Results 1 to 14 of 14

Thread: If Not an RTA, Than What?

  1. #1
    Senior Member
    Join Date
    Jul 2003
    Location
    Orange County, CA
    Posts
    3,597

    If Not an RTA, Than What?

    Certainly we've come a long way since the Behringer DEQ 2496, 1/6th octave RTA.

    But which of these newer programs will allow for REAL TIME measurement using a laptop computer running Windows 7?

    If it's not in real time, I'm not interested. I want to be able to move the Heil even a fraction of an inch and instantly know whether this is helping or hurting a #dB dip/peak at #kHz. The Behringer allows for this in real time as I move something and watch the display without having to set up and run a sweep frequency each time. So what else runs in real time with a higher resolution?

  2. #2
    Administrator Robh3606's Avatar
    Join Date
    Apr 2003
    Location
    Rocinante
    Posts
    8,170
    Why don't you look into REW?? I think that would do it for you not sure about real time. It's free and well supported. Ask on AK about it have to be some users there or go to;

    https://www.roomeqwizard.com/

    Rob
    "I could be arguing in my spare time"

  3. #3
    Senior Member
    Join Date
    Jul 2003
    Location
    Orange County, CA
    Posts
    3,597
    Thanks.

  4. #4
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,719
    Every measurement system I have used will give you some form of realtime capability along with MLS or other tone bursts that are typically more useful. You would be able to use the RTA or other "realtime" measurement for the experiment you described and then you can follow up with a more detailed analysis once you think you are close.

    I have used REW for room analysis and for free software it is outstanding. I have not used any of their measurement software.

    I own and use CLIO which is quite good, but mine is an older version and doesn't offer all the features I would like. My favorite measurement software these days is by Studio Six Digital and is iOS based. I use it on an iPad with a calibrated laboratory grade mic with an ACO Pacific capsule.

    Below are a couple of quick and dirty plots I just did on my phone using its built in mic. For this example I used FFT measurements with 1/24 octave smoothing. I used a downloaded pink noise signal as the source. I used both "realtime" and averaged measurements to show how using pink noise in realtime will give you a variety of measurements, where averaging the same signal over a minute gives a more accurate measurement. See below, two snapshots off my phone in "Realtime" aka FAST, and one averaged measurement.


    Widget
    Attached Images Attached Images    

  5. #5
    Senior Member
    Join Date
    Jul 2003
    Location
    Orange County, CA
    Posts
    3,597
    Thanks,

    A nice feature of the Behringer is the ability to control the "averaging time" to basically do what you did over the 1 minute period. There is no need to follow the "bouncing ball" and try to make a determination from that.

  6. #6
    Senior Member
    Join Date
    Aug 2017
    Location
    WDC USA
    Posts
    310
    Smaart is the pro audio solution and is priced accordingly. https://www.rationalacoustics.com

    OSM has the core FFT function of Smaart but few bells and whistles https://opensoundmeter.com

    REW has similar tools but is a pretty busy interface so a bit more learning curve.

    The RTA is a pretty limited display of data. It can tell that level changed in a fractional octave band but nothing else. Just as important when measuring sound is seeing what changed in the time domain and for that you need a display of phase as well as level.

  7. #7
    Administrator Robh3606's Avatar
    Join Date
    Apr 2003
    Location
    Rocinante
    Posts
    8,170
    Quote Originally Posted by toddalin View Post
    Thanks,

    A nice feature of the Behringer is the ability to control the "averaging time" to basically do what you did over the 1 minute period. There is no need to follow the "bouncing ball" and try to make a determination from that.

    In CLIO as an example you can select the number of samples that are averaged so essentially the same thing. Just up your sample size using a software measurement system.

    Rob
    "I could be arguing in my spare time"

  8. #8
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,719
    Quote Originally Posted by Riley Casey View Post
    Smaart is the pro audio solution and is priced accordingly. https://www.rationalacoustics.com

    OSM has the core FFT function of Smaart but few bells and whistles https://opensoundmeter.com

    REW has similar tools but is a pretty busy interface so a bit more learning curve.

    The RTA is a pretty limited display of data. It can tell that level changed in a fractional octave band but nothing else. Just as important when measuring sound is seeing what changed in the time domain and for that you need a display of phase as well as level.
    You can get a single channel version of SMAART from Studio Six Digital for about $70. I haven’t tried it, but based on the modules I have tried I bet it’s pretty good.

    https://m.youtube.com/watch?v=ZjmCyGo-KR4


    Widget

  9. #9
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    Hi Todd,

    I understand what you are looking for.

    However real time RTA measurements use pink noise as a source which in it self is a constantly varying signal as it’s noise. Using smoothed pink noise such as 1/3 octave is what JBL used back in the 70’s. Those measurements were averaged. With averaging you will see a more precise change but not an instantaneous change because average takes a number of samples. The resolution is limited due to the nature of pink noise. So this kind of measurement is best suited to overall representation of the full frequency range for a front of house PA sound engineer.

    If you want to examine the effect of moving the Heil driver are you talking about a back and forth movement?

    That requires a move detailed measurement analysis to define X, Y, Z changes in the location of the driver.

    I defer to Rob’s reference of REW software which has both RTA and IMP type measurements.

    That kind of movement is most notably going to effect what happens at the crossover point where phase becomes sensitive to movement in the Y and Z location of the driver. The higher the crossover point the sensitive it will be to movement.

    I offer a boarder and more in-depth perspective on the impact of movement of drivers below.

    The crossover point is defined by the attenuation at which point two drivers frequency response intersect. The attenuation is defined by the crossover filter type and is typically -3dB or -6 dB.

    It should be pointed out that only certain filter types will in fact sum flat without manipulation of the electrical phase or the crossover points. This is the acoustical response not the electrical response of the filters.

