The No.1 Website for Pro Audio
How good is 'good' and how 'good' can it be?
Old 2 weeks ago
  #1
How good is 'good' and how 'good' can it be?

I feel like the title of this thread seems cryptic, but I really don't mean it that way. I've got a room 10'x12' that's a now a primary mix space for me. I did a lot of measurements and built and applied treatment as what seemed appropriate to address the issues, covering all first reflection points, and the 'easy' corners (there are a door and window that come right up to the two other vertical room corners), as well as a decent cloud. I've done a few mixes with the space like this, and the clients are super happy and they seem to translate quite well.

So this is where the questions come in. How 'good' is a 'good room' in terms of its Frequency Plot, Waterfall, ETC, RT60 and IR?

Next question, at what point is it no longer worth pursuing getting it 'better'? I realize that's an opinion based on a lot of factors, mostly driven by budget I'm sure, but how far would YOU go if this were a dedicated mix space that puts out commercial records?


I still have a bunch of extra wood, Guilford of Maine, and R80, so I CAN keep going, but I'm not sure it makes sense, and at this point I think I need tuned traps at particular frequencies to clean up some of the low end nulls in the room.



Here are the files that go with my room. The overlays you see are two different sets of monitors, I did take traces of left and right, but they're pretty much identical so I didn't see much value in having those included in this measurement set rather than having both monitors in there.

Average SPL (Red Monitor Set 1, Green Monitor Set 2, Blue Average)

http://ryanojohn.com/RoomTuning/2020...AverageSPL.png

Waterfall (400ms window)

http://ryanojohn.com/RoomTuning/2020...0Waterfall.png

IR ETC (I don't honestly know how to properly read this)

http://ryanojohn.com/RoomTuning/20200619%20IR%20ETC.png

RT60 (or this)

http://ryanojohn.com/RoomTuning/20200619%20RT60.png

SMAART Transfer Function of Monitor set 1

http://ryanojohn.com/RoomTuning/20200619%20BFMM35.png

SMAART Transfer Function of Monitor set 2

http://ryanojohn.com/RoomTuning/20200619%20CP42.png

Last edited by ryanojohn; 2 weeks ago at 12:28 AM.. Reason: Added Images
Old 2 weeks ago
  #2
Quote:
Originally Posted by ryanojohn View Post
I

So this is where the questions come in. How 'good' is a 'good room' in terms of its Frequency Plot, Waterfall, ETC, RT60 and IR?
The most basic criteria is to get your room to meet the EBU Tech 3276 standard for a stereo listening room:
https://tech.ebu.ch/docs/tech/tech3276.pdf
Old 2 weeks ago
  #3
That seems like a pretty loose standard with everything is measured in 1/3rd octave... the time domain part all seems reasonably detailed, but it seems my decay numbers are already all below that. Although this does also state that the minimum size for the room is about twice that of my room.
Old 2 weeks ago
  #4
Lives for gear
 
avare's Avatar
 

Quote:
Originally Posted by ryanojohn View Post
That seems like a pretty loose standard with everything is measured in 1/3rd octave...
To use your terminology, your room does not meet that loose standard.

There are early reflections within 15 ms.

The nominal RT60 is greater than 0.16 s.

The low end rise exceeds the mask

There is a significant dip ~70 Hz.
Old 2 weeks ago
  #5
No need to be snarky... As I said, it SEEMS like a loose standard only because 1/3rd octave smoothing can’t even see that 70Hz dip at all. It just seems surprising that it would be such wide smoothing for F.

I’ve also mentioned that my room is HALF the square footage as deemed “minimum” by their requirement... so it will never adhere to that requirement set.

How do you determine “Nominal” RT60? An average across the F? Is there any weighting for particular octaves? As I said the time domain stuff seems quite particular so I’m sure that’s specified clearer, I just need to re read it.

