Reference level and calibration

TB Rich

Prominent Member
Joined
Jan 17, 2019
Messages
2,487
Reaction score
1,642
Points
672
Age
40
Location
Bournemouth
I've got an old AVR that has no auto setup, so it rely's on pink noise generation and manual adjustment of the channels. I decided just now to check with my UMIK-1 and REW how good a job I had done by ear balancing them all - turns out spot on! However to get ~75dB of pink noise referenced in REW, my AVR master volume is -5dB.

So, if I had auto calibration on a modern AVR, am I correct in thinking it would calibrate the channel trims so that they would be putting out ~75dB with the master volume at 0dB?
If so then I believe that means my listening level of -22dB is actually equivalent to -17dB on a system calibrated against reference?

I appreciate the only thing that matters are the channel balances being level, and that master volume is simply a means to an end. But I've always felt I listen at a decent enough volume and have often wondered how on earth people tolerate reference listening! This would at least explain I'm closer to reference than I thought.

Cheers
 
If getting a measurement of 75db at -5db master volume then you'd technically be getting 80db reference at 0dnb master volume. The 0 setting on the master volume scale shouldequate to what is relative to reference and would ordinarilly be the master volume setting you'd calibrate the speakers to so that they measurer 75db relative to. 0db Master volume equates to the master SPL your setup is calibrated in association with.

The speaker levels would be adjusted so that each speaker measures 75db as measured from your primary listening location. This location should be where the calibration mic in placed initially. The receiver would output a test tone at 0db master volume and the auto calibration would adjust the receiver's speaker levels in order to attain a 75db measurement from each of the speakers.

This gives you the assurance that you'd be correctly attaining reference if setting the AV receiver's master volume to 0db (75db). This is below the reference used during the prodiction of film soundtracks and would be less than the reference used in movie theatres, but is used for home theatre setups. Being able to attain this level and knowing where it is gives a user the ability to easilly know at what volume they need to set the AV receiver in order to portray content at reference. This will portray the content with roughly the ame dynamic range as perceived by the person who mixed the film's soundtrack.


 
If getting a measurement of 75db at -5db master volume then you'd technically be getting 80db reference at 0dnb master volume. The 0 setting on the master volume scale shouldequate to what is relative to reference and would ordinarilly be the master volume setting you'd calibrate the speakers to so that they measurer 75db relative to. 0db Master volume equates to the master SPL your setup is calibrated in association with.

Ok thanks that answers what I was asking and assuming, and my listening therefore isn't quite as quiet as I thought it was.
I have taken out 5dB from each channel now, so that 0dB master vol now gives pink noise at 75dB in the listening spot.

Doesn't really change a lot, other than when I watch a movie later I'll need to use -17dB and not -22dB! But it's nice to have it accurate now in relation to other people which is of course the point of the relative volume scale.

So when modern AVR's run to set distances and levels - do they do it with pink noise or sweeps? Because I'm fairly sure if I ran a sweep now generated from REW and the AVR set to 0dB master vol - it'd probably blow my socks off and be way past 75dB at my seat!
 
The actual tones used are all pink noise in nature, but different systems output the tones differently. The calibration only really needs a 20 to 20.000Jz test toneto be able to calibrate the levels and the distances. The tone itself isn't 75db and the AV receiver is the device that controlls how much or how little the signal is amplifide or reduced in order to attain reference. Sweeps are more relevent to EQ adjustments than they are in relation to the distance and or level settings.

The actual volume at which the test tone or sweep is portrayed is dependant upon the AV receiver and the master volume setting. REW has its own output setting that controls the level you output the associated test tone. This can vary the voltage associated with the output signal, but doesn't determine where you locate reference onboard the AV receiver. If the test tone seems excessive then reduce the REW output or ensure it is set to its default.

Pink noise portrayed at 75db wouldn't actually sound inordinately loud. 75db is loud though if pot=rtraying an action movie and most home theatre users don't actually portray audio at reference. It would be tolerable, but maybe not very condi=ucive with what your neighbours or other family members would like to hear?
 
Last edited:
The test tone noise the AVR generates is the static noise variety, and at 0dB on master vol (and channels now adjusted), REW is recording this as being 75dB at the seat (using REW's SPL meter tool, and a calibrated UMIK-1).

If I use REW to generate a 10-20k signal sweep (going in to a L/R input on the AVR), then the master volume needed to hit 75 dB at the seat is only -12dB. (That's of course an average 75dB if you were to draw a line through the response).

Does this mean simply the signal REW generates is therefore likely too high (I have the sweep level at -12dBFS)?
However it seems to coincide with the levels at which actual audio is portrayed from my sources too. I.e, if I ran my AVR at 0dB playing a BluRay it would be deafening and easily beyond 75dB, but at -12dB it would be loud for sure but not crazy (esp given my new default/comfortable volume now equates to -17dB).
 
You are using REW to modify the EQ and not to balance the speakers. You could use the output setting on the REW device to adust the level so it conrisponded with the AV receiver's own calibratyed state, but this isn't structly nescessary if only using REW to analyse the EQ curves.

If using the AV receiver's own test tone to balance the speakers then the receiver's levels should be correct and it would be an issue with the source if not getting audio portrayed at the correct level if the receiver is set to 0db relative to reference. It would then be a matter of adjusting the source to correspond with the correctly level output of the AV receiver.

Note that there's also the matter of the quality of the SPL meter you are using compared to the accuracy of the UMIK mic you are using for REW? It is unlikely that both would give you the exact same measurements.
 
The Auston Jerry REW guide is also a good reference point to ensure you're set up properly

 
I was just surprised that a closed loop system like REW - i.e the same system being in charge of output level and monitoring level - was not giving the same levels as the AVR calibration tone levels.

The only thing that matters are the sweep results, post EQ, which I am more than happy with how to do this.
Got very much side tracked this afternoon though with tweaking main speaker placement seeing as I had the gear out again, massive rabbit hole that was! Long story short is they were in the best place, but I've manage to just align them slightly better to pick up a few db for a more linear response out from a few k to 20k. I'll give it a music test in a mo :)

But this means that my 2nd BK that's due to arrive this week, definitely won't be going in the other corner now because pulling out the speaker to make room for it very much upsets things.
 

The latest video from AVForums

TV Buying Guide - Which TV Is Best For You?
Subscribe to our YouTube channel
Back
Top Bottom