VST AutoEq....

[Paul Henry]

Has anyone tried the MathAudio Auto EQ VST?

http://mathaudio.com/

They claim:

Corrects acoustic imperfections of speakers, headphones and earphones.
Corrects deficiencies of room acoustics (multipoint compensation).
Corrects both amplitude and phase components of frequency response.
Quells resonance peaks of frequency response while leaving the deep notches to prevent large excursions of the speaker diaphragm. Avoids the muddy sound of conventional linearizing equalizers.
Contains built-in frequency responses of ideal speakers, headphones and earphones and uses them as references.
Allows drawing one's own reference frequency response.
Manually adjustable level of compensation allows one to reach the maximum transparency of the sound.
Perfect for use with standard measurement microphones. Even a cheap measurement microphone like Behringer ECM8000 ensures the accuracy of +/- 2 dB which is enough for professional musical applications.
A patented method of the frequency response correction ensures maximum accuracy of the compensation.

[Andreas]

To mee It seems, that this plugin has no own user Interface and uses a simple Interface of the DAW its used in.
Those Plugins do not run under SAC. (Tested with several mda plugs)
But, there is a free demo, you can test it yourself.

Best Wishes
Andreas

[Jeff Scott]

Looks like it runs in SAC. Didn't shut it down anyways...The user interface opens up when you start clicking on the "Make Correction File" button. Seems straight forward. $100.00 to buy. Maybe those who've got a lot of experience using Room EQ Wizard can do a comparison for us. Room EQ Wizard is free after all...

[KarelNoon]

That does not sound like a realtime application to me, so is it actualy doing something when you insert it?

[Paul Henry]

It runs a fast sweep tone one channel at a time and uses a calibrated mic to catch the freq response, it then calculates and inserts an eq curve that you can tweak...

Or so they say, haven't had a chance to try the demo yet.

[NoFear13X]

Very interested in the results from this (I literally came onto the SAC forums just to see if someone else had spotted this product). I would be very impressed if I could find a software AutoEQ for my rig that I could apply to multiple channels. I wouldn't even bother with an external speaker processor.

Any updates, has anyone tried this yet?

[NoFear13X]

Played with the demo for a bit just with my laptop and earphones, and the results were impressive, to say the least. I think I'll be purchasing the VST within the next few days and try it out on the rig. This is the exact plug-in that I've been looking for for years!

[Donnie Frank]

Please keep us informed. I'm going to keep a close eye on this thread. This sounds like something I would be very interested in, especially for those situations when someone else is providing racks-n-stacks or a room has horrible acoustics (which is most of the time).

Edit: Interesting is the Pickup Corrector. This would be a god-send for several of my clients who play acoustic guitars with piezo pickups.

[NoFear13X]

So I did some testing today after service. Installation was easy enough, typed in the registration once and it worked. (For $100, I figured I'd just buy it). My first thoughts were a bit worrisome. Things that you do and click on don't 'seem' like they're working, but they are. It just doesn't have any feedback, if that makes sense. It's like a dead-fish handshake. It counts, and it's uncomfortable. Again, everything 'did' work though.

The software is actually two programs. The first is a VST plug-in. Essentially, the only thing it does is apply the EQ curve. It takes nearly no processing power ( <1% draw for each instance ) and the only thing of relavence is the on/off switch and load a curve. Simple enough.

The second part of the software is a little more involved, and not immediately user-friendly. First up is to run a sweep through each speaker which is measured by the mic. This is COMPLETELY outside of the SAC software, which doesn't even need to be running. It uses window's default input and output devices for selection, so I set FOH to my main out and a pair of unused channels to my default recording input, hooked to a Behringer ECM8000 ref mic. Then it played a mindnumbingly loud low-to-high sweep. After turning the amp to about a third of the volume, I ran the test again, and it allows you to take additional measurements, so I ran it about 8 times in different locations in the room. The next step is to apply a preferred curve to the reading. I went with flat, for simplicity. Then it processes the curve and outputs a file with a curve in it tailored to give my speakers a flat sound -- typical AutoEQ stuff. Back in SAC, I loaded the curve in the VST and hit ON. Done, night and day difference.

