The No.1 Website for Pro Audio
Impulse Response Interpretation
Old 4 weeks ago
  #1
Gear Head
 
EastNashRocker's Avatar
 

Impulse Response Interpretation

Hello,

I've been making some room improvements and things have come a long way except that I still have four spikes on the impulse response that jump above -20. They are happening at 1, 3.5, 15.5, and 19m. It's totally possible I'm not evaluating this correctly since I'm not as familiar with this measurement.

A little about my room.. I have corner traps floor to ceiling in all corners, all first reflection points are covered with 703, there is a significant cloud above the listening area, and the back wall is nearly covered with 703. My desk is 2ft wide by 1.5ft deep and is angle about 10%. The floor is carpeted with padding underneath. The desk covers the floor bounce or at least it passes the mirror test. My speakers are close to the front wall. After extensive testing with REW that really seemed to be where they work the best. A month ago I took all the insulation out (there was too much) and started with speaker placement and brought treatment one piece at a time and measured the room to ensure improvement. My speakers are on the short wall and I've followed the 38% recommendation for placement of the listening area. The Carl Tatz room mode calculator has been helpful. My room is about 15x13x10ft

I've done my best to measure the distance with a piece of string to figure out what could be causing the spikes on the impulse response but I'm STUMPED. I've taken extra soundproofing and put it in all different places and the REW impulse measurements don't get any better.. on top of my desk, all over the floor, all four walls, etc.

What does the spike at 1m mean?
Is it possible that the spikes are still coming from the first reflections by bouncing off the 703?
How bad are the spikes that I have?
What am I missing??

Thanks for any thoughts or suggestions!
Attached Thumbnails
Impulse Response Interpretation-screen-shot-2020-06-10-9.41.35-pm.jpg  
Attached Files
File Type: mdat June 2020 11.0.mdat (14.03 MB, 13 views)
Old 4 weeks ago
  #2
Lives for gear
 
akebrake's Avatar
 

ETC, two things

Quote:
Originally Posted by EastNashRocker View Post
... It's totally possible I'm not evaluating this correctly since I'm not as familiar with this measurement.
Hello and welcome back

You’re in good company because most people don’t realise what's not shown on the (usually presented) ETC

1. The Soundwave have launched from the speaker some time before it appears as the Direct spike on the Impulse plot!
That Depends on the speaker to mic distance.

EDIT: Example here
(Reflections create peaks and nulls in the FR. PLD Shows the first interference Null)

Say you mic is 4ft from the speaker. That means the first spike (Direct) appears @ 4ms, not 0ms... The first 4ms are invisible on that ETC unless you have set your T0 (T0= Reference Time) at an earlier time.
You have several choices in the Measure window. Pic 1

Then, say you spot a strong reflection 11ms after Direct. REW can help translate that time (ms) to distance in meter or ft. (Zoom info)
Remember you have to add the Time of Flight (TOF) 4ms to that ETC figure because the reflected path is also longer.

2. The Envelope Time Curve (ETC) is ”frequency blind” e.g. the spikes you see ”favours” the highest frequencies.
Lower frequencies are lurking in the deep, spread out in time. To watch them you have to filter the Impulse.

Example:June 6 Left 800Hz 1/3 Oct filtered. Reflection 11 ms after direct only 9 dB down. Pic 2

Quote:
What does the spike at 1m mean?
It means a high frequency reflection 1 millisecond (ms) after the Direct sound.
Usually the desk or another object close to the speaker or mic.

Hope that helps

PS Show us a link to your earlier thread with photos etc.
Attached Thumbnails
Impulse Response Interpretation-set-ref-time.jpg   Impulse Response Interpretation-zoom.jpg   Impulse Response Interpretation-june-6-r-11ms-800hz-filt.jpg  

Last edited by akebrake; 4 weeks ago at 01:48 PM.. Reason: Link & Info added
Old 4 weeks ago
  #3
Just a couple things to add when looking at the ETC:

1) Use % over dbfs for studying and looking at ETC.
It makes it easier to see and compare peaks.

2) Select zero phase filtering if you use filtered ETC.
When you filter the ETC there can be a phase shift based on the EQ filter, if you select zero phase shift it gives you a response of zero phase shift.

REW says;
"This may be useful when comparing the locations of reflection peaks in filtered impulse responses or the ETCs derived from them as the peaks will not be shifted in time by the effect of the filter's phase response".

3) Select show data panel.

