AB neutrality of in-room response?

[dasdoing]

I think it should be possible,
you put another speaker above your speaker, create several in-room corrections for your speaker. then you play pink noise on the upper speaker and record it at LP (convolve the recording with the correction curve of the mic). you than AB the pink noise from above speaker to the recorded pink noise reproduced via the corrected speaker. you should be able to come to a clear conclusion which correction came closest to the original event,
am I wrong?

 

[Philbo King]

Pink noise is very difficult to subjectively analyse and compare. Perceptual illusions/delusions tend to skew the results. It may sound different, but how do you know which is more accurate?

A better way to A/B test your room EQ is to simply switch room correction on and off while listening to wide band music.

 

[FrankW]

you should be able to come to a clear conclusion which correction came closest to the original event,
am I wrong?

You have no clue what the "original event" were there one, sounded like

 

[RayDunzl]

am I wrong?

Below I what I do,

You have no clue what the "original event" were there one, sounded like

The clue I use...

I compared corrected playback as seen by the UMIK-1 with the original files, and don't see any obvious failures (other than black hole in mono bass at 48Hz not present below), using RTA peak as the reference for both the uncorrected source and the corrected in-room.

Top: Corrected In-Room RTA, with a "flat" curve applied
and below that
Left and Right channels of an RTA of the source

Maybe I don't get whatever sound originally existed but I surely seem to get a pretty accurate presentation of what they put into the recording.

An uncorrected in-room would have higher bass levels (boomy) and not follow the "curve" of the souce as closely.

 

[FrankW]

The clue I use...

I compared corrected playback as seen by the UMIK-1 with the original files



Maybe I don't get whatever sound originally existed but I surely seem to get a pretty accurate presentation of what they put into the recording.

An uncorrected in-room would have higher bass levels (boomy) and not follow the "curve" of the souce as closely.

Hmmm, I don't think it quite works that way Ray. I presume you took a spectral snapshot (electronic) of a track at a time point for your "reference", then an acoustic snapshot with a pressure mic at one point in space in your room. I don't believe those two things are the same at all. Could be wrong tho...
Anyway, unless classical concert or some live rock recording, there is no "event"

 

[RayDunzl]

Hmmm, I don't think it quite works that way Ray. I presume you took a spectral snapshot (electronic) of a track at a time point for your "reference", then an acoustic snapshot with a pressure mic at one point in space in your room. I don't believe those two things are the same at all.

I played the tune while monitoring the electrical source, and the mic was at the Listening Position, which is where I listen.

The peak traces are the RTA for the entire tune.

The instantaneous traces (black) don't align due to slight display delays between the three instances of REW in use.

Works for me as a check on the room correction, YMMV.

 

[FrankW]

If it sounds good to you, rock on

 

[Blumlein 88]

Pink noise is very difficult to subjectively analyse and compare. Perceptual illusions/delusions tend to skew the results. It may sound different, but how do you know which is more accurate?

A better way to A/B test your room EQ is to simply switch room correction on and off while listening to wide band music.

I disagree. FR trends between two sources that aren't terribly clear with wide band music are thrown into obvious relief with pink noise in about 2 seconds. True it might take a bit more to pin down what is different, but it would only be worse with music.

People who setup up speakers for musical events in large spaces or outdoors will use pink noise to figure out where they stand with things.

Now an issue with two speakers in physically different locations will be how the reflected sound differs. MMM measures probably will get you pretty close. You cannot them use both together as there will be comb filtering.

 

[dasdoing]

simply switch room correction on and off

thinking about this, actually the second speaker isn't really needed.
the speaker can reproduce the noise, and the recorded noise. why not?
to make things easier the recorded noise can be convolved with the various room correction filters. you than compare those to a pure pink noise in a simple audio player.
will reply the others later

 

[Thomas_A]

Not sure what the questions is. But...
First question: During the original recording event, how were the microphones arranged?
Second question: Should I reproduce what those microphones captured at LP?
Third question: Or should I add a distance/room between me and what the microphones captured?

 

[Philbo King]

I disagree. FR trends between two sources that aren't terribly clear with wide band music are thrown into obvious relief with pink noise in about 2 seconds. True it might take a bit more to pin down what is different, but it would only be worse with music.

People who setup up speakers for musical events in large spaces or outdoors will use pink noise to figure out where they stand with things.

Now an issue with two speakers in physically different locations will be how the reflected sound differs. MMM measures probably will get you pretty close. You cannot them use both together as there will be comb filtering.

True for freq response. But not so clear an advantage for spectrum decay, stereo imaging and group delays. (Keeping in mind that 99% of concert PAs are set up in mono so audience seating position doesn't skew what one hears).

 

[abdo123]

Pink noise is very difficult to subjectively analyse and compare. Perceptual illusions/delusions tend to skew the results. It may sound different, but how do you know which is more accurate?

A better way to A/B test your room EQ is to simply switch room correction on and off while listening to wide band music.

I completely disagree.

Pink noise is by far the most powerful thing that exists out there for discerning differences between speakers and room correction methodes.

Music is way too distracting in comparision.

 

[Thomas_A]

IMO: Pink noise is very sensitive to hear relative differences but for absolute hearing it is not easy. I find sweeps more useful in absolute terms, e.g. to pinpoint resonances.

(Music can be good as well but...I can hear when something is wrong with voices or bass notes, but to pinpoint the exact frequencies I often need measurement or sweep/tones.)