    A point l will also make is that you can have a flat summed response on axis between the two drivers at a specific x y z listener location. But if you stand up the summed response may not be flat. It might be a peak or a notch. This depends on the order of the crossover filters, the type of filter and the x y z position of the drivers.

    Typically in crossover design the Linkwitz Riley 24 dB crossover is used because it sums flat with a narrow overlap of the drivers. Where the drivers are not aligned in the Z axis we can introduce a delay in the tweeter signal to align with the woofer. At high frequencies this is often limited to an inch less at 10,000 hertz.

    From a solution perspective it’s best to determine your reference point for a summed flat response. This is your x y z listening point. Then set up the HEIL to be on axis with your listening point if possible or angled towards your listening point. The closer the HEIL driver is located in the Y plane to the woofer or other driver the more likely you will succeed in a flat summed response. A simple electronic crossover network maybe of assistance where you can manipulate the crossover points of each driver independently to reach a summed flat response (ref Douglas Self - The Design of Active Crossovers) where the driver z y z location cannot to modified for a flat summed response.

    In summary l would give REW a go. All you need is the Minidsp mic in the case of a Mac laptop which can be ordered online. It’s very well supported.

    I hope this is of some assistance.
    Last edited by Ian Mackenzie; 01-21-2022 at 08:21 PM. Reason: Typo

  10. #10
    Senior Member
    Join Date
    Jul 2003
    Location
    Orange County, CA
    Posts
    3,597
    Thanks Ian.

    I simply set the mic in the listening position because that is the only position that really matters to me. Yes, the 10" and Heil sum at the crossover point and proximate frequencies and it is very easy, even only using 61 bands, to tell the preferred "Z" for the Heil simply by moving it back and forth and tilting it up/down for a slightly off-axis response. Sometimes, the nexus lies in adjusting the slopes/overlap, and the textbook formulas don't always yield the most desireable balance because drivers don't follow textbook impedience/phase/volume curves across their band. This is not only influenced by the components in use, but also the acoustics of the room.

    It is also very easy to look at these bands and adjust the crossover frequency/slope to see what provides the smoothest response..., based on its limitations. That's why I like RTA.

  11. #11
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,719
    Quote Originally Posted by toddalin View Post
    I simply set the mic in the listening position because that is the only position that really matters to me. Yes, the 10" and Heil sum at the crossover point and proximate frequencies and it is very easy, even only using 61 bands, to tell the preferred "Z" for the Heil simply by moving it back and forth and tilting it up/down for a slightly off-axis response. Sometimes, the nexus lies in adjusting the slopes/overlap, and the textbook formulas don't always yield the most desireable balance because drivers don't follow textbook impedience/phase/volume curves across their band. This is not only influenced by the components in use, but also the acoustics of the room.

    It is also very easy to look at these bands and adjust the crossover frequency/slope to see what provides the smoothest response..., based on its limitations. That's why I like RTA.
    Do you keep the mic in the same location or average a number of locations? I find it most useful for "in-room" measurements to use an average of several mic locations. For measurements focused in the midband there will probably be less variation, but it is still a useful practice with varying results depending on the room and listening position. I usually move the mic randomly in 2' by 2' by 2' cloud at the primary listening position.

    Of course even after going through all that, the ultimate judgement is made by listening.


    Widget

  12. #12
    Senior Member
    Join Date
    Jul 2003
    Location
    Orange County, CA
    Posts
    3,597
    Quote Originally Posted by Mr. Widget View Post
    Do you keep the mic in the same location or average a number of locations? I find it most useful for "in-room" measurements to use an average of several mic locations. For measurements focused in the midband there will probably be less variation, but it is still a useful practice with varying results depending on the room and listening position. I usually move the mic randomly in 2' by 2' by 2' cloud at the primary listening position.

    Of course even after going through all that, the ultimate judgement is made by listening.


    Widget
    Location can vary a bit, but unless doing nearfield testing, is proximate to where my head would be located at the couch. I know that minor variation in the crossover region can manifest itself in different ways depending on proximity to furniture as well as verticle off-axis mic/ear placement. As expected, there is a lot of HF roll off and variation between what comes out of the Heil and the receptor location at the couch cushions.

  13. #13
    Administrator Mr. Widget's Avatar
    Join Date
    Apr 2003
    Location
    San Francisco
    Posts
    9,719
    From our earlier conversations, I assume the goal is not to make the "perfect" speaker, but rather the best possible one with the available components for your listening room as heard at the listening position.

    This is why I suggested taking several measurements in a 24" cube (or sphere) at your listening position. I think this will give you the closest approximation to what you actually hear at that location.

    A single mic located near the center of where your head will be will not accurately account for the reflected and out of phase sound that we hear with two ears and mentally process it to become a single sound. I have come up with averaging a "cloud" of measurements as the most representative measurement of my subjective experience. I haven't experimented with binaural measurements, but I doubt it will give us a representative measurement either... though it would be interesting to give it a go.

    Now, if the goal was to make a more generally representative measurement of the speaker, it would be great to have access to a large anechoic chamber or quiet wind free outdoor space. I have typically used gated measurements at 2m in a large room away from early reflections. Depending on how high your ceiling is and how far off the floor you can get, this will get you a usable measurement down somewhere between 200Hz and 1000Hz. Then I use close mic techniques below that.


    Widget

  14. #14
    Senior Member
    Join Date
    Jul 2003
    Location
    Orange County, CA
    Posts
    3,597
    Thanks.

    Yes, the goal is the best possible sound, to taste, using what's available.

    I agree that some measurements would be better obtained over a slightly larger area. But some measurements can reveal room and speaker characteristics so obvious that even a "pert near" measurement reveals their presence and cleaning these up can do a lot of good.

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •