This is topic Proper B Chain in forum Film Handlers' Forum at Film-Tech Forum ARCHIVE.


To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=1;t=004575

Posted by Tom Wienholt (Member # 1510) on 02-19-2003, 07:14 PM:
 
What is the proper procedure for EQing a sound system?

I recently set up a CP50 using a pink noise generator and a RTA. I tried to get each channel to a flat response up to 2000 Hz and then rolling off. In order to get this flat response, I had to turn the bass control down quite far and the treble up substantially. Then I adjusted the 1/3 controls to finish the EQ. The result looked fairly nice on the RTA but when I ran a movie through there seemed to be TOO much high frequency and not enought bass. What is the problem? I only used one microphone to do the EQ (placed off center 2/3 back) but i don't think that was the problem. Please give me some suggestions. Thanks.
 
Posted by Manny Knowles (Member # 1171) on 02-19-2003, 08:45 PM:
 
There are other people here who can probably be more helpful.

I had a particular room that was very difficult to EQ because it wasn't soundproof and was near a well-travelled main road. The only time I could EQ that room properly was very late at night...and we had to be sure to turn off the AC and the water pump (yes, they actually have a water pump behind that screen).

I'd like to hear more about this theatre. How big is it? What about the soundproofing? Is it well made? What speakers are you using?

Or is this a "noisy" room? Did you take a look at the RTA readings prior to generating the pink noise?

What kind of RTA are you working with? Is the microphone calibrated to that particular analyzer? They have to be a matched pair.
 
Posted by Gordon McLeod (Member # 33) on 02-19-2003, 09:11 PM:
 
Also what type of pink noise generator did you use and how was it injected into the system

Also did you check the aiming of your speakers
 
Posted by Tom Wienholt (Member # 1510) on 02-19-2003, 09:17 PM:
 
The auditorium is 45 feet long and 30 feet wide. The RTA is an AudioControl with a mic that came with the analyzer. It is soundproofed quite well with no sound leaking into other auditoriums. The speakers are brand new JBL 3377's. It just seemed like I had to cut the bass and boost the treble quite a lot to get a flat response. So, it sounds is too "bright" with too much treble. The response on the RTA looks good but to me it sounds like there is not enough bass. There is no subwoofer in the auditorium but that is not the problem. Dialogue sounds very clear and easy to understand but some of the high frequency just seems to "loud". Any help would be appreciated.
 
Posted by Tom Wienholt (Member # 1510) on 02-19-2003, 09:20 PM:
 
I used a Component Engineering generator that plugs right into the CP50. What could be wrong with the aiming of the speakers? Thanks
 
Posted by Daryl C. W. O'Shea (Member # 1303) on 02-19-2003, 09:27 PM:
 
In the future, please use the edit feature instead of posting twice.

 -
 
Posted by Manny Knowles (Member # 1171) on 02-19-2003, 09:43 PM:
 
I know the AudioControl 3051 unit and I'm okay with it (an Ivie would be nicer, but this chain has me on a tight budget so that kind of luxury will come later). The AudioControl can do the job.

First things first -- Did you get an A-C weighting filter for your RTA? Don't use that when you're setting your EQ because it will roll off the high frequencies and trick you into boosting them. That's one sure way to end up with a bright-sounding room. (But you will need that A-C weighting attachment when you set your levels.)

If that's not the problem, then let's move on and talk about the settings on your RTA...

How loud are you playing your pink noise? It should be about 85dB and you should then set the RTA for the 80dB range. You can use the "calibration" knob to raise or lower the LED's if you want them to line up with the line on the display.

How many "dB per step" are you using? You should use 3dB per step because that seems to be the threshold where we can hear an increase/decrease in sounds. But 4dB/step is too much. If you use that setting, you could very easily end up with a nice, neat little line and that will surely mislead you.

And then there's the response setting: You should use the medium or slow response. I think the manual says to use the slow setting when working with pink noise but, personally, I find the slow speed to be too slow so I use the medium setting.

How far off the ground is your mic?
What is holding it in position?
What kind of cable are you using to connect your mic to the RTA?

Also, what Gordon said about aiming the speakers should be looked into. My tech advised me to place them as close to the screen as possible (without touching the screen). The horns are angled inward toward the centre-line of the auditorium and downward toward the seats about 2/3 the distance to the back wall. Repositioning the speakers and aiming the horns went a long way toward getting a better result.

Tell me -- did you find that you had to tweak a lot of the EQ dials, or just a few? For the most part, whould you say that you were boosting or cutting frequencies?

I'm also curious as to what you listened to when you tested your system.
 
Posted by Jack Ondracek (Member # 1466) on 02-19-2003, 10:10 PM:
 
Tom, before you get into a panic here, it would be a good idea to make sure you've got the basics covered.

The standard Dolby alignment procedure works pretty well, as far as our experiences have shown. If, as their alignment procedures indicate, you have to move the eq settings radically from band to band, you should be looking at other parts of your system for possible problems. Phasing, placement, cone & driver condition can have huge effects on your alignment efforts, any acoustic issues notwithstanding.

If you have to move neighboring bands radically opposite from center to get your RTA to look like you think it should, you should probably make sure you don't have other issues impacting your efforts to get ahold of your equalization efforts. If you're sure you're ok with all your physical & electronic equipment, center all your bands & give it a fresh try.
 
Posted by Manny Knowles (Member # 1171) on 02-19-2003, 10:16 PM:
 
That's very good advice from Jack.
 
