djandydee Posted May 8, 2009 Share Posted May 8, 2009 Having spent all day EQing a new system to "look" flat, I am dissapointed with the outcome. This is a 4 way horn loaded system with aux fed subs crossed over at 60Hz, bins from 60 to 200, mids from 200 to 2K and tops from 2K up. Soundweb is being used to control the system. The subs are 21", bins are 15", mids double 10", and HF double 1" 60x30 and double 1" 120x20 Each element can easily be flattened without too much effort and taking into consideration room accoustics and time alignment I can get a near flat response. This however sounds dreadful....I mean, slice you in half and turn your eyelids insideout. Fletcher Munson curve is probably the key here but what do off the shelf systems look like when measured? My ears are telling me to put a 6dB dip in from about 2K to 6K and add 15dB of sub. Can this be right? Any help appreciated. Link to comment Share on other sites More sharing options...
Rob_Beech Posted May 8, 2009 Share Posted May 8, 2009 Yes, this can be right, and probably is. Those 2 things on the side of your head are very good, as they work in a similar way to the similar things on the side of most peoples heads. Listen to the system, do what you need to get the best sound over the whole area, make sure this isn't overdriving any part of the system, and sit back and enjoy. The eq curve might look a mess, but if your ears says it's right, and nothing is being damaged, it's right. With that many 1inch drivers in the system its not surprising really that you need to cut alot of those hi mids. Rob Link to comment Share on other sites More sharing options...
djandydee Posted May 8, 2009 Author Share Posted May 8, 2009 Thanks Rob Don't panic about the HFs - This system is in a fairly big hall - 2 of the 1" drivers are on 20x60 horns (20x120) to cover the first 6 rows and are throttled back the other 2 are bolted on a Y adaptor to a 60x30 ASS horn for the remaining 10 rows further back. The coverage is excellent. I wanted plenty of headroom. My concern is that I thought "flat" was the ultimate aim. I would like to see a plot of an LA or EV rig running about 110 at the mix point and what the EQ is like with a well mixed and produced CD set to give a deep chesty low, nice bight and sparkle. I hope that makes sense. This is in a church youth centre seating about 600 -1000 people...not the biggest venue. The music type is guitar lead worship with plenty of drum kit and bass. Obviously vocals are important and need to be intelligable. Link to comment Share on other sites More sharing options...
waster Posted May 8, 2009 Share Posted May 8, 2009 What is the phase trace like? Did you set delay before taking measurements? What is the coherence of your measurements like? Link to comment Share on other sites More sharing options...
dbuckley Posted May 9, 2009 Share Posted May 9, 2009 I find flat systems obnoxious, and end up with a little dip in the mid, which is pretty much where you are. Dont worry about subs being flat to the rest of the system. The level of kick and bass guitar that is de rigeur in a live context is anything but contextual to the rest of the mix. Link to comment Share on other sites More sharing options...
Simon Lewis Posted May 9, 2009 Share Posted May 9, 2009 Andy, I'd want to know more about how you were measuring the system... Are you measuring each pass band on its own? Where are you placing the mic? Are you summing responses across the coverage area of that device? What settings are you using. Presumably you have the basics (delay etc. as mentioned above) set correctly? The Rationale acoustics guys would be the first to say that Smaart is a tool to help you understand what the system is doing (and take action) not an optimisation tool in itself. Flat lines can be achieved quite easily, but eq'ing doesn't fix other problems in the system! Simon Link to comment Share on other sites More sharing options...
TimmyP1955 Posted May 10, 2009 Share Posted May 10, 2009 In addition to looking at the phase trace, you need to look at the coherence to see if your measurements are valid. Indoors, there's a lot of scope for problems in this regard. Link to comment Share on other sites More sharing options...
noiseman Posted May 10, 2009 Share Posted May 10, 2009 Bottom line is does it actually sound good? If it does then job done.Smaart is a great TOOL. Don't let it rule your ears. Link to comment Share on other sites More sharing options...
djandydee Posted May 10, 2009 Author Share Posted May 10, 2009 Just to clarify my measurment technique - The measurment mic is on a long mic stand and is connected to Smaart CH1 with a long lead. The Pink noise generator is connected to Smaart CH2 and desk input channel using a Y lead. Whenever a measurment is being made the Auto LG delay is used. I generally run several passes of auto and also check if it is close by walking the distance before inserting the delay. This hopefully puts both inputs in the same time domain. I have measured each section of the PA at 2M on axis without any processing and as an experiment can flatten the required pass band with very little EQ. Coherance is solid.I have time alligned each driver to the front grills by inserting the required delay into the processor and have put in the crossover settings to use the required part of each section. I have adjusted processor gains so that each section is matching in level. Each stack has been measured at about 4M and the coherance is solid. The system/room measurment is done just in front of the mix point at audience ear height and some other locations to get an average. The coherance could be better but that is to be expected with untreated walls. Link to comment Share on other sites More sharing options...
Simon Lewis Posted May 10, 2009 Share Posted May 10, 2009 Try the measurement mic either on the (hard) floor with its cable under the XLR so that the capsule just touches the floor, or, put a large flightcase lid on the top of some chairs and get teh mic capsule close as possible to the flat surface. Either technique will help stop nearby reflections causing problems. You might also want to look at AutoSm for delay? Lastly, what settings have you got for smoothing, averages and FFT time? Are you using FPPO? Simon Link to comment Share on other sites More sharing options...
mackerr Posted May 10, 2009 Share Posted May 10, 2009 I have time alligned each driver to the front grills by inserting the required delay into the processor and have put in the crossover settings to use the required part of each section. I have adjusted processor gains so that each section is matching in level. Each stack has been measured at about 4M and the coherance is solid.Getting "flat" response with Smaart, or any other system, just means that what comes out of the speaker is an accurate representation of what went in. It doesn't mean it sounds good. When I think a system sounds good I generally find the frequency response curve is gently sloping down over the whole band from 200Hz to 20kHz. When you are aligning the various sections of the PA to each other with delay you don't want to just use the impulse response, which is what you get with the auto sm and auto lg buttons. You want each crossover region to have both speakers in phase with each other at the crossover point. When you take an impulse measurement you are reading the highest frequencies of each pass band. What you care about is the highest frequency of the lower pass band, but the lowest frequency of the higher pass band. The best way to make these corrections is with the phase trace in Smaart. You can change the scale of the display so you are only looking at the area of interest, then take a snapshot of the trailing speaker. Then, without changing the delay setting in Smaart, look at the phase trace of the the leading speaker, and add delay to the speaker till the phase curves line up at the crossover frequency. If the phase traces have the same slope, but do not lay on top of each other, you may have a polarity inversion. In any case, your ears should be the final arbiter. Any analysis system is only a tool to help you find problems. How to solve those problems is not what the analyzer tells you. To expand a bit on what Simon said, not only can you save time by using auto small instead of auto large, but you can get a better look at what is happening in the time domain by looking at the actual impulse response curve. This will show all the reflections, and just what is happening around the highest peak. You also want to use more averages when doing phase measurements to keep the trace from jumping around. Mac Link to comment Share on other sites More sharing options...
Charlie Jeal Posted May 11, 2009 Share Posted May 11, 2009 Nice to see you posting here Mac. Nice well informed posts as ever. Charlie Link to comment Share on other sites More sharing options...
mackerr Posted May 11, 2009 Share Posted May 11, 2009 Nice to see you posting here Mac. Nice well informed posts as ever. Charlie I don't have 7000 posts here, but I've made myself known once in a while. Mac Link to comment Share on other sites More sharing options...
djandydee Posted May 12, 2009 Author Share Posted May 12, 2009 Thanks to everyone who has helped out on this matter. I am confident that each section of the system is working in harmony, I was just surprized how "unflat" it needs to be to sound good. Most of the responses have indicated that a boost at LF and a dip in the mid is required. Over to Fletcher Munson . I would still be keen to do a measurment of something with a capital L or E in the name. Andy Link to comment Share on other sites More sharing options...
Simon Grainger Posted May 12, 2009 Share Posted May 12, 2009 One thing I learned from my broadcast days - a flat EQ always sounds harsh and discordant and always trust your ears over technology :) Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.