Hi
I just did some measurements on my speakers, and using the files it got me, I made REW calculate EQ to flatten the response.
The response is now next to flat when measuring, but I actually enjoy the music better without the EQ.
This got me wondering - Sound I aim for a flat EQ? Perhaps there are certain areas that I actually prefer louder? What do you think?
In REW the default "full range" setting that comes up when EQ'ing the speaker isn't flat either
The speakers are used for both parties and background listening in a 4,5x10m area
I just did some measurements on my speakers, and using the files it got me, I made REW calculate EQ to flatten the response.
The response is now next to flat when measuring, but I actually enjoy the music better without the EQ.
This got me wondering - Sound I aim for a flat EQ? Perhaps there are certain areas that I actually prefer louder? What do you think?
In REW the default "full range" setting that comes up when EQ'ing the speaker isn't flat either
The speakers are used for both parties and background listening in a 4,5x10m area
Hi, this is the room acoustics forum, have you tried to EQ the room? Why Can't I Fix All my Acoustic Problems with EQ?
I am not able to do anything about the room. It's my school and the speaker system is only here till summer.
I looked through the subforums and this one seemed to be somehow suited, but I guess I was wrong 🙂
IMG-20190320-203618 — imgbb.com
It's very messy right now.
The speakers are on the window shelf. Coaxial 12". Tham 12" for sub. Crossed around 95 hz The tops are mono (I only have a 2 channel amp around).
I took 3 measurements from 100hz and up (w/o sub) on three different spots along the corridor. All the measurements have about the same distance to the walls and at ear-level
Measurements can be found here. I made a flat EQ from the averaged response:
1 — imgbb.com
2 — imgbb.com
1 — imgbb.com
2 — imgbb.com
Did you follow the REW advice, basically that is to cut the peaks rather than boost the dips? Also it's advisable not to use too much EQ
The EQ is only performed using the auto EQ-ing tool from REW. I'll try and mess around with the target dB. I can see that I achieve a very different predicted response when adjusting it
Here's some things to remember about in room measurements and trying to EQ flat:
1) the speakers are not radiating all sounds equally in all directions.
2) the room is not equally reflective at all frequencies in all directions.
3) the microphone is only showing you the sum sound pressure at the point in space the microphone diaphragm is located, it is ignorant of the two above points.
This a topic with some complexity attached, but the takeaway is that a flat room curve almost always sounds bad.
1) the speakers are not radiating all sounds equally in all directions.
2) the room is not equally reflective at all frequencies in all directions.
3) the microphone is only showing you the sum sound pressure at the point in space the microphone diaphragm is located, it is ignorant of the two above points.
This a topic with some complexity attached, but the takeaway is that a flat room curve almost always sounds bad.
In typical domestic listening rooms, yes. Flat doesn't sound very good. REW has a falling response curve included in the filter section. That's a good place to start.
In different acoustics, different curves can sound more balanced.
In different acoustics, different curves can sound more balanced.
Qualified no.
Best systems have a response curves following human ears equal loudness curve:
Be careful, that curves are varies depending on SPL, so it’s better to optimise your sustem based on actual preferred loudness.
It all depends on your preference. To hear an approximation of what the recording/mixing engineer heard then your would aim for a flat in-room response.
Last edited:
The problem here is the environment, school classroom and corridor. There are so many modes and reflections that good eq result is impossible. REW shows typically response that has 500ms gating and 1/12 octave smoothing.
It is better to eq only full octave bands and just listen to the sound.
It is better to eq only full octave bands and just listen to the sound.
Having spent a good deal of time on this subject, I find that these answers aren't very satisfying. In fact, I believe them to be misrepresentations of what is really occurring. For most people, they just want to "fix it" and move on. If the reader is one of those, then please ignore the following...