Switching on and off made apparent differences in frequency response -- and volume. The software works by attentuating frequencies that are too loud, so my resulting curve was about 10db lower than without the plug-in, and no built-in way to boost the levels back up. I added 10db of gain to the levelizer plug-in before it in line and continued on my way.

Now when I tried the same thing on monitors, I had terrible static with the plug-in on both the monitors AND the house. Long story short, it didn't work with more than one instance. Solution? Copy and paste the VST in the plugin directory with different names. AutoEQ1, AutoEQ2, AutoEQ3, etc... restart SAC and load different instances where I need them, and problem solved.

My first instinct is that I'm going to really enjoy the plugin, but this Friday night is a youth rally, so that will be the real test of it's curve and results. My singers have been complaining of poor intelligibility on stage (spoiled brats, 5 years ago they didn't even OWN monitors), so we'll see if they notice a difference as well. If it matters, all the speakers are JBL JRX series. Nothing special, but not a bad entry level system either.

I'll post back after Friday's results, but so far, so good. Fingers crossed nothing blows up :rolleyes:

[Jeff Scott]

Great initial observations. I'll be watching for your job report...

[Donnie Frank]

Great initial observations. I'll be watching for your job report...

Ditto.

[gdougherty]

Do you get to see the EQ points it recommends so that you could set things outdoors, then transfer to an EQ other than theirs as a baseline EQ, allowing you to use another EQ for room adjustments where necessary?

[Guitarkeys.com]

Do you get to see the EQ points it recommends so that you could set things outdoors, then transfer to an EQ other than theirs as a baseline EQ, allowing you to use another EQ for room adjustments where necessary?

I've been waiting for the answer to this all day... Come on NoFear13, quit holding out on us.

:)

[gdougherty]

I've been waiting for the answer to this all day... Come on NoFear13, quit holding out on us.

:)

From the looks of the plugin, that answer is no, unless it's in some human readable format for the data file.

[Guitarkeys.com]

After reading the instuctions, it will display the results curve, so you could just keep applying EQ until you get the results curve to read flat.

Sounds like an afternoons work, but doable.

[NoFear13X]

Sorry for the delay, I've been out of town for 2 weeks.

The short answer some of you are looking for is that the EQ curve worked great. Multiple singers went out of their way to mention that the monitors and overall sound was much clearer and intelligible, and they were easily able to identify their own voices in a mix of 8 people as opposed to the semi-muffled EQ I had before. As it turns out, I also killed a horn in FOH during that initial test, so that kinda blows (some pun intended), but I'm looking forward to running the sweep again tomorrow with new compression drivers installed.

The software allows you to record the pink noise 'response', which you could either do outside for general cabinet EQing, or inside for room EQing, which applies to my case being a permanent install. Once you have that response, you can drop a line on top of it for how much attenuation you want to apply to get the response to be flat. That calibrated curve is what you would insert into the channel to make SAC output calibrated sound, if that all makes sense. So yes, you could save a preset from an outdoor test, and then add an additional instance for minor tweaks inside, or use a simple EQ on top of the calibrated curve for room changes. You can also drag points on the graph if you don't want a flat curve, if that helps for some of your cases.

@Guitarkeys, You only need to apply the EQ once to get a flat response. Additional instances would be overkill, and would probably start sounding audibly digital after enough layers of EQ. Currently, I'm running one instance of this EQ for each speaker, plus an instance of the SAC EQ as a high-pass filter. That's about it. Maybe I'll shoot a video or something to help clarify things, but for $100, I would strongly suggest just buying it. I would definitely purchase it again. Well worth the money for me.

