SPL and Waterfall Interpretation

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
I did some additional measurements using a wider speaker spread (60") and moved the mic forward, backward, and sideways. The center position was 60" forming an equilateral triangle. Each movement about the center was roughly 6 inches. Here are the results.

There are some subtle differences including a decent reduction of the null in the 120-130hz range. Again, though, there are tradeoffs. This includes the introduction of another null at roughly 75-80Hz.

It also seems that I've added more paper to the dumpster fire and increased the number of variables at play. It seems that I'm looking at speaker spread, distance from the speakers, and distance to the front wall as factors in finding a sweet spot or at least reducing nulls. I wish I could narrow this list somehow. But I need to find an answer to the equilateral triangle issue and rule out the idea of moving the speakers back towards the front wall.
 

Attachments

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
Also here is an example measurement of what I’m currently listening to. This is 100% review gear that I setup in less than an hour in my dedicated room, so I think it’s looking pretty good. You can see remaining evidence of my one problem area, ceiling reflection interference.

B10C5CFD-7894-4DCC-99BE-84D88A66EA1E.jpeg
 

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
I did some additional measurements using a wider speaker spread (60") and moved the mic forward, backward, and sideways. The center position was 60" forming an equilateral triangle. Each movement about the center was roughly 6 inches. Here are the results.

There are some subtle differences including a decent reduction of the null in the 120-130hz range. Again, though, there are tradeoffs. This includes the introduction of another null at roughly 75-80Hz.

It also seems that I've added more paper to the dumpster fire and increased the number of variables at play. It seems that I'm looking at speaker spread, distance from the speakers, and distance to the front wall as factors in finding a sweet spot or at least reducing nulls. I wish I could narrow this list somehow. But I need to find an answer to the equilateral triangle issue and rule out the idea of moving the speakers back towards the front wall.
Can you leave the speakers in this 60” spread but move the mic back a full foot or so from the position you used for this. I just want to see what is going on.

Also, did you ever indicate if these were measurements of only one channel or both?
 

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
This might be of interest. I moved the measurement mic forward and backward. The red line shows the mic at 48" with the speakers spread at 48" (an equilateral triangle). The green line represents the mic at 36" with the speakers again located 48" apart. Finally, the blue line shows the mic pushed back to 60" with the same 48" speaker spread.

The mic movement has reduced the null but with some tradeoffs. In particular, a dip at 80hz when the mic is at 60" and a further boost at 150hz when the mic is at 36". Moving the mic also raises that question about the importance of the equilateral triangle. I may need a variance from the rule granted by one of the home studio gods. Or at least an authority on room acoustics.
How did you get that Green measurement now? Moved the mic in closer to the speaker? If so, wow that is a workable response. You can simply apply eq to knock down the 150hz. That’s not a problem at all.
 

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
One more set of runs, despite my warning about adding paper to the fire...

I set up the speakers at 52" tweeter-to-tweeter and took a baseline measurement with the mic at 52." I then moved the mic forward in 12" increments until I was at the mixing desk. In other words, the mic was move forward from 52" by 12" and then 24."
 

Attachments

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
Also, did you ever indicate if these were measurements of only one channel or both?
As I mentioned somewhere in the weeds above, these are measurements using both speakers. I started out by measuring each speaker individually, but they were essentially the same.
 

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
How did you get that Green measurement now? Moved the mic in closer to the speaker? If so, wow that is a workable response.
I need more color separation. There is green and light green. The dark green indicates the mic moved 6" forward. The light green represents moving the mic 6" to the right.
 

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
Can you leave the speakers in this 60” spread but move the mic back a full foot or so from the position you used for this.
I think that set up would put the mic in the exact center of the room. Is that what you're looking for? It sounds like a way of looking at the worst case scenario to me.
 

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
As I mentioned somewhere in the weeds above, these are measurements using both speakers. I started out by measuring each speaker individually, but they were essentially the same.
Measure only one speaker at a time. Your adding variables we don’t want to deal with. Better to measure the left and then right speaker. It’s true that at LF’s it shouldn’t be that different but still best to avoid inadvertent cancelations.
 

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
One more set of runs, despite my warning about adding paper to the fire...

I set up the speakers at 52" tweeter-to-tweeter and took a baseline measurement with the mic at 52." I then moved the mic forward in 12" increments until I was at the mixing desk. In other words, the mic was move forward from 52" by 12" and then 24."
Is the purple graph a viable listening position or is that your head over the mic desk itself?
 

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
Ok. As requested, here is the FR using a 60" speaker spread with the mic placed at 72" or one foot behind the previous set of measurements. That places the mic at the exact center of the room.
 

Attachments

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
Is the purple graph a viable listening position or is that your head over the mic desk itself?
The purple line represents "elbows on the table" with my head at the mic position close to the front of the desk. So it's a viable position unless the recording gods strike me down for violating the equilateral triangle maxim. I keep coming back to that and I'm wondering if it's a hard and fast rule or just something that became hard and fast after being repeated on the internet until it became truth.
 

