image description


Implanting an RTA?
  Forum Index » Live Sound 
Author
Message
sjc193
Presonic
[Avatar]

Joined: 30/03/2012 20:51:58
Messages: 568
Location: Warren, PA
Offline

chrisatrational wrote:
For system tuning, an RTA is inferior to more advanced measurement tools; namely the transfer function, or FFT analysis. The reason being, and the common misconception, is that an RTA is able to view its middle name...Time. In fact, that is where an RTA falls short. The RTA can't compare itself to anything, doesn't know where it's signal has been, or where it's going, so it can't discern between direct and reflected energy. Dual channel measurements compare the output of the console, to the output of the system. Because they are matched in time (a process in Smaart called 'delay tracking', or 'find/insert delay' - one of the steps that the wizards do for you), the analyzer is able to show exactly what happens to the signal as it passes through the system and interacts with the environment. The wizards coming in 1.7 incorporate dual channel FFT measurement by way assisted automated processes. You will be able to set a system Eq with a reasonable amount of precision that was previously unachievable without purchasing a stand alone FFT analyzer.
C


That's the best explanation of what FFT is that I've read yet, Thanks! Many people have thrown the word FFT out here many times but they never seemed to explain what it was, just that it was better. Can't wait for 1.7!

EQ has always been something I use to smooth the sound, I never end up cutting just one band to kill feedback by ringing out the room, I know how to do it but I usually don't have time and I hate to ring the room when people are in it (there are always people) and often I can crank it without getting feedback anyways, I end up making very broad stokes that can seem to tame the sound some (hard to explain). I have a really old consumer grade EQ that a friend gave me called a Sound Shaper, stereo 10 band EQ, it's kind of a cool little vintage piece, anyways, that name really sums up how I use an EQ, I shape the sound with it by ear. I really would love to actually know what knobs should be cut instead of just using my trial and error method, again, can't wait for 1.7!
StudioLive 16.4.2 in 10U + slant top OSP case
rackmount Windows 8 PC Quad core 8G ram
ASUS RT-N66U Dual Band Router
Ipad2 Wifi Griffin Survivor Case
Studio One 2.5 Pro
2 Turbosound TMS-1's (Mains)
2 OHM MR450D Subs with Kilomax 18inch drivers
2 Yamaha S115V's (for floor monitors),2 Yamaha BR 15's (for floor monitors)
3 Behringer 12" floor monitors
3 Behringer EP2500 amps, 1 QSC Amp
1 DBX Driverack 260
1 DBX 231S Graphic Eq
10 58/57 mics, 1 beta52 kick mic, 20 mic stands
2 e609 guitar amp mics
2 Radial J48 Active DI's
2 Blizzard Puck FAB 5 led lights
1 DMXIS usb to dmx controller
1 Temor 1 Alto 1 Soprano 1 Clarinet
1 Sheraton semi hollowbody electric
1 pedalboard with lots of pedals favorites are javaboost into tube screamer
1 Fender Deluxe Amp
2 Acoustic guitars, 1 elec bass, 1 acoustic bass, 1 bazooki
3 mixers and lots of other stuff I don't use in the closet
1 beer holder that clamps to the side of my mic stand with my stage name printed on it "Swingin' Steve"
Jerryd
Presonic

Joined: 28/07/2010 16:01:24
Messages: 698
Offline

Monolithent wrote:For those who may not have realized it, Chrisatrational is one of the Rational Acoustics gurus.

Had a very good chat with him at Presonusphere and a few other chats at the bar in the Indigo.

Very nice guy and very very knowledgeable.


This is exciting to see two of my most favorite companies come together. I have learned a ridiculous amount off of Rational and the Smaart program. I cannot say enough great things about that company. Presonus is doing great as well..... I had a rough start with them but their constant dedication to Listening to the customer and making changes based off that has been exceptional. I feel this is only the beginning of something spectacular in the near future.

