Sub 'In-Room' Measurements, do room modes affect measurements right by or 'inside' the sub cabinet?

pedwar

Established Member
Joined
Jun 15, 2001
Messages
380
Reaction score
37
Points
79
Location
London
Hi Guys,

Thought I'd ask here as I'm sure there are people here with masses more knowledge than me on this subject... :laugh:

I've got an SVS PB200 Pro Ported sub. My room is about 16ft x 11ft (with no door at the moment), no traps/treatment yet and has carpet and soft furnishings....

I've EQ'd things so I get a reasonable response at my MLP from a single sub (had to pull down (-12db) a large room mode at 37Hz and there's another I pickup at about 71hz in some areas but at MLP actually gives me a tiny dip so I EQ that up ~3dB).

My question is... Why EVEN when I put the UMIK-1 Mic ~4" from centre point between the cone and ports I still get a similar response curve to what I see in my room?

I also see the same if I put the mic 4" - 5" inside the port on the subwoofer.

Are room modes really that strong that cancellation is occurring 'inside' the sub's cabinet? I mean I guess the cabinet is only 25mm or so MDF so I would expect waves to bounce back through it, itself.

I see the same in other locations in the room as well.

I guess the term 'in-room' response literally means that in this case? i.e. it's not just performance at different points in the room... it's literally that units performance 'in-room' ? If so, I'm glad I went with the ported unit instead of sealed as I was actually worried it would give me too much low end :laugh:

I was quite surprised I don't see any room gain in the lower region (although I probably need a bigger room for the very low frequencies which is why I'm seeing a 37Hz peak).

This is the curve I see from my MLP (neighbours were home so ignore the low dBs.. the curve is the same at higher volumes).

Yellow = Ported Mode, Green = Sealed mode (with bungs).

PB2000Pro.jpg



I mean I can EQ things up a bit with DynamicEQ or a house curve, but was surprised I don't see more naturally.
 
I also see the same if I put the mic 4" - 5" inside the port on the subwoofer.

If you got the same measurement with the microphone inside the port as at the MLP then my bet is you were accidentally recording the signal from a different microphone instead e.g. one built into a laptop you were using perhaps?

The correct UMIK-1 measurement when it's inside the port will have a peaked response centered around the port tune frequency and will fall off above and below this. The red curve below is an example of a measurement I made with a UMIK-1 just inside the port of a BK Monolith to give you the idea. (The blue curve is the response with the microphone close to the driver.)

Monolith nearfield.jpg
 
If you got the same measurement with the microphone inside the port as at the MLP then my bet is you were accidentally recording the signal from a different microphone instead e.g. one built into a laptop you were using perhaps?

The correct UMIK-1 measurement when it's inside the port will have a peaked response centered around the port tune frequency and will fall off above and below this. The red curve below is an example of a measurement I made with a UMIK-1 just inside the port of a BK Monolith to give you the idea. (The blue curve is the response with the microphone close to the driver.)

View attachment 1458887

Oh no, I do get a different response from the MLP.. there's a huge bump at where my room modes ae (37Hz and ~71Hz).. so I think they're pushing the curve up in the middle, where as my curve should be similar to yours for my PB2000 Pro.

I'll grab a measurement in the port and show you. Maybe that is the case, that I am getting the flat response but the rise in the middle from the room gain in the middle frequencies is making the 20Hz region look lower.
 
Yes, post all your measurements to show what is concerning you. As per my example, room modes should not be dominating nearfield measurements. On the other hand what actually matters is what you get at the MLP so I wouldn't get too hung up on this.
 
Last edited:
Your scale is at 20dB as well, that'll be masking a lot of variation too, although probably isn't the cause here. The gap between the two responses at 20Hz is close to 10dB. From 60Hz to 20Hz you fall 20dB, which is a lot.

However, the shape should still be visibly different between the MLP, inside the port, and nearfield to the driver, even with a 20dB scale.
 
Ok, here's a measurement inside the port with the UMIK-1, I made sure the mic wasn't touching the inside of the port so there was no vibration.

(top line is from inside the port)

PB2000Pro.jpg


See what I mean.. :) It's like the room modes are still measurable 'inside' the sub box. I would have thought I could get a 'factory' looking measurement that close to the sub, but it looks like it can never perform that way even in my environment? I mean it still sounds great, it just intrigued me.

This is the 'factory' measured curve:

SVS_Image.JPG
 
(also bear in mind that measurement I took above should have a 12db dip from PEQ at 37Hz which isn't visible at all with the mic there which again makes me think the gain it must be getting from the room at that frequency is overpowering the 'flat' measurement I would normally get from the sub if it was in an open field?)
 