[NoFear13X]

https://www.youtube.com/watch?v=GO4R__gYQgw

Sorry in advance if the video makes you dizzy. I'd also recommend fast forwarding the middle, lol. It's just a quick ride-along on how I use the AutoEQ function, it's far from an instruction video and is definitely not one of my proudest videography moments, but you get what you pay for, and this video is very much 'free'. :rolleyes:

[Yogi]

first off this EQ in not the normal type of EQ, it is a linear phase EQ. This from their website...
Corrects both amplitude and phase components of frequency response.
In the project I'm doing with Dorton arena (it's NC State fair week this week) I have just witnessed something I previously thought impossible. Martin Audio now has a system (they are careful not to label it a line array but that's exactly what it looks like) that is the best sounding system I've ever heard. It also is highly aimable (more so than what I've witnessed with most line arrays). It utilizes a VERY complex system of IIR and FIR filters (DSPs control each box in the line array) to produce a very accurate signal that is definable in 1 foot squares. From the front to the back of the floor of the arena I could not hear any changes in frequency response and signal level was within 2 db over a 200 X 15 foot throw area. Sound levels on the walls were a full 12 db below sound level on the floor. In short this is one amazing system, and I don't throw those types of terms around. The point is all that DSP power is using the IIR and FIR filters to do pahse manipulation on the signals. The result is uncanny. I just purchased this auto EQ and i hope it may give just a tiny bit of what that martin system does.
As far as using the curve from this and applying it to another EQ...if they are indeed doing phase compensation like they described they are doing applying that curve to anything other than their filter probably won't work. I will be doing some testing with impulse responses to see where their response line lines up with an impulse response.

[Jeff Scott]

Very, very Interesting Yogi. Both the Martin Audio description and your comments on the VST EQ plugin. Looking forward to seeing what you discover with the EQ.

[davidss1]

i would love to try this plugin, but have my speaker dsp controlled in sac utilising various plugins on the outputs. can anyone think of a way to do the initial testing through sac ?

[davidss1]

@ Yogi,
looking forward to seeing a before and after trace if u can post them.

[Yogi]

Here is how I plan to test it within SAC. My setup is not what you would term "normal" in any way. I have two motu 2408s and I run one bank on one as a loop around for outputs to the speakers so I can do quasi speaker management. The loop around allows me to have one output channel for what is fed to the mains and then I use a monitor mixer as the point to distribute that audio to 4 output channels. One is for subs, one is for installed speakers set above the stage, and two are for biamped speakers on each side of the stage. Each one of those feeds have been rung out with impulse response testing. The overall sound is pretty darned good. There are some areas that the above stage speakers and the side stage biamped speakers share coverage. The response is so close that unless you are really paying attention you can't tell which sound source you're listening to. That being said what I'd like to accomplish is better, cleaner signals. All of the speakers have an eq curve to bring them up to flat and then the overall main curve feeds all of them with an equal loudness coloration curve. I've tested this system extensively and it is pretty good. I have a great deal of volume before feedback in a relatively small room (60 X 60). Yesterday it was rocking at almost 100 db. What I have seen through constant monitoring is that at times I have phase problems with different types of signals. Most of these problems contribute to time smear which I'm hoping this vst can help me correct. I'm going to be doing the testing this week.

Testing setup will be through SAC. Since I have a loop around I already have the windows sound inputs and outputs set to channels of my motu. We can play CDs on the computer's DVD player (or any other material for that matter which can play in media player) through two channels of the loop around. I plan to set each of the speakers up on their own auto eq vst and measure them at 1 meter. Once they are all flat (or close to it) I'll then set up an auto eq vst on the main output. I also plan to eq the monitors (we have 4 monitors running as two stereo pairs) and we have wired IEM that I plan on eqing the earphones. For the individual speakers the 1 meter test should go quickly. For the main feed I plan on testing at least 20 locations throughout the room. For the stage monitors I'll probably do every location that someone is standing in.

[Jeff Scott]

Yogi: Interesting that you are looking to EQ the Earphones. How do you do measure that? Do you position the mic really close to the earphone and then place it in the middle of the room away from wall or ceiling reflections? Or do you contain the sound somehow like when you calibrate a Measurement Mic for SPL?

[Yogi]

The mathaudio site has a testing description. You more or less just wrap something around the microphone and the earplug to do it.

[soundchicken]

How could this be of any realistic relevance?

