Need advice with home studio

[bkspero]

First, I would like to thank everyone who contributes so much time to this forum, and particularly to the gentleman who wrote REW. I am amazed by the power of the software and the amount of information on this site. My son and I have been lurking here and at a few other sites for about a week, and I think that we understand enough to ask reasonable questions.

I will paste some photos into a later note in this thread when I figure out how to shrink them enough. For now, please indulge my wordy description. We have a small single-room home recording studio that is also used for mixing and mastering. And we're trying to use REW to help optimize the sound during playback. The room is 11.8 ft x 13.5 ft with an 8 ft ceiling. The mains are along one of the longer walls about 1.5 ft to the right of center. There is a single window opposite the speakers behind the listening position, a double window on the wall to the left, and a double sliding door closet to the wall to the right. Where the back wall goes up to meet the ceiling it doesn't go straight up into a right angle, instead it slants into the room at an angle for about 10" before it reaches the ceiling. We have 2 inch foam covering much of the exposed walls, and Lernd type corner bass traps along most wall ceiling junctions (except the angled portion). We have a foam cube and Lernd traps in the triple corner at the ceiling to the left of the mains (the door is down a small hall in the right front corner. For all the discussion below, we also placed double thickness of portable fiberglass panels made by Clearsonic about 20 inches from each vertical corner (a total thickness of 3" of fiberglass board). There were also 1.5" Clearsonic panels 18 inches in front of the window opposite the speakers (covering the back first reflection point). Finally, there was a 3" thickness of Clearsonic panels about 18" in front of the closet door (including the 1st reflection point). There was 2" of foam at the left 1st reflection point. Finally, the room has a full carpet with 1/2" foam pad over hardwood flooring. There is nothing on the ceiling. The room is amazingly quiet and "dead" (in a good way, I hope).

Speakers are Mackie 624 mains and a Mackie 12" sub (downfiring active and front firing passive driver). Mains are set about 50" apart and the back of the speakers are about 10 inches from the front wall (just ahead of 2" foam panels). The sub is located on top of 1" heavy foam on the floor just to the left of the left main (without the foam the floor resonates like crazy at 20-25 Hz). The mains are on stands with the tweeters about 53" above the floor. The mains are angled to meet just behind the head of a listener located midway between them and 38% away from the front wall.

We have 2 problems. First, the bass response has some severe valleys that we can't seem to correct no matter what we've done with sound absorption or speaker placement. Second, frequency response is very sensitive to small shifts in listening position, even movements of less than a foot can have drastic effects.

I have attached some graphs showing the problems. All include both the sub and mains. The crossover is set at about 90 Hz. The first graph shows the sensitive dependency on position. It compares four positions within a foot of each other. Red is the normal listening position (38% from front wall midway betw. speakers), green and blue are 1 ft behind and 1 ft ahead of that along the midway axis, purple is about 8 inches to the left of blue. Is this normal? Any suggestions on how we can widen the listening region.

The 2nd graph shows the problem we are having smoothing the peaks and valleys at even the normal listening position. We thought that the valleys were nulls and the peaks resonances. So we thought that if we changed the sound level we would be able to change the peaks without changing the valleys (our reading here taught us that nulls don't respond to sound level, since both the primary and reflected waves change in the same proportion they continue to cancel to the same degree...that's why one shouldn't try to Eq up a null). But the 2nd graph shows that a 4 dB increase in sound level raised all portions of the spectrum about equally. We have a dozen more graphs showing different sub positions, speaker positions, sound absorption configurations, etc. without getting anything better than this. We would appreciate guidance as to what could cause this, and how to fix it. We would prefer treat the room and not use a PEQ, because the room is used for recording as well as playback, and the PEQ won't help with the recording.

[JohnM]

On the effect of volume on dips, if you have a true notch (as deep as you can measure) it will not be affected by volume or EQ as it is caused by the reflection having the same level as the direct sound so they cancel perfectly. It is more common for the reflection to be a little lower in level, so the cancellation is partial - if the reflection is 90% of the direct sound that means 10% remains after the direct and reflected sounds mix, so raising the volume or applying EQ lifts that 10% just as it lifts everything else.

The sensitivity to position is fairly typical of small spaces. you may find by moving even further than you have you find a better spot, the RTA view can be useful to quickly try different positions. Treatment at those frequencies needs to be very thick to have any effect, but you might find an angled rigid panel helps to break up the reflections more.

[Wayne A. Pflughaupt]

Welcome to the Forum!

This might be better suited for our Acoustics Forum, but we'll see where the discussion takes us before deciding whether or not to move it there.

We have a dozen more graphs showing different sub positions, speaker positions, sound absorption configurations, etc. without getting anything better than this.

I would think that the listening position is the only one to be concerned about. What does it matter if a foot or so over the response is bad if that's not where you're listening from?

The 2nd graph shows the problem we are having smoothing the peaks and valleys at even the normal listening position. We thought that the valleys were nulls and the peaks resonances.
But the 2nd graph shows that a 4 dB increase in sound level raised all portions of the spectrum about equally.

That tells me that what you're seeing there aren't nulls and therefore should be equalizable.

We would prefer treat the room...

Well, the folks who sell bass traps will tell you that if you put enough of them in a room, it'll pretty much eliminate depressions in response (or at least greatly reduce them). Of course the problem there is that you'll have no more room in the room...

...and not use a PEQ, because the room is used for recording as well as playback, and the PEQ won't help with the recording.

Not following you there. Why would PEQ to optimize response at the listening position have any effect on recording? Typically electronic instruments are direct-input. Few acoustic instrument are going to get down to where the 40 Hz problem is.

