which weighting A,C or FLAT

[ian hatch]

Hi All

I know this subject has come up before but the answers seemed a little confused, to me anyway.

What is the best weighting A or C and also would I use the FLAT setting on the meter (PAA3) to do room EQing.

Cheers

Ian

[Simon Lewis]

A and C weighting were designed to provide a figure that more closely represented how the human ear would perceive a sound. Technically, you'd use A weighting for sounds at around 40 Phon. However, despite the obvious technical errors (i.e. the perceived lack of bass at low levels changes with increasing sound pressure level), A weighting gets used for just about all hearing and environmental measurement, irrespective of actual SPL.

 

In your case, you'd be better off with the flat scale. However, I'd suggest that you may not be starting from the correct point. Firstly, it's not possible to "eq the room" as such (although you can equalise the sound system so as not to exarcerbate room resonances and other aberrations), and secondly the real time analyser will give you information only about the variation in sound pressure at the microphone, and no information about the arrival time or phase of the wanted audio signal. To get around this, you would need to use and understand Smaart, EASERA etc.

 

Sorry if that's not the answer you need!

 

The manual tells you how to use the device, and states, "The EQ setting of audio system will vary from one venue to another. The ultimate goal in setting your EQ is to create a completely flat response and to eliminate every possible source of feedback. Though the PAA3's EQ SETTING function, even a novice user can successfully accomplish this.

 

It then goes on to tell you how to take a measurement, and then apply an inverse eq until you achieve a completely flat line. No disrespect to Phonic, but this isn't right. It ignores the fact that some issues can be equalised, and other issues are acoustic "black holes" where any attempts to boost signal by equalisation will only make matters worse.

 

A quick search on Blue Room for 'RTA' and 'Smaart' will expand on this topic considerably.

 

Simon

[ian hatch]

A and C weighting were designed to provide a figure that more closely represented how the human ear would perceive a sound. Technically, you'd use A weighting for sounds at around 40 Phon. However, despite the obvious technical errors (i.e. the perceived lack of bass at low levels changes with increasing sound pressure level), A weighting gets used for just about all hearing and environmental measurement, irrespective of actual SPL.

 

In your case, you'd be better off with the flat scale. However, I'd suggest that you may not be starting from the correct point. Firstly, it's not possible to "eq the room" as such (although you can equalise the sound system so as not to exarcerbate room resonances and other aberrations), and secondly the real time analyser will give you information only about the variation in sound pressure at the microphone, and no information about the arrival time or phase of the wanted audio signal. To get around this, you would need to use and understand Smaart, EASERA etc.

 

Sorry if that's not the answer you need!

 

The manual tells you how to use the device, and states, "The EQ setting of audio system will vary from one venue to another. The ultimate goal in setting your EQ is to create a completely flat response and to eliminate every possible source of feedback. Though the PAA3's EQ SETTING function, even a novice user can successfully accomplish this.

 

It then goes on to tell you how to take a measurement, and then apply an inverse eq until you achieve a completely flat line. No disrespect to Phonic, but this isn't right. It ignores the fact that some issues can be equalised, and other issues are acoustic "black holes" where any attempts to boost signal by equalisation will only make matters worse.

 

A quick search on Blue Room for 'RTA' and 'Smaart' will expand on this topic considerably.

 

Simon

Thanks Simon for the reply.

 

I take your point about the room EQing, about being many factors that can change in the time you have done the EQ settings.

I really got the device as a SPL meter as I have come across a couple of venues that have mentioned limits etc.

I wasn't sure which weighting to use as the booklet states 'A'=50-110 db & 'C'=70-130db and as all PA's run above 70db anyway this was the bit that confused me.

I work quite abit with a band that has a singer that walks/runs about with her radio mic, alot of the time in front of the PA stacks, so alot of times we are getting the dreaded feedback.

I have tried to teach her,please face away from the speakers when you are running around but it doesn't always work.

So could I use to PAA3 to try and EQ out the problem IE:EQing (sort of) the area of the room that see is likely to move around in, or would I be better of standing her radio mic in front of the speakers and try ALL 31 dips on the Graphic to see which ones ring and cut these?

 

Any advice most welcome.

Ian

[sleah]

I'm no expert by any stretch of the imagination, but have worked with a few live acts in similar circumstances to what you describe.

To be honest it sounds like you may be over-complicating the issue!

 

Such instruments have a real place in theatrical environments where positioning and movement are much more of a known quantity. i.e. the actors do pretty much the same thing every show, they also plan the movements beforehand, partly to avoid feedback. They will be mostly using lapel or headset mics.

 

There's no real replacement for an engineer riding the faders when working with musicians :P they tend to be unpredictable and do what they want, as you have found out :unsure:

 

It could be worth ringing out the room using an EQ as you describe (yes you do have to sit and go through all 31 bands!) to see if it helps, but you risk 'muddying' the sound and your singer may not like it :P

[ian hatch]

I'm no expert by any stretch of the imagination, but have worked with a few live acts in similar circumstances to what you describe.

