Best test signals other than music?

[Dennis Moore]

Okay, what are the best test signals to use in listening
tests other than music? For the purpose of differentiating
between components.

Pink noise?

[Nousaine]

(Mr 645) wrote:

><<<Okay, what are the best test signals to use in listening
>tests other than music? For the purpose of differentiating
>between components.
>
>Pink noise?
>
>>>>>
>
>In my opinion Not sure what use pink noise, or other non musical sound would
>be
>for other then balancing volumes or comparing other volume or power related
>things. The bottom line is to listen to real music, well recorded and base
>decisions on that.
>

Actually full band uncorrelated pink noise is possibly the single most
revealing source for uncovering differences of any magnitude and practically
every sound quality (some transient areas may be hard to test with pink noise.)

It's main flaw is that it's overly sensitive to many sound quality differences
because, as I like to say: It contains every sound ever made and every note of
music ever played and heard by humans all at once.

Conversely no human has ever really 'heard' pink noise played over loudspeakers
or headphones because none of them are perfect. Rushing water might be as close
to a natural pink noise as one may find.

But in general its the single most useful recorded sound available for
evaluating audio playback equipment.

[Dennis Moore]

Okay, Mr. Nousaine, pink noise is a good test signal.

So I have a question. Why does masking not limit its effectiveness
versus some other type signals? Seems every sound possible
all at once would also come close to masking everything possible
all at once until differences were pretty large.

Dennis

"Nousaine" > wrote in message
news:1UP6b.383524$YN5.254602@sccrnsc01...
> (Mr 645) wrote:
>
> ><<<Okay, what are the best test signals to use in listening
> >tests other than music? For the purpose of differentiating
> >between components.
> >
> >Pink noise?
> >
> >>>>>
> >
> >In my opinion Not sure what use pink noise, or other non musical sound
would
> >be
> >for other then balancing volumes or comparing other volume or power
related
> >things. The bottom line is to listen to real music, well recorded and
base
> >decisions on that.
> >
>
> Actually full band uncorrelated pink noise is possibly the single most
> revealing source for uncovering differences of any magnitude and
practically
> every sound quality (some transient areas may be hard to test with pink
noise.)
>
> It's main flaw is that it's overly sensitive to many sound quality
differences
> because, as I like to say: It contains every sound ever made and every
note of
> music ever played and heard by humans all at once.
>
> Conversely no human has ever really 'heard' pink noise played over
loudspeakers
> or headphones because none of them are perfect. Rushing water might be as
close
> to a natural pink noise as one may find.
>
> But in general its the single most useful recorded sound available for
> evaluating audio playback equipment.
>
>

[Thomas A]

"Dennis Moore" > wrote in message >...
> Okay, what are the best test signals to use in listening
> tests other than music? For the purpose of differentiating
> between components.
>
> Pink noise?


A sin2 pulse perhaps.

Thomas

[Mark Wilkinson]

"Dennis Moore" > wrote in message >...
> Okay, what are the best test signals to use in listening
> tests other than music? For the purpose of differentiating
> between components.
>
> Pink noise?

I use pink noise and sine waves to measure and listen to speakers.

Pink noise is good for typical RTA analysis, and if you listen to it
enough while calibrating your ears with measurements, you begin to get
a sense of what it should sound like. As many have said, it's easier
to hear changes in pink noise than music.

IMO, sine waves are very revealing. It's almost shocking to fully
experience boundary cancellations that routinely occur. It isn't
difficult to find spots of near silence in my listening room, when
measuring 90db in other spots (for freq under 100hz). To minimize
these cancellations, sine waves are great for fine tuning speaker
placement and determining listening position (or coverage if, like me,
you refuse to sit still)-- IMO, they waay surpass the info from even a
1/24 octave RTA. The only problem is that they are tedious to use.

Sine waves with a high resolution RTA are also good for getting a
handle on THD. I'm convinced bass, particularly sub bass, is the most
difficult thing to get right in audio, mostly due to room
interactions, but also due to THD. Get away from 1 watt fantasy specs
and you're lucky to keep harmonics 15 db below level. Push sine waves
through your bass drivers, and compare the primary level with the
levels of the harmonics -- eye opening. I doubt anybody would ever
fail a DBT of speakers using single sine waves <100hz as the signal.
Ever seen any sub manufacturer's full output THD specs????

One more neat thing about sine waves, is determining what I call TRD,
or total room distortion :~) As you dial through the frequencies,
the room starts talking to you -- buzzes and rattles, rather damn loud
at times, that you never knew were there. You usually can't hear them
with full range pink noise or music, they are masked. Sometimes you
can hear them when running pink through a sub alone. I'm surprised
how often the room noise is at a harmonic of the test freq.

[Marco Stanzani]

"Dennis Moore" > wrote in message >...
> Okay, what are the best test signals to use in listening
> tests other than music? For the purpose of differentiating
> between components.
>
> Pink noise?

Try the MATT : it is supposed to test the time domain behaviour of the
room by checking the retention capabilities at various frequency by
sending interspaced bursts of increasing frequency. The interspacing
is ~ the fligh time of a reflected wave when bouncing in the room
before reaching the listenr

[[email protected]]

(Nousaine) wrote in message news:<1UP6b.383524$YN5.254602@sccrnsc01>...
> ><<<Okay, what are the best test signals to use in listening
> >tests other than music? For the purpose of differentiating
> >between components.

HF reproduction: shaking keys, bagpipes, percussion (castanet type
stuff).

Midrange: any good acoustic recording of piano or flute

bass: organ, percussion, double bass, cello

pink / white noise is good for level matching prior to tests, and for
hearing comparative differences in overall frequency response

[Nousaine]

"Dennis Moore" wrote:

>Okay, Mr. Nousaine, pink noise is a good test signal.
>
>So I have a question. Why does masking not limit its effectiveness
>versus some other type signals? Seems every sound possible
>all at once would also come close to masking everything possible
>all at once until differences were pretty large.
>
>Dennis

There is no masking mechanism present in pink noise that is not also present in
musical programs. Indeed part of its beauty is that IF you cannot hear a
response error with pink noise you surely will not hear it with music.

Also it's wonderful for predicting staging and imaging. Playing uncorrelated
pink noise simultanously in all driven channels will tell you nearly instantly
where holes and alleys (poor image/ambience coverage) will occur with regular
programs.

For example if you play pink noise into a 2-channel system and you don't get
sound 'fuzz' in the center you'll have a weak, unstable and vague center
vocalist with ordinary music programs. Likewise 'blank spots' in the frontal
stage will tell you exactly where images 'won't' form.

With multichannel systems a blank spot, lacking acoustic fuzz, between the
listener and the frontal stage tells you that moving sound that travels from
the screen into the room will tend to "jump" to either the surround speaker
locations or to the rear of the listener.

To avoid the inevitable flames I've never said it's the ONLY signal one will
EVER need; but simply its the MOST useful of any given signal for revealing
subjective sonic qualities

>
>"Nousaine" > wrote in message
>news:1UP6b.383524$YN5.254602@sccrnsc01...
>> (Mr 645) wrote:
>>
>> ><<<Okay, what are the best test signals to use in listening
>> >tests other than music? For the purpose of differentiating
>> >between components.
>> >
>> >Pink noise?
>> >
>> >>>>>
>> >
>> >In my opinion Not sure what use pink noise, or other non musical sound
>would
>> >be
>> >for other then balancing volumes or comparing other volume or power
>related
>> >things. The bottom line is to listen to real music, well recorded and
>base
>> >decisions on that.
>> >
>>
>> Actually full band uncorrelated pink noise is possibly the single most
>> revealing source for uncovering differences of any magnitude and
>practically
>> every sound quality (some transient areas may be hard to test with pink
>noise.)
>>
>> It's main flaw is that it's overly sensitive to many sound quality
>differences
>> because, as I like to say: It contains every sound ever made and every
>note of
>> music ever played and heard by humans all at once.
>>
>> Conversely no human has ever really 'heard' pink noise played over
>loudspeakers
>> or headphones because none of them are perfect. Rushing water might be as
>close
>> to a natural pink noise as one may find.
>>
>> But in general its the single most useful recorded sound available for
>> evaluating audio playback equipment.