When reading the data the ideal is an R of -1. The closer the better.
Anything outside -.98 of the range will show as orange and may not be reliable when comparing decay figures.

When looking at the curvature 0-5% is what you normally want. Higher than 10% can be feedback or a 2 stage decay curve which is what you don't want.
Negative figures have to be viewed with caution.
Attached Thumbnails
Impulse Response Interpretation-screen-shot-2020-06-11-1.27.02-pm.jpg   Impulse Response Interpretation-screen-shot-2020-06-11-1.27.58-pm.jpg   Impulse Response Interpretation-screen-shot-2020-06-11-1.28.17-pm.jpg  
Old 4 weeks ago
  #4
Gear Head
 
EastNashRocker's Avatar
 

Quote:
Originally Posted by akebrake View Post
Hello and welcome back

You’re in good company because most people don’t realise what's not shown on the (usually presented) ETC

1. The Soundwave have launched from the speaker some time before it appears as the Direct spike on the Impulse plot!
That Depends on the speaker to mic distance.

EDIT: Example here
(Reflections create peaks and nulls in the FR. PLD Shows the first interference Null)

Say you mic is 4ft from the speaker. That means the first spike (Direct) appears @ 4ms, not 0ms... The first 4ms are invisible on that ETC unless you have set your T0 (T0= Reference Time) at an earlier time.
You have several choices in the Measure window. Pic 1

Then, say you spot a strong reflection 11ms after Direct. REW can help translate that time (ms) to distance in meter or ft. (Zoom info)
Remember you have to add the Time of Flight (TOF) 4ms to that ETC figure because the reflected path is also longer.

2. The Envelope Time Curve (ETC) is ”frequency blind” e.g. the spikes you see ”favours” the highest frequencies.
Lower frequencies are lurking in the deep, spread out in time. To watch them you have to filter the Impulse.

Example:June 6 Left 800Hz 1/3 Oct filtered. Reflection 11 ms after direct only 9 dB down. Pic 2



It means a high frequency reflection 1 millisecond (ms) after the Direct sound.
Usually the desk or another object close to the speaker or mic.

Hope that helps

PS Show us a link to your earlier thread with photos etc.
Thanks for the help, I was measuring all wrong b/c I wasn't factoring in the distance between the speaker and the microphone. I've got some more tests to run!
Old 4 weeks ago
  #5
Gear Head
 
EastNashRocker's Avatar
 

Quote:
Originally Posted by thethrillfactor View Post
Just a couple things to add when looking at the ETC:

1) Use % over dbfs for studying and looking at ETC.
It makes it easier to see and compare peaks.

2) Select zero phase filtering if you use filtered ETC.
When you filter the ETC there can be a phase shift based on the EQ filter, if you select zero phase shift it gives you a response of zero phase shift.

REW says;
"This may be useful when comparing the locations of reflection peaks in filtered impulse responses or the ETCs derived from them as the peaks will not be shifted in time by the effect of the filter's phase response".

3) Select show data panel.

When reading the data the ideal is an R of -1. The closer the better.
Anything outside -.98 of the range will show as orange and may not be reliable when comparing decay figures.

When looking at the curvature 0-5% is what you normally want. Higher than 10% can be feedback or a 2 stage decay curve which is what you don't want.
Negative figures have to be viewed with caution.
This is exciting because I've never used the ETC this way before and I can tell I'm in just a little over my head but it's starting to make sense. The curvature of the graphs you showed were 4.0%, -3.0%, and -9.9.. so one out of those three is good. How does a negative % happen and why should I be cautious of interpreting them?

Some people would think that I already have too much absorption but I'm happy with the frequency response of my room. I couldn't define a 2-stage decay curve but I'm wondering if this is part of improving the ETC reading for my room. Acoustics is addicting, thanks for your help!

Last edited by EastNashRocker; 4 weeks ago at 04:56 AM.. Reason: grammar
Old 4 weeks ago
  #6
Gear Head
 
EastNashRocker's Avatar
 

Quote:
PS Show us a link to your earlier thread with photos etc.

Help Needed With Room Null


Good idea, this link has pictures of my room. Keep in ming all the treatment has been rearranged but it gives you an idea of what I'm working with. Lol, I'm nervous to look at the .mdat in that old thread because I'm sure it looked rough
Old 4 weeks ago
  #7
Lives for gear
 
akebrake's Avatar
 

Quote:
Originally Posted by EastNashRocker View Post
..Some people would think that I already have too much absorption but I'm happy with the frequency response of my room.
Yes much better than most small DIY spare bedrooms.

