• 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!

Determining safe level of headphone volume with umik-1

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,874
Likes
6,672
Location
UK
Yes, I mean A weighted (or so I trust the manufacturer). My sound meter only has the one dB(A) unit, as the C weighting option is only for its higher cost sibling. I recorded the averages using the max and min functions It has, which correspond to those fluctuations yes.
Right ok, you recorded the range of dB(A) values it was reading out. Well I think that's ok, and afterall I have a higher 1kHz point than you anyway and I was seeing 80dB(A-weighted) average across my loudest track, so I'd expect you to be under that, and given the UK Safety Regs re noise then it falls within the green zone on that table I showed earlier.
 

Aerith Gainsborough

Addicted to Fun and Learning
Joined
May 4, 2020
Messages
853
Likes
1,280
I think it can't be trusted too much to average real music
Such a signal can be a tough cookie to average, yes. That's why I always looked at the instantaneous Z-weighted peaks.
It's easier to let a program such as MusicScope handle the averaging of music files and do the rest via math in your head.
 
OP
Wendigo79

Wendigo79

Member
Joined
Apr 27, 2020
Messages
50
Likes
24
Back again, haven't been participating to this thread which I started. Been bit busy. Much of this stuff is quite advanced and so I gave up halfway of this thread when reading it now (did read it, but gave up understanding) :) But today postman had brought me my new FiiO BTR3K bluetooth dac/amp. I noticed when studying it, that it has the ability to calculate the SPL for the headphones which are connected to it. It estimates the output by using the Ohm and sensitivity values.I tried it first with AudioTechnica m50x. I entered the values given by AT, but when I used the UMIK to compare, I got ~10db lower values. Then I entered those values that were in the @amirms review, I got pretty accurate reading that corresponded.AT says "impedance 38ohm, sensitivity 96". @amirm " 39ohm and sensitivity 86".

But then I tried the same method for Sennheiser 560s and got completely bonkers values. With both, @amirms values and Sennlheirsers.

I also found an Android app called "HearAngel". It does the same thing, but you cannot enter values manually. It has small database of headphones. And luckily ATH-m50x is supported. It gives reasonable values. Did not compare that with UMIK though. And it also works only with phones own 3,5mm output. No support for external dac. But interesting project anyways
 

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,874
Likes
6,672
Location
UK
Back again, haven't been participating to this thread which I started. Been bit busy. Much of this stuff is quite advanced and so I gave up halfway of this thread when reading it now (did read it, but gave up understanding) :) But today postman had brought me my new FiiO BTR3K bluetooth dac/amp. I noticed when studying it, that it has the ability to calculate the SPL for the headphones which are connected to it. It estimates the output by using the Ohm and sensitivity values.I tried it first with AudioTechnica m50x. I entered the values given by AT, but when I used the UMIK to compare, I got ~10db lower values. Then I entered those values that were in the @amirms review, I got pretty accurate reading that corresponded.AT says "impedance 38ohm, sensitivity 96". @amirm " 39ohm and sensitivity 86".

But then I tried the same method for Sennheiser 560s and got completely bonkers values. With both, @amirms values and Sennlheirsers.

I also found an Android app called "HearAngel". It does the same thing, but you cannot enter values manually. It has small database of headphones. And luckily ATH-m50x is supported. It gives reasonable values. Did not compare that with UMIK though. And it also works only with phones own 3,5mm output. No support for external dac. But interesting project anyways
That sounds interesting re your FiiO DAC/amp that tells you the SPL - is that a number that changes as the music is playing, or does it just tell you the max theoretical output for any given "volume knob position"?

For your measuring of SPL, if I look back at your initial posts, you might want to try what @Aerith Gainsborough suggested you do to try to improve your accuracy of measurement.
 
OP
Wendigo79

Wendigo79

Member
Joined
Apr 27, 2020
Messages
50
Likes
24
It shows it like that. Value changes with dac/amp volume setting. The phones volume does not affect. So it needs to be at max or else the amp shows it wrong. I'm currently listening with it and it sounds OK. Actually quite good. With LDAC it can be compared to wired dac. It can also be connected directly to the phone or pc by usb. Haven't tried that yet.

I measured it just with holding my hand over the mic on the "cup". I think it simulated head and ear quite well :)
 

Attachments

  • Screenshot_20220412-193358.jpg
    Screenshot_20220412-193358.jpg
    73 KB · Views: 60

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,874
Likes
6,672
Location
UK
It shows it like that. Value changes with dac/amp volume setting. The phones volume does not affect. So it needs to be at max or else the amp shows it wrong. I'm currently listening with it and it sounds OK. Actually quite good. With LDAC it can be compared to wired dac. It can also be connected directly to the phone or pc by usb. Haven't tried that yet.

I measured it just with holding my hand over the mic on the "cup". I think it simulated head and ear quite well :)
How do you mean it shows it wrong? If you run it below max output then it will lower it's quoted SPL figure, so how do you know it's wrong?

About your HD560s, the Sennheiser specs are 110 dB (1kHz, 1 Vrms) for it's sensitivity, and I've just measured my Unit#2 of HD560s on my miniDSP EARS and got 110.7dB (1kHz, 1Vrms), so the reported values from Sennheiser I think are very accurate. Here's my screenshots of left & right channels at 1kHz and 1Vrms:
HD560s Sensitivity Unit#2 - left channel.jpg HD560s Sensitivity Unit#2 - right channel.jpg
So I think you can be quite confident about the figures you're putting in for the HD560s.
 

kthulhutu

Active Member
Joined
Mar 15, 2022
Messages
132
Likes
113
I've done some reading around to try to guage what are safe volumes to listen to. The main source I'm gonna go off here is the UK Health & Safety Executive where they've investigated this area & come up with some guidance for employers, find the whole pdf file attached at end of post.