I can accept that you could "flatten" the sound coming out of the earpiece, but there are sooo many other factors that contribute to the sound of IEMs to the listener that the profile created by the autoeq could actually be hurting the sound quality. When setting up an IEM mix I always attempt to set a good "studio" mix and then allow the recipient(or do it for them) to adjust the over all eq'ing.

I'm really not trying to be an ass, I'm just curious about your reasoning since it's so far away from my SOP.

[Yogi]

Ben, rarely do we consider that an earphone is not close to flat. We normally just try and get a mix into the earphone that sounds good. Most times no one has considered a method to get the earphone "right" in the first place. This procedure does that and it only has to be done once for a pair of earphones so why not?

[Jeff Scott]

I actually did this with some cheap headphones I had in my studio before I purchased extra pairs of my Sony's. I got the idea from Dave Rat when he used an RTA to check out the frequency response of a whole bunch of different headphones.

It did make a difference being able to apply a little 7 band SAC EQ to the output channel of the phones.

[NoFear13X]

i would love to try this plugin, but have my speaker dsp controlled in sac utilising various plugins on the outputs. can anyone think of a way to do the initial testing through sac ?

I don't follow your question, why would you have a problem using it in sac? Just plug it in, not much to it... :confused:

[davidss1]

I don't follow your question, why would you have a problem using it in sac? Just plug it in, not much to it... :confused:

"The second part of the software is a little more involved, and not immediately user-friendly. First up is to run a sweep through each speaker which is measured by the mic. This is COMPLETELY outside of the SAC software, which doesn't even need to be running. It uses window's default input and output devices for selection"

as you stated, this is outside of sac ,,which is a problem for me when i am using plugins within sac to control bandpasses etc,,,or am i still missing something ?

[NoFear13X]

Oh, ok. I can definitely clarify that. Actually, did you watch the video? I know I wasn't recording the screen, pretty much at all (sorry about that), but combined with an explanation, it might make sense.

It's a 2 part software. The VST plugin invokes a second piece of software. That software, outside of SAC, runs the sweeps, tests, curves and final EQ plot, which is saved to a file. That test is ran on the same PC, but you don't need SAC open to do it. You just run that to get the EQ curve.

Then, you take the curve, and you import it into the VST plug-in back in SAC. Done. Ultimately, the curve affects your output channels of SAC just like any other VST, but you "build" the curve with the second piece of software, which is just a windows executable that it invokes.

[davidss1]

yes seen the video, and thanks for that and all other info here.
i am thinking i can do it using a second pc/laptop to feed the signal through sac on my host machine to preserve my xover'd outs, capture the response on the laptop, create the file then use a thumbdrive and transfer it and insert it in sac on my master group fader.
i was just hopeing for a solution to make it all happen on my host machine.

[NoFear13X]

Well I just bypassed my other plug-ins during testing. You could output the sweep to a random output and feed it back into SAC? But I don't think you want to do a sweep with X-Overs enabled otherwise it's going to try to boost the frequencies you've cut. I think you want to run the test flat.

[davidss1]

it shouldnt be a problem if all sections of the xover are open.
please clarify ,, i thought this program only utilised cuts to excessive frequencies above a defined "flatline" ? ( an exception would be using the draw function to boost)

[NoFear13X]

Yeah, I suppose that's true. I guess it wouldn't really affect it afterall. It's 3am, I should probably just reply tomorrow while I'm thinking clearly. :rolleyes:

[Andy Hamm]

yes seen the video, and thanks for that and all other info here.
i am thinking i can do it using a second pc/laptop to feed the signal through sac on my host machine to preserve my xover'd outs, capture the response on the laptop, create the file then use a thumbdrive and transfer it and insert it in sac on my master group fader.
i was just hopeing for a solution to make it all happen on my host machine.

If it uses the default windows audio settings, then it will come out of your default sound card. Plug that into a channel of your interface just as you would do to play music through the system and push up the fader. This also means that you need a mic pluggedinto your default audio interface to take the measurements, so most onboard won't cut it.

