There's strong evidence that they do. Does it matter in the end if the result is the same?
At the Highend show in Munich a Lyngdorf man said their device "is something like an equalizer with infinite resolution". Can this mean anything else than convolution? I also strongly doubt that such an algorithm could have enough "intelligence" for adjusting parametric EQs. Especially as in the German forum there was mentioned the importance of the Q-value for killing resonances also in the time domain.
Really, really nice and intelligent technology.
But as I assumed FIR filtering based on the inverse room response seems to be what is usually done:
Digital room correction - Wikipedia, the free encyclopedia
Could be a killer for systems that rely on reflection like the CFS.
But as I assumed FIR filtering based on the inverse room response seems to be what is usually done:
Digital room correction - Wikipedia, the free encyclopedia
Could be a killer for systems that rely on reflection like the CFS.
I also think it is nice and i would like to play around with it.
Influencing excitation of LF room modes seems very viable to me,
but i guess the higher you go in frequency, the more the result
will be just an optimization for a very small listening area.
But you need not use all possible features of room correction at one
time, and to have a setting optimized for a single place of listening
may be OK sometimes. For other purposes you may have different
settings ...
Looks as if this is one of those things which have to be tried out ...
Influencing excitation of LF room modes seems very viable to me,
but i guess the higher you go in frequency, the more the result
will be just an optimization for a very small listening area.
But you need not use all possible features of room correction at one
time, and to have a setting optimized for a single place of listening
may be OK sometimes. For other purposes you may have different
settings ...
Looks as if this is one of those things which have to be tried out ...
good to know
I have to admit Gentlemen that I am really confused
physical measures-electronic corrections, manual-automatic, parametric-FIR filtering, digital-analog, Behringer-Dspeaker, sufficiently precise-imprecise, PC-Mac, good-bad, try-stay out
Last edited:
good to know
I didn't say I know. In the Wikipedia article they say the goal is to reverse the room response, but they also say this can't be done radically because the right listening position would be a matter of millimeters then. So these devices can't cancel the ceiling reflection completely. The question is: Do they leave enough of it to preserve the "reach out and touch effect"? And what do they do if the treble comes more from the ceiling reflection than from the speaker, as in a pure CFS with a highly directive fullrange driver?
I would try to find out how the Sony works or borrow one, or both.
Waste of time. You can't correct the room within the statistical region. What can be achieved is a better matching of each speaker. That won't be room correction but speaker correction.
Hi,
Interesting test. Reasonably well done.
Even though the music chosen (all artificial studio productions mostly with fairly low production values) may not be very revealing.
The number of participants is neither diverse enough nor numerous enough (IMHO) to draw statistically significant conclusions.
The bottom line, a sloping in room response sounds best (did J Gordon Hold say that before I was born?) and equally, badly used equalisers are worse than not using them.
If only more discriminating music recordings and larger numbers would have been used...
Ciao T
Interesting test. Reasonably well done.
Even though the music chosen (all artificial studio productions mostly with fairly low production values) may not be very revealing.
The number of participants is neither diverse enough nor numerous enough (IMHO) to draw statistically significant conclusions.
The bottom line, a sloping in room response sounds best (did J Gordon Hold say that before I was born?) and equally, badly used equalisers are worse than not using them.
If only more discriminating music recordings and larger numbers would have been used...
Ciao T
Other posts of Sean address exactly that. A song by Tracy Chapman was found to be as revealing as Pink Noise (!) and the number of participants would NOT improve results.
Hi,
I may have been one of the people that popularised the use of pro audio EQ's for "room correction" (and other uses).
Interestingly I also ended up broadly with a curve like the most preferred one in the Harman test for listening.
In the end the limitations of this kind of system pushed me progressively more towards speakers with better controlled directivity than average and trying to solve as many problems in different ways before applying EQ and towards using analogue EQ's for programme re-equalisation.
At the same time, sensibly set up** such EQ is a clear improvement...
** sensibly set up is with precise and high resolution parametric EQ's in the bass and with only general adjustments of tonal balance and speaker matching, but without attempting to produce "ruler flat" response AND with the same slope as in Harmans test for the room sound (a lot of if's and buts, I know)
The measurement mikes in these systems generally are omni's, so the mike does not "know" if the treble came direct, via the ceiling or via New York, London and Tokyo. It will simply react to the local sound pressure.
Ciao T
I may have been one of the people that popularised the use of pro audio EQ's for "room correction" (and other uses).
Interestingly I also ended up broadly with a curve like the most preferred one in the Harman test for listening.
In the end the limitations of this kind of system pushed me progressively more towards speakers with better controlled directivity than average and trying to solve as many problems in different ways before applying EQ and towards using analogue EQ's for programme re-equalisation.
At the same time, sensibly set up** such EQ is a clear improvement...
** sensibly set up is with precise and high resolution parametric EQ's in the bass and with only general adjustments of tonal balance and speaker matching, but without attempting to produce "ruler flat" response AND with the same slope as in Harmans test for the room sound (a lot of if's and buts, I know)
The measurement mikes in these systems generally are omni's, so the mike does not "know" if the treble came direct, via the ceiling or via New York, London and Tokyo. It will simply react to the local sound pressure.
Ciao T
Hi,
Interesting. I do occasionally use pink noise as diagnostic signal (in diagnosing speakers), but it is ill suited to low frequencies.
I was not suggesting a larger and wider listener selection would improve the "results". It merely would increase the certainty that results are reliable and MORE CRUCIALLY that they are widely applicable.
I may throw a "fair coin" five times. I may get tails all five times. These are the results. They do not change if I do it another five times.
But if I get tails again I have a much better grounds for suspicion that coin may not be so fair. In fact, if I do it 100 times and still get all tails one may wish to investigate if the coin has "tail" on both sides...
Ciao T
Interesting. I do occasionally use pink noise as diagnostic signal (in diagnosing speakers), but it is ill suited to low frequencies.
I was not suggesting a larger and wider listener selection would improve the "results". It merely would increase the certainty that results are reliable and MORE CRUCIALLY that they are widely applicable.
I may throw a "fair coin" five times. I may get tails all five times. These are the results. They do not change if I do it another five times.
But if I get tails again I have a much better grounds for suspicion that coin may not be so fair. In fact, if I do it 100 times and still get all tails one may wish to investigate if the coin has "tail" on both sides...
Ciao T
But they know what comes first and what comes last. So they can add the lacking treble to the direct signal and subtract it from the reflection (no matter where it comes from). In principle, but limited by the wavelength in relation to the variation of the head position. So probably Olive is right when he says room correction is only possible up to 500 Hz and above it's only speaker correction (traditional EQing, just finer).
But only if the impulse is gated to exclude room reflections from the measurement. Otherwise the direct signal will be deformed.