How can I establish a reference level?

With now 350+ hours on my new rear end (I hope it is OK to call the new speakers that, given the common use of the expression front end for the source) I am trying to understand gain and how it relates to listening level. The terms reference level, anchor level, gain structure, dB, dBU, crest factor, etc. all form an ill defined blur. My interest is in protecting one of my five senses. No amount of money and equipment swapping will ever reclaim lost hearing! Now with oodles of distortion free headroom I need to be careful.

Specifically -
JBL 4367 - 94 dB, 300W
Benchmark AHB2 in bridged mono 380W
SONY XA5400ES Compact Disc Player

What I find is -24 dB set on the preamp (with 0 dB being no voltage gain or cut with respect to the source signal) is too loud on most recordings - especially Pop and Jazz. On the 1964 SONY Classical recording of Petrushka, Ormandy (SBK 47664) a gain of -24 dB is pretty realistic as it is on many other classical recordings.
Some recordings sound loud no matter what the volume. Take Jimmy Smith with Kenny Burrell (Phono 870267) for example (listening to him now @ -34 dB)
Much of my listening is far-field, though the speakers are only about 6' apart in an open floor plan of about 1200 sq. feet.

The inverse square law relating to how loudness decreases with distance from the source - how is it affected by a stereo pair? Get on axis near-field with the speakers and just try and keep your mouth shut. Awesome!

So I have a new definition of LOUD and I want to be careful and have some consistency.
I feel -24 dB is a good reference level and am wondering how that relates live sound and the recording process.

The quick answer is that it is very complicated and impractical to calculate the sound pressure level at the listening position based on calculations. You should measure it using a dB meter, this will allow you work out a safe listening level. General advice is that prolonged listening at greater than 70dB SPL could cause damage... this is probably quieter than most of us are used to.
The slightly longer but very generalised answer is that the loudness at the listening position is dictated by the output level of the amplifier, the efficiency of the speaker and the size of the room. The dynamic range of the source material is also a factor in this as heavily compressed recordings will sound louder.
I'd recommend understanding these relationships if you're trying to spec a system to meet your needs i.e. realising you don't need a 300W amp if you have an average sized room and reasonably efficient speakers. But if you want to protect your hearing you won't get close enough to the correct answer without measuring.
Waste of time. For the simple reason every recording is at a little bit different level. Some a lot different. Some compressed real bad. 

But that's the least of it. You're right, your understanding of these terms is a blur. As in, there is none. Pick one or two. Right now I could write a book trying to address all your misconceptions. A page per word at the very least. Seriously. Baby steps. Take one. Which one do you want to work on first?
When I read “new rear end”, I thought you were going to say you had a butt lift.😂
Your approach is a complicated way to get to a simple answer and will likely not work out. Assuming your primary concern is protecting your ears, first look at what OSHA has adopted for guidelines.
They use science and know. Second, you must understand that while you are listening to music there is a fairly constant change in volume. Maybe some extremely compressed metal music has stretches of fairly constant volume? Third, live music can be much louder than what the OSHA guidelines deem acceptable.

Read the OSHA guidelines, buy a dB meter, play a variety of your favorite music and take some notes about the volume levels (-xxdB) that comply. Repeat as needed. Short peaks are not an issue. Measure at your listening position.

Just sounding (being) uncomfortable or sounding irritating doesn't necessarily mean that you are playing your system too loud. Certain types of distortion are very irritating even at low volumes.
I have pressed the Radio Shack meter into use.
C weighting, slow response.

I set +68 dB up close to one speaker and figure that gives +74 dB of SPL into the room (because there are two speakers and because of room boost).  I hope that this is safe for near-field listening. It seems plenty loud.

As for far-field listening, like when cooking in the kitchen, +88 dB at the speaker seems reasonable.  It is easy to rock the house with higher SPL, but again I want to be careful with my hearing.

No one asked - the input sensitivity on the AHB2 is set to low; 22 dBU (9.8 Vrms).

Empirical results with the RS meter are fine, but I was hoping someone could use the numbers to give me a reasonable idea of loudness from the speakers with respect to the original recording. That would help me gain insight into the process.

+1 @arion, with high volumes being one thing to be aware of when listening, repetitive lower level exposure is cited as a hazard as in industrial settings.
A good HiRez playback System like a Live Music Hall event draw in and demand ones attention.
An enthralling symphony concert can last 2-3 hours at very high levels and extreme peaks without putting anyone in the hospital.
Imagine being in the orchestra playing!
Consider physiological and psychological factors in the evaluation as well.
How your system is integrated into the room can reinforce frequencies in various positions.
Making the system work well is the goal.  
I felt heartened by the OSHA Guidelines. They are pretty liberal or hearing is pretty robust.
Duration of exposure is certainly an issue with home HiFi.
Mike, There really is no direct correlation between the -dB reading on your equipment and the loudness (dB) at your listening position. The levels set during the recording process are all over the map. There are many factors along the audio chain that also affect the final volume.
The audio industry typically measures at one meter as a standard. I suggest that you measure at your listening positions. Room gain varies. The loudness measured at one meter will be considerably louder than at three meters or at your listening position.
The RS meter is robust, durable and pretty accurate- use it as many here have directed.learn to read both the fast and slow for peak and average. Yes the average drives most hearing damage. The OSHA standard may sound conservative but I know and worked w many coworkers who might beg to differ - now.... they may have contributed, but hey we all know better than the scientists....

congrats on your system, enjoy !!!!!


After listening to a lot of music these past few months, I believe that -17dB as set on the preamp is closer to a true reference level.

Lots of music is really loud! I mean how many times can you listen to, say, Aretha singing RESPECT at a realistic level and not go deaf? Or take Aerosmith at full volume? No.

