After First Reflection

[SAC]

...and then you will have a program calibrated to an SPL meter (which was calibrated how, exactly? I know for a fact that my RS SPL meter is as accurate as they come ) instead of to a reference voltage or sensitivity.

Bottom line, by that method, its still relative. But at least you won't be arguing amongst 'yourself' over any differences! (At least not until you discover the non-linearity of the SPL meter!)

For our acoustic purposes we are concerned with the relative differences in dB SPL between various energies arriving with respect to time in the same measurement (or group of measurements of the same 'session'.).
If you need absolute levels (which I would assert few do, as I would suspect that few have ever encountered such a device), calibrate it correctly to a true voltage reference source.

For those dealing with digital and hardware based 'hard' limits, stick with the optimal dbfs scale.

[liscio]

Originally Posted by DanDan

Nice one Chris. Hardly glaring, we saw it occasionally but it was of no consequence to us.

It's glaring to me, now that I know the problem.

I should have known that the 'log squared' calculation removed the need for scaling the log operation by 20, rather than 10. You live, you learn...

Interesting that Lupo noticed it. Ironically, he didn't have or use FM at the time!

Indeed.

Since you have ETC now, plus that viewing option, I suggest you include the 'piece of string' in the manual. That is clever, simple, and effective.

Not until I apply, and understand it fully, myself. Once something's in the manual, or I suggest folks use a particular method, I'd better be ready to walk them through the process.

If you do implement dBSPL and even calibration, I strongly suggest you take a look at John REW's work, Nice presentation of choices. Good effort on the level setting routing, but not successful in my experience. I believe the HF overload issue is where the problems lies and maybe the sweep need consideration. I would be happy to do some practical real world tests for you to help you get it working well.

Hardly going to be helpful until you're on a more modern Mac, though.

I've got a more wider-sweeping change I need to deal with in FM before I play any more with the different scale options.

For instance, I distinguish the SPL frequency graph from the Frequency Response graph, solely because their dB units are different.

Instead I should allow the user to set the scales on either axis.

Once this is done, I can present similar options in the time domain, and also allow the user to specify alternate time units (or, more importantly, distance).

It seems like a simple fix, but I wasn't as smart a lad when I started working on FuzzMeasure and its graphing back-end in 2004.

Cheers,

Chris

[liscio]

Originally Posted by SAC

...and then you will have a program calibrated to an SPL meter (which was calibrated how, exactly? I know for a fact that my RS SPL meter is as accurate as they come ) instead of to a reference voltage or sensitivity.

Bottom line, by that method, its still relative. But at least you won't be arguing amongst 'yourself' over any differences! (At least not until you discover the non-linearity of the SPL meter!)

This is why the only method of calibration in FuzzMeasure 3.x is to use a 94dB SPL calibrator. They're definitely not cheap, but doing stuff properly rarely is…

Chris

[SAC]

Thanks Chris!

That is quite sufficient!
And with both scales, just about everyone should be happy!

(Oh, and the inclusion of BOTH the log squared and the ETC is wonderful too!)

[liscio]

Originally Posted by SAC

Thanks Chris!

That is quite sufficient!
And with both scales, just about everyone should be happy!

(Oh, and the inclusion of BOTH the log squared and the ETC is wonderful too!)

I aim to please…

I just wish that I could do more in less time. With so many projects on the go right now, a new baby in the house, and WWDC in approx. 3 weeks, I'm pretty swamped.

The likelihood of seeing any of my aforementioned improvements over the next 3 months is fairly low. But we'll see—sometimes I need to escape one project by working on another one for a few days, so little bursts of work on FuzzMeasure sneak in here and there.

(Not counting bugfixes/etc. Those generally cause me to drop all new development until they're solved.)

[DanDan]

Calibrators are very very expensive. I really don't think that's a runner Chris. I would much rather see the money spent on a DPA or Earthworks.
Calibration by comparison works fine in my experience.
I have a very useful AZ SLM. €125.
I used to check it by bringing to college occasionally, where I could compare it to B&K's and even use Calibrators. I have never found it to be more than a dB out.
DD

[liscio]

Originally Posted by DanDan

Calibrators are very very expensive. I really don't think that's a runner Chris. I would much rather see the money spent on a DPA or Earthworks.
Calibration by comparison works fine in my experience.
I have a very useful AZ SLM. €125.
I used to check it by bringing to college occasionally, where I could compare it to B&K's and even use Calibrators. I have never found it to be more than a dB out.
DD

As you said yourself, calibration by comparison is doable without any extra UI. However, I refuse to call it calibrated—sorry.

If a user wishes to align the values so that they're lining up with their outboard SPL meters, that's fine. It's even easier than you suggest, because the built-in level meter does not report dBFS. It just behaves as an uncalibrated meter. So you'd adjust until FM's meter said 80dB in your example.

Chris

[DanDan]

Real Calibration is done by comparison, with a Calibrator.
That Calibrator is calibrated or verified occasionally by comparison with a better one back at the mother ship. It's just a pecking order.

Absolutely No big. Let's call it Aligned :-)

DD

[DanDan]

Chris, you may know this in which case please don't take offence at my stating the bleedin obvious. However, just in case you are not familiar with the world of Building Acoustics and all that stuff we spoke of before.

Calibrators are seriously expensive. I bought a B&K one with my B&K SLM.
I use this periodically to 'Calibrate' the SLM. The next level up from this, required for some legal purposes, is to have the Calibrator, and SLM, checked and calibrated back at the mother ship. Hugely expensive and required legally every 24 months. Personally, for the occasional situation I hire a currently certified SLM.

To the point. A Calibrator fits snugly over the mic. Unless your mic is 12mm you cannot use it. This rules out the use of the majority of mics. In particular it rules out the extremely useful DPA 4060 and 4090.



DD