So you remember Sean's blind tests as confirming that all room EQ is a dramatic improvement? That's not the way I remember it. And no I am not going to search it out for you.
Dumptruck
Thanks for that. I don't know where Markus has been but that article is pretty clear. And pretty much exactly what I said as well.
Thanks for that. I don't know where Markus has been but that article is pretty clear. And pretty much exactly what I said as well.
Measurement tools and lots and lots of listening experiments and how to correlate the two.
Patience, and a lot of it. ...Plus improving our measurement tools, always.
* I fully agree with Mitch's earlier post (great post by the way - #180); parametric equalizers are not what will permit our speakers in our rooms to sound right.
They won't solve or restore the proper acoustic phases.
Sean Olive tested 5 systems (3 commercial products and 2 Harman prototypes). 3 were reported as producing a 'better' (i.e. more preferred) result than the unequalised speakers, with one being the same and one worse. I doubt that the test was intended to be particularly representative of the market even back in 2009.
This is what Olive said:
And this is what you've said:
If that's "pretty much exactly" the same then my English is worse than I've thought...
What phase? Between the channels? To some extend they do and at low frequencies only impractical amounts of absorption (think non-environment design) will eliminate the detrimental effects the room has on the signal so EQ can improve a situation that can't be improved otherwise.
I just read Olive's musings. To me, the most important lines are right here:
It would appear that in a round about way, he's simply saying that whichever product was able to best do it's job of cleaning up larger response holes and smoothing out bass irregularities were the ones that received good ratings.
I would have to be in the same ballpark that Earl is coming from, I can imagine thinking: Well, if we multisub this we can clean up the bass, and having a response hole at 2khz should be something that we fix in the crossover.
So, the question (and perhaps, a new double blind test we should suggest or try) would be to use a pair of speakers with the absolute best performance we can muster, paired with a bunch of large, powerful subwoofers, dial everything in until it's already flat and clean sounding, and only then start introducing correction devices.
Starting out with a broken reproduction environment and double-blind testing which black box can make a silk purse out of it doesn't seem to have yielded any usable information, aside from "People seem to prefer smooth extended response" which we don't really need to do any testing for.
It would appear that in a round about way, he's simply saying that whichever product was able to best do it's job of cleaning up larger response holes and smoothing out bass irregularities were the ones that received good ratings.
I would have to be in the same ballpark that Earl is coming from, I can imagine thinking: Well, if we multisub this we can clean up the bass, and having a response hole at 2khz should be something that we fix in the crossover.
So, the question (and perhaps, a new double blind test we should suggest or try) would be to use a pair of speakers with the absolute best performance we can muster, paired with a bunch of large, powerful subwoofers, dial everything in until it's already flat and clean sounding, and only then start introducing correction devices.
Starting out with a broken reproduction environment and double-blind testing which black box can make a silk purse out of it doesn't seem to have yielded any usable information, aside from "People seem to prefer smooth extended response" which we don't really need to do any testing for.
Careful, Markus won't like you for that! I agree however that the worse the speaker design the more attractive the appeal of a room EQ box becomes. What would be interesting would be to compare a room EQ correction of a poorly designed loudspeaker to a non-EQ'd well designed system. Also which costs less to the same end result? I know that cost is seldom a consideration around here, but I think that it is a key parameter.
Per the slide show/paper, the systems were: Anthem Statement D2, Audyssey Room EQ, Lyngdorf DPA-1, and two attempts from a Harman prototype using six and then one mic. The top two were Harman, per the blog comments.
That isn't an answer; patience and listening practice do not help measure phase accurately to a small fraction of a degree at high frequency resolution.
Earl,
I've tested very poorly designed speakers (full range in a cardboard box) with EQ and I couldn't believe the results at first. Definitely not just like putting lipstick on a pig. Key is a better understanding how stereo and our hearing works.
I've tested very poorly designed speakers (full range in a cardboard box) with EQ and I couldn't believe the results at first. Definitely not just like putting lipstick on a pig. Key is a better understanding how stereo and our hearing works.
Markus
As I said, the uglier the pig, the better the lipstick works. But beautiful women don't need lipstick to look good. Many look better without it.
And good loudspeakers don't need room EQ to sound good. In fact there is a 40% chance (according to Olive) that they won't sound any better or even worse. That's not an insignificant factor.
I am not completely against EQ, I use it of course. But it should be done in the speakers above the modal region independent of the room. And, of course, it has to be done in-situ for the modal region. Nothing new there I suppose so maybe this thread isn't for me! 😉
As I said, the uglier the pig, the better the lipstick works. But beautiful women don't need lipstick to look good. Many look better without it.
And good loudspeakers don't need room EQ to sound good. In fact there is a 40% chance (according to Olive) that they won't sound any better or even worse. That's not an insignificant factor.
I am not completely against EQ, I use it of course. But it should be done in the speakers above the modal region independent of the room. And, of course, it has to be done in-situ for the modal region. Nothing new there I suppose so maybe this thread isn't for me! 😉
Last edited:
Well, if all of this isn't new to you then why not contribute? For example, how exactly do you determine the upper limit where EQ is OK? What needs to be measured? How to derive EQ settings from those measurements?
That would still leave a 60% chance of better sound, probably higher if you get the right kind of EQ and not one of the duds tested.
Markus, wouldn't the Harmon room correction software (tested by Olive) be somewhat similar to the one used in the JBL MS8 car processor? Lots of info on that one at diymobileaudio.com.
Markus, wouldn't the Harmon room correction software (tested by Olive) be somewhat similar to the one used in the JBL MS8 car processor? Lots of info on that one at diymobileaudio.com.
This is the kind of discussion where my comments can come off as disingenuous and I get accused of just marketing. What I do is all custom software so its not available to others and many of my techniques are proprietary as well. So I can tell people what to do, but they can't do it themselves. That's not an ideal situation.
Briefly, there is no upper limit for when EQ can help as long as it benefits the free field listening axis and polar response of the system (but not one at the expense of the other. If you find that happening then the speaker needs to be fixed.) In-situ EQ where the room reflections are considered is highly tenable, except below 100-200 Hz where it is essential. If you must do in-situ EQ then do it only on the first 10 ms. or so. In a vehicle we found that this sounded more natural than using the full steady state response. When doing in-situ you also have to deal with the statistical uncertainty in the measurements. That isn't a problem when one EQs based on free field data.
What needs to be measured is, of course, the listening axis response and the polar response.
I get the EQ settings from custom software so I can't help out very much there.
Last edited:
I hardly think that people will want to hear that they have a little bit better than a 50-50 chance that their new expensive box is going to make their system sound better. That's not what the advertising claims.
If you read the comments from Sean Olive below that blog entry you'd bet on the Harman solution (if available, it was a prototype) and have a better than 60% chance 😀.
No way of knowing that wasn't marketing on his part though!
No way of knowing that wasn't marketing on his part though!
Hindsight and blind listening tests are hard to beat, but without those my point was that it's hit or miss with room EQ.
- Status
- Not open for further replies.