"Flat doesn't sound very good..." Basically, there are three reasons for this:
- Your loudspeakers don't have the same controlled directivity vs. frequency across the band 200-7000 Hz, so they paint the nearfield of where they're placed with bandpassed acoustic energy which upsets the overall timbre balance of the sound reproduction. If you have loudspeakers that control their polars to ~90 degrees included angle (-6 dB point) from 200 to 7000 Hz, then flat on-axis response by far sounds the best. This excludes almost all direct radiating loudspeakers, unfortunately.
. - Room acoustics--even loudspeakers with full-range directivity will need some help in the extreme nearfield (within 1 metre) to control these nearfield reflections from just around the loudspeakers. Not only do these very nearfield reflections upset the timbre balance of loudspeakers having poor directivity control within the 200-7000 Hz band, these early reflections will also significantly detract from the soundstage imaging, particularly the phantom center image in stereo. The solution is easy: add absorption within 1 metre all around the loudspeakers. Some people try to pull the loudspeakers away from room walls and floor--which has a disastrous effect on bass frequency response. Instead of doing that, just add absorption around the loudspeakers to control midrange and mid-bass early reflections.
. - Recordings - perhaps you might not know that the recordings that you play have been timbrally altered from their as-recorded condition. The best place that I know to read about this and what to do about it is at this link.
"In different acoustics, different curves can sound more balanced"-- I think by now, you will see that house curves are a "Mr. Fixit" approach to something that is more systemic.
Chris
Last edited:
More to the OP's direct question...see https://www.diyaudio.com/forums/mul...at-vs-accurate-hi-fidelity-4.html#post4258668
I have been experimenting the speaker / room correction based on both objective and subjective effects.
1. Sean Ovive's study that Cask05 mentioned is a really good reference. I started my correction based on this study, and I found sloped correction is better than flat correction.
2. Lower frequency is impossible to measure correctly in the room, and the room mode is unavoidable. Bad resonant frequencies at the listening position are notch filtered "by ears" with sine wave generator playing semitones. This method actually works very well for confirming the actual higher frequency response at the listening position. You can clearly hear the resonance or notch at many different frequency points by many different reasons including directivity issue.
3. I was still not 100% satisfied with the result. Currently, a little BBC dip and slight 10K boost are added. I'm pretty sure that most of the commercial mastering were done expecting the listening environment with BBC dip and Sean Olive slope.
4. Coclusion: Flat response is a wrong goal.
1. Sean Ovive's study that Cask05 mentioned is a really good reference. I started my correction based on this study, and I found sloped correction is better than flat correction.
2. Lower frequency is impossible to measure correctly in the room, and the room mode is unavoidable. Bad resonant frequencies at the listening position are notch filtered "by ears" with sine wave generator playing semitones. This method actually works very well for confirming the actual higher frequency response at the listening position. You can clearly hear the resonance or notch at many different frequency points by many different reasons including directivity issue.
3. I was still not 100% satisfied with the result. Currently, a little BBC dip and slight 10K boost are added. I'm pretty sure that most of the commercial mastering were done expecting the listening environment with BBC dip and Sean Olive slope.
4. Coclusion: Flat response is a wrong goal.
Flat response is the only way of hearing what's actually been recorded. Anything else is an effects box.
FWIW, I usually try to aim for flat response, and my preference lies there.
Chris
yes. Because you did not EQ Near-Field. Near-Field is the only way to use EQ. If you attempt to EQ in Far-field... way back in room, it is wrong way and will never sound good or right.
THx-RNMarsh
Hi Chris, when you see the response of Danlavy or B&W speakers, you'll find those speakers are not flat on and off axis response, but in real world, those speakers are two of the most popular reference speakers among classical engineers and mastering engineers historically. Just look at the off axis response of those speakers, you should notice that the response at the listening position of those speakers should have BBC dip in the room.
If you listen to the music mastered with those speakers, with flat on/off axis response speaker in your room, you know what would happen. It is not as bad as playing vinyl without RIAA effect box, though. 🙂
Last edited:
- Status
- Not open for further replies.