ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2  All   Go Down

Author Topic: FFT vs RTA for live sound applications  (Read 6926 times)

walker rosewood

  • Newbie
  • *
  • Offline Offline
  • Posts: 3
FFT vs RTA for live sound applications
« on: August 20, 2015, 10:44:02 pm »

I've been reading up on FFT measurements and most of the information seems to be geared toward home theater.   Can this be applied to live audio?  Specifically, if every night I was in a different room with a different PA, would an FFT set up provide any benefits over an RTA in dialing in a PA.  Enough to justify spending some money on hardware?

To keep cost down while learning the ins and outs of FFT measurements I've been considering the following USB mic and software. 
http://www.minidsp.com/products/acoustic-measurement/umik-1
http://www.roomeqwizard.com/
Logged

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 6396
  • Audio Plumber
Posting Rules
« Reply #1 on: August 20, 2015, 10:49:23 pm »

I've been reading up on FFT measurements

Please go to your profile and change the "Name" field to your real first and last name as required by the posting rules displayed in the header at the top of the section, and in the Site Rules and Suggestions in the Forum Announcements section, and on the registration page when you registered.

Mac
Logged

Doug Fowler

  • Member since May 1995, 2nd poster on original LAB, moderator on and off since 1997, now running TurboMOD v1.826
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1676
  • Saint Louis, MO USA
Re: FFT vs RTA for live sound applications
« Reply #2 on: August 21, 2015, 02:04:11 pm »

I've been reading up on FFT measurements and most of the information seems to be geared toward home theater.   Can this be applied to live audio?  Specifically, if every night I was in a different room with a different PA, would an FFT set up provide any benefits over an RTA in dialing in a PA.  Enough to justify spending some money on hardware?

To keep cost down while learning the ins and outs of FFT measurements I've been considering the following USB mic and software. 
http://www.minidsp.com/products/acoustic-measurement/umik-1
http://www.roomeqwizard.com/

You'll need training in any case.  RTA has no timing information, and can be very misleading.  OTOH FFT measurements can be very misleading as well.  It's easy enough to make squiggly lines, but correctly interpreting the squiggly lines is not intuitive.
Logged
"It's got electrolytes.  It's got what plants crave."

Hayden J. Nebus

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 637
  • Richmond, VA
Re: FFT vs RTA for live sound applications
« Reply #3 on: August 21, 2015, 02:55:25 pm »

  It's easy enough to make squiggly lines, but correctly interpreting the squiggly lines is not intuitive.

I still see people regularly with dual channel FFT software who only use its RTA function ... I suspect solely for this reason.

The "analyzer", whatever its form, really just shows the input data juggled in a way that's been designed to be human-readable. The analysis is performed when a human sits in front of it and makes a decision based on what's displayed.
Logged

Peter Kowalczyk

  • Full Member
  • ***
  • Offline Offline
  • Posts: 162
Re: FFT vs RTA for live sound applications
« Reply #4 on: August 21, 2015, 07:49:43 pm »

While single-channel RTAs have their uses (e.g., detecting the frequency of a squeal of feedback), a Dual-Channel FFT analyzer is a much more powerful tool, and is highly suited to measurements of systems for live sound reinforcement.  Many professionals who encounter a different PA each night use an FFT analyzer precisely as you describe.

As others have said, with any powerful too, there is a learning curve.  Furthermore, an FFT analyzer is likely to reveal all sorts of 'problems' in a sound system that cannot be solved by signal processing.  A training class taught by an experienced professional is worth vastly more than the cost of the hardware itself.
Logged

walker rosewood

  • Newbie
  • *
  • Offline Offline
  • Posts: 3
Re: FFT vs RTA for live sound applications
« Reply #5 on: August 21, 2015, 09:52:02 pm »

As others have said, with any powerful too, there is a learning curve.  Furthermore, an FFT analyzer is likely to reveal all sorts of 'problems' in a sound system that cannot be solved by signal processing.  A training class taught by an experienced professional is worth vastly more than the cost of the hardware itself.

Noted.  I'll look into any classes within reasonable distance. 

So what sort of things could I tell from an FFT that I can fix with signal processing?  What information does an FFT give me that an RTA can't, and how could I apply the results to make the room sound better?

About half of the time I am using my own (well, the company's) PA and and have access to things like a driverack, lab gruppen PLM, etc.   The other half I am walking into clubs with all sorts of interesting pieces cobbled together and might only have access to a house graph and if I'm lucky a crossover or delay. 
Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 8752
  • Atlanta GA
Re: FFT vs RTA for live sound applications
« Reply #6 on: August 22, 2015, 09:22:53 am »

Noted.  I'll look into any classes within reasonable distance. 

So what sort of things could I tell from an FFT that I can fix with signal processing?  What information does an FFT give me that an RTA can't, and how could I apply the results to make the room sound better?

About half of the time I am using my own (well, the company's) PA and and have access to things like a driverack, lab gruppen PLM, etc.   The other half I am walking into clubs with all sorts of interesting pieces cobbled together and might only have access to a house graph and if I'm lucky a crossover or delay.
All any measurement system can tell you is what it is seeing AT THAT POINT IN SPACE.

Now what is causing the problem (a dip for example) is up to YOU and YOUR KNOWLEDGE and UNDERSTANDING.

A dip in the response could be an actual dip in the freq response of the loudspeaker-and a little eq could help that.

OR it could be due to cancellations due to offset drivers and different time arrivals.  EQ will do nothing to fix that (altough it might "look" like it does on the computer screen.  Signal delay is the tool that will fix that.

OR it could be reflection.  Moving the mic will "fix" that.

No measurement system will tell you the problem-only the RESULT of the system and what you have done (or attempted to do)

The big difference between a DUAL FFT (NOT single channel FFT) and an RTA is that an RTA cannot tell the difference between a truck going by and the response of your subs.

Different systems and how each system is set up (windowing-averaging etc) will determine how much "other" signals contribute to what you are seeing on the screen.

Understanding this is NOT something that is learned in a few hours-but rather years of looking at kinds of different situations and understanding what you can (and more importantly) what you CAN'T fix with ANY sort of processing.

Learning your way around the actual measurement platform is pretty easy-IF you understand measurements-how to read them (INCLUDING PHASE-which many people simply ignore).

Learning what the measurement is telling you is the hard part.

I would start by looking at some of the videos and reading the tutorials from Rational Acoustics and Smaart.

While it may not be the platform you use-the ideas/concepts apply to ALL measurement systems.

I use a couple of different systems, depending on what I am doing and what type of result I am looking for.

Some are faster-some are more accurate-some give information the others can't and so forth.

They all have their own advantages and disadvantages.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

Kevin McDonough

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 305
Re: FFT vs RTA for live sound applications
« Reply #7 on: August 23, 2015, 07:38:03 am »

Hey

Yeah really what everyone has said.

RTA is a "dumb" measurement. It hears sound and shows you on screen exactly what it hears at that point in space, but it has no clue 'why' it sounds like that. Problems could be caused by an actual dip in the frequency response, a time alignment problem between two (or more) speakers causing a cancellation, a driver or speaker out of phase, or a room reflection.

FFT measurements, by comparing to the original signal with their 2 channel system, are able to have some understanding of time as well as frequency. This means impulse response measurements can be used to set delay times between different speakers, subs and tops, delay speakers etc. Phase graphs and group delay measurements allow you to set up much more accurate crossover points in your system, and when you have frequency problems out in the room the phase graph and others may be able to give you some idea of why you're getting the problem, not just that it's there. Plus you also have the RTA function built in as standard for the more basic identifying feedback frequencies etc.

But as said, having the graphs and information is only part of the solution, you need to be able to understand what they're telling you. After messing about with a few less than legal copies of Smaart over the years, friends were offering the Systune training course and I decided to bite the bullet and invest in a more comprehensive and professional measurement set up and go for it. Was a great decision.

In the past I kind of knew what all the graphs meant and what I was doing but now have a MUCH better understanding, and feel much more confident that the results I'm getting are accurate and that I have all the equipment set up properly.

However much like a driving test, the course got me to a good level of competency but is just the start of the journey, and the coming years of experience and practice will be what really teaches me the fine detail and (hopefully) takes my sound work to a new level  :D

K
Logged

Lyle Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1539
Re: FFT vs RTA for live sound applications
« Reply #8 on: August 26, 2015, 06:07:03 am »

While it may take some years to "master" dual channel FFT, one can make good use of it after only a couple of hours.

Make a measurement.  Come up with a hypothesis about what you are seeing.  Change something (mic position, EQ, delay, whatever.) Do the new measurements support your hypothesis about the earlier measurements?  If not, come up with a new hypothesis about what is happening.
Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 8752
  • Atlanta GA
Re: FFT vs RTA for live sound applications
« Reply #9 on: August 26, 2015, 07:24:13 am »

While it may take some years to "master" dual channel FFT, one can make good use of it after only a couple of hours.

Make a measurement.  Come up with a hypothesis about what you are seeing.  Change something (mic position, EQ, delay, whatever.) Do the new measurements support your hypothesis about the earlier measurements?  If not, come up with a new hypothesis about what is happening.
Agreed.

Teaching YOURSELF, goes A LONG way towards learning.  What you teach  yourself is often more valuable than what somebody else teaches you.

You can tell somebody the stove is hot.  But until they touch it themselves-they don't realize HOW hot and how much it hurts.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!
Pages: [1] 2  All   Go Up
 


Page created in 0.063 seconds with 23 queries.