Jerryd
Presonic

Joined: 28/07/2010 16:01:24
Messages: 698
Offline

sjc193 wrote:
chrisatrational wrote:
For system tuning, an RTA is inferior to more advanced measurement tools; namely the transfer function, or FFT analysis. The reason being, and the common misconception, is that an RTA is able to view its middle name...Time. In fact, that is where an RTA falls short. The RTA can't compare itself to anything, doesn't know where it's signal has been, or where it's going, so it can't discern between direct and reflected energy. Dual channel measurements compare the output of the console, to the output of the system. Because they are matched in time (a process in Smaart called 'delay tracking', or 'find/insert delay' - one of the steps that the wizards do for you), the analyzer is able to show exactly what happens to the signal as it passes through the system and interacts with the environment. The wizards coming in 1.7 incorporate dual channel FFT measurement by way assisted automated processes. You will be able to set a system Eq with a reasonable amount of precision that was previously unachievable without purchasing a stand alone FFT analyzer.
C


That's the best explanation of what FFT is that I've read yet, Thanks! Many people have thrown the word FFT out here many times but they never seemed to explain what it was, just that it was better. Can't wait for 1.7!

EQ has always been something I use to smooth the sound, I never end up cutting just one band to kill feedback by ringing out the room, I know how to do it but I usually don't have time and I hate to ring the room when people are in it (there are always people) and often I can crank it without getting feedback anyways, I end up making very broad stokes that can seem to tame the sound some (hard to explain). I have a really old consumer grade EQ that a friend gave me called a Sound Shaper, stereo 10 band EQ, it's kind of a cool little vintage piece, anyways, that name really sums up how I use an EQ, I shape the sound with it by ear. I really would love to actually know what knobs should be cut instead of just using my trial and error method, again, can't wait for 1.7!


It could not be any simpler with the Spectrograph. In fact -- you can catch hot spots before they even turned into an issue. It is good to be able to cut on hotspots if needed. If the system isn't tuned very well -- those hotspots will cause trouble or at least make the system sound poor.

Now here is another issue I have noticed with "HotSpots" on my sound system recently....... Out of at least 20 to 30 gigs -- I have YET to either have a Hotspot or a full feedback frequency land on the prescribed ISO Center. They are so close to the middle of the two frequencies that it is rediculous when I observe it.

It is like I want to hit a button that re-locates those ISO centers to the half way mark in between the standard frequencies. It is crazy!

Anyone else out there actually land on the exact slider chosen for you???? Maybe we can design a "BAR EQ" made specifically for dumpy bars where feedback and hotspots never land on the frequencies chosen for you..... lol

Actually this has made me think about my Tuning strategy on my system. I may have to take these observations into consideration and do some more fine tuning on my rig. I only see good things out of this....
chrisatrational
Presonic
[Avatar]

Joined: 12/12/2011 19:02:02
Messages: 138
Location: Putnam, CT
Offline

"It is like I want to hit a button that re-locates those ISO centers to the half way mark in between the standard frequencies. It is crazy! "

Jerryd - I believe what you are talking about is that the feedback frequencies don't match up exactly with your graphic Eq sliders, correct? This is not surprising, and simply repositioning the sliders wouldn't be a proper solution. Graphic Equalizers are fixed frequency and bandwidth, the only control you have is gain. Therefore, if you are experiencing feedback at 2.35kHz, you are out of luck with a Graphic Eq, because that lands right in the middle of two bands, 2kHz and 2.5kHz. The Smaart Spectrograph is a visual representation of all frequencies, over time and amplitude - not just the frequencies that you can attenuate with the graphic Eq.

The tool you are after is a Parametric Eq; inserted on the main L/R. With a parametric (which has variable Q and Frequency), you have a surgical tool in your arsenal for feedback mitigation. A graphic, although will get the job done as well, may require two slider adjustments until the desired result is heard. It is worth mentioning that these situational "Mix Equalizers" are separate, and should be considered aesthetic, from the dedicated parametric system Eq.

Chat with me on Skype = chris.tsanjoures
[Email] [WWW]
Jerryd
Presonic

Joined: 28/07/2010 16:01:24
Messages: 698
Offline

chrisatrational wrote:"It is like I want to hit a button that re-locates those ISO centers to the half way mark in between the standard frequencies. It is crazy! "

Jerryd - I believe what you are talking about is that the feedback frequencies don't match up exactly with your graphic Eq sliders, correct? This is not surprising, and simply repositioning the sliders wouldn't be a proper solution. Graphic Equalizers are fixed frequency and bandwidth, the only control you have is gain. Therefore, if you are experiencing feedback at 2.35kHz, you are out of luck with a Graphic Eq, because that lands right in the middle of two bands, 2kHz and 2.5kHz. The Smaart Spectrograph is a visual representation of all frequencies, over time and amplitude - not just the frequencies that you can attenuate with the graphic Eq.