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
After looking at pictures of the Washington monument and the back of the dollar bill, I'm beginning to think that the equilateral triangle rule is much deeper--perhaps a Masonic thing. Did the founders of our Nation have mixing desks?
 

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
One more set of measurements.

Speakers at 60"
Mic positioned 36" from speakers just off desk front where I would normally sit to edit.
Two measurements done using left then right speaker (Left speaker in blue)
 

Attachments

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
One more set of measurements.

Speakers at 60"
Mic positioned 36" from speakers just off desk front where I would normally sit to edit.
Two measurements done using left then right speaker (Left speaker in blue)
Ok well let’s forget everything so far and go with this. Those are good measurements. You can fix that peak with eq. The null is gone. If you can send the mdat file I can check one last thing and I would say you are good.
 

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
Here is the mdat file.

I'm a little unclear on the EQ issue. The recording world seems to fear room EQ almost universally. I not sure of the exact argument but it's almost universally condemned. Perhaps it's less of an issue if you are looking at a LF frequency boost. But it appears that addressing a null with EQ is viewed as hopeless and pointless. Sort of like trying to fill a black hole.

The other issue is how to address the boost. I can identify the frequency, but I'm not clear how much would need to be cut. Is that something REW can determine? And where would the actual correction take place. I assume that would happen in my DAW software (Logic ProX) by adding an EQ filter to the final stereo output while mixing. Then removing the filter when the track is complete and ready to be output or bounced. Does that sound about right?
 

Attachments

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
Here is the mdat file.

I'm a little unclear on the EQ issue. The recording world seems to fear room EQ almost universally. I not sure of the exact argument but it's almost universally condemned. Perhaps it's less of an issue if you are looking at a LF frequency boost. But it appears that addressing a null with EQ is viewed as hopeless and pointless. Sort of like trying to fill a black hole.

The other issue is how to address the boost. I can identify the frequency, but I'm not clear how much would need to be cut. Is that something REW can determine? And where would the actual correction take place. I assume that would happen in my DAW software (Logic ProX) by adding an EQ filter to the final stereo output while mixing. Then removing the filter when the track is complete and ready to be output or bounced. Does that sound about right?
I’ll take a look later tonight and show you how to figure out the eq to be used. Yes REW can calculate it.

Eq shouldn’t generally be used to fill mills because they are typically caused by cancelation. Making a source louder that is being canceled doesn’t change the output because the cancelation wave becomes stronger. Very low q troughs can be eqed however.

In this case we are removing a bump in the response, not trying to fill a null. It’s totally ok. Acoustics are acoustics. There are no differences in studio and home acoustics. Only differences in small and large rooms. As such, given the vast research and evidence supporting the use of eq to address room mode peaks suggests anyone against it doesn’t understand acoustics.
 

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
Here is the mdat file.

I'm a little unclear on the EQ issue. The recording world seems to fear room EQ almost universally. I not sure of the exact argument but it's almost universally condemned. Perhaps it's less of an issue if you are looking at a LF frequency boost. But it appears that addressing a null with EQ is viewed as hopeless and pointless. Sort of like trying to fill a black hole.

The other issue is how to address the boost. I can identify the frequency, but I'm not clear how much would need to be cut. Is that something REW can determine? And where would the actual correction take place. I assume that would happen in my DAW software (Logic ProX) by adding an EQ filter to the final stereo output while mixing. Then removing the filter when the track is complete and ready to be output or bounced. Does that sound about right?
Your approach sounds fine. I’m less an expert in the specifics of studio software. If there is no way to insert an eq in the monitor out only, then likely your suggested approach is the best option.

I use a Motu 828x for my room acoustics work, since I have it i use it as an 8 channel DAC as well. I apply eq correction to that using various softwares that let me choose the output and the eq is only applies to that output. So if I use the monitor out, I can choose just the monitor out as my output device and eq is only applied to that. As such, if I record something on the inputs, no eq is applied. Is it not possible to do similar with your setup? Apply a PEQ filter to just the monitor output?
 

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
I worked my way through the EQ section using the following setup. I'm not sure about all the settings because it's based on a random internet tutorial. You never know what you're getting in cyberspace.