The crux of their measuring is based around a daily allowable Noise Dose which is expressed in percentage, with 100% being your maximum permitted Noise Dose for the day. They use the following two terms/variables to calculate/determine the Noise Dose: Lep and Lcpeak:
View attachment 198892
Lep is an average dB that you have been exposed to for that set time period, whereas Lcpeak is a peak single upper limit value (C-weighted) that is not to be crossed (so that's not an average).

They have two thresholds, an upper & lower limit set on these two previously mentioned variables whereby employers have to take certain actions to lower hearing damage risks. Here's the Lower Threshold:
View attachment 198894
And here's the Upper Threshold and an important thing to note is that the Upper Threshold is equivalent to a 100% Noise Dose for your whole day (24hr period I believe):
View attachment 198895

They end up collating/boiling everything down that we've talked about here into one table that you can reference to determine your percentage daily Noise Dose that you've received from any given Lep (dB A-weighted), this seems highly useful and I think I'm going to endeavour to remain firmly in the green zone:
View attachment 198897
In the above table the previously mentioned Lower Threshold sits on the bottom edge of the Yellow Zone, whilst the previously mentioned Upper Threshold sits on the bottom edge of the red zone. You can see that a 100% Noise dose is associated with the Upper Threshold value that they previously mentioned which is a Lep of 85dB for an 8hr work day - you can read that off the above table.


So with this in mind, and me wanting to stay in the Green Zone then how do I interpret the sound level graphs that I captured from the tracks I listen to. Well here's a graph I created this morning based on A-weighted measurement which is the standard applied to the noise guidelines we've been talking about, following is a measurement of probably my loudest track I listen to recorded at my typical loud listening level:
View attachment 198899
In the above graph the Lep is 80.5dB which is the lower purple line in that graph. If we look in the previous table we can see a Lep of 80dB over a 4hr period is still in the green zone, so according to that I'd be ok listening to that track at the volume level for 4hrs as it's below their Lower Threshold. However, they do cite an Lcpeak (C-weighted peak) as maximum level that would not be crossed at any point, so I did a C-weighted measurement of the same track, which is in the following graph:
View attachment 198904
To be honest it looks like REW doesn't record Lcpeak, but it does record Lzpeak, with Lzpeak being Z-weighting rather than C-weighting - however we know that C-weighting is a bit more forgiving than Z-weighting because Z-weighting is literally just a flat line (no roll off in frequencies at any point), and as we can see from the following graph the C-Weighting rolls off in the bass:
View attachment 198905
So given we know the Lzpeak in my recorded tracks are happening in the bass, and the C-weighting rolls off the bass then we can loosely interpret that we can shave off around 3dB from Lzpeak to roughly arrive at Lcpeak. So my Lzpeak was 102dB in that graph, so that's likely around 99dB for Lcpeak which is a long way below the 135dB Lcpeak of the Lower Threshold (yellow area on the table).


So in conclusion, I'd say firstly to make sure to keep your Lzpeak (below 135dB), and probably actually a fair bit lower because Lzpeak will be reached repetitively and often in music by the nature that it would be the bassline of the track so would happen every few seconds perhaps, whereas in the workplace you might expect your loudest noise events to be happening maybe say 3 times an hour, so I'd say be careful about Lzpeak.....most of my tracks I've measured don't go above 104dB Lzpeak, so that should be fine I'd say. Secondly I'd say use that table I posted earlierView attachment 198897 and stay in the Green Zone. I'll probably aim for about half the threshold of the Green Zone, so 80dB Lep for 2hrs even though they have 4hrs as a maximum to stay in the green zone for that level......I suppose I have a bit of noise exposure at work and during the day, so I don't want to max out my daily dose with 4hrs of headphone listening at 80dB Lep whilst adding work exposure onto that too.....because afterall these are based on Daily Dose criteria. I also don't listen to headphones for 4hr in total anyway......although I sometimes could do that if gaming, so keep your headphones on low volume if you're doing long gaming sessions.



EDIT: just an additional piece of information, the same body that created these guidelines (HSE UK), also researched the long term effects of damage from noise exposure (or found it somewhere else, they don't cite a source), and that was actually their introduction to their presentation, here's the graph:
View attachment 198893
I'm assuming the units of the x-axis there is Lep for an 8hr period (even though they don't say). So I interpret that as an 85dB Lep for 8hr everyday gave a 5% increase in chance of someone suffering 30dB hearing loss at age 60.....although that says nothing for other lesser amounts of hearing loss or for hearing loss potential at younger ages. 85dB Lep for 8hrs is a 100% Daily Noise Dose, as seen in that colourful red/yellow/green table - just sitting in the lower edge of the red zone.

EDIT #2: I wonder if there is any other hearing damage noise exposure research out there that has been done that is contrary (contradicts) the information I've shown here, sometimes it's good to get other perspectives, anyone else found different research/guidelines that contradict the info here?
This is quite reassuring honestly. Even if was exposing myself to a high dose of 85 dba every day (which I don't) I'd only have a slightly increased chance of hearing loss. With the clarity of my system, I find even 80 db (z weighted) to be unnecessary so most of my listening is kept under.
 

Robbo99999

Master Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
6,874
Likes
6,672
Location
UK
This is quite reassuring honestly. Even if was exposing myself to a high dose of 85 dba every day (which I don't) I'd only have a slightly increased chance of hearing loss. With the clarity of my system, I find even 80 db (z weighted) to be unnecessary so most of my listening is kept under.
Yeah, it's quite hard to research this kind of thing (I mean googling it & and coming up with a conclusion).....that was the best I could do in terms of finding some standards and applying them to music listening, as the guidelines are not intended for music listening per se, but I'm sure there's some crossover that can be made.
 
Top Bottom