Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Lynn Oliver]

When I set up my PA a while back, I ran Auto EQ on my DR260 on each type of speaker in order to provide a flatter effective speaker response.  The method I used was to place each speaker outside on a speaker stand and point the reference mic at the speaker from about 15 feet away. 

This week I reran Auto EQ indoors, placing the speaker on the floor tilted forward a bit, and the reference mic at floor level approximately at the point the speaker was aimed at.

The outside Auto EQ runs took a long time (some did not finish at all) and were not very consistent from run to run.  The indoor runs completed fairly quickly and were pretty consistent.  The indoor results were significantly different from the outdoor ones.

To my ear the PA sounds better after EQ, although I have not been able to test with the newer settings yet.  Is there any objective way to validate what I am doing?  I'm concerned that the two methods do not produce at least similar results, that all the results are well outside the published specs for the speakers, and that the results for the Yorkville NX550P are flatter than those for the RCF ART-310A, even though user comments would indicate the opposite.

Any suggestions?

[Dalton Brand]

First off- real names are required on this board.

Secondly- the first rule of audio measurement is "what are you trying to measure?". Are you trying to use auto EQ to find the correct EQ curve for your cabinets or for the the cabinet/room combination? And at what point in the room?

Auto EQ rarely gets the results the marketing department would have you to believe. You're better off starting with the DSP settings provided by the manufacturer for the cabinets and use your ears from there. Search the old forum for your cabinets and you may find what you're looking for as a starting point.

Don't expect auto EQ to optimize the system for you; there are far too many variables, not the least of which is time, which auto EQ won't address.

[Lynn Oliver]

Are you trying to use auto EQ to find the correct EQ curve for your cabinets or for the the cabinet/room combination?

As I stated in the OP, I am trying to EQ the speakers (cabinets).

Auto EQ rarely gets the results the marketing department would have you to believe.

I've always understood that Auto EQ was not very effective for its stated purpose.  The idea and methodology for using it with speakers come from other DR users, not marketing.

You're better off starting with the DSP settings provided by the manufacturer for the cabinets and use your ears from there. Search the old forum for your cabinets and you may find what you're looking for as a starting point.

The manufacturers do not provide settings for these speakers, nor are there any posted in the old forum.  That's why I'm trying to validate what I am doing with Auto EQ.

Unless someone already has good DSP settings for these speakers, what is the best way to proceed?

[Brad Weber]

As I stated in the OP, I am trying to EQ the speakers (cabinets).

You are using packaged, powered, full range speakers, they are already tuned by the manufacturer so what are you actually trying to do?
 

I've always understood that Auto EQ was not very effective for its stated purpose.  The idea and methodology for using it with speakers come from other DR users, not marketing.

Maybe I'm missing something but are you saying that you know it's not very effective but you're relying on it anyways and then wondering why it isn't working that well?  With the measurement methods you used, especially indoors, you are not removing the environment from the measurement.  You may have done some things to reduce the impact of the environment but you are still addressing the processor, amp, cables, speaker, environment and mic rather than just the speakers.

The manufacturers do not provide settings for these speakers, nor are there any posted in the old forum.  That's why I'm trying to validate what I am doing with Auto EQ.

Unless someone already has good DSP settings for these speakers, what is the best way to proceed?

Again, what are you actually trying to do?  Anything you do is going to be to try to achieve some specific result in a particular situation and since factors such as the desired result and environment are unique to each situation, there is no predefined solution.

The best way to proceed is probably to start searching the Study Hall, archives and past dicsussions to try to learn more about the general topic.

[Dave Bigelow]

The Auto EQ on those is garbage, use your ears

[Dalton Brand]

Dave said it more succinctly than I did but he's essentially correct. I've messed around with enough cabinets that I had known factory DSP settings and auto-EQ rarely comes close to finding the correct settings.

It probably has something to do with my lack of an anechoic chamber.

The auto-EQ is an amateur function on an otherwise professional processor. It may have been made mildly more useful if it included the ability to take a bunch of measurements and average the results but even then it's beholden on the user to know how to place the measurement mike.