Last edited by ryanojohn; 2 weeks ago at 04:32 PM..
Old 2 weeks ago
  #6
Lives for gear
 
avare's Avatar
 

Quote:
Originally Posted by ryanojohn View Post
No need to be snarky... As I said, it SEEMS like a loose standard only because 1/3rd octave smoothing can’t even see that 70Hz dip at all. It just seems surprising that it would be such wide smoothing for F
Huh? The bandwidth of a theoretical one third octave is f±2^(1/6) which is just about 70 Hz for the 63 and 80 Hz bands. Monitor 2 shows a ~14 dB dip at ~70 Hz. So with one third octve band smoothing that will become only an 11 dB dip in the 63 and 80 Hz band. Hardly disappearing.
Old 2 weeks ago
  #7
Monitor 2 is also a set that high passes at 100Hz. (Pelonis Model 42), so I pretty much ignored everything below that.


Quote:
Originally Posted by avare View Post
Huh? The bandwidth of a theoretical one third octave is f±2^(1/6) which is just about 70 Hz for the 63 and 80 Hz bands. Monitor 2 shows a ~14 dB dip at ~70 Hz. So with one third octve band smoothing that will become only an 11 dB dip in the 63 and 80 Hz band. Hardly disappearing.
Old 2 weeks ago
  #8
Lives for gear
 
avare's Avatar
 

Quote:
Originally Posted by ryanojohn View Post
Monitor 2 is also a set that high passes at 100Hz. (Pelonis Model 42), so I pretty much ignored everything below that.
Your SPL graph shows response down 35 to Hz. Something does not make sense.
Old 2 weeks ago
  #9
Lives for gear
 

For nominal RT 60(T60) read page 6 in the EBU-link which thrillfactor posted; How good is 'good' and how 'good' can it be?

The reference room is 100m³, your rom is smaller so it should have a shorter decay than the reference room. Calculate the volume of your room and insert the value in the math formula.

As in my room, your decay will not be even over the frequency range. As it is small and modes will dominate, decay changes with position. In REW measurements, with the mic at a position with stronger mode influence the red Schroeder integral-curve tend to deviate away from the black straight regression line, but addition of sub(s) and their position(s) plays a role too as can be seen in the pictures. Unfilterded with mic at exact same position. With added subs the Schroeder integral-curve starts to deviate away around -43 dB and the unfiltered decay time also increases versus unfiltered without subs.
Attached Thumbnails
How good is 'good' and how 'good' can it be?-t60-without-subs.jpg   How good is 'good' and how 'good' can it be?-t60-subs-crossover-80-hz.jpg  
Old 2 weeks ago
  #10
Lives for gear
 
avare's Avatar
 

Quote:
Originally Posted by Adhoc View Post
The reference room is 100m³, your rom is smaller so it should have a shorter decay than the reference room. Calculate the volume of your room and insert the value in the math formula.
Or read post 4. That is how I calculated the 0,16s for RT60 that I wrote.
Old 2 weeks ago
  #11
Lives for gear
 

What I wrote was an answer to ryanojohns' question in post #5 : "How do you determine “Nominal” RT60?" Your post #4 , while it states results of your calcualtions, doesn't show how you came to those results. He hasn't written out his room height, so the room volume it's a bit "open" for guesses.
Old 2 weeks ago
  #12
Lives for gear
 
avare's Avatar
 

Quote:
Originally Posted by Adhoc View Post
while it states results of your calcualtions, doesn't show how you came to those results. He hasn't written out his room height, so the room volume it's a bit "open" for guesses.
Thank you for pointing out my assumption. I falsely assumed the room height was 8 feet. I did not detail how I got the number because the context of post was all dealing with EBU 3276.
Old 2 weeks ago
  #13
Quote:
Originally Posted by avare View Post
Your SPL graph shows response down 35 to Hz. Something does not make sense.
You're dead on! After a billion or so measurments (give or take), it looks like I still accidentally had the console routed to a sub that shouldn't have been on that set at all. I'll retake them...

But my question ultimately still stands. In a room that can never adhere to that guideline (room size limitations), what is a 'good' response for a room like this?



