Help with 35 year old listening room

Discussion in 'Room Acoustics and Treatments' started by Jeff Weissert, Oct 14, 2018.

  1. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.
    I need help on my listening room that I made thirty five years ago…

    I made it but then due to the WAF it sat mostly as a museum. Now that I am retired I want to tune the acoustics of this space.

    With the advent of REW now it has showed modes and other ugly things that would have been theories when I made the room! Life was simpler before I saw the REW output!

    I have a Helmholtz resonator back wall that I can bet is not tuned to the right frequency and not using the correct absorbent material… I used 7” of pink insulation back then.

    I am using 4 Altec Voice of the Theater cabinets and two Altec 511B horns. There is also a Velodyne ULD-18 subwoofer. The Voice of the Theater cabinets were just updated to use GPA 416-16b woofers.

    The Room is 19’x 13’x 7.5’. Two walls are double wall frame construction covered with 5/8” particle board and ½” drywall over the particle board.

    The walls behind the Helmholtz and the bookshelves are block.

    There is a small room behind the racks and one behind the speakers. The room space behind the speakers is 4’ deep.

    The ceiling has Auralex Acoustics 2 inch Wedge Metro Studio Foam Sound Absorption Panels


    I saw a web reference indicating that the woofers are “too big” for the room… well if that’s true I sure knocked it out of the park.

    I was considering taking some of the Helmholtz wall down and installing a large bass trap… After reading posts from Matthew Poes it sounds like it would be hard to have too much bass absorption. I could take down 8-10’ of the Helmholtz wall and use the space for a trap.

    The current wall face is 9” from Styrofoam insulation glued to the outside basement block wall.

    My listening position is such that the back of my head is 14” from that Helmholtz wall… if I lean forward it does sound better.

    I am uploading the .mdat

    Measurement #1, L+R and Sub

    Measurement #2, Left and Sub

    Measurement #3 Right and Sub

    Measurement #4 L+R with Sub turned down all the way, Note though the woofers/amp were driven from the High pass in the Velodyne amp so there is that roll off.


    Any input would be appreciated….



    Jeff
     

    Attached Files:

    Todd Anderson and tesseract like this.
  2. Sonnie

    Sonnie Senior Admin
    Staff Member

    Joined:
    Apr 2, 2017
    Messages:
    2,046
    Likes Received:
    433
    Location:
    Alabama
    Looking at your L+R w/ Sub is the one that I would work with fixing if possible. I am not sure you will be able to do much with room treatments to fix that dip at 100Hz. Have you tried moving the sub around? Adjusting phase on the sub? Adjusted crossover settings?

    I think it will be a lot of trial and error, which is about the only way to figure it out short of some sort of correction system like Dirac Live, which would quickly solve your issues.
     
  3. tesseract

    tesseract Senior Admin
    Staff Member

    Joined:
    Nov 12, 2016
    Messages:
    976
    Likes Received:
    209
    Location:
    Lincoln, NE
    Wow, such a beautiful space! I love the cloud treatment, too.

    Sonnie is right, Dirac Live helps so much with room anomalies. I'd block off the Helmholtz resonators to mitigate their contribution, then measure again with REW. Good on you for finding us and REW, we are here to help!
     
  4. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.
    So I have tried moving the sub... multiple places not much better... I even put masking tape over the slots on the Helmholtz wall... didn't see much difference on the REW plot... maybe masking tape isn't enough...
    I looked at the Dirac Live web page. Is it a software driver, in the end, that would be essentially in series with my JRiver Media center... or anything else playing out my DAC?
    I fount the MINI DSP page and saw that I can buy hardware... the DDRC-24 or the DDRC-88A. It would be nice to run the DDRC-88A as and active crossover also but it seems that it is only for 7.1 use.
    I am thinking that the room is just too "tight" and the standing waves are killing me... Short of winning a lottery and buying a closed church I don't have too many options...
     
    #4 Jeff Weissert, Oct 17, 2018
    Last edited: Oct 17, 2018
  5. Matthew J Poes

    Matthew J Poes Staff Writer
    Staff Member

    Joined:
    Oct 18, 2017
    Messages:
    1,729
    Likes Received:
    349
    In some ways I think you are right. Particle board, chip board, and OSB are all much more rigid than drywall. Block is even more rigid and fairly dense/heavy. When these exist in the walls of a sound room, it tends to trap more bass in the room. It does sound like your particular construction led to a room with pretty rigid/solid walls. Not good for bass. That doesn’t mean it can’t be fixed (but as others have mentioned, DIRAC will do a lot with a lot less work).

    I’m not at my computer to look at your file, I’ll try to look later, but before going too nuts I would try to track down the sources of the various problems. If you measure your rooms dimensions and look at modes, which of the peaks and dips do they align with?

    What about SBIR (Speaker boundary interference effects)? Have you tried calculating those?

    Beyond that, I would also try turning various compliments on and off (maybe your measurements showed this already?) and see which ones the peaks and dips are most prominent in.

    If the prominent 100hz dip shows up in all speakers, it may be a ceiling effect or a length SBIR effect.

    Masking tape will have no effect on a Helmholtz resonator trap of the style you seem to have. At LF’s it would not stop it from working, it would actually slightly lower the tune and damp it’s effect, but not stop it.

    The insulation you used is likely the perfect insulation. How deep is the panel? Anything over 6” and that is really ideal. My max for the higher density stuff is 8”-10” and only when used as a stand-alone velocity absorber. Otherwise it gets too dense and is less effective than the pink stuff.

    If it is thinner, 4”-6”, then I would change it, but it won’t be a night and day difference. A tiny difference at best.
     
  6. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.
    The Helmholtz wall is 9.5" from the face of the board to the back wall. As best as I can feel through one hole is around 6" of insulation. The boards are 7" wide and 3/4" thick.

    I started some new baseline measurements before I started moving stuff and measuring again but ran into one of the issues that I had before. If I measure the room say 6 times in a row while sitting in the same place and only clicking on the mouse to initiate another run the displayed response varies from run to run below 1K... is that normal? I attached a demo file of 6 runs with differences.
    I also ran with the sub and sub-amp high pass bypassed... same 100 hz dip. Not sure if more sub moves are in order. Next idea of a possible point of sub move based on Matthews article on SBIR will be painful on room layout.
     

    Attached Files:

    #6 Jeff Weissert, Oct 18, 2018
    Last edited: Oct 18, 2018
  7. jtalden

    jtalden Member

    Joined:
    May 22, 2017
    Messages:
    89
    Likes Received:
    20
    Location:
    Arizona, USA
    The impulse response looks abnormal for all measurements. This also impacts waterfall, spectrogram and others. I did not see anything in the file that pointed to a reason for the issue. I suggest you investigate to correct this situation before making any decisions on other changes.
    jaCapture.JPG
     
  8. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.
    Well... the issue was REW was showing corrupt data! It turns out that the Focusrite Scarlett 2I2 driver was not playing nice in the sandbox. It also was causing some issues in JRiver's Media Center...
    Focusrite tech support suggested to unload their driver and let Windows use its WASAPI driver... Bingo!
    I wonder how Dirac would work with a non AISO driver.
     
  9. DanDan

    DanDan Member

    Joined:
    Mar 10, 2018
    Messages:
    39
    Likes Received:
    6
    Cool looking room. Dirac is brilliant. I am on Macs and I have never seen a bug or crash. Other systems have options of using Linear or Minimum Phase Filters etc. etc. so I don't quite know why, but Dirac seems to take control of the low end in a very commanding way. But let's park it there, I recommend you leave that till last. I think the Demo period is quite generous.
    With rare exception, horn loaded speakers don't do it for me. There are quite a few of the best speakers all looking really similar these days. Soft Dome HF AND MF. I like sealed subs but anything engineered well....
    Looking forward to valid .mdats.
    Let's consider the ceiling. That foam must be looking a tad tired. But in any case is is way too thin. Generally we would like to see a minimum of 4" fibre, plus 4" air gap. Perhaps not on the whole ceiling, but certainly at least in the usual Cloud location.
    DD
     
  10. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.
    I love horn speakers... super efficient... and therefore less "Doppler". At extreme levels the cones hardly move let alone at regular listening levels.

    I though this thread died like a rock in the ocean! Deep and fast!

    I uploaded the latest mdat... the foam in the ceiling was replaced a few years ago... the original Sonex was turning into mush!
    I tried the Dirac and there was only some HF added... I like the horn curve and find it easy to listen too... when I flatten it out I don't like the sound...
    I haven't messed with this for a while... working on rebuilding a HK Citation II for a friend...
     

    Attached Files:

  11. Matthew J Poes

    Matthew J Poes Staff Writer
    Staff Member

    Joined:
    Oct 18, 2017
    Messages:
    1,729
    Likes Received:
    349
    I'm sorry if we killed your thread. I travel for work and have been on a bit of a speaking tour if you will, so I was out for a bunch of talks and not able to be as active. I'll be disappearing again at the end of the month for about a week or so.

    In any case, I just looked at your new file. The impulse looks generally ok to me. Ok I just edited this to remove the comment about distortion. That was caused by the cancelation dips in the response. Sorry about that, I wasn't looking carefully and had it set to normalised. Woops!

