ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2  All   Go Down

Author Topic: Smaart Queries  (Read 9267 times)

Dave Gould

  • Marketplace
  • Newbie
  • *
  • Offline Offline
  • Posts: 16
Smaart Queries
« on: January 25, 2011, 01:52:34 pm »

Hey guys,

Over the past fews days I have been trying to get to grips with Smaart v6 by testing various aspects of sound systems and rooms. Although I do a lot of work around my local area I am yet to bump into any individual who is totally confident with the program to give me help.

Today I managed to do tests on a 300 cap. venue kitted out with a 4-way TurboSound Floodlight system. To test the system I used the following equipment:
Berhinger ECM8000 measurement microphone (all I could get my hands but I did manage to get my hands on a pro calibrator!)

Sony Vaio Laptop
Smaart v6
Lexicon Omega
Yamaha M7 (I know it would have been better to bypass the desk but using the M7 was only the way I could possibly patch everything correctly with the audio interface I was using)

For me the confusion begins when I then begin to analyse the results that I have captured. I have attached some screen shots and if there is any advice you guys can offer on how to best read the results it would be much appreciated.

Example 1
The first screen shot is of the sub and low mid crossovers and the phase between the two. Am I right in saying that the crossover happens at 190Hz and that they are about 120 degrees out of phase with one another? I know from reading about phase - which I am slowly getting my head around - that when two signals are 120 degrees out of phase either acoustically or electronically then this is part of the phase cycle known as cancelation.
The second jpg shows the magnitude results with all drivers on at once with a null in the response where the x-over point appears to be. Am I correct in saying this?

Example 2
Impulse responses are also another aspect of Smaart that is confusing. To do the IR test I am using the signal generator sweep within Smaart. I have great difficulty getting the IR to look anything like the ones that I have thus far in books, instead it looks like a massive block of noise and it is hard to see the RT60 of anything useful from the figure. I have attached a screen shot of the IR where you can see the setting I used.

I have read both the manual and several chapters of the brilliant Bob McCarthy book still I feel very confused.

Any help or advice would be much appreciated and I am sorry for the purely written post it has been a long day being exposed to pink noise and computer screens and it sure does make one tired!

Logged

Dave Gould

  • Marketplace
  • Newbie
  • *
  • Offline Offline
  • Posts: 16
Re: Smaart Queries
« Reply #1 on: January 25, 2011, 01:53:48 pm »

And here are the IR screen shots!!! 8)
Logged

Arthur Skudra

  • SR Forums
  • Full Member
  • *
  • Offline Offline
  • Posts: 160
Re: Smaart Queries
« Reply #2 on: January 25, 2011, 02:28:47 pm »

I'd remeasure the system with averaging turned on, set it to 32 averages for starters.  This will give us coherence data from which we can better evaluate what you have on the screen.  With the screenshots you captured an instantaneous moment in time, noise, reflections, anomalies included, so we really can't do much to evaluate things accurately based on what we see here.  Make sure your time delay is set correctly for each measurement.

Also, decrease the smoothing to 3 points or less, anything higher smooths out narrow band cancellations and/or peaks, giving you a false sense of how narrow or wide your problems are.  Increased smoothing is only good for spotting overall trends in response, and does little to help you evaluate smaller problems that will bite you.

The Impulse response you captured looks "normal", and in Smaart v6 you won't be able to easily figure out RT60 without using a ruler overlaid on top of your screen and a calculator. Maybe you could improve the signal to noise ratio of the impulse response measurement by measuring the system louder (or eliminating any background noise in the venue)?  Really you're best off saving the impulse response to a file and using another acoustical analysis program to figure out whatever acoustical measures you're looking for.  Without knowing more about the venue and how you conducted your tests, it's hard to make any conclusions that would be helpful!  Yes, you seem to have a clear direct field of the loudspeaker, and yes, you have reflections, and yes there is background noise, and apart from changing things in the physical realm, what can you do about the information you gather from the impulse response?  Move the mic?  Move the speaker?  Acoustically treat a wall?  Tell the custodians to turn off their vacuum cleaners?
« Last Edit: January 25, 2011, 02:31:36 pm by Arthur Skudra »
Logged

Dave Gould

  • Marketplace
  • Newbie
  • *
  • Offline Offline
  • Posts: 16
Re: Smaart Queries
« Reply #3 on: January 25, 2011, 03:20:57 pm »

The screen shots you are seeing are reloads of .ref files that were all done using averaging between 32 and 64 and even 128 just so I could make sense of the readings. The Cohearance of readings were always at 100% which I thought was very bizarre. Is there something I'm failing to do here? I enabled the coherence but still it never changed.

The smoothing i used was a little much and will keep it down next time.

The room is very lively, of quite large volume, with the ceiling almost being the same dimensions as the width.

Yes each of the screen shots are a moment in time, though I did take 5 readings of each measurement. Would you suggest combining these as one single average reading?

Cheers
Logged

Arthur Skudra

  • SR Forums
  • Full Member
  • *
  • Offline Offline
  • Posts: 160
Re: Smaart Queries
« Reply #4 on: January 25, 2011, 05:09:19 pm »

Ah ok, the screenshots you posted show "Inst" averaging, so that threw me off on an assumption on my part.  Make sure that you have coherence enabled in Options:Freq Resp.  Then make sure you click on the trace in the legend to show the coherence associated with the trace when it was captured and post a screenshot of each trace, with both magnitude a phase in a dual screen view.

So you're saying that coherence is 100% all the way across?  Very odd, sounds like something is awry!  Typically in very live rooms, the coherence trace should be all over the place depending on conditions (reverberance, noise, etc), but you should see very good coherence in the upper frequencies if your time delay is set correctly.  Check your setup, make sure the right inputs are selected.
Logged

Dave Gould

  • Marketplace
  • Newbie
  • *
  • Offline Offline
  • Posts: 16
Re: Smaart Queries
« Reply #5 on: January 25, 2011, 05:30:28 pm »

I think my lack understanding of Smaarts settings, mic posistioning and the inability to translate the readings are my main downfalls! And thinking about it they are things that you got to know to get the program running correctly!!!

Also I was wondering how the measurements are generally taken. Personally my
Goal today was analyse as much of the system as I possible could using Smaart. So I decided to test both speakers stacks together then separately L/R and then each individual element by muting the correct driver. The tests were done at various locations - mix position, centre, left, right and back of venue - with the microphone pointing at the stage at hear height and I took each reading 5 times. I thought thatthis would give me enough info on the system to determine phase issues as well as xover points and any peaks or nulls in the system.

Is this generally the approach that you would take? Or do you go about in a different way?

Logged

Arthur Skudra

  • SR Forums
  • Full Member
  • *
  • Offline Offline
  • Posts: 160
Re: Smaart Queries
« Reply #6 on: January 25, 2011, 06:38:22 pm »

Indeed, a lot of thought needs to go into *how* you make your measurements, then of course making sense of the squiggly lines, separating out truth from fiction.  A classic example of misapplication of a measurement is the floor bounce causing a big comb filter in your response curve due to the mic being on a stand several feet above the floor.  If you have opportunity to take a measurement class, by all means do so, it will be an excellent investment in time.

With regards to testing strategy, start with individual components, make sure they are operating correctly and all the correct polarity, level, and time delay, then eq.  Once you get one speaker right, then add the other components in the array, then measure the entire array, then combine with ancillary systems.  If your system/venue has symmetry, then leverage that to your advantage and save yourself some time, comparing left to right to make sure things are tracking correctly.

In addition to a wealth of posts here at PSW, there are lots of good materials, resources, and information on Smaart found here:
http://www.rationalacoustics.com/
http://www.rationalacoustics.com/forums/
Logged

Rasmus Rosenberg

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 55
Re: Smaart Queries
« Reply #7 on: January 26, 2011, 05:08:07 am »

Hey Dave,
+ 100xxx to what Arthur said, and take a class. The easy way to make sure you have the basic settings are to trash the config file. There is a nice description on the Rational forum about that. Also as said make sure you turn on the Coherence trace. The Floodlights are, as fare as I remember pr factory crossed at 125-180 depending on the subs. My experience with measuring them are, its hard because of the high cross over point of interest is "flooded" with reflections from walls and from ground bounce. A good visual way to see if you have a ground bounce reflection is to take an Impulse response and zoom in. The reflections will show as "spikes" after the initial arrival, the Ground bounce proberly some where between 2-8 ms.(But as every thing it depends). To eliminate it, move the mic,  use a baffle lager than 1.5x1.5 mtr or measure using the ground plane (put the mic on the floor) Just make sure that the HF actually hits the mic.
The first step is to make "good" measurements for your purpose, then you can start to analyze them. If I was you I would start by measuring a monitor (you said your in a club, so I assume you have one) And go from there, build confidence... instead of jumping into the lions mouth with a 4 way system.
Logged

Dave Gould

  • Marketplace
  • Newbie
  • *
  • Offline Offline
  • Posts: 16
Re: Smaart Queries
« Reply #8 on: January 26, 2011, 12:38:02 pm »

Hey Rasmus thanks the reply. I'm yet to check ou the rational acoustics forum but I will do once I get proper access to the Internet on CPU rather than an iPhone!

When I zoom into the impulse response I do get some very early reflections which could be affecting the readings that I am trying to take. I don't have access to a baffle but the floor is a good option. If I was to place the microphone on the floor how would I go about getting hi  frequencies to the mic other than by physically tilting the speakers?

Do you have any tips on how to interoperate the phase response that intact posted?

Does any one reading this have any idea about Smaart courses in the UK?
Logged

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 6871
  • Audio Plumber
Re: Smaart Queries
« Reply #9 on: January 26, 2011, 01:31:30 pm »

Does any one reading this have any idea about Smaart courses in the UK?

Smaart training schedules are kept pretty up to date on the Training Schedule web page. It shows none currently scheduled in the UK, but 2 in Spain in March and April.

Contacting them directly may get you a better idea of when the last one in the UK was, and how much longer before the next one.

Mac
Logged

ProSoundWeb Community

Re: Smaart Queries
« Reply #9 on: January 26, 2011, 01:31:30 pm »


Pages: [1] 2  All   Go Up
 



Page created in 0.043 seconds with 23 queries.