The tool you are after is a Parametric Eq; inserted on the main L/R. With a parametric (which has variable Q and Frequency), you have a surgical tool in your arsenal for feedback mitigation. A graphic, although will get the job done as well, may require two slider adjustments until the desired result is heard. It is worth mentioning that these situational "Mix Equalizers" are separate, and should be considered aesthetic, from the dedicated parametric system Eq.



Hi Chris - yes I know this. Just sympathizing with the user stuck on the graphic. I have 12 parametrics on the Outputs and 9 on the input side of my DSP Plus 6 notch filters. I usually use the notch filters but at times when I am stuck out front - the SL graphic is vey handy. I rarely cut anymore than 3db on a frequency and it is fine. Very singer dependant type of thing....... Ie. some just whale on it and some are very soft breathy singers...... However - I am finding consistency with the offending frequencies to be in between the graphic sliders. Yeah it is time to fine tune the system utilizing the parametrics.

Thanks for commenting - don't want others to think that the key to great sound lies in the graphic eq.
Jerryd
Presonic

Joined: 28/07/2010 16:01:24
Messages: 698
Offline

Hey Chris - do you guys have a Presonus SL up at your place? If so - you should measure the response on the graphic. I discovered that the Q isn't like a typical 1/3 graphic. To me - it seems that the Q is much narrower. I'm okay with that....
chrisatrational
Presonic
[Avatar]

Joined: 12/12/2011 19:02:02
Messages: 138
Location: Putnam, CT
Offline

I understand completely! sometimes, a tight notch on the offending vocalists channel strip is just what the doctor ordered as well.

We do have a a couple SL's, and you are right about the graphic Eq. If you read the owners manual (GASP!) for the SL 24.4.2 on page 57 it says....

"The StudioLive Graphic EQ features an innovative design that sets it
apart from traditional graphic EQs. Traditionally, a 31-band graphic EQ uses 31 second-order shelving filters with fixed frequencies in order to simulate a curve set by the user via 31 front-panel sliders. A well designed graphic EQ creates an output frequency responses that corresponds as closely as possible to the curve displayed graphically by the sliders.
In an analog EQ, this is achieved by carefully choosing the bandwidth of the filter and deciding how or if it varies with the gain and how the filters are summed
or cascaded. In general, narrower bandwidth signifies a higher quality EQ But
in traditional graphic EQ designs, the center frequency of each band is fixed.
PreSonus took a different approach with the StudioLive. The StudioLive Graphic EQ is a pool of shelving filters from which coefficients like cutoff frequency, bandwidth, and gain are extracted through a process of curve-fitting The curve entered by the user is first oversampled. The system then works with an internal curve made up of 128 bands to find coefficients for the first shelving filter
that, when subtracted from the user’s curve, will produce the flattest possible response: 0 dB. The resulting response is then used to find coefficients for the second shelving filter through the same optimization process. Coefficients for all available shelving filters are found through a recursive process
Unlike conventional designs, the frequency and bandwidth of the “bands” depends on the curve entered by the user. This allows for much tighter matching of that curve. Because of this innovative design, the accuracy of the StudioLive EQ might feel “wrong” at first. The curve fitting process is capable of very steep transitions and, unlike conventional analog graphic EQs, what you see is what you get. With a carefully drawn, smooth curve the StudioLive EQ will have almost no frequency ripple. The StudioLive EQ is also capable of very sharp notches. To widen a notch or a bump on a particular band, just bring the adjacent bands closer to that band‘s gain and the bandwidth will change accordingly. Making drastic adjustments too quickly, may create an audio artifact as the new curve is redrawn and crosses over an audio peak. If you are hearing audio artifacts while setting up your Graphic EQ, try creating a smoother curve"
Chat with me on Skype = chris.tsanjoures
[Email] [WWW]
 
Forum Index » Live Sound
Go to: