pink noise, correlated vs uncorrelated
1 - 17 of 17 Posts
isn't correlated where they are in phase, or am I wrong?
I listened shortly to them on a laptop computer and it seemed the uncorrelated was out of phase. If that's the case, then I'd shoot for the correlated measurements.
*just guessing here*
Now that I think about it, maybe the uncorrelated is designed to center with.
I'd like to know the answer myself.
I listened shortly to them on a laptop computer and it seemed the uncorrelated was out of phase. If that's the case, then I'd shoot for the correlated measurements.
*just guessing here*
Now that I think about it, maybe the uncorrelated is designed to center with.
I'd like to know the answer myself.
Joined
·
17,148 Posts
Joined
·
346 Posts
i know one is stereo, thats why i put that in there
but my question is, should i use one more than the other for eq ing or just split the difference, didnt know if a car environment leaned more towards one or another
the cd tracks that play left and right channels seperatly are un correlated, so im just trying to guestimate, thanks
but my question is, should i use one more than the other for eq ing or just split the difference, didnt know if a car environment leaned more towards one or another
the cd tracks that play left and right channels seperatly are un correlated, so im just trying to guestimate, thanks
Joined
·
346 Posts
actually the left and right tracks just say "pink noise 20hz to 20khz left channel only"
hey bro no need to appoligize for that, i understand completely, priorities,in fact i was just about to go to home depot to pick up some concrete for 15 sections of fence im putting up and noticed how quiet it is in my house, hey the baby is sleeping and the older one is in school, hmmm, i can still make it into work by 4, hopefully the wife is not napping, ....ummm, bye
hey bro no need to appoligize for that, i understand completely, priorities,in fact i was just about to go to home depot to pick up some concrete for 15 sections of fence im putting up and noticed how quiet it is in my house, hey the baby is sleeping and the older one is in school, hmmm, i can still make it into work by 4, hopefully the wife is not napping, ....ummm, bye
Joined
·
346 Posts
yeah oh well, at least i made it to home depot and work on time
ok so Chad likes correlated, anyone else?
i am measuring left and right seperatly but wondered when i play both sides with correlated vs uncorrelated which is more accurate to whats really going on, or should i just stfu and split the difference,thanks
ok so Chad likes correlated, anyone else?
i am measuring left and right seperatly but wondered when i play both sides with correlated vs uncorrelated which is more accurate to whats really going on, or should i just stfu and split the difference,thanks
Joined
·
718 Posts
From Kevin7909 on ECA:
It does make a difference whether it's in stereo or mono. A large difference actually. Stereo pink noise is 'uncorrelated'. Therefore it is not associated with the pink noise on the other channel. This minimizes problems with standing waves, etc. Mono pink noise does not share this attribute. This is easy to test. Play stereo pink noise and try to set your time alignment. You can't.
The reason is that the left and right speakers must have a similar signal for us to compare the distance between them based on time differences. We will place the soundstage using SPL with stereo pink noise. Mono pink noise has the same signal through both speakers at the same time so we can compare the time differences between the two sources. Therefore, we can hear the soundstage moving across the vehicle with mono pink noise while changing the TA settings on our processors.
Basically we set frequency response using uncorrelated (stereo) pink noise. We set time alignment by ear using correlated (mono) pink noise. (bold emphasis mine)
Kevin
It does make a difference whether it's in stereo or mono. A large difference actually. Stereo pink noise is 'uncorrelated'. Therefore it is not associated with the pink noise on the other channel. This minimizes problems with standing waves, etc. Mono pink noise does not share this attribute. This is easy to test. Play stereo pink noise and try to set your time alignment. You can't.
The reason is that the left and right speakers must have a similar signal for us to compare the distance between them based on time differences. We will place the soundstage using SPL with stereo pink noise. Mono pink noise has the same signal through both speakers at the same time so we can compare the time differences between the two sources. Therefore, we can hear the soundstage moving across the vehicle with mono pink noise while changing the TA settings on our processors.