Posted by Tom Wienholt (Member # 1510) on 02-20-2003, 01:09 AM:
 
I dont think that my analyzer has a C weight filter. But, since this is not necessary for room equalization, this should not be the problem. Can I still set sound pressure levels without the filter?
 
Posted by Michael Schaffer (Member # 1204) on 02-20-2003, 06:17 AM:
 
You will a lot of answers in the documents that Manny pointed to in his "Non-standard practices" topic. Especially in "toss_meters". Check p.6, it comes very close to your problem except for the dialogue quality.
 
Posted by Manny Knowles (Member # 1171) on 02-20-2003, 07:24 AM:
 
Tom wrote:
quote:
I dont think that my analyzer has a C weight filter. But, since this is not necessary for room equalization, this should not be the problem. Can I still set sound pressure levels without the filter?
Sadly, the weighting filter is an OPTIONAL accessory from AudioControl.

My understanding is that you cannot set SPL levels properly without the weighting filter. The reason is that the C curve compensates for how we perceive loudness at different frequencies. Until I got my A-C filter, I had to go down into the theatre and check with a hand-held Radio Shack meter. This involved a lot of stair climbing. As upset as I am about the filter not being included with the RTA, I have to say that it was worth the $40 it cost.

Maybe someone knows a work-around: Is there a formula to accurately set SPL in a movie theatre with an unweighted signal?

I can already tell you -- from experience -- that it seems to involve more than subtraction. I found that the difference between the RTA and the hand-held meter always differed from speaker to speaker. So, it wasn't as easy as saying "OK, the meter says 85dB and the RTA says 90dB so I'll subtract 5dB when I set the other speakers." Somehow that logic did not apply. I don't really understand why.

Anyway...

You've said that you believe the room itself is okay. Likewise for the system components and the measurement device. If so, that seems to rule out everything...except you!

I think you should just try it again. Something you did wasn't quite right. If it comes out bad a second time then maybe we'll need you to provide some answers to the questions we've asked.

Michael wrote:
quote:
You will a lot of answers in the documents that Manny pointed to in his "Non-standard practices" topic. Especially in "toss_meters". Check p.6, it comes very close to your problem except for the dialogue quality.

Yes, I've been converted by those documents but since your topic is entitled "Proper" B-Chain I wasn't going to go there.
 
Posted by Dave Macaulay (Member # 813) on 02-20-2003, 08:28 AM:
 
I'm not sure what weighting your RTA uses without the optional filter. If it's "flat", not so bad... if it's "A", big problem.
The actual curves and a simple conversion table are here http://www.prosoundweb.com/studyhall/lastudyhall/slm/slm_3.shtml you can click on the curve and get a bigger version.
Your problem suggests that you have the A curve applied, which really ignores low frequencies. If you have any way to get a flat response (close to the C curve excepting the extremes), this will give you a better starting point then adjust the final settings by ear.
As has been said, any extreme EQ settings suggest other problems. In particular, ensure all your speaker line polarities are correct.
 
Posted by Manny Knowles (Member # 1171) on 02-20-2003, 10:15 AM:
 
The AudioControl in its natural state is unweighted. The attachment has a switch for either A or C weighting.
 
Posted by Tom Wienholt (Member # 1510) on 02-20-2003, 06:05 PM:
 
Manny
Where did you get the A-C filter for your audiocontrol microphone? Let me know.

[ 02-20-2003, 10:10 PM: Message edited by: Tom Wienholt ]
 
Posted by Gordon McLeod (Member # 33) on 02-21-2003, 12:47 AM:
 
I have used the audio control rta's and you need to use the slow setting and set it in 2 db per division to start with
aim the mic at the ceiling straight up
also turn the sub amp off when doing the eq
Also check the CE generator on the side of it is a slide switch marked red and pink make sure it was in the pink mode not red mode

[ 02-21-2003, 12:10 PM: Message edited by: Gordon McLeod ]
 
Posted by Manny Knowles (Member # 1171) on 02-21-2003, 12:25 PM:
 
Gord, thanks for the tip. I will certainly try that next time. I find that if you use the slow setting (at 3dB) then the LEDs just don't move. I suppose that it could be an equipment failure but I doubt it since my unit is brand new and works fine in all other respects. I will try using slow and the 2dB/step setting.

As for the A-C filter, you can probably get it through your equipment supplier. It is made by AudioControl and I think the product number is AC-10. (I'll verify that later.)
 
Posted by Frank Angel (Member # 248) on 03-06-2003, 12:34 AM:
 
Ok, this is getting a bit confusing. Dave is saying that the A curve was a big problem, which I assume is not a good thing. Yet the author of the article says "....it later became evident that the weighting curves chosen for SLM measurements at certain levels do not correlate very well to how we actually hear complex sounds. More recently it has been found that the A-weighting curve provides a realistic approximation of how we hear over a fairly wide range of sound level measurements."

So will EQing with the A filter get the more accurate response?

Frank
 
Posted by Manny Knowles (Member # 1171) on 03-06-2003, 07:26 AM:
 
No. Use the C filter.

The filters are designed to offset the Equal Loudness Contours. Each contour applies to a specific range of loudness levels. Movie theaters are set to 85dB and that falls within the C range. Hence, the "C" in the expression: 85dBC.
 
Posted by Gordon McLeod (Member # 33) on 03-07-2003, 07:33 PM:
 
Eq should be down with no weighting but a flat response
the spl is measured as DBC weighting
 




Powered by Infopop Corporation
UBB.classicTM 6.3.1.2