[Yogi]

The really easy way to do this is to pick an output channel and input channel that isn't being used if possible and loop it around. Then set the windows default OUTPUT to be that ASIO output channel. Pick any real SAC input channel and set it's input to the ASIO input channel you looped the card (raydat or motu or whatever) output into the card input. On windows default input pick any ASIO channel that you'd be plugging your calibration mike into. At this point you have the windows system output going out on an ASIO channel and coming back into an ASIO channel that is assigned to a SAC channel. You can now easily control the output volume through the input channel. Now on the SAC input channel you can select whatever SAC outputs you want to send a test signal to be it your mains, subs or monitors and you can save the result to a file to be loaded into the EQ vst for that SAC output channel. The latency added for the loop around is moot.

Remember, if you are using more than one iteration of the vst they must have unique names so you would have to copy and rename the plugin.

[davidss1]

The really easy way to do this is to pick an output channel and input channel that isn't being used if possible and loop it around. Then set the windows default OUTPUT to be that ASIO output channel. Pick any real SAC input channel and set it's input to the ASIO input channel you looped the card (raydat or motu or whatever) output into the card input. On windows default input pick any ASIO channel that you'd be plugging your calibration mike into. At this point you have the windows system output going out on an ASIO channel and coming back into an ASIO channel that is assigned to a SAC channel. You can now easily control the output volume through the input channel. Now on the SAC input channel you can select whatever SAC outputs you want to send a test signal to be it your mains, subs or monitors and you can save the result to a file to be loaded into the EQ vst for that SAC output channel. The latency added for the loop around is moot.

Remember, if you are using more than one iteration of the vst they must have unique names so you would have to copy and rename the plugin.

thanks Yogi, that works .

i have fooled around with system tunings using FFT for quite a few years now , and must admit i am humbled by what this plugin seems to do in mere seconds :o
the trial limitation (nasty 1khz tone every 20 seconds) is making it a bit annoying to test, but its looking like it would be a solid investment for $100

[Yogi]

I finally was able to do some testing with this VST last evening. NoFear, I'm not sure how you did it but no renaming of the vst could result in me being able to assign differing response files to different vsts. Example, I have two different types of speakers for mains. Two that are mounted from the ceiling and two that are on pole stands on each side of the stage. I'm also running a sub. I run a loop around from one output channel to an input channel and then I run that to 3 outputs to feed the three sets. I tried putting the auto eq vst on all 3 channels (each named differently). No matter what I did if I changed the loaded file for one they all changed to that loaded file. What I had hope to accomplish was each of the outputs would have a curve optimized for that particular speaker (measured at 1 meter). Then have an overall auto EQ that would compensate for the room itself. That's actually how I do it presently. I did wind up trying the auto eq in place of the overall eq and that worked somewhat better. One thing I discovered was less is more here. If you move the line down too far you wind up killing the dynamics and the sound winds up being muffled. What worked best for me was if there were dips in the middle I moved the line down such that only the peeks were affected and some dips were left unchanged.
That being said I still don't understand why with the vsts renamed that changing the input file for one changes it for all of them.

[NoFear13X]

Maybe I can clarify? I copy and pasted the file multiple times, and renamed the VST plugins in the VST folder to AutoEQ1, AutoEQ2, AutoEQ3. They then show up under the FX Window in SAC under the multiple names. I loaded 1 to FOH, and 2 and 3 to sets of monitors. Within each of those plugins, I loaded the appropriate curves. If that doesn't help, or work, I can video myself doing it again so there's no confusion, but that's about it. Worst case, I can use LogMeIn or something to check out your PC, but I think you can figure it out.

[Yogi]

Jesse, I did exactly the same thing. Each one was named something different however if I loaded a curve into any one of them all of them had the same curve.

[Yogi]

I had emailed MathAudio about the problem I was seeing and they have fixed it (they say). I'll know tonight. Supposedly you do not have to rename the dll to make it work with different correction files for each instance.

[davidss1]

I had emailed MathAudio about the problem I was seeing and they have fixed it (they say). I'll know tonight. Supposedly you do not have to rename the dll to make it work with different correction files for each instance.