(and the ceiling height is 8' for HALF of the room, and sloped down starting halfway through the room to 7' at the opposite wall)
Old 2 weeks ago
  #14
Lives for gear
 
avare's Avatar
 

Quote:
Originally Posted by ryanojohn View Post
But my question ultimately still stands. In a room that can never adhere to that guideline (room size limitations), what is a 'good' response for a room like this?
You are begging the question. Boggy designed many rooms wi that size. You are also ignoring that the (frequency) response is just one of many criteria for the room acoustics.
Old 2 weeks ago
  #15
Using the formula there, it maths out to 0.16 as a Tm exactly as @ avare pointed out. The tolerance limits in the spec show it as a time over F, I don't fully understand the relationship of T20 vs T30 vs Topt, and which is ultimately relevant for what. Is an ideal to have all three measure about the same?

At what point is a filtered IR's level considered no longer relevant? The graphs you posted there measure to -70, is it considered generally acceptable to kind-of ignore the signal below that level?

In the case of my room where I've clearly got some specific low F's decaying much slower than others, and what I can only assume is nulls due to reflection, is the only way to address them (in a room I can't reshape) to put in some resonators tuned to those Fs? Or if it lined up with room modes add low f absorption to spots that line up with the math for the Axial or Tangential areas that are 'causing' this buildup?

And realistically without building a room from scratch, how much can you tame some of those resonances/long decays?

Quote:
Originally Posted by Adhoc View Post
For nominal RT 60(T60) read page 6 in the EBU-link which thrillfactor posted; How good is 'good' and how 'good' can it be?

The reference room is 100m³, your rom is smaller so it should have a shorter decay than the reference room. Calculate the volume of your room and insert the value in the math formula.

As in my room, your decay will not be even over the frequency range. As it is small and modes will dominate, decay changes with position. In REW measurements, with the mic at a position with stronger mode influence the red Schroeder integral-curve tend to deviate away from the black straight regression line, but addition of sub(s) and their position(s) plays a role too as can be seen in the pictures. Unfilterded with mic at exact same position. With added subs the Schroeder integral-curve starts to deviate away around -43 dB and the unfiltered decay time also increases versus unfiltered without subs.
Old 2 weeks ago
  #16
Well I've never mentioned only frequency response when asking things, if that's all that mattered, then something like Sonarworks would be the obvious and easy solution. But since that's only a piece of the pie...

I sadly can't rebuild this room from scratch, instead, I'm trying to understand what I CAN do with it as it is built, and what to REALISTICALLY aim for as targets for not just F, but everything time domain also.

I do not have an acoustics engineering degree, and some of the graphs that come out of the various software that exists for testing spaces aren't that clear to me... hence the questions.


Quote:
Originally Posted by avare View Post
You are begging the question. Boggy designed many rooms wi that size. You are also ignoring that the (frequency) response is just one of many criteria for the room acoustics.
Old 2 weeks ago
  #17
Lives for gear
 

The decay rate in a room varies with the frequency of interest. One is intersted in how long time it takes to decrease down 60 dB from original. RT60 = Reverberation time for 60 dB decrease. Reverberation includes that the room shall have a diffuse sound which is not possible in small rooms where room modes and early reflections is a fact. It is therefore more correct to talk about T60 (no real reverb).

A 60 dB decline = 1 million times lower than original. If the room environment is noisy, because of traffic, noisy machines around etc, it is not practical or healthy to try to measure a 60 dB decay if surrounding noise is say 45 dB. 105 dB is very loud … The decay rate inside the room is quite linear though, so one can calculate those 60 dBs by just measuring time for the first 20 dB and then multiply by 3. For a less noisy room T30 = the first 30 dB and multiply by 2. -So, ignore when you reach the room’s noise floor. You can increase the SPL of the sound sweep if you have a very noisy environment. The decay measurement in REW starts at -5 dB. For T20: -5 to -25 dB is measured, for T30: -5 to -35 dB is measured.

