn't fault you for doing it your way though. The process above is pretty imperfect. My post was just meant to be a caution about the pitfalls...
 
And for those still keeping score at home:
This would be a case where using (A) weighting would be useful. But here we're trying to actually determine what the sound will be like at the ear. When trying to determine what the sound will be like at the output of a DAC upstream - not so useful. Sorry, had to sneak one in there. I'm done now. 

 
 
Thanks Grumpy,
 
I shall certainly look into this method when I have a little time free.

 

[PETEBULL]

Why bother doing measurements that could as well be erroneous. Have you taken into account natural ear resonance at 3000-4000Hz, that amplifies resulting loudness? You will have to reconstruct a true rubber ear model and fit mic into it just like here https://www.youtube.com/watch?v=DhtLFDjbf0E
Because of such as you, most smartphones are equipped with nannyware that doesn's allow you use high volume for extended period of time or bugging you with messages. Just find the volume at which you hear ALMOST everything and be done with it.

 

[CraftyClown]

Why bother doing measurements that could as well be erroneous. Have you taken into account natural ear resonance at 3000-4000Hz, that amplifies resulting loudness? You will have to reconstruct a true rubber ear model and fit mic into it just like here https://www.youtube.com/watch?v=DhtLFDjbf0E
Because of such as you, most smartphones are equipped with nannyware that doesn's allow you use high volume for extended period of time or bugging you with messages. Just find the volume at which you hear ALMOST everything and be done with it.

Funnily enough, some of us care about our hearing and maintaining it into old age.

"Just find the volume at which you hear ALMOST everything and be done with it"

and what if that volume is excessively loud to the point of damaging your hearing?

My efforts may be flawed but at least I'm going to be in the right ballpark.

"Because of such as you, most smartphones are equipped with nannyware that doesn's allow you use high volume for extended period of time or bugging you with messages"

OK, so because of me people are being suggested to protect their hearing. I guess by that same logic, because of you people are going deaf?

 

[PETEBULL]

and what if that volume is excessively loud to the point of damaging your hearing?

Maybe there is tolerance after you "hardened" your eardrums so that you could bear more db without degrading further but this is more of a medical research. Personally I can sense when it's time to stop rotating volume knob. Moreover if frequency balance is bad, microphone may tell that there are few db but your ears are in pain and vice versa. For instance sub bass does almost no negative impact.
 

My efforts may be flawed but at least I'm going to be in the right ballpark.

If your efforts are 20db flawed, you may get false sense of security.
 

 OK, so because of me people are being suggested to protect their hearing. I guess by that same logic, because of you people are going deaf?

I mean this turns into a social issue. You won't get powerful enough portable just because of forced regulations. That does not solve the problem but forces you to use higher sensitivity-lower quality headphones.