Hi Yogi any update about the multiple instance problem?

[Yogi]

I'm still waiting on another fix. Something I didn't figure out until this afternoon. you can't use these in series. IOW you can't have one that corrects the speakers to flat and then another that corrects the room itself. Since I have two different sets of mains this presents a real problem for me. HOWEVER, I did put one into place for each set of speakers and did correction with the mike at 1M for each set. Then I used the parametric from reaper (reaeq) which I like along with the EASERA to set the room eq from impulse responses (which is sorta what the autoEQ does anyway). I have to say it sounds much better than what I'd done before (I had done a backup scene so I could A-B and listen to the results). If they can fix the series deal I can say it was definitely worth the 100 bucks.

[LarryBfaderJockey]

The really easy way to do this is to pick an output channel and input channel that isn't being used if possible and loop it around. Then set the windows default OUTPUT to be that ASIO output channel. Pick any real SAC input channel and set it's input to the ASIO input channel you looped the card (raydat or motu or whatever) output into the card input. On windows default input pick any ASIO channel that you'd be plugging your calibration mike into. At this point you have the windows system output going out on an ASIO channel and coming back into an ASIO channel that is assigned to a SAC channel. You can now easily control the output volume through the input channel. Now on the SAC input channel you can select whatever SAC outputs you want to send a test signal to be it your mains, subs or monitors and you can save the result to a file to be loaded into the EQ vst for that SAC output channel. The latency added for the loop around is moot.

Remember, if you are using more than one iteration of the vst they must have unique names so you would have to copy and rename the plugin.

I guess I am little stupid here. I am trying to get it to work in SAC but I can not get it to output the sweep and it tells me to increase the vloume and try agin (DUH!) I tried the winamp thing and after a couple of hours I finally got it to sort of work but after a couple sweeps it blue screens me with a memory error. Very consistent about it too. Maybe I need a "auto EQ guide for Dummies" Yogi, I am not sure what you mean. When you say to loop around is it a hardwire loop from an input to an output. I want to see how Auto EQ compares to the EQ wizard on my DR260 and SMAART. Thanks for any help. Looks like a cool plugin if I can get it to work. Would be nice to leave my SMAART laptop home for the small gigs.

[Tim A]

I have a dbx 260 that has the RTA feature that will flatten the speaker response. I never was real happy with the results. does any one think this VST autoeq would do a better job?

Tim....

[Yogi]

Well I've gotten it to work. Although it is a bit screwy in the way I did it. My hardware loop around is adat to adat (of course I have 6 banks to play with since I run motu 2408s). I have that set up so I can push one set of outs to an input channel that I use on a monitor mixer to drive my speakers by pushing that input to 4 output channels. It's my method of doing speaker management. with the auto eq I put it on an output channel that I'm not even using as output. That gives me a method of bringing up the tester screen. I can then select one output channel at a time and run the test on them. I decided to use the calibrated mike at about 10 feet from each output speaker set. YMMV. I ran each one saving the output response file. When I had all the response files I needed for each output channel I loaded up the autoeq into each output and loaded the response file. I was REALLY impressed. I did do some tweaking by running it again on one set of speakers but over all it did a really good job. Surprisingly my subs were really sweet. I run my subs up to 180 hz with a 96 db crossover. This thing made them thump, and really clean. After the combination of all the channels out I set an reaeq on the main feed channel ad did just some touch up tweaking by ear. Mostly wide Q and nothing was more than just a few db higher mostly in the high mids to give it some presence. We normally run right at 95 db at FOH on Sundays but yesterday I inched it up to about 98 and lots of people were raving after the worship time about how great it sounded. I have to admit I was pretty darned happy. It was very clean sounding for lack of a better way to describe it.

[davidss1]

hi Yogi. do u still have an issue running instances in series?

[Yogi]

For the moment but they are working on it. It still sounds darned good with how I set it up though. Since this is a hybrid type of filter and by what I'm told it is based on linear phase eq vs. minimal phase I see how it can work better. One of the things