Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Mark Wilkinson]

It's all good until windowing is applied.   SysTune is extremely powerful but the user still has to choose to make a windowed or unwindowed measurement.   And the user must choose the windowing parameters.

A few years back a (very) controlled test was done between five analyzers.  Each produced identical results for unwindowed measurements (I would hope so...) but when windowing is applied, basically five different results are produced.

That's what turned me off about Systune...the need to use windowing at times vs not at others. 
Imho, I believe Systune has fallen behind with the trend to build in some type of fixed-PPO scheme, to solve its windowing requirement.

I've compared Smaart and REW many times using default settings, and they always appear spot-on with each other.
My hope is there has been similar convergence with most other measurement programs.

[Chris Grimshaw]

I'd never try to use REW is this noisy situation....the sine sweep can't average.

Actually, if you run multiple sweeps under the same measurement (there's a drop down - 1,2,4,8), REW looks at the correlated data, thus discarding background noise.

As an interesting example, I've measured a speaker that was inaudible in-room. I used a Beyerdynamic MC930 positioned close to the speaker, and ran sweeps at decreasing levels until I couldn't hear the sweeps any more. I was sat at around 0.5m from the speakers. It was an investigation into what happens when speakers are used very close to their zero-point, and the answer is "nothing unexpected - they just get quieter".

I usually use 2-4 sweeps in most situations, 1x for electrical loop-through and 8x when there are noise problems.

I should probably get around to Smaart for the Armonia+ integration if nothing else, but the pricing is prohibitive at the moment.

Chris

[Mark Wilkinson]

Actually, if you run multiple sweeps under the same measurement (there's a drop down - 1,2,4,8), REW looks at the correlated data, thus discarding background noise.

Yep, I've used the multiple sweep option many times.....works great.  There's so much to like about REW. 
Heck, the excellent trace arithmetic and windowing export capability can even generate FIR filters (via simple impulse inversions).

But, the annoyance of multiple sine sweeps whooooooping it up gets me in trouble with nearly everybody,......... at home, in the neighborhood, or at an installation or gig site.
Whereas nobody seems to mind some steady pink noise.

[David Junius]

It's all good until windowing is applied.   SysTune is extremely powerful but the user still has to choose to make a windowed or unwindowed measurement.   And the user must choose the windowing parameters.

A few years back a (very) controlled test was done between five analyzers.  Each produced identical results for unwindowed measurements (I would hope so...) but when windowing is applied, basically five different results are produced.

Ok, I was following along fine until you brought up windowed and unwindowed measurements. I’ve been researching on the links provided so I’ll see if I can figure that out! I feel this subject area goes from the kiddie pool area to 20,000 leagues real fast but I’ll get there one day!

David

[Dave Garoutte]

Whereas nobody seems to mind some steady pink noise.

Plus you can mix in some recorded music with the pink.
It's way less annoying and works perfectly.

[Tim McCulloch]

For those playing along at home and wondering what the talk of windows is about:

https://en.wikibooks.org/wiki/Digital_Signal_Processing/Windowing

[Russell Ault]

[...] And the real fun begins once you allow for the fact that smoothing in the frequency domain can use used to give the same result as windowing in the time domain....

Wait, what? How does that work?

-Russ

[Michael Lawrence]

Wait, what? How does that work?

-Russ

In broad, non-mathematical strokes:
Shorter time window = lower frequency resolution, suppresses influence of late arriving energy.
Frequency domain smoothing = lower frequency resolution, suppresses influence of late arriving energy.

Fine print: it must be complex smoothing, and it must be linear frequency smoothing, not fractional octave smoothing as is far more common. Without going too far down a very tedious rabbit hole, frequency-domain smoothing can give you a comparable result at HF while avoiding the LF data point loss that comes with time-domain windowing. There's actually a "hidden" feature in v8 that exists pretty much only to demonstrate this point: in Transfer Function options, check "Enable FTW" under "Advanced." That enables "Frequency-Domain Time Windowing", which is a frequency-domain linear complex smoothing function that's mathematically equivalent to windowing in the time domain. Input the half-window length in ms into the FTW field in the TF control bar, and you can compare that against a measurement taken without the setting but using "regular" smoothing, and also to a time-windowed measurement taken in IR mode if you choose. Details here, pages 6-8.

[Mark Wilkinson]

Plus you can mix in some recorded music with the pink.
It's way less annoying and works perfectly.

Heck yeah, works great !!

Speaking of annoyances....i found even when using music with background pink as the reference....
..when my habit was i'd turn the whole music/pink caboodle mix on & off  when making measurements....

that a number of people have come up to me and said "would you please let the songs play through without turning them off?"

So now i just keep music playing continually during all measurements..

[Frank Koenig]

And the real fun begins once you allow for the fact that smoothing in the frequency domain can use used to give the same result as windowing in the time domain....

In rough terms, yes, but it depends. Multiplication by a function, such as a window, in one domain (time or frequency) is equivalent to convolution by the Fourier transform of that function in the other domain. A window, by definition, is of finite length (zero outside some range) so its Fourier transform has ripples that go out to infinity, which may not be the best smoothing kernel. (This is a manifestation of the Gibbs Phenomenon.) Furthermore, in audio, we often like frequency-domain smoothing to be frequency-proportional (such as 1/6 octave). Multiplication of the impulse response by a window results in constant-frequency smoothing (such as 100 Hz).

The most common reason for windowing the impulse response is to remove reflected sound so as to achieve a pseudo-anechoic measurement. The goal is to include as much of the direct sound as possible without causing excessive artifacts in the frequency response due to the sharpness of the window's edges. This is at odds with good smoothing.

I believe what Smaart does in "MTW" mode is to apply longer windows to lower frequencies to allow greater frequency resolution (in Hz, not octaves) at those frequencies at the expense of including more reflected sound.

Smaart is great for quick-and-dirty measurement in the field -- system health, sub alignment, and so forth. It's also nice for what I might call acoustic exploration where you move things around and can immediately see the effect. For detailed speaker measurements under more controlled conditions, as well as room acoustic measurements, where you really want to know what you're getting, I think there are better tools.

--Frank

[ProSoundWeb Community]