The loud knob and the bleeding edge

[Charles Jenkinson]

On the back of escape’s limiter testing: http://billfitzmaurice.info/forum/viewt ... =4&t=20370

…and as a noob, I figured I’d broach the issue of where the bleeding edge is when turning the loud knob up - I get nervous (and hot) when I have to turn it up for real, so I need to explore how far I am from the edge by intelligent means rather than just peering from a distance and nervously hoping for the best. Now, I should say the context is tops (the limits are set for subs), and full range one’s (J12L’s) at that, and Bill has mentioned that the power limit at distortion is 150watts for such (…perhaps that’s all I need to know, but it doesn’t take the fight or flight feelings away ). So, using 150watts as the bleeding edge, if I plug into Crown’s power calculator ( http://www.crownaudio.com/elect-pwr-req.htm ) for example: listener distance = 10m,desired level at listener distance = 100dBSPL, sensitivity =102dB/W/m (EQ’d to be flat, everything pulled down below the knee), Amplifier headroom = 3dB (not sure I should have this in), it equates to a power figure of 126watts. Bearing in mind that the figure I want as the OUTPUT of the calculation is the dBSPL value (i.e. the calculated watts is very sensitive to the dB figure inputted so 126watts is round abouts), if we ignore amp head room (my real amp is 300w/chan so head room is reasonably present) one gets '103dB at 10m for 126watts'. And this would be for outdoor conditions (See ‘Other Considerations’, in How much power do I need: http://www.crownaudio.com/how_much_power.htm ) . It would typically read 6dB higher indoors.

So, as a secondary check, if I was playing a commercially produced/mastered CD I could put my DMM across the amp terminals and read of the order of 32volts (128 watts at 8ohms, as per Ian’s look up tables http://speakeressentials.com/pages/8-oh ... okup-table ).

Are these scenarios the ‘real world’ numbers I would get, assuming my DMM and sound meter are reading ok?

[Radian]

Good Morning Charles,

For "real world" keep all your math entirely in dB and work from a target SPL you'd like to see at FOH or mid-crowd, and work back into the system from there.

The only time you should be taking into account real power (watts) is to spec what'll be required out of the wall or from a generator set.

For example:

32V from 2.83V is 21 dBV

Subtract a dB or two for power compression and add that to the "sensitivity" value in the Crown calculator to see if you have enough cab for the job.

Once you get used to working entirely in dB, you'll be amazed at how hard you can comfortably drive your system. No sweat.

[Charles Jenkinson]

I don't have a crowd or target SPL Radian, ...or a method for working the system out how you would...?

[Radian]

A good target SPL at crowd center would be 100 dB SPL peak from the tops and 110 dB SPL peak from the subs.

Start there.

Say the center of my crowd is 7 meters out.

So for the tops you desire 100 dB SPL at 7 meters

Using the inverse square law calculator:
3.2 ft reference distance
23 feet new distance

Put in an arbitrary value for sound level at reference and click "calculate".
Keep bumping up the sound level at reference number and clicking "calculate" until the sound level at new distance value is 100 dB.

It would take a source capable of 117 dB SPL peak to put 100 dB SPL peak onto your target, barring any other acoustic factors.

Working out the subs would be the same way.

*edit* Fixed a number error. Thank you Charles.

[Charles Jenkinson]

Ok, thank you.

The part about referencing it to 1m (though never measured at 1m) seems to make sense due to the SPL reference curves being referred to 1m. Having put the numbers into it, I know I'm not at the B. edge for the scenario we/you are proposing, re:Below. The point is I've referenced it back to watts, which seems intuituve to me since that was the parameter given that defined where the BE is, in this case for tops (150watts / distortion), when I was choosing an amp.

Therefore, is this the question...:
Is there a 'keeping it all in dB' definition of the BE at the 1m reference position? (such that if one is above or below it one needs another cab or doesn't need another cab).

Apologies if this is really simple/banal - I've not measured SPL at a gig, ever.

Edit: Rad's, your first post above changed in the meantime. Just hang on while I follow it.
Edit2: That's not an accusation! Just an observation, .

[Radian]

Is there a 'keeping it all in dB' definition of the BE at the 1m reference position?

Yes. The limiting voltage of the design (thermal or displacement), which is either in the plans or can be provided by Bill.

That's why the charts are referenced to 2.83V @ 1m

1m gives you a foundation for inverse square law calculations (distances & dB SPL), and 2.83V is used as the other half of the voltage ratio to determine dBV.

Example:

For a random sub, the displacement limited voltage is let's say 50V. Using the dB voltage ratio calculation along with 2.83V yields 24.9 dB above baseline in the chart.

The caveat here is power compression...which above 20 dB of gain conservatively is ~1.5 dB. So subtract that from 24.9 to get a final "Bleeding edge" output of 23 dB SPL over baseline...(once again) barring any other acoustic factors.

I've not measured SPL at a gig, ever.

Better get a meter soon. It is an essential tool of the trade. Once you get one, you'll hear how conservative those target numbers I quoted above actually are.

[Michael Murphy]

The pieces of the puzzle are coming together

Yes. The limiting voltage of the design (thermal or displacement), which is either in the plans or can be provided by Bill.

That's why the charts are referenced to 2.83V @ 1m

1m gives you a foundation for inverse square law calculations (distances & dB SPL), and 2.83V is used as the other half of the voltage ratio to determine dBV.

Example:

For a random sub, the displacement limited voltage is let's say 50V. Using the dB voltage ratio calculation along with 2.83V yields 24.9 dB above baseline in the chart.

The caveat here is power compression...which above 20 dB of gain conservatively is ~1.5 dB. So subtract that from 24.9 to get a final "Bleeding edge" output of 23 dB SPL over baseline...(once again) barring any other acoustic factors.

What converter are you using, this one does not equate to 85v = 30db
http://www.sengpielaudio.com/calculator-volt.htm

There is no direct relation ship between volts and dB. You have to convert to power. 2xPe=3dB, 10xPe=10dB, 100xPe=20dB, 1000xPe=30dB.

what you guys are saying is that you take the max wattage of the driver convert it into volts

Maximum thermal power is moot, it's displacement limited voltage that counts.
So, what's next on the agenda, how many angels can dance on the head of a pin?

So rad thanks for helping me figure that one out.

[Radian]

So rad thanks for helping me figure that one out.

Your welcome.

[88h88]

Better get a meter soon. It is an essential tool of the trade. Once you get one, you'll hear how conservative those target numbers I quoted above actually are.

I ran the ECM8000 setting up for my last gig and the levels were comfortably about 110db at the back of the room. The problem with Bill designs is they sound stupidly clean and distortion free so although you can be running them exceptionally loud they don't 'sound' like they're at ridiculous levels. I'll make sure to get some video of the RTA and such next time.

[Charles Jenkinson]

If running separate subs then, what signal is used to drive the subs and how is the subs dB-SPL peak level measured? Is it just a case of muting the tops with a suitable CD playing, and setting the meter to 'fast & peak (C weighting)' and measure at the crowd-centre distance.

[Michael Murphy]

Well yes and no

I ran the ECM8000 setting up for my last gig and the levels were comfortably about 110db at the back of the room

In this instance, I pretty sure 88 was measuring his total rig and got the 110db across the room on the opposite side of his rig. If you are interested in overall SPL thats whats was measured.

If you are interested in your subs output only then yes the reading is taken only with them playing, C weighted yes.

Distance where to measure is not a defined point, if you are interested to find out what the SPL is at the crowd point or listening point well yes then put yourself at that point. If you are documenting you gears ouput well then its the whole outdoors thing measurements taken @1m/1w or 10m/10w.

If you know the distance from the point of measuring from the source then the 1m/1w can be worked out. I have to sit and write it down but im sure Grant or MC or Rad or RyanA could give us a good explanation using 88's 110db and lets say it was 30m from the source.

The db meter is a good tool that can be used many ways.

Ps regarding the signal to be used, if you are just interest in an average output level as in the case of the listioning position the your fav track is fine. If documenting you will need a specific value to refer to so it would be anywhere from 20Hz to 120Hz or some other defined input.

[Grant Bunter]

Hi Charles,
If A and C weighting are your only options, then go with C.
Ideally though, IMO, no weighting is best, because the A B and C curves are meant to replicate some correlation to human hearing.
I would much prefer to know maximum SPL, not some modified version.

Sure thing, mute the tops, and use music if all you're after in the bandpass you're measuring is an average figure.
And, as always, if you want specific information about system response, then, outdoors, plenty of clear space, tone generator and a reference mike is the go.

[Michael Murphy]

From a PM

"Each doubling of distance has a drop off of 6dB in volume. So you have to measure at 1 metre, or 2 metres, or 4 metres, 8, 16, 32, 64 etc.

Since you don't want to expose yourself to 120+ dB, you don't measure at one metre. Measure at 16 metres, and add 24dB to calculate your 1 metre reading. Or 8 metres and add 18dB, or 32 metres and add 30dB, and so on."

So in the theoretical scenario I posted

If you know the distance from the point of measuring from the source then the 1m/1w can be worked out. I have to sit and write it down but im sure Grant or MC or Rad or RyanA could give us a good explanation using 88's 110db and lets say it was 30m from the source.

If we modify the suggested 30m to 32m the SPL at 1m would have been 110db + 30db = 140db

Ideally though, IMO, no weighting is best, because the A B and C curves are meant to replicate some correlation to human hearing.
I would much prefer to know maximum SPL, not some modified version.

So that the purpose of the A B C or no weight? Human ear factors?

[Charles Jenkinson]

I worked out Radian's 7 metres for the hell of it:

dB = 10 log (1/49) = -16.9dB

i.e. 17dB down at 7m, on the value at 1m.

I get the idea of no weighting, but the C scale is pretty close to flat enough down the bottom.

[Radian]

If running separate subs then, what signal is used to drive the subs and how is the subs dB-SPL peak level measured? Is it just a case of muting the tops with a suitable CD playing, and setting the meter to 'fast & peak (C weighting)' and measure at the crowd-centre distance.

The calculations are used for figuring out how much rig to bring to the job. Using the numbers the other way around (measuring the system and comparing that data directly against the equipment specifications) requires some due diligence.

If you deploy a system then proceed to simply measure it without any situational awareness, you could run smack into the caveat I kept referring to in my previous posts...."other acoustic factors"

In short, anytime you test or sample something, you'd better be damn sure of the circumstances under which everything is taking place and the limitations of the equipment.