Basically we set frequency response using uncorrelated (stereo) pink noise. We set time alignment by ear using correlated (mono) pink noise. (bold emphasis mine)
Kevin
Apologies for a thread from the dead, but I think there is an issue with "Correlated Pink Noise" on the Autosound CD.
I wanted to make a longer noise track, 1:40 wasn't enough. I loaded the track into Audacity to copy and paste a longer track. This is what I found:
They don't match. The L & R waves are the same but they are misaligned by a ~0.02ms.
Is this enough to be significant? It's the same throughout the whole track.
![]()
I made this by copying the left channel onto the right. Was it a waste of time?
![]()
I wanted to make a longer noise track, 1:40 wasn't enough. I loaded the track into Audacity to copy and paste a longer track. This is what I found:
They don't match. The L & R waves are the same but they are misaligned by a ~0.02ms.
Is this enough to be significant? It's the same throughout the whole track.
I made this by copying the left channel onto the right. Was it a waste of time?
I copied this from Home Theater Shack, but it is the explanation that makes most sense to me...
Correlated versus uncorrelated applies to signals coming from multiple speakers at the same time while measuring. If correlated, then the same measurement signal is coming from the speakers in phase. This is not good for determining EQ settings, you will end up with phase cancellation at some frequencies, phase reinforcements at other frequencies, and will get a totally crazy frequency response curve. Move the measurement mic a few inches, and you get a totally different crazy frequency response curve.
By using uncorrelated pink noise, the measurement signal coming out of each speaker is random and completely unrelated to all the others, so there will be no phase cancellations or reinforcements, and you get a sane looking frequency response curve.
Which one to use when? Always use uncorrelated. The only use for correlated pink noise for taking measurements would be when you want to play a trick on someone who doesn't know any better and have them waste a bunch of time. That would not be nice.
Of course, you only run one speaker at a time while equalizing, so none of this applies to the normal measurement/equalization process. It can be useful/interesting to do it after equalization is complete, to get a curve showing how all the speakers are working together.
Read more: Measuring Car Frequency Response - Page 7 - Home Theater Forum and Systems - HomeTheaterShack.com
Correlated versus uncorrelated applies to signals coming from multiple speakers at the same time while measuring. If correlated, then the same measurement signal is coming from the speakers in phase. This is not good for determining EQ settings, you will end up with phase cancellation at some frequencies, phase reinforcements at other frequencies, and will get a totally crazy frequency response curve. Move the measurement mic a few inches, and you get a totally different crazy frequency response curve.
By using uncorrelated pink noise, the measurement signal coming out of each speaker is random and completely unrelated to all the others, so there will be no phase cancellations or reinforcements, and you get a sane looking frequency response curve.
Which one to use when? Always use uncorrelated. The only use for correlated pink noise for taking measurements would be when you want to play a trick on someone who doesn't know any better and have them waste a bunch of time. That would not be nice.
Of course, you only run one speaker at a time while equalizing, so none of this applies to the normal measurement/equalization process. It can be useful/interesting to do it after equalization is complete, to get a curve showing how all the speakers are working together.
Read more: Measuring Car Frequency Response - Page 7 - Home Theater Forum and Systems - HomeTheaterShack.com
By bad. I didn't read through to end carefully.
I don't have that CD, but looks like copying left onto right fixed the problem.
I could be wrong here but isn't .02 mS less than 3-tenths of an inch error in driver placement? or half a wavelength of 20KHz? Would it matter?
I use both. It's especially important if you are tuning a system with an upmixer that does ambience and center channel processing. Correlated pink noise will be steered to the center mostly and uncorrelated will be played by all the speakers. For non-upmixer systems, it can also be helpful to use both. MUCH of the information in music is recorded in mono, so the measurement of the correlated signal will be approximately the frequency response of your center image and the response of the uncorrelated signal will be the frequency response of the stereo information.
Thanks for the reference. The paper referenced in Post #14 was a good read too.
1 - 17 of 17 Posts
Join the discussion