To be honest it sounds like you may be over-complicating the issue!

 

Such instruments have a real place in theatrical environments where positioning and movement are much more of a known quantity. i.e. the actors do pretty much the same thing every show, they also plan the movements beforehand, partly to avoid feedback. They will be mostly using lapel or headset mics.

 

There's no real replacement for an engineer riding the faders when working with musicians :P they tend to be unpredictable and do what they want, as you have found out :unsure:

 

It could be worth ringing out the room using an EQ as you describe (yes you do have to sit and go through all 31 bands!) to see if it helps, but you risk 'muddying' the sound and your singer may not like it :P

Cheers Sleah for the advice.

I take your point about the cuts, might make the overall sound abit strange. I'll see what happens.

Going back to your points about the theatrical enviroment, I went to see the touring Scooby Doo Stage show the other week, (with the kids I might add) and in one scene the actors come off stage and walk along the front of the stage, in the process they walk within 2 feet of one of the speaker stacks to get back onto the stage whilst still singing without a hint of feedback. (I was abit sad to be thinking of this at the time)

Any ideas how this achieved, link to my singer problem.

[nlbrown]

Hello,

 

Age old problem!

 

A tip that I use time & time again is that if the singer wanders in front of the left stack, Pan the channel right & vise versa - you still have a reasonable level in the room, but no feedback! you might also find that it stops them going there as they won't be able to hear their vocal in front of the stack! (unless they are wearing IEM)

 

As another poster said, there is no substitute for an engineer riding the faders!

 

Neil

[DangerousMark]

Can't rule out the possibility that they may have just been lipsynching either, which solves a whole barrel of problems so long as they can do it just well enough to convince the kiddies :unsure: :P (and the clowns who dub most cheap-filler cartoons manage that)

 

I mean, it's Scooby Doo, not Whitney Houston or Aida or whatever.

 

PS if you manage to make the reverse EQ sound nice and convincing, I've got some old sound files which picked up wierd EQ effects during the recording process I'd love you to wave a magic wand over because I are fail at it (tho that may just be cooledit's FFT graph and eq filters being dodgy).

[ian hatch]

Can't rule out the possibility that they may have just been lipsynching either, which solves a whole barrel of problems so long as they can do it just well enough to convince the kiddies :unsure: :P (and the clowns who dub most cheap-filler cartoons manage that)

 

I mean, it's Scooby Doo, not Whitney Houston or Aida or whatever.

 

PS if you manage to make the reverse EQ sound nice and convincing, I've got some old sound files which picked up wierd EQ effects during the recording process I'd love you to wave a magic wand over because I are fail at it (tho that may just be cooledit's FFT graph and eq filters being dodgy).

Very good point, I wish I could do that with my problem singer.

[Shez]

I went to see the touring Scooby Doo Stage show the other week, (with the kids I might add) and in one scene the actors come off stage and walk along the front of the stage, in the process they walk within 2 feet of one of the speaker stacks to get back onto the stage whilst still singing without a hint of feedback. (I was abit sad to be thinking of this at the time)

Any ideas how this achieved, link to my singer problem.

 

I remember reading a while ago about some big name singer (think it was Robbie) who at one point wanted to be able to walk in front of one of the stacks. Being a large scale show with a significant number of cabs covering different areas of the audience, with matrix style routing, the engineer simply dipped his vocal out of that stack at the appropriate time which if course didn't affect the rest of the audience covered by any other cabs.

 

The panto I do each year always involves a couple of the cast wandering out in to the auditorium and always walking right in front of the main speakers. Despite them wearing omni headset mics, I've never once had a problem with feedback. Good quality kit doesn't suffer nearly as much from feedback as budget kit, all other things being equal.

 

Feedback will always occur on whatever frequency has the highest round trip gain. If the mic has a response peak at 4k, and the speaker also has a peak at 4k, (apart from sounding horrible) it'll be 4k that feeds back first. Good quality kit doesn't have these peaks to nearly the same extent and therefore there isn't any significant difference in the loop gain vs frequency.

[ian hatch]

I went to see the touring Scooby Doo Stage show the other week, (with the kids I might add) and in one scene the actors come off stage and walk along the front of the stage, in the process they walk within 2 feet of one of the speaker stacks to get back onto the stage whilst still singing without a hint of feedback. (I was abit sad to be thinking of this at the time)

Any ideas how this achieved, link to my singer problem.

 

I remember reading a while ago about some big name singer (think it was Robbie) who at one point wanted to be able to walk in front of one of the stacks. Being a large scale show with a significant number of cabs covering different areas of the audience, with matrix style routing, the engineer simply dipped his vocal out of that stack at the appropriate time which if course didn't affect the rest of the audience covered by any other cabs.

 

The panto I do each year always involves a couple of the cast wandering out in to the auditorium and always walking right in front of the main speakers. Despite them wearing omni headset mics, I've never once had a problem with feedback. Good quality kit doesn't suffer nearly as much from feedback as budget kit, all other things being equal.

 

Feedback will always occur on whatever frequency has the highest round trip gain. If the mic has a response peak at 4k, and the speaker also has a peak at 4k, (apart from sounding horrible) it'll be 4k that feeds back first. Good quality kit doesn't have these peaks to nearly the same extent and therefore there isn't any significant difference in the loop gain vs frequency.

Cheers Shez for the reply.

I can see how this worked for the Robbie? gig, good thinking on the engineer's part, but I'm only using 2 stacks and I would guess it might sound a bit strange if I dropped one side of the PA.

Any suggestions how I could limit the problem.

The kit that is being used are as follows:

2 stacks of Mackie 1232's on top of Mackie 1801 subs (all active) Sennheiser G2 (345?)radio mic, LA Audio EQ's, A&H GL2400

this might help you with any suggestions.

Any advice much appreciated.

Ian

[Shez]

I would probably start with putting the mic in front of the speaker where it's likely to be most troublesome, slowly ramp up the gain (be very careful!) until you hear it on the verge of feeding back, take out some of that frequency on the graphic, edge the gain up a bit more (still being careful!) until you hear a different frequency on the verge and knock that back too.

 

I've never been much of a fan of mackie speakers; there may not be all that much you can do about it at the end of the day. The frequency response plot in the manual is labelled as being a 1/3 octave response so it's possible that it looks rather smoother than it actually is in real life.

 

A proper parametric EQ would likely be more useful than the graphic as you can home in more precisely on the problem frequencies.

[ian hatch]

I would probably start with putting the mic in front of the speaker where it's likely to be most troublesome, slowly ramp up the gain (be very careful!) until you hear it on the verge of feeding back, take out some of that frequency on the graphic, edge the gain up a bit more (still being careful!) until you hear a different frequency on the verge and knock that back too.

 

I've never been much of a fan of mackie speakers; there may not be all that much you can do about it at the end of the day. The frequency response plot in the manual is labelled as being a 1/3 octave response so it's possible that it looks rather smoother than it actually is in real life.

 

A proper parametric EQ would likely be more useful than the graphic as you can home in more precisely on the problem frequencies.

Shez

Do you have to have the faders on the desk (main LR and the mic channel) turned up to the levels you would use for the gig, then adjust the mic gain, so when it starts to feedback it will be loud.

Ian

[Shez]

Do you have to have the faders on the desk (main LR and the mic channel) turned up to the levels you would use for the gig, then adjust the mic gain, so when it starts to feedback it will be loud.

 

It doesn't make a huge amount of difference to be honest. If you turn the master faders down 10dB, turn the channel fader down 10dB and push up the gain 20dB, the net result is exactly the same (assuming nothing is near to clipping). Feedback will occur when the loop gain of the entire system (mic-> desk-> amp-> speakers-> room-> mic) is at a certain point so it doesn't make a huge amount of difference which part of the signal chain creates that gain. It's the loop gain that causes feedback, not the system volume.

 

I would probably use the channel fader to creep up the levels - it's easier to be precise with that than with the gain pot.

[ian hatch]

Do you have to have the faders on the desk (main LR and the mic channel) turned up to the levels you would use for the gig, then adjust the mic gain, so when it starts to feedback it will be loud.

 

It doesn't make a huge amount of difference to be honest. If you turn the master faders down 10dB, turn the channel fader down 10dB and push up the gain 20dB, the net result is exactly the same (assuming nothing is near to clipping). Feedback will occur when the loop gain of the entire system (mic-> desk-> amp-> speakers-> room-> mic) is at a certain point so it doesn't make a huge amount of difference which part of the signal chain creates that gain. It's the loop gain that causes feedback, not the system volume.

 

I would probably use the channel fader to creep up the levels - it's easier to be precise with that than with the gain pot.

Shez

Double checking that I'm going to go about this the right way.

Set the gain on the mic as per normal (0db) main LR faders say -10db then just push the mic channel up until it's right on the limit before feedback, then adjust one at a time the 31 EQ faders, boost each one if no feedback set back to the middle or if feedback occures say +5 on the fader then you set the fader to -5 and so on.

Is this right?

Thanks

Ian

[MarkPAman]

Set the gain on the mic as per normal (0db) main LR faders say -10db then just push the mic channel up until it's right on the limit before feedback, then adjust one at a time the 31 EQ faders, boost each one if no feedback set back to the middle or if feedback occurs say +5 on the fader then you set the fader to -5 and so on.

OK up to the last bit, if that's the method you want to use.

 

Say you wanted to have 6dB headroom before feedback, then the slider that feeds at +5 needs to go to -1, & the one that feeds at +3 get set at -3. If you get to +6 or better then leave that one flat, don't boost anything.

 

I tend to push the mic channel until I just get feedback, then pull right slider on the graphic just enough to to get rid of it. Then do it again & again until I'm happy. The trick here is knowing where to stop, which (assuming it's a vocal mic) is best done by talking/shouting/singing through it