Pink noise for evaluating speakers

[r042wal]

How does a pink noise source help evaluate the performance of a speaker? I just read about this yesterday. Someone suggesting a pink noise source to take to you local speaker dealer when you are in the market for new speakers.

Thanks

 

[sweetsounds]

In A/B blind tests pink noise is the most reliable sound to distinguish speakers.

Reason is, that any deviation from a flat frequency response results in a coloration of the noise. Due to the uncorrelated nature of the sound the brain is not distracted by listening to a musical pattern or stereophonic effects.

Downside is, that either you use it only for A/B comparisons or you have to train your ears in advance, how a neutral pink noise sounds.

 

[SSS]

Pink noise and also white noise are useful for evaluating speakers regarding difference sound. But the question is where to get the true neutral sound as reference? There is no real reference transducer for this, neither speaker nor headphones. Further, beside differences in the f-response the noise does not give a clue how the speaker will sound with music.

 

[Blumlein 88]

With some experience pink noise can be helpful. Moderate volumes please.

Just to see what you think you can play pink noise over what you have. Then adjust or EQ different bands of the noise to compare to non-EQ'd pink noise. Get some feel for how peaks or dips sound. Peaks are easier to hear.

You do have the issue if you try it with two or more speakers which one is the more correct one? If one is a good measuring speaker that Amir has reviewed that is useful as something of a reference.

 

[Yorkshire Mouth]

The other method I’ve heard, is when using a sub and monitors.

Get an SPL meter, play each monitor without the sub, check the level, turn off the monitor, turn on the sub, set it to the same level.

If I have that wrong, someone please correct me, but it’s something like that.

 

[Holmz]

In a frequency domain sense, then pink noise is sensible.
There is also the time-domain, which is a thing.

 

[dualazmak]

Interesting thread; I agree pink noise is useful for tentative/quick subjective A/B comparison of SPs; in some situation white noise is also useful...

We should be careful enough, however, about the QC (quality control) of the pink noise and/or white noise we would use in such tests. I always use/bring "Sony Super Audio Check CD (ref. here)" for this purpose which has very well prepared/QC-ed pink noise and white noise tracks. The liner notes (booklet) of the CD (English translation by myself) is attached herewith. (If you would be seriously interested in all the intact tracks of this CD, please simply PM me writing your wish.)

Furthermore, for Fq response measurements, I always stick to "cumulative (recorded) white noise averaging method", and usually do not like "rapid sine sweep method" with too much smoothing (especially psychoacoustic smoothing). If you would be interested, please refer to my post here and my summary post here.

My very recent post here and here would be also for your reference, I assume.

EDIT: And my post today (Aug 22) here also discussing the pros of the "cumulative (recorded) white noise averaging method".

 

[fineMen]

How does a pink noise source help evaluate the performance of a speaker?

Could be useful, but only in a room with highly attenuated, or at least decidedly controlled reverberation. P/n will not reveal the 'direct sound', hence would not cover what, to my understanding, is an important factor in speaker assessment.
Following this musings another signal (with no infomation) might be invented for this kind of task. Like a random series of plopps and clicks. Anyway, I doubt that many would keep a stable understanding ofn how good real p/n shall sound actually--without frequently recalibrating the impression against a true neutral source.

 

[r042wal]

Where could I download the Sony Super Audio CD Check? Thanks in advance for any help

 

[dualazmak]

Where could I download the Sony Super Audio CD Check? Thanks in advance for any help

OK, I will soon PM you, hopefully within one hour!

Spoiler

EDIT: Done it at 9:01 AM today Japan time!

 

[levimax]

OK, I will soon PM you, hopefully within one hour!

REW (free) has a "generator" tool which can create many types test tones including pink and white noise. You can play them directly or create a file with it. No need for a CD.

 

[dualazmak]

REW (free) has a "generator" tool which can create many types test tones including pink and white noise. You can play them directly or create a file with it. No need for a CD.

Thank you, of course I know well your point since I too occasionally use REW (e.g. recent ref. here).

I still very much like "Sony Super Audio Check CD", however, since it contains so many really useful/important and well QC-ed validated test signal tracks as well as excellent reference/sampler music tracks; please refer to my posts on my project thread:

Excellent Recording Quality Music Albums/Tracks for Subjective (and Possibly Objective) Test/Check/Tuning of Multichannel Multi-Driver Multi-Way Multi-Amplifier Time-Aligned Active Stereo Audio System and Room Acoustics; at least a Portion and/or One Track being Analyzed by Color Spectrum of Adobe Audition in Common Parameters:
____[Part-15] Again, CBS/Sony's "Super Audio Check CD": Analyzed by Adobe Audition 3.0.1 and MusicScope 2.1.0: #651
____[Part-04] Bimmel Bolle Antique Orgel; Extremely High-Energy High-Frequency Sharp Transient sound: #592

Five "real air-recorded transient-sound tracks" of Sony Super Audio Check CD played and analyzed by MusicScope 2.1.0: #760

Dancing video of my IEC 60268-17 compatible large glass-face DIY 12-VU-Meter Array
_____Part-1: with "High Frequency Linearity Check Track" of Sony Super Audio Check CD: #750

 

[Travis]

Agree will all of the great comments, great for A/B to tell difference. Floyd/Olive are very clear that pink noise was the best discriminator of difference between speakers, followed by various genres of music.

However, to use it as a means of choosing a preference, you probably need years of use/experience with speakers you know measure flat.

If you currently have active speakers, as suggested , use DSP to play around with the pink noise on one side, compare to other side, A/B. Then play a know reference track, left, and right, which gives you the most useful information. Pink noise or the track?

 

[dualazmak]

Furthermore, as I recently wrote here,,,

Each of us (hopefully audiophiles, audio enthusiasts) should always have/bring his/her own consistent "reference/sampler music playlist/tracks" throughout our long-year audio exploration journey including tuning of our audio system at home and the audition sessions at HiFi audio shops dealers and manufacturers.

In my case, I always have/bring my "reference/sampler music playlist intact 60 tracks" for which you can find the summary posts here and here on my project thread. (I also started an independent thread on it.)

 

[dualazmak]

Hello friends,

A little bit to be out of the scope of this thread, but for your possible reference (and interests), I just posted on my project thread;
- Reproduction and listening/hearing/feeling sensations to 16 Hz (organ) sound with my DSP-based multichannel multi-SP-driver multi-amplifier fully active stereo audio system having big-heavy active L&R sub-woofers: #782

 

[witwald]

Reason is, that any deviation from a flat frequency response results in a coloration of the noise. Due to the uncorrelated nature of the sound the brain is not distracted by listening to a musical pattern or stereophonic effects.

But isn't it white noise that has a flat frequency response?

 

[witwald]

Further, beside differences in the f-response the noise does not give a clue how the speaker will sound with music.

But won't changes in frequency response cause a commensurate change in the sound with music?

 

[Blumlein 88]

But isn't it white noise that has a flat frequency response?

It is, but pink noise has a spectrum closer to music on average and it has a defined slope. Listening to white noise is more bothersome than pink noise, but not necessarily more revealing of basic FR differences.

 

[dualazmak]

But isn't it white noise that has a flat frequency response?

I hope the PDF booklet (liner notes) attached herewith of "Sony Super Audio Check CD (ref. here)" would somewhat give you the clue.

 

[witwald]

It is, but pink noise has a spectrum closer to music on average and it has a defined slope.

So, which way does the slope go? Up or down? And what's its approximate value?