Regards,
Wayne

[bkspero]

Thanks for the quick response (as well as for the program). One point of clarification:

Treatment at those frequencies needs to be very thick to have any effect, but you might find an angled rigid panel helps to break up the reflections more.

By rigid panels, do you mean solid reflecting surfaces? Or do you mean more of the rigid fiberglass panels? If its solid surfaces, we have large (2' x 6'), 1/4" (but heavy) hinged, Clearsonic acrylic panels that we could locate in the room as a test. They are used to isolate drum kits when performing and recording. Would they provide at least a material amount of the desired effect?

[bkspero]

Thanks again for the thoughtful response.

I would think that the listening position is the only one to be concerned about. What does it matter if a foot or so over the response is bad if that's not where you're listening from?

Our thought was that even though we measured the graphs at 1 ft intervals, we still would have some material degree of null between those spots. So we thought that even a few inches of movement could affect those notes. And mixing and mastering requires that the listener move back and forth between the listening position and the console constantly. Often for hours at a time. It's usually only to lean forward and back, but it's tough to do it absolutely reproducibly. Do you think we're overanalyzing? How much of a difference would you expect to hear from the spots 1 ft apart?

Not following you there. Why would PEQ to optimize response at the listening position have any effect on recording? Typically electronic instruments are direct-input. Few acoustic instrument are going to get down to where the 40 Hz problem is.

My son told me I didn't do a good job with that section of the note. Sorry. Bass drum is, I think, the biggest issue. It is recorded with microphone and can have information in that frequency range. He sets a low cut-off of about 25-30 Hz in the computer and uses the information above that.

The microphone directly on the drum may be ok, since it is so close to the drum (often inside of it). If I am analyzing this correctly, the relative amplitudes of the primary and reflected waves are proportional to the inverse square of the distance they travel (minus some extra scatter of the reflected wave lost in the reflection). If so, since the primary wave distance from the bass drum in only a few inches to the mic, and the reflected wave has to travel at least about 8 ft (out and back), the primary wave would swamp the reflection and the mic would only record the primary signal. Does that make sense?

That leaves the 2 high microphones on the drum kit that are placed up and separated to provide stereo imaging and overall sound. I described Clearsonic acrylic panels in the response to JohnM above. They are often used to surround the drum kit during recording (in about an 8 ft circle). Do you think that they are a type of structure that would attenuate and scatter the secondary waves to protect the high mics from reflections?

Thanks again to all.

[Wayne A. Pflughaupt]

I suggest we stick to the playback/listening position issues on this thread, as that is the focus of this particular Forum. For your recording-specific questions I suggest you open a new thread in our Multimedia Production Forum. We have some professional folks there who are well-versed on the topic. You could post your graphs there, too, and get some feedback (no pun intended) on what your best options are.

Regards,
Wayne

[Wayne A. Pflughaupt]

And mixing and mastering requires that the listener move back and forth between the listening position and the console constantly. Often for hours at a time. It's usually only to lean forward and back, but it's tough to do it absolutely reproducibly. Do you think we're overanalyzing? How much of a difference would you expect to hear from the spots 1 ft apart?

I suggest playing pink noise through the speakers and lean around as you would regularly. Pink noise is a particularly unforgiving signal source; you will hear a change in timbre if response is changing at those positions. I suggest doing this little test with one speaker only, as timbre will change if your position in relation to the two speakers changes - i.e., moving slightly closer to one or the other.

Something else you could do, possibly in conjunction with the pink noise test (especially if it does show timbre change) would be to take a few REW readings at the various places where your head will be during mixing and mastering, to see at what frequencies response changes are taking place. Again, I suggest using a single speaker for this to avoid comb filtering from being off-centered between the speakers, since I presume that you aren't always located prefectly between the two.

Once you see in REW where potential problem areas are, you could play some music you're familiar with and see if what you're finding with the pink noise and/or REW tests transfers to listening tests. For instance, if an REW reading you took in the leaning-back position showed increased bass in the 60 Hz range, listening tests may show that it's not nearly as audible as it looked on the graph. Or, it could sound just as bad as it looked.

Hope this helps.

Regards,
Wayne

[brucek]

I suggest playing pink noise through the speakers and lean around as you would regularly. Pink noise is a particularly unforgiving signal source; you will hear a change in timbre if response is changing at those positions. I suggest doing this little test with one speaker only, as timbre will change if your position in relation to the two speakers changes - i.e., moving slightly closer to one or the other.

Something else you could do, possibly in conjunction with the pink noise test (especially if it does show timbre change) would be to take a few REW readings at the various places where your head will be during mixing and mastering, to see at what frequencies response changes are taking place. Again, I suggest using a single speaker for this to avoid comb filtering from being off-centered between the speakers,

or simply use the RTA feature of REW while moving around with the microphone.

brucek

[bkspero]

Thanks for the continued input.

I suggest you open a new thread in our Multimedia Production Forum. We have some professional folks there who are well-versed on the topic.

Sorry for the naive question, but do I just manually start a new thread (cut and paste the original note and graphs), or is there some way to shift this thread there?

This morning we set up the acrylic panels along one side wall and half of the back wall in an accordian configuration. It was a big help with widening the listening position. The two big nulls at the +- 1 ft positions are gone. There are still peaks and valleys, but no 20+ dB dips.

Plans now are to build some fiberglass bass traps and also some diffusers. And treat the ceiling as well.

[Wayne A. Pflughaupt]

Sorry for the naive question, but do I just manually start a new thread (cut and paste the original note and graphs), or is there some way to shift this thread there?

You can cut and paste from here what you want to have them discuss. No need to upload the images again, you can drop the URL's between "IMG" tags.

Or, I can move the whole thread over there if you like.

Regards,
Wayne

[bkspero]

I will repost. Thanks.