Read some Bob McCarthy to really find out how many variables one is up against in the field of measurement.

[Lynn Oliver]

You are using packaged, powered, full range speakers, they are already tuned by the manufacturer so what are you actually trying to do?

These are not exactly expensive cabs; I spent less than $2500 total for a pair of each model.  At that price point the manufacturers have to make certain compromises, so it is reasonable to look at the frequency response and see if it needs compensation.

Maybe I'm missing something but are you saying that you know it's not very effective but you're relying on it anyways and then wondering why it isn't working that well?

The DR260 manual shows how to use the Auto EQ function to EQ a venue.  I don't think I would get any argument in saying that it is a little more complex than just plunking a mic down out in the middle of the audience and running an RTA.  To be fair, dbx does advise taking several measurements at different points, but in my experience it isn't worth the effort.  What I am doing instead is to use the RTA processing to evaluate the frequency response of a speaker cabinet.   

With the measurement methods you used, especially indoors, you are not removing the environment from the measurement.  You may have done some things to reduce the impact of the environment but you are still addressing the processor, amp, cables, speaker, environment and mic rather than just the speakers.

The processor and mic are the measurement system; for better or worse they are what they are.  The amp is part of the powered cab, so I want that included.  Cables I'm not too worried about as I'm using good quality mic cables and any effect they might have is minimal.  The speaker is also part of the powered cab, so I want that included. 

The environment is the biggest concern; outdoors where there are no surfaces to reflect sound (other than grass) seems like an obvious choice if you don't have access to an anechoic chamber.  The second method attempts to approximate a boundary mic closely enough that the comb filtering effects are above the frequencies of interest.

What I've seen with the outdoor method is that Auto EQ processes for a long time trying to balance the system, and repeated runs are not always consistent.  With the indoor runs Auto EQ completes in less time and the results are consistent.  These experiences are similar to what other DR260 users have reported. 

Again, what are you actually trying to do?  Anything you do is going to be to try to achieve some specific result in a particular situation and since factors such as the desired result and environment are unique to each situation, there is no predefined solution.

When I first set up my PA I was simply trying to flatten the response of the system.  I took a number of Auto EQ runs and came up with a starting EQ setting based on the most consistent readings in each frequency band.  Then I set up the PA and fine-tuned the EQ by ear.  This worked better for me at that time than doing it completely by ear.

Now I'm looking at the frequency response of individual speaker cabs relative to each other.  Lets say I plug a guitar processor into my studio monitors and fine-tune the sound.  How will it change if I plug it into a speaker cab?  Is it possible to compensate the speaker cab so that the response is more consistent with my studio monitors?  Or if I plug into a speaker cab and fine-tune the sound, how will that relate to the FOH sound on some arbitrary PA? 

The best way to proceed is probably to start searching the Study Hall, archives and past dicsussions to try to learn more about the general topic.

I've been through all that material.

The Auto EQ on those is garbage, use your ears

What specifically is the problem?  Is it the pink noise generator, the mic preamp, the processing algorithm, what? 

Dave said it more succinctly than I did but he's essentially correct. I've messed around with enough cabinets that I had known factory DSP settings and auto-EQ rarely comes close to finding the correct settings.

At least this supports the idea that if you find the correct settings it can improve the response of the system. 

It probably has something to do with my lack of an anechoic chamber.

It would make things easier.

The auto-EQ is an amateur function on an otherwise professional processor. It may have been made mildly more useful if it included the ability to take a bunch of measurements and average the results but even then it's beholden on the user to know how to place the measurement mike.

I agree.

[MARK PAVLETICH]

The DR series auto-eq seem to be based on a simple RTA type system. This is never a substitute for proper impulse response analysis of a rig/room combination. When the DR series came out I spent some time in a venue experimenting with its capabilities including taking several positions of measurement. The curves that it produced were extreme and suggested many huge boosts of up to 12 dB !!!.

Unfortunately many of the users of this system would not yet know of the uselessness of trying to correct phase/time issues with EQ, ( sort of like trying to chop down a tree with a screwdriver....)