In the diagrams in post #7 , you can see that impulses decays down about -50 / 55 dB, then it evens out, as I reach my noise floor. The measumeremts were made at around 75 dB across the frequency 5-20 000 Hz, so I have a pretty well isolated room / quiet environment. Non filtered in the diagrams means it shows the ”average” over the frequency range, If I had filtered out say 50 and 2000 Hz at 1/3 octave the diagrams would show different slopes and decay times, longer for low frequencies and shorter for high freqencies. The goal is to have as even as possible over the frequency range and ”appropriate” versus the room volume and one’s personal taste. Personally I think the Spectrogram in REW is easier to evalutae than the Waterfall diagram. In enclosed picture one can see I have room modes at 40 Hz and 20 Hz, the decay is quite quick though. The ”streak” around 150 Hz is some vibrating / resonating stuff I have to hunt down. 174 ms is lowest recommended decay over the frequency in my room according to the EBU document.
Attached Thumbnails
How good is 'good' and how 'good' can it be?-filtered-50-hz-1-third-octave.jpg   How good is 'good' and how 'good' can it be?-filtered-2000-hz-1-third-octave.jpg   How good is 'good' and how 'good' can it be?-spectrogram-10-4000-hz.jpg  

Last edited by Adhoc; 2 weeks ago at 12:36 AM.. Reason: spelling
Old 1 week ago
  #18
Quote:
Originally Posted by Adhoc View Post
so one can calculate those 60 dBs by just measuring time for the first 20 dB and then multiply by 3. For a less noisy room T30 = the first 30 dB and multiply by 2.
That's a great explanation thank you. So ultimately the 30 would be more accurate if you haven't yet hit the 'noise floor' of a space at -30, which of course one would hope you haven't haha.


Quote:
Originally Posted by Adhoc View Post
In the diagrams in post #7 , you can see that impulses decays down about -50 / 55 dB, then it evens out, as I reach my noise floor. The measumeremts were made at around 75 dB across the frequency 5-20 000 Hz, so I have a pretty well isolated room / quiet environment. Non filtered in the diagrams means it shows the ”average” over the frequency range, If I had filtered out say 50 and 2000 Hz at 1/3 octave the diagrams would show different slopes and decay times, longer for low frequencies and shorter for high freqencies. The goal is to have as even as possible over the frequency range and ”appropriate” versus the room volume and one’s personal taste.
In those F specific graphs, is it safe then to assume that the "humps" and "nulls" are the comb filtering of standing waves/reflections into the room?

Quote:
Originally Posted by Adhoc View Post
Personally I think the Spectrogram in REW is easier to evaluate than the Waterfall diagram. In enclosed picture one can see I have room modes at 40 Hz and 20 Hz, the decay is quite quick though. The ”streak” around 150 Hz is some vibrating / resonating stuff I have to hunt down. 174 ms is lowest recommended decay over the frequency in my room according to the EBU document.
In the case of these ringing out frequencies, when do you determine that it's something that should be addressed with a F specific 'resonator' versus some broadband, or semi broadband absorption.
Old 1 week ago
  #19
Quote:
Originally Posted by ryanojohn View Post
That's a great explanation thank you. So ultimately the 30 would be more accurate if you haven't yet hit the 'noise floor' of a space at -30, which of course one would hope you haven't haha.

Topt is optimized for small room measurements. Especially when you are trying to surpress the early reflections below -30 db.

When the early reflections are below -30 it becomes basically a cross between the EDT and T30 measurements.

By the way EDT is an important measurement to study if you are looking to see how effective the early reflection treatment is.


Quote:
Originally Posted by ryanojohn View Post
In those F specific graphs, is it safe then to assume that the "humps" and "nulls" are the comb filtering of standing waves/reflections into the room?
Most of the time this is correct, unless its issues in the electronics/speaker components.

Then it becomes a challenge on tracking down are the phase issues in the measurements modal, SBIR or something else(if a tweeter is wired out of phase for example).

The Group delay/excess phase window in REW is helpful in narrowing things down.



Quote:
Originally Posted by ryanojohn View Post
In the case of these ringing out frequencies, when do you determine that it's something that should be addressed with a F specific 'resonator' versus some broadband, or semi broadband absorption.
When its modal( on a water fall it looks like a peak at specific frequency then descending decay that is steady going down), then treating the corresponding surface becomes the choice.

If you are building a membrane, building it as wide an effective as you can makes it easier without having to tune it in place.

If they are vibrations or noises by other things in the room( which can show up as branches peering off a decay , decay that dies down and gets strong again or a steady line in the Spectogram), a different approach is required.
📝 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