Originally Posted by SteveCallas
In regards to EQing the top end, how exactly would that work?
I was simply describing how Audyssey works (without app intervention)--it EQ's the speakers full range. The "Reference Curve" that most seem to prefer aims for a significant rolloff at the top end. A speaker with a hotter top end will be rolled off more.
If using Audyssey in the receiver, any EQ in the higher frequencies is only going to be relevant to keeping your head within a very tight range - moving in your seat a little bit this way or that way could throw it off, and it is only applicable to one specific seat in the room.
While it tries to optimize for the MLP, the correction affects all seats. Think of it as a glorified tone control. It won't affect dispersion characteristics. With a good controlled directivity speaker, the EQ change will sound similar over a very wide area. For a speaker with poor directivity, it can't fix that and make it sound the same for all seats so if you optimize it for one it might sound worse in other locations.
If you know of a different way this is being accomplished, please let me know, I would be interested.
The higher level room correction systems (Dirac, JBL, etc) have allowed this for a long time through customizing target curves. Now with the App, Audyssey can let you do some of this within limits (if a speaker has a really large rolloff it won't let you EQ it flat as the amount of boost is limited). While harder to use and not as powerful as some of the other systems, the app does allow you to try many different curves when before you were limited to two.
I’m looking at the graph of the VR3 and I’m not seeing a hump or bump on the top and that you are referring to. With the exception of a dip at 6 to 7 kHz, the VR3 is essentially flat the entire top end with no rise or rolloff.
It's a hump (centered at about 9K) in the anechoic
graph on the first page. In your in-room response
that translates to a "lack of rolloff" that begins in the 6-7K region by rising again, where a flat speaker would generally have a flat line sloping downward. From the look of the picture, I had assumed the mic was off axis which would result in a more significant rolloff at the top end with a typical dome tweeter.
That would be classified as a flat or neutral speaker......Additionally, I don’t believe most people want a rolled off top end, they want accuracy, which is flat. The Revel Salon Ultima 2, reguarded as the best speaker available today by most after beating the JBL M2 in a listening comparison, has undergone many listening tests with a variety of listeners, casual and hard-core in its development to see what most people prefer. The resulting top end frequency response is flat. When you have a rolled off top and, there is an airiness and sense of realism and clarity that disappears.
Yes, flat anechoically
. Not in-room response
. A speaker that measures flat anechoically
won't measure flat in a room. How far off from flat depends upon the room, of course, and other aspects of the speaker. The expected in-room
response of a "flat loudspeaker" will generally show some amount of boost in the bass region and a line that gradually slopes downward toward the top. How much it slopes down will obviously depend on the size and reflectiveness of the room. There are many example curves showing this in the work of Floyd Toole, Harman, etc, and is reflected in the "target curves" of many higher-end room corrections.
To guess which speaker people would like based on those graphs is not a good practice, as my GF, who represents the absolute most casual listener possible, quickly reached the conclusion that the Titans sound muffled in comparison and preferred the Bostons. There is a percievable difference in realism/clarity in the sound when you compare the two speakers, but that will only be perceived when you compare the speakers.
I don't doubt that at all. Everybody
will hear more detail with the treble turned up. What I'm trying to say is that many will eventually find that fatiguing and want to turn it down. Some people say they do prefer the Audyssey "Flat" curve for music--which will "turn up the treble" quite a bit for most speakers in order to get them as close to flat in-room
as possible. A lot of other people find that setting way too bright. That's where the personal preference comes into play. The research such as that done by Toole and others shows most people will find that too bright--instead preferring the sound a flat loudspeaker (anechoically) which will give an in-room response that is gradually sloping downward.
And that's just for music. For anybody more concerned with movie watching, music testing is important but leaves some things out. For example, widely dispersing sound/dipole/rear firing tweeters, etc, can add a spaciousness to music that people find very pleasing...but also harm things like speech intelligibility, accuracy of directional cues, etc, which can be more important in movie watching when the spaciousness is supplied by the surround channels.