What's going on here?

[89-300ce]

I just ran this test last night and have no experience with reading a RTA.

The hump betwen 2k - 8k im guessing is horn hump. I can fix that, but what is going on on the bottom?

I've tried moving the speakers about, changing the porting, and opening doors with little effect.

Would this be the room, the cabinets, or the crossover, or am I running the tests wrong?

The room is 22 x 14 x 8 with acoustic ceiling, wood paneling on studs with fiberglass insulation, and a short commercial style rug.


Jorg

ps. For this test the box is tuned for 35Hz.

[Hoerninger]

Some detailed information about your the speakers would be helpful:
homebrew, chassis, network in detail ...
position of mikrofon ...
___________
Peter

[89-300ce]

Homebrew.

JBL K-145, Altec 511b with BMS 4552nd, and homemade Altec N1201 crossover. The box is 4.8 cu.ft MDF tuned to 35hz, 26" wide x 22" high x 15.5" deep ID's. I think the box is adequately braced, but there is minimal fiberglass stuffing, 18x18x2 suspended verticaly just behind the driver.

Microphone is on a boom tilted down (off of a tripod that is too high) to get it at ear height, and pointing at the centre between the speakers.


Jorg

[Hoerninger]

Wondering whether the network is a good choice, there seems to be hardly an improvement except for high passing.
The room seems to be dominating, others may judge the bass tuning.
http://audioheritage.org/vbulletin/attachment.php?attachmentid=4287&stc=1&d=1103031148

___________
Peter

[JBL 4645]

Was the microphone pointing downwards, instead of upwards at a 45° angle?

[89-300ce]

I should mention that for this test I was running R3 at 0 ohms (full bypass), and R1 at -2db.

I was planing to incorporate a fixed lpad before the comression driver

http://www.audioheritage.org/vbulletin/showpost.php?p=191703&postcount=2

to further attenuate the horn and deleting the R5 resistor entirely.

Jorg

[89-300ce]

Was the microphone pointing downwards, instead of upwards at a 45° angle?

Yes, down, but not quite 45 deg.

Jorg

[Andyoz]

Looks like room mode behaviour to me below about 150Hz.

The large "suck-out" around 80Hz may be due to boundary interference effects, i.e. phase cancellation between the direct sound and the reflected sound off the rear wall or floor. Is the distance from the woofer to the floor about the same as the distance from the rear wall. Try moving the speakers up against the rear wall and the centre frequency of the dip should move a bit.

[89-300ce]

Looks like room mode behaviour to me below about 150Hz.

The large "suck-out" around 80Hz may be due to boundary interference effects, i.e. phase cancellation between the direct sound and the reflected sound off the rear wall or floor. Is the distance from the woofer to the floor about the same as the distance from the rear wall. Try moving the speakers up against the rear wall and the centre frequency of the dip should move a bit.


Yes, the centre of the woofer cone is about 18" off the ground and about 18" from the wall. I have tried moving them out a bit (6") but didn't see a big change. I can't move them back much. I will try again tonight and look closer for any change.

Jorg

[Andyoz]

At a guess, I think the suck out around 170-180 Hz correlates better with the 18" speaker to wall and floor spacing.

The suck-out at 70-80Hz would be consistent with the speakers being about 4 feet from the side wall, with the right speaker being further away from the side wall when compared to the left speaker.

As with anything, this could be completely wrong.:)

[89-300ce]

At a guess, I think the suck out around 170-180 Hz correlates better with the 18" speaker to wall and floor spacing.

The suck-out at 70-80Hz would be consistent with the speakers being about 4 feet from the side wall, with the right speaker being further away from the side wall when compared to the left speaker.

As with anything, this could be completely wrong.:)

That's a pretty good guess. The right speaker is also about 4 foot from the side wall, but the large opening to the fireplace might make it seem like the wall is further back.

So, if I can confirm this to be true where does that leave me? The Tact fixes all of this, but it's just borrowed from a friend and my guess is he will want it back :(. Should I be looking at diffusers to break up those reflections? The wall paneling actualy flexes when pushed on.

Jorg

[89-300ce]

Was the microphone pointing downwards, instead of upwards at a 45° angle?

I guess I'm a little slow. What is the correct orientation of the microphone?

Jorg

[Andyoz]

Not much that can be done unfortunately. Diffusion won't work at these frequencies. Also, trying to correct for it electronically causes more problems than it solves.

I guess this shows why soffit mounting is so good for the low-frequency response. These boundary interference effects no longer apply.

Post some plots with the spekers moved say 2 feet closer to the side wall. Tracking the effects won't be as simple as it sounds as the speakers will interact with the room modes in a different manner with every relocation.

[89-300ce]

Post some plots with the spekers moved say 2 feet closer to the side wall. Tracking the effects won't be as simple as it sounds as the speakers will interact with the room modes in a different manner with every relocation.

Nuts! I'm stuck between a wall and a wall unit.

It took me a while to figure out that as I move away from the side walls I'm getting closer to the wall unit, which just amplifies the 60Hz problem again. If I try to move away from the back wall the hump at 120Hz starts to climb.

I'm guessing the dip at 800-1000 is the woofer coil and cap. I'm going to try removing some capacitance to see if I'm on the right track.

I'm starting to get a feel for it, through trial and error. Thanks for the heads up on the walls.

Jorg

[Robh3606]

Those are your raw room curves?? They look pretty good. Have you tried the correction?? Do you know what the smoothing is and what it uses to measure MLS or Sine?? This is a system curve for my current DIY set-up. With smoothing of course. One is Sine the other MLS. The measurements are reasonably close.