All SPL>average L and R speaker
Open EQ
Set EQ to Generic
Target settings=Full range
Set target level
Match range=45-10,000
Max boost=0
Overall max boost=0
Flatness target=1
Match response to target
Open EQ filters or export filter setting as text
under filter tasks

Results:
Equaliser: Generic
Average 1

Filter 1: ON PK Fc 148 Hz Gain -6.3 dB Q 8.021
Filter 2: ON PK Fc 192 Hz Gain -2.1 dB Q 7.838
Filter 3: ON PK Fc 525 Hz Gain -1.6 dB Q 5.000
Filter 4: ON PK Fc 629 Hz Gain -2.8 dB Q 5.000
Filter 5: ON PK Fc 833 Hz Gain -3.7 dB Q 5.000
Filter 6: ON PK Fc 1355 Hz Gain -7.1 dB Q 5.000
Filter 7: ON PK Fc 4195 Hz Gain -2.3 dB Q 5.000
Filter 8: ON PK Fc 5629 Hz Gain -5.0 dB Q 1.758
Filter 9: ON PK Fc 8745 Hz Gain -2.9 dB Q 5.000
Filter 10: ON PK Fc 9971 Hz Gain -4.0 dB Q 5.000

My plan from here is to enter the data into a ten band EQ in Logic and place it on the stereo out during mixing. Once I make my mixing decisions, I'll bypass or perhaps remove the EQ and bounce the finished track. I'm not sure if Logic has a stock 10 band EQ. So I may have to purchase a new plugin.
 

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
I worked my way through the EQ section using the following setup. I'm not sure about all the settings because it's based on a random internet tutorial. You never know what you're getting in cyberspace.

All SPL>average L and R speaker
Open EQ
Set EQ to Generic
Target settings=Full range
Set target level
Match range=45-10,000
Max boost=0
Overall max boost=0
Flatness target=1
Match response to target
Open EQ filters or export filter setting as text
under filter tasks

Results:
Equaliser: Generic
Average 1

Filter 1: ON PK Fc 148 Hz Gain -6.3 dB Q 8.021
Filter 2: ON PK Fc 192 Hz Gain -2.1 dB Q 7.838
Filter 3: ON PK Fc 525 Hz Gain -1.6 dB Q 5.000
Filter 4: ON PK Fc 629 Hz Gain -2.8 dB Q 5.000
Filter 5: ON PK Fc 833 Hz Gain -3.7 dB Q 5.000
Filter 6: ON PK Fc 1355 Hz Gain -7.1 dB Q 5.000
Filter 7: ON PK Fc 4195 Hz Gain -2.3 dB Q 5.000
Filter 8: ON PK Fc 5629 Hz Gain -5.0 dB Q 1.758
Filter 9: ON PK Fc 8745 Hz Gain -2.9 dB Q 5.000
Filter 10: ON PK Fc 9971 Hz Gain -4.0 dB Q 5.000

My plan from here is to enter the data into a ten band EQ in Logic and place it on the stereo out during mixing. Once I make my mixing decisions, I'll bypass or perhaps remove the EQ and bounce the finished track. I'm not sure if Logic has a stock 10 band EQ. So I may have to purchase a new plugin.
Ah sorry, I should have explained this before. I do not suggest applying eq above around 300-500hz and since this was largely based on a single measurement point, I would only focus on that stuff around 150-200hz.

I wouldn’t apply eq in other higher areas for now. I suspect based on your response that the speakers response is non-constant with angle.
 

dwillis60

Member
Thread Starter
Joined
Jan 2, 2019
Messages
42
Ah sorry, I should have explained this before. I do not suggest applying eq above around 300-500hz and since this was largely based on a single measurement point, I would only focus on that stuff around 150-200hz.
Thanks. That thought had occurred to me, too. And on some level, it simplifies the task. There are only two adjustments in that LF range. So I won't need to purchase a full ten band EQ plugin to insert into my signal chain. The stock EQ in Logic should do the trick.

Thank you for all your help. It's been a great learning experience and you've been very insightful. It's great to see that people are so unselfish with their time and experience, and so willing to help the uninitiated! You deserve a giant pat on the back for helping me and for adding value to so many posts in this forum.

Thanks again
 

Matthew J Poes

Staff Writer
Staff member
Joined
Oct 18, 2017
Messages
1,830
Thanks. That thought had occurred to me, too. And on some level, it simplifies the task. There are only two adjustments in that LF range. So I won't need to purchase a full ten band EQ plugin to insert into my signal chain. The stock EQ in Logic should do the trick.

Thank you for all your help. It's been a great learning experience and you've been very insightful. It's great to see that people are so unselfish with their time and experience, and so willing to help the uninitiated! You deserve a giant pat on the back for helping me and for adding value to so many posts in this forum.

Thanks again
Well, REW is an important part of what we do here at AV NIRVANA. I charge customers who ask me to do this for them. I see no reason to charge when someone wants to learn. The entire point of REW is to empower everyone to have this valuable capacity. I think it’s especially important to help people learn the right way to use REW.

I find a lot of hard and fast rules or absolutes stem from research that was far more nuanced. Where the right answer depends on an understanding of the factors at play. I understand that in trying to simplify the concepts for the uninitiated it made it possible for these ideas to be accessible. I just don’t like when it’s taken to the ridiculous and people begin fighting over who does it right. Better to teach people the complex science in easy to understand terms rather than develop inaccurate rules of thumb.
 
Top Bottom