There are still some L/R differences between 100 Hz - 1kHz.
How did you deal with the fireplace protrusion to the right?

BTW Did you try any other LP than the "38%" one?
E.g. What happens when you move L/R, up/down/front & Back?

Best
Old 3 weeks ago
  #8
Gear Head
 
EastNashRocker's Avatar
 

Quote:
Originally Posted by akebrake View Post
Yes much better than most small DIY spare bedrooms.

There are still some L/R differences between 100 Hz - 1kHz.
How did you deal with the fireplace protrusion to the right?

BTW Did you try any other LP than the "38%" one?
E.g. What happens when you move L/R, up/down/front & Back?

Best
I tried moving the speakers every which way and I've tried moving the listening position closer and further away.. but I haven't tried moving the listening position left or right, which would be off the center axis of the room. I'm game to try it though

The fireplace has much less absorption up that before. Before I tried to "eliminate" as much of the fireplaces as possible with maximum absorption but when I took everything out of the room and moved things back in one at a time I ended up with two 2x4ft panels of 2 inch thick safe n sound.

I did an experiment the other day and got the strangest response! I tried putting a sheet of 703 everywhere in the room to reduce some of the impulse response peaks and had no luck for a loooong time. I have tons of extra treatment now so it's easy to just throw insulation at a problem to test and see if it helps. I'm testing with one speaker at a time. With just my left speaker, if I hold a piece of 703 flat (parallel to the ground) about 12inches above the speaker I get a dramatically improved impulse response? I then tried moving it three feet above (like a low hanging cloud) and it didn't have any of the improvements as 12inches above.

Any thoughts on that? I can retest and post an .mdat here but I don't even know how I'd permanently mount something like that.. if I could first get over how weird it looks
Old 3 weeks ago
  #9
Quote:
Originally Posted by EastNashRocker View Post
if I hold a piece of 703 flat (parallel to the ground) about 12inches above the speaker I get a dramatically improved impulse response? I then tried moving it three feet above (like a low hanging cloud) and it didn't have any of the improvements as 12inches above.

Any thoughts on that? I can retest and post an .mdat here but I don't even know how I'd permanently mount something like that.. if I could first get over how weird it looks
Probably diffraction from the front baffle design of your speaker. You see this in the ETC when the peaks are close to or near zero.

Another reason why soffit mounting is suggested for speaker as you extend the baffle.
Old 3 weeks ago
  #10
Gear Head
 
EastNashRocker's Avatar
 

Quote:
Originally Posted by thethrillfactor View Post
Probably diffraction from the front baffle design of your speaker. You see this in the ETC when the peaks are close to or near zero.

Another reason why soffit mounting is suggested for speaker as you extend the baffle.
I'm not very familiar with diffraction. I have my speakers close to the front wall, do you think the diffraction could be bouncing off the front wall?

My speakers are Genelecs with rear firing ports, so no soffit mounting for me unfortunately. Any other thoughts on how to deal with diffraction?
Old 3 weeks ago
  #11
Lives for gear
 
akebrake's Avatar
 

Quote:
Originally Posted by EastNashRocker View Post
I tried moving the speakers every which way and I've tried moving the listening position closer and further away.. but I haven't tried moving the listening position left or right, which would be off the center axis of the room. I'm game to try it though
It was not ment to say "you should sit off center axis" but rather a check if small movement makes a huge difference. Which in turn means you probably need to treat a side wall better.

Moving the microphone (or speaker) while simultaneously watching REW RTA is an alternative way to observe what happens to the FR in different positions. No sweep (Very educating IMHO) See this set up example (LINK)

Of course you've got no time information (because the measuring signal is "steady state") but LF response is quite close to the long (REW default 0.5s) IR window.

One speaker at a time.

Best
Old 3 weeks ago
  #12
Quote:
Originally Posted by EastNashRocker View Post

My speakers are Genelecs with rear firing ports, so no soffit mounting for me unfortunately. Any other thoughts on how to deal with diffraction?
Check the tweeter dispersion for your particular monitor.

Also the shape of the baffle can sometimes diffract the hi's.
📝 Reply
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
🖨️ Show Printable Version
✉️ Email this Page
🔍 Search thread
♾️ Similar Threads
🎙️ View mentioned gear