[89-300ce]

Those are your raw room curves?? They look pretty good. Have you tried the correction?? Do you know what the smoothing is and what it uses to measure MLS or Sine?? This is a system curve for my current DIY set-up. With smoothing of course. One is Sine the other MLS. The measurements are reasonably close.

Yes they are raw cuves. Yes I have tried the corrections, I'll try to get some out. They are smooth as silk. I'm not sure what it uses. The unit is designed more for plug and play than DIY. If it was mine I'd just plug it in and let it fix things. It's that good at it.

http://www.stereophile.com/roomtreatments/437/index3.html

Alas, it's not mine, so out comes the soldering iron, etc. , but when it's all done I think I'll have to learn to live with some anomalies.

Jorg

[toddalin]

Yes they are raw cuves. Yes I have tried the corrections, I'll try to get some out. They are smooth as silk. I'm not sure what it uses. The unit is designed more for plug and play than DIY. If it was mine I'd just plug it in and let it fix things. It's that good at it.

http://www.stereophile.com/roomtreatments/437/index3.html

Alas, it's not mine, so out comes the soldering iron, etc. , but when it's all done I think I'll have to learn to live with some anomalies.

Jorg

Welcome to the club!

I've been trying to achieve a flat response in my room and know what you are up against. :biting: This stuff can drive you crazy and I too may have to live with some anomalies. The idea is too get it as good as it can be using what's at hand or readily available. I feel your grief. :blink:

[Robh3606]

Are you trying to adjust your crossover based on these room curves?? If you are why don't you take a couple of close in measurements to minimize the room. You should be able to get a better handle on what the changes are doing. If you run the pots through their ranges where do the adjustment ranges lie?? One looks like an overall the other a trimmer.

Rob:)

[89-300ce]

Welcome to the club!

I've been trying to achieve a flat response in my room and know what you are up against. :biting: This stuff can drive you crazy and I too may have to live with some anomalies. The idea is too get it as good as it can be using what's at hand or readily available. I feel your grief. :blink:

My mistake was letting this thing through the front door. If it where an RTA I would have played with it, gotten it as close as I could, and then consoled myself with the belief that some of the room effect is for the most part not correctable. Unfortunately this thing corrects them. Ignorance is bliss.

Jorg

[89-300ce]

Are you trying to adjust your crossover based on these room curves?? If you are why don't you take a couple of close in measurements to minimize the room. You should be able to get a better handle on what the changes are doing. If you run the pots through their ranges where do the adjustment ranges lie?? One looks like an overall the other a trimmer.

Rob:)

That's an excellent idea. I've noticed that speaker placement does have an effect on that dip.

The R3 pot seems to be primarily for controlling the horn hump, roughly 2KHz-4KHz. I've already switched to the 8ohm terminals of the pot for finer granularity but am still forced to run it at 0 ohms. I'm going to try removing the R5 2 ohm resistor.

The R1 resistor looks like it should be overall high end, but the L1 coil just upstream seems to cause it to effect primarily the upper top end, 4Khz and up. I'm sure it has some effect on the mid but it is far more pronounced on the top. I'm running it cut -2db on it's scale, so also close to minimum. I was planning to use a fixed lpad to bring it to 0db which should also help to bring the mid in line.

You're right, I should get the mic right in there and get a real idea of what the crossover is doing before the room messes it up. Place the mic about midway between the LF and HF and about 4' back?

Jorg

[toddalin]

That's an excellent idea. I've noticed that speaker placement does have an effect on that dip.

The R3 pot seems to be primarily for controlling the horn hump, roughly 2KHz-4KHz. I've already switched to the 8ohm terminals of the pot for finer granularity but am still forced to run it at 0 ohms. I'm going to try removing the R5 2 ohm resistor.

The R1 resistor looks like it should be overall high end, but the L1 coil just upstream seems to cause it to effect primarily the upper top end, 4Khz and up. I'm sure it has some effect on the mid but it is far more pronounced on the top. I'm running it cut -2db on it's scale, so also close to minimum. I was planning to use a fixed lpad to bring it to 0db which should also help to bring the mid in line.

You're right, I should get the mic right in there and get a real idea of what the crossover is doing before the room messes it up. Place the mic about midway between the LF and HF and about 4' back?

Jorg

Even at just 1 meter with the mic located midway up the cabinet, I still see room effects. :blink:

[Robh3606]

Even at just 1 meter with the mic located midway up the cabinet, I still see room effects.

Yes you do and you will. Most of the time the room effects are prominent under about 300Hz or so. Unless you can get a gated MLS measurement you won't be able to remove the room artifacts. Even then the shorter the gate the more resolution you loose and you can easilly see this when you compare the gated vs the ungated measurement. He should be able to do his horn compensation and adjustments for his compression driver with less influence from the room close in. It won't be perfect but it should help make things a bit easier. 2-4 ft should work fine.

Rob:)

[89-300ce]

I quickly did some crossover test last night. I didn't quite match the mic position for measuring from left to right as can be seen by the different levels and response. I do believe I'm going to adjust my woofer coil and cap to bring it up a bit and lpad the horn to bring my controls to mid-point. I'm not sure what's happening at 300Hz. Could that be the back of the box which is about 15.5" deep?


Jorg

[Earl K]

Jorg,

- To remove room effects from contaminating the response curve generated by the woofer / move the measuring microphone to within 1/4" of the woofers cone ( near the dustcap ) . The LF info coming from the ports won't be captured / but in this case you're really testing for smoothness of response above 100 hz .
- You can use an 8 ohm resistor on the HF section of your crossover as a dummy load to completely remove any contributions from the horn/driver , to this new curve .

- Once you have this woofer reference curve , then you can decide what to do regarding the anomalies generated by the rooms surfaces .

:)

[89-300ce]

Thanks Earl,

I'll try that next.

Jorg