I came to the conclusion that it added no more value than what I could get from tuning to a known CD and vocal mic, assuming that the system was already impulse response FFT tuned either by the manufacturer or end user. Of course the Driverack can't do that job ,so overall I see it as a next to useless feature.

 However, the Wizard suggested settings for many popular boxes do seem to provide a good starting ( and often finishing ) point. I look forward to the next series of Driverack which hopefully features automatic time alignment and impulse/sweep based automatic EQ adjustment. Now THAT would be cool.....

[Douglas R. Allen]

Its been quite a few years but I did have a chance to work with the DRPA and DR260. Before that the peavey Autograph/mate.
These do give repeatable results if you take the time to know what it is you want from them.
When you do an rta sweep and it doesn't finish it could be.
1. RTA pink noise not loud enough.
2. Speakers freq. response not wide enough to give wanted     results. If your shooting for flat response from 20hz to 20khz and the speakers even with a 12db boost at 20khz won't go that high the unit will "hang up" if you will. Unable to give the target response. Also the distance the reference mic. is away from the speaker will come into play here as well.
3. Time alignment issues linked to 2 above. A sign of out of alignment drivers could look something like this. At a horn to mid crossover point of 1.2k as an example. At the next freq below you might see full cut. The next freq. above it full boost.
Its trying to work around the cancelation caused by the drivers being out of alignment. The plus side to this is I used to use this to my advantage. I would add delay until I seen the least amount of boost/cut at the crossover point. This was before I had a proper measurement tool.

A handy tool in the Autograph was you could enter a target freq. response that you wanted the speaker being tested to read on its output. Knowing the high freq. limit of say 16.5k of the horns I would cut everything above this. Of course the same with the low freq. range of the driver/sub.

Again its been awhile but I don't remember being able to have a target curve put in a DRPA but the DR260 may have that feature. I am pretty sure it does have target preset curves. The one where the output decreases as the freq. increases used to sound good for live sound use. Look for a user preset area. If it has one enter a curve that is within the freq. range of your speakers with the normal "sound" your looking for.

One last thing. Be sure to go in a look at the final curve. It may have a big boost/cut somewhere that you may want to bring down/up. Could be caused by room reflections or the speaker in test just not able to give a flat response at 20hz....

I found the rta system to be fine if you could enter a target curve and new the limits of the system.

Douglas R. Allen

EDIT: I never did both PA stacks at the same time the way it is stated by DBX. I found 1 stack gave the best results.

"Where is spell check?"

[Lynn Oliver]

Thanks for the responses.

I spent some time with it again today and I think with my speakers, which are at least OK to start with, the GEQ resulting from the indoor Auto EQ does more harm than good.   

Maybe I'll run an impulse response on the speaker and work from that.

[Royce Covington]

since you do not have the dsp settings for your speakers, you don't have a good starting point for your system set-up.  so, you're trying to use the auto eq to optimize the speaker sound, right?

well, ordinarily when setting up the driverack the speakers/amps are tuned with the crossover and parametric filter settings, etc.  then, once the 'system' configuration is established, the main/graphic eq is used to optimize the relationship of the 'system' to the room.

the auto-eq is notably one of the worst features of the driverack and by no means a reasonable tool for any practical purpose.

R~

[Lynn Oliver]

Royce, that is correct, I'm using it as a starting point. 

I may have given up a bit prematurely, as I just got a copy of the specs for the 310a's and the published frequency response is actually pretty close to what Auto EQ came up with.  Of course it would have been simpler just to pull the settings off the graph.

What I did originally with the NX550Ps was to take the Auto EQ settings and translate them to the PEQ.  Then I tuned the PEQ by ear, and that left the GEQ available to deal with the room.  The ART 310A's Auto EQ settings were too messy to translate readily to PEQ, and since those were being used as stage monitors I ended up just working with the settings in the GEQ.  So I guess I'm pretty ordinary in that respect. 

[Royce Covington]

kinda the long way around, but i see what you're trying to do now...

R~

[Brad Weber]

Maybe I'll run an impulse response on the speaker and work from that.

I agree that the AutoEQ seems to be using an RTA and graphic EQ approach.  However, for what you are trying to do that has several limitations, one of which is that since it will measure everything that hits the mic and thus you cannot remove the environment.