    There are some acoustic issues in the room too. Decay times are all fine, but there appear to be some strong reflections showing up at 1-2khz. They are strong enough to shift the peak energy curve and look like an acoustic mirror of some kind. Like what happens when you place a speaker against a wall or have some reflecting surface right next to the speaker. Across all frequencies I see this showing up between 3 and 6ms. That reflects a path length difference in the reflections of around 3-6 feet, roughly. If you create a triangle between your speaker, you, and a reflecting surface, what location near your speaker can account for that 3-6ms delay? It may be your offending surface.

    As for response curve. I've been reviewing all of the literature on in-room measurements and the relationship between anechoic and in-room. Going beyond what Toole says to get at where these ideas originated and what is known. I realized that some generalizations that Toole made created confusion in the industry. A proper room curve is not only not flat, it is not 1dB per octave rise from 20khz to 20hz. That is the natural rise in a speaker whose free-space measurements are flat and whose DI is flat at 3dB or so. If the DI is elevated from there, the response curve is more tilted. The narrower the dispersion at high frequencies, the more the tilt in-room will be naturally. The room curve that Dirac applies depends on the speaker's directivity. It tries to guess at this, but you can also use your own knowledge of the speaker's directivity to help ensure. What you should never do is try to fight the natural response shape with an artificial curve. A flat response, especially with horns (which are narrow dispersion high DI speakers), because it will actually make a very bright speaker.

    I have people argue this point with me a lot and it stems from what has become a huge misunderstanding in the industry. Our ears can filter out early reflections. A microphone cannot. Our ears are directional, they don't hear reflections from certain directions as strongly and they can tell which direction the reflections are coming from. An omnidirectional measurement microphone cannot. When you listen to a speaker, what you hear is mostly the direct sound where the period of the tone is substantially longer than the time delay in the reflections. As the period of the tone and the reflection time become similar, you can't tell the difference so much. A microphone can never tell the difference. A microphone measurement in a room combines all of the sound energy equally from all directions. So a wide dispersion speaker with a flat DI at 0 will actually measure flat in a room. It will also sound flat and neutral. A narrow dispersion speaker whose response is very directional will have a tilted response because at mid and high frequencies the reflections are very down in level, they aren't adding much to the energy in the steady-state response. The LF's still have a lot of room contribution causing a rise in the LF's. Add to that the fact that many directional speakers are also more directional at high frequencies and what you find is that the in-room response of such a speaker is very tilted at high frequencies. However it isn't really that tilted, that is just a function of the direct to reflected energy being captured and combined to create the steady-state response. It is critical that you don't try to compensate for that since that isn't what you hear.
     
  12. DanDan

    DanDan Member

    Joined:
    Mar 10, 2018
    Messages:
    39
    Likes Received:
    6
    Nice to see action down to and even below 30Hz. My download shows only one Measurement in the .mdat. The mic is very rarely precisely between the two tweeters, so for HF testing it is best to do L only, R only, L+R.
    DD
     
  13. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.
    I am guessing that the hard polyurethane coated boards of the Helmholtz resonator behind my listening position is the culprit… It is 24" behind the measuring mic and the wall is 10-12" deep behind that surface... I re-did the measurements and propped up a 2'x4' piece of 2" thick auralex (which has a piece of paneling glued to it to keep it straight) behind the measuring mic and against the resonator to compare. If I look or measure from the listening position mic there isn't anything within the 3-6' range except that back wall and the ceiling which is 4' above the mic. The ceiling should absorb mid to highs... as it is squares of the 2" auralex panels. Maybe some absorptive panels should be hung on the back wall... my head is way too close to the wall but unless I move to a different house I am stuck there! Wish that I was younger and could buy a used rural church... a sanctuary would be a great space!
     

    Attached Files:

  14. DanDan

    DanDan Member

    Joined:
    Mar 10, 2018
    Messages:
    39
    Likes Received:
    6
    Do you have two speakers running during your sweeps?
    DD
     
  15. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.
    Yep... after earlier work to check each side was done I have done both together... Other posts told me that it averages some of the acoustical anomalies.
     
  16. Allen Rumbaugh

    Allen Rumbaugh Moderator
    Staff Member

    Joined:
    Oct 28, 2017
    Messages:
    30
    Likes Received:
    4
    Do you have a picture of your rear wall? I would really like to see that.
     
  17. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.
    Here is the back wall.... all the surfaces are in separate pictures attached to the original post...
    Thanks for looking...
     
  18. DanDan

    DanDan Member

    Joined:
    Mar 10, 2018
    Messages:
    39
    Likes Received:
    6
  19. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.
    Whoops.... I came back to my PC and found that there was an error uploading the photo of the back wall... its an older picture. I have since removed the pictures hanging on the wall.
     

    Attached Files:

  20. Jeff Weissert

    Thread Starter

    Joined:
    Sep 2, 2017
    Messages:
    12
    Likes Received:
    2
    Location:
    White Oak, PA.

    Attached Files:

Share This Page