I would have thought I could get a 'factory' looking measurement that close to the sub,
No this is not what to expect. See my post above with the graph :) .

I don't believe your measurements are accurately showing you what is really going on but I can't quickly think what may be causing the odd looking results.

Unless there is a setting on the sub or somewhere else that may be significantly affecting the output?

How does it sound?
 
Try running the real time analyser and playing back some content, even if that's test tones.
As it's playing move the mic closer to the sub and the further away. The level should change.

You also really need to change that scale. You're showing about half the SVS graph on the horizontal axis (frequency) and twice as much on the vertical axis (SPL). Even if you got a matching measurement it wouldn't look like the SVS chart, it would be much flatter.
 
No this is not what to expect. See my post above with the graph :) .

I don't believe your measurements are accurately showing you what is really going on but I can't quickly think what may be causing the odd looking results.

Unless there is a setting on the sub or somewhere else that may be significantly affecting the output?

How does it sound?
Yeah it sounds ok and for an non-ideal smaller room I don’t think the MLP graph is too bad, I almost went sealed as I was worried about bitten end boom, I have the opposite issue (I guess hard work for it forming a wave long enough for 20Hz in there but it does ok.

it is situated in a corner at the moment but I get similar measurements for MLP elsewhere, not tried a near field measurement to it with it all pulled out yet.

My settings are as you would expect for a test:
Crossovers 80Hz
Speakers Small
LFE set to normal and 120Hz
DynamicEQ off
Audyssey correction Off
Tone, Bass all at 0
LFC of course off

I bet it’s sound waves bouncing around in that corner and resonating in/around the box overpowering the mid and mid-high bass?

maybe I should try a sweep at very low volume inside the port to see what I get 🤔
 
Yeah it sounds ok and for an non-ideal smaller room I don’t think the MLP graph is too bad
The response in the graph looks pretty rubbish actually but if it sounds OK I think the graph is wrong.
 
The response in the graph looks pretty rubbish actually but if it sounds OK I think the graph is wrong.
😄 well that’s why I was hoping for a naturally flatter response (and why I measured near field on I know the problem corner I’ve got that needs some absorption/trapping in there, I probably need to sort that next.

I haven’t got enough PEQ to pull down the 70Hz area at the moment if I up the level of the sub too much, it is more level when I run the gain hotter, but yeah for such a small room with that sub I had dreams a lovely house curve without going too nuts on the EQ
 
Try running the real time analyser and playing back some content, even if that's test tones.
As it's playing move the mic closer to the sub and the further away. The level should change.

You also really need to change that scale. You're showing about half the SVS graph on the horizontal axis (frequency) and twice as much on the vertical axis (SPL). Even if you got a matching measurement it wouldn't look like the SVS chart, it would be much flatter.
It’s 100% the right mic measuring.
I even went through all the laptop settings to make sure there was no vocal mic boost funky was going on or other Maxxbass rubbish, the independent USB interface on the mic does it’s job to steer clear though.
 
😄 well that’s why I was hoping for a naturally flatter response (and why I measured near field on I know the problem corner I’ve got that needs some absorption/trapping in there, I probably need to sort that next.
I think the priority should be trying to sort out what is going wrong with the measurements actually. Bass traps have to be BIG to be much use at all.

If you use the real-time analyser (RTA), what does the spectrum you see look like with no sound playing?
 
Something doesn't look right there..🤔

This is my PB2000 pro.
 

Attachments

  • IMG-20210203-WA0004.jpeg
    IMG-20210203-WA0004.jpeg
    231.7 KB · Views: 46
  • 20210203_140027.jpg
    20210203_140027.jpg
    259.6 KB · Views: 45
Something doesn't look right there..🤔

This is my PB2000 pro.

Was that taken from inside the port? / near-field?

Also is your unit corner loaded at all?

I'll crack out the mic again tomorrow and have another play.
 
Was that taken from inside the port? / near-field?

Also is your unit corner loaded at all?

I'll crack out the mic again tomorrow and have another play.
No mines at MLP and it is corner loaded. I was playing with crossover selection 100hz best in my room.
 
No mines at MLP and it is corner loaded. I was playing with crossover selection 100hz best in my room.
Can I ask what size room you have? (does door open/closed make a difference if it's a regular shaped room).

Did you need much PEQ to get that response?

I think my 37Hz -12db Room mode cancellation is pulling down my lower region due to the width of the Q (it's not too wide but it's reasonable).

I'll wipe all the corrections tomorrow, pull it out of the corner and get a measurement inside the port... as at the moment I get a better measurement at MLP than right by it so I'm sure the room is helping that. If the neighbours are out I'll up the measurement volume to a better level as well as it gets a little flatter then (especially as it gets closer to the 70-80HZ crossover region).
 
I'd thought the graphs above were with no EQ applied. Don't move the sub. In the first instance if suggest three measurements:

  1. The background noise RTA test I suggested above.
  2. A frequency sweep at the MLP with no EQ applied.
  3. A frequency sweep measurement with the microphone just inside the port.
Post these tomorrow and let's take it from there. Please don't apply any smoothing to the graphs.
 
Can I ask what size room you have? (does door open/closed make a difference if it's a regular shaped room).

Did you need much PEQ to get that response?

I think my 37Hz -12db Room mode cancellation is pulling down my lower region due to the width of the Q (it's not too wide but it's reasonable).

I'll wipe all the corrections tomorrow, pull it out of the corner and get a measurement inside the port... as at the moment I get a better measurement at MLP than right by it so I'm sure the room is helping that. If the neighbours are out I'll up the measurement volume to a better level as well as it gets a little flatter then (especially as it gets closer to the 70-80HZ crossover region).
My room is 5m x 4.5m in a shallow L shape. My living door open or closed has no effect. I have a 35hz room gain and 70hz null, so I have to have a high crossover due to the null. I've not used the SVS peq. All has been done with dirac.

My sub faces the wall too and not firing into the room.
 
I would expect a door being open or closed to have some effect in every room to be honest but there are bigger issues here :) .
 
My room is 5m x 4.5m in a shallow L shape. My living door open or closed has no effect. I have a 35hz room gain and 70hz null, so I have to have a high crossover due to the null. I've not used the SVS peq. All has been done with dirac.

My sub faces the wall too and not firing into the room.
That sounds VERY similar to my room mode and my null. just about to post new sweeps below :)
 
Ok guys... the neighbours went out.... the Mrs popped to the shops so I set to work rattling the house with the BEEEEWWW noises we've all come to love and you were all right!! 😀

  1. So I did sweeps with the UMIK-1.. with and without calibration file (in case the file was screwing things up).
  2. I swapped to the Audyssey calibration Mic (and hats off to the mic it was actually pretty similar to the calibrated Mic).
  3. Checked and double checked the MaxxBass rubbish was all off on the laptop, all 'speech enhancements' etc. etc. all off and it was. dug deep into the drivers... everything off
  4. Still the same..
Then thought.. there's nothing wrong with the measurements but my ears are telling me different....

I picked up my GFs laptop (another W10 DELL).. downloaded REW, plugged in the UMIK-1, ran some sweeps... and BAM... There is my 20Hz Curve I was looking for and could hear!!! haha

Check it out...

In port sweeps:

Green = GF Laptop No EQ
Blue = GF Laptop with 3x PEQ bands applied
Red = My newer laptop No EQ
Yellow = My newer laptop with PEQ

I even had to lower the volume on for the in-port sweeps on my GFs laptop as the input signal was clipping badly.. which makes me believe the driver on the laptop is doing something dumb to reduce clipping and cutting the input signal

In_Port_Sweeps.jpg



Here are my MLP sweeps (all UMIK-1)

I'm sure you can figure it out... but legend is:

Pink = My PEQ'd curve from GF Laptop
Blue = No-EQ curve from GF Laptop
Yellow = PEQ'd curve from my laptop
Red = No-EQ curve from my laptop

MLP Sweeps.jpg


That would explain why it sounded really bloated when I tried a bump at 20Hz or enabling DynamicEQ (which I try to avoid most of the time because of what it does to the surround channels).

THAT is what I was expecting from a ported sub of this caliber in my space 😎

Thank you all for coming back to me and making me question the measurements, apart from the dip at 80Hz I'm happy with that for now.

This afternoon's task will be removing all the audio drivers on the (only recently built DELL 7400 Latitude.. ) reinstalling the drivers and sitting there going through device manager properties with a fine tooth-comb to make sure there are NO enhancement 'things' in there.

I'm interested to put the bungs back in and see what I'm getting below 15Hz at MLP now :laugh: (although the Neighbors are home again for the moment)
 
BTW - I figured out it was the input side on the laptop not output by running the sweep output sound from my GFs laptop and recording the sweep on my laptop (just clicking measure on both at exactly the same time with them wired up accordingly).
 
Not sure what is going on with your laptop but your newer measurements look much more sensible/believable :) .
 

The latest video from AVForums

Is 4K Blu-ray Worth It?
Subscribe to our YouTube channel
Back
Top Bottom