Trifonov playing "The Carnegie Recital" sounds very realistic at that level. Turn it down and it is just like moving further back in the hall. Same with Gould playing Contrapunctus 1-9 on Organ.  WOW!

As an aside, there was a Tip someone posted over in the Analog Forum for us old geezers about trimming our ear hair for better fidelity. And that reminded me that I need to cut my hair!

Happy Holidays!



After pushing the 4367s around for awhile with the powerful SS amps (AHB2's in MONO configuration) i decided to try some lower power tube amps.

First in was a pair of CAD-300SE. They are 300B SET with 11W output. I found that I must be much closer to the speakers and keep the volume way down. First impression was they do a lot of what the AHB2's do but quieter. Then I thought "I guess I don't care what those singers are saying - it sounds so beautiful." Then a mechanical transformer hum became annoying, especially since the volume is down and I had to sit so close. Also, the tube amps are much more finicky WRT line voltage. A Variac was pressed into service for regulation. These amps need 117V not the 121V which is more usual here.

Next I put in the ARC Ref 75SE. That was a bust. V3 flamed with a runaway bias. I called ARC service and they are so backlogged that can't take the unit for 5 months. I am on a waiting list for the repair. Go figure. At least it didn't fry a speaker.

Next was a single AHB2 and it just has a little less jump than the pair of AHB2's.

Next is 25W of glory in the form of a Luxman MQ-88uC.   I LOVE, LOVE, LOVE it! It too needs the voltage down to 115V. Push Pull KT-88 in triode. Beautiful and relaxing sound on the JBL4367 with NO Fatigue! And just enough power to turn it up a bit.

On the way is a Pass Labs XA-25. I can't wait to hear how it plays with the big speakers.

As for reference level, Volumes equivalent to realistic levels are just to tiring though possible with these speakers and the AHB2's

So there you have it. I hope this is entertaining to some readers. It certainly is to me.


Very interesting. i found that when I optimized my playback chain with low noise, low distortion gear (Benchmark DAC3 B + HPA4 + AHB2), I'm able to listen without fatigue at higher volume levels and for as long as I want.


I've owned gear in the past that output higher distortion and they always sounded good at certain volume levels or with certain genres but the Benchmark system lets me enjoy everything I want to listen to without problem.


In any case, if you're interested in how some studios may set their reference levels, here's an article from Sound On Sound - 'Establishing Project Studio Reference Levels'. Maybe you've already come across it yourself. Personally, I listen to music around 72 dB average SPL C weighted when measured with my RadioShack meter. The volume knob setting to achieve this varies with source material (e.g. -20 dB on most classical, -26 on most jazz, and -30 on contemporary pop music). I call it comfortably loud.

FYI: I set my volume level at the minimum acceptable level. I turn it up and I slowly lower the volume to the point where I don’t lose anything. Your long-term hearing health will thank you.

Yage- thank you for the link. I remember seeing that article awhile back and the reread was definitely worthwhile. To be clear, my room is small. The fatigue I speak of with levels 85dB+ has more to do with that than the equipment. Lots of room gain. I have done what I can to treat it but still, high levels are tiring. 

72dB is a good recommendation. Thank you for your response.


In my post on 4/30 I mentioned there was a Pass Labs XA-25 OTW.

Well, it's here and has been up and running for about 3 months.

If you have speakers that SUCK power forget it, but for the JBL 4367s the amp is great. Compared to the AHB2s the XA-25 draws you in. At times with the AHB2s I felt I needed to put up the yellow caution tape, "Do Not Approach." Not so with the XA-25. If I want louder, moving from mid to near-field does the trick. Near-field with the AHB2s is a bit much.

But for a party on the back porch I can open the windows and turn it way up with the Benchmarks. They are more convincing at loud levels.

I'm glad you started this thread, sometimes the importance of ear drum preservation gets "lost in the mix". The problem is some music starts softly, peaks loudly, etc. But it is something that needs attention once I moved from a condo with neighbors to a private dwelling, no limits if you know what I mean.


In house now is a ML No. 534. OMG!!!

This amp and speaker combo effortlessly produces the whole thing.

Like a tapestry.

Noticeable are the dynamics. Quiet and LOUD! 

The JBL 4367 and ML No. 534 play exceptionally well together.


Having not had any parties on my porch for some time, I have switched back to the XA-25 in front of the JBL 4367. I'm glad I did. The 94dB JBLs get as loud as I want with the XA-25 and sound absolutely fantastic. The higher powered amps, ML534 and Benchmark AHB2, seem to prefer higher volumes and this is tiring. On the JBL 4367 the XA-25 gives me fatigue free listening and sounds more like music at all volumes.

Noticeable are the mid-bass textures and overall glare-free presentation. Those 15" woofers are handled with aplomb by the XA-25. It is amazing that 25W will do it, but these are a very special 25W. To paraphrase Dick Olsher, if the first watt is no good who wants the rest?




It is time for a correction. After playing around for a while with the low power amps I have come to the conclusion that they do not sound "more like music at all volumes." Instead each puts its own spin on the sound and gets to sound less like music at higher volumes.


Switching back to the Harmon duo, specifically the ML 534 paired with the JBL4367, noticeable are the MACRO dynamics and KICK. Back is the house filling sound!


So, if you don't feel too silly sitting near-field to these speakers the lower powered amps are fun... Unfortunately though, the first glorious watt (or 40 watts) may ultimately not be enough for the 4367.

I forgot to mention,

Also noticeable with the change is the HUGE soundstage. With the right recordings I am immersed in a sound field that extendsr beyond the boundaries of the room. Be it Alan Lomax "Southern Journey" or Barbie "The Album" the effect is of being surrounded by sound.