The price of ears

Posted by: Huge on 06 October 2015

How much to pay for a calibrated microphone to find out if you're playing music too loud for too long?

 

 

See this from Jan-Erik (who knows rather a lot about hearing damage!):

https://forums.naimaudio.com/to...64#49295590448231664

 

and this for the limits you should keep within:

https://forums.naimaudio.com/to...73#49295590401702973

 

 

Alternatively £75 for miniDSP UMIK-1 which you can also use to calibrate the acoustic properties of your room to see what adjustments you need and to help position speakers.  I originally bought one to finalise the acoustic treatment of my room, but soon realised that checking my listening levels was another very useful function it could perform.

 

 

I found that 'normal' listening for me was about 78dB - 80dB and I consider anything over 85dB average to be very loud.

Posted on: 07 October 2015 by Mike-B
Originally Posted by joerand:
........  I'm reading this thread with interest while wondering about the accuracy of apps from various devices or digital meters from places like Radio Shack. Do these include some sort of calibration to an ambient level?  ..................  
Great for fatcat's hearing if that's where he listens, but how much can we trust these readings? 

 Any meter of any sort in an app is not going to be that accurate,  ditto the 30-50 buck sound meters from places such as Radio Shack.  However for general domestic ambient & music level sound level checks as we are discussing here,  they are as good as they need to be.

Having checked my own RS against my pro-level SPL meter & also per the various www forums on the same subject, the RS units are good for accuracy in the 100-4000Hz range [+/-1dB]  but loose it outside that,  reading up to 5dB low around 32Hz & 2dB high above 4000Hz.  

For measuring specific frequencies its another story, there are correction tables & PCB modifications published for both the RS units. But that is not for this discussion.     

Posted on: 07 October 2015 by VladtheImpala
Originally Posted by Innocent Bystander:

Microphone capsules in calibrated measuring instruments are selected for their level frequency response and consistent sensitivity, as opposed to size and cost which are primary considerations in mobile devices ( which is why proper measuring instruments cost as much as they do) - so don't expect accurate results unless your phone+app combination is calibrated with a reference source (and then valid only over the frequency range calibrated).

 

When I wanted to study something a few months ago and so assessed what was possible with phone or iPad and a free app, I found between 4 and 9 DB difference between simultaneous readings on an iPhone and IPad with the same app, and up to 13dB difference between readings with different apps on the same iPad with a constant source. those are very significant differences, and show that such apps/devices cannot be relied upon to give meaningful absolute readings. They appear possibly to be able to give comparative level readings (same app and individual device), though that is by no means certain if the frequency content varies. 

 

As for sound levels and hearing damage, it depends in part on the period of time as well as the sound pressure level. And it depends on whether you're looking at peaks or averages, and in that regard different music can be very different, with classical orchestral music not infrequently having a huge difference between soft and loud passages, such that the average can be very low even with peaks well over 100dB, while some rock music is may have only small variation all the way through.

 

For reference, IIRC UK H&S controls for workplace noise only start to kick in above 80dB (A-weighted average) based on 8 hours' exposure a day, though with specific controls on instantaneous peaks above 135dB. That would suggest that with an average sound level of 80 and peaks less than 135 you shouldn't have to worry too much unless you're listening like that for more than 8 hours a day. And higher average levels would be tolerable for shorter periods.

 

The first occupational exposure action limit is indeed 80dB(A) in the UK, with mandatory measures applicable at 85+dB(A). However, this is not a "no effect" level as a small percentage of the population will suffer noise-induced hearing loss when exposed to 80 dB(A) over many years. And noise is also a statutory nuisance for environmental purposes!

Regards,

Vlad 

Posted on: 07 October 2015 by Huge
Originally Posted by ChrisSU:

Sonos apparently have faith in the accuracy of the iPhone's mic, as they've just released a digital room correction thing that uses it. It seems that we live in a world where, if your phone can't provide a service, it ain't gonna happen.

Unlike SPL measurement, Digital Room Correction doesn't require absolute accuracy, it just requires a reasonable estimate of the frequency response of the microphone in use.

 

As there are relatively few iOS devices available, and as they're all made for Apple, there are relatively few variants of microphone capsule in use, so knowing which device is in use you can get a reasonable estimate of the frequency response (manufacturing tolerance of maybe +/- 3dB away from nominal calibration, 30Hz - 1kHz) even if the absolute sensitivity still has wide variance.

 

This still isn't ideal for DRC, but Sonos is about convenience, style and achieving a reasonably good standard of sound, rather than the pursuit of optimal sound quality; so the compromise of using an iOS device fits in with their philosophy.

 

But an interesting observation none the less.

Posted on: 07 October 2015 by SB

slighty OT, but I gave my phone app a run out when clay shoooting.

Shotgun blast is > 145dB. App didn't even flicker, so couldn't handle the short peak.

Posted on: 07 October 2015 by Innocent Bystander
Originally Posted by SB:

slighty OT, but I gave my phone app a run out when clay shoooting.

Shotgun blast is > 145dB. App didn't even flicker, so couldn't handle the short peak.

You hadn't shot it by mistake?

 

Posted on: 07 October 2015 by fatcat
Originally Posted by Jan-Erik Nordoen:

Mike, change the weighting to *A* as this more closely represents the ear's frequency response. The reading should drop a bit, say 4 to 6 dB. Now you can turn it some more

 

Jan

Jan

I have a meter very similar to Mikes, an RS33-2055, a digital model that replaced the 2050.

The instructions advise C weighting for music

A-weighting has A-curve frequency characteristics.
This setting causes the meter to respond mainly to frequencies in the 500-to-10,000 Hz range, which is the human ear’s most sensitive range.

C-weighting has C-curve (flat) frequency characteristics. This setting causes the meter to respond mainly to frequencies in the 32 to 10,000 Hz range. Select C-weighting to measure sound levels of musical material.


The 2055 is a clever piece of kit. It displays the real time values of average and maximum sound levels. Plus, it can be set to display the above values taken over a predetermined time.