If you break it down you are often trying to tune a system for several factors including the response of a box or array, the effects of nearby surfaces on the response, the effects of other speakers, the effects of the room and subjective goals.  It sounds like what you a are trying to do is address just the first issue.  Shooting for a flat or known response eliminates the subjective aspect, limiting it to only one device eliminates the effects of other devices and a ground plane mic setup, etc. will help minimize or eliminate specific effects.  However, with an RTA based system you cannot avoid the effects of some nearby surfaces or of the room in general.  And measuring at just one location means you are then basing teh results on the effects of those surfaces and the room at that one point.  If you have not already done so you might want to read http://www.bennettprescott.com/downloads/devil_with_rta.pdf.

If you have Smaart, SysTune, SIM, Praxis, TEF or similar, then using a windowed Transfer Function would be much better, but as Mark noted, there are still limits on what can potentially be corrected with equalization, especially with a graphic EQ.
 

Royce, that is correct, I'm using it as a starting point. 

I may have given up a bit prematurely, as I just got a copy of the specs for the 310a's and the published frequency response is actually pretty close to what Auto EQ came up with.  Of course it would have been simpler just to pull the settings off the graph.

Do you mean that the stated response is close to the inverse of what the AutoEQ derived?  If you are compensating for the speaker then the EQ applied might resemble the inverse of the speaker response.

[Lynn Oliver]

Do you mean that the stated response is close to the inverse of what the AutoEQ derived?  If you are compensating for the speaker then the EQ applied might resemble the inverse of the speaker response.

Yes.  Accounting for the inversion, all three plots have the same basic shape (indoor Auto EQ, outdoor Auto EQ, manufacturer's specs).  Auto EQ tends to boost the mids and upper mids a bit more than the specs would indicate. 

I've learned a bit going through this exercise, but I've probably spent more time on it than the Auto EQ deserves. 

[Scott Bolt]

I have been using the DRPA for quite a few years. 

First, you need to make sure that your basic config is setup well.  The cross-over point(s), amp configuration, and time delay between your different drivers need to be configured.  There are rules of thumb you can use for the time alignment (for instance, my folded horns get ~8mSec delay to account for the distance traveled through the horn respective to the tops).

Second, there is a sensitivity setting on the DR's to allow a less precise eq.

Third, if your sub can't handle the lower frequencies (~30Hz) you will need to allow less precision for it to finish.

Fourth, make sure that the LP and HP are used on your sub.  Don't send signals to your sub that it can't reproduce.  You just end up working the amp and speaker harder and wasting headroom for power that you can actually use.

Fifth,  outdoors is a bass sponge. 

Sixth,  ensure that your sensitivity and input gains are set properly or you risk clipping or loss of utilized power.

I have seen the saw tooth effect mentioned above and have always manually corrected it.  It hadn't occured to me that it was comb filtering at work.  I will try doing a single stack at a time to see how good the results are.

My system is all passive with Cerwin Vega LR36's on the bottom and Klipsch KP301's on the top.

If I bypass the auto-eq, I find that my system does not sound as good as it does without it and it would require quite a bit of work at the instrument rack eq to get it there.

I only use 2 presets.  One for indoors, and one for outdoors.  I don't like using the auto-eq on a gig because of the disturbance to the venue while I do it.  I do fine adjustments with the rack graphic eq.

Most sound pros I have spoken to agree with the assessment of those here.  No "real" sound-man would use auto-eq.  Perhaps my talent doesn't extend as far as the "real" sound engineers   I kinda like the crutch.

As an aside, if you are using a matched pair of powered speakers, I would also agree that it is likely that the factory DSP is likely better than what you will be able to achieve with the DR; however, once you start mixing and matching, I am not sure this is still the case.

Many of these new powered speakers are DSP'd to sound good at low to mid volumes in a stand alone audition in order to attract buyers based on their "out of the box" sound of a CD.  They typically have the smiley face eq going on which usually needs some correction in order to sound better with a live PA setup over a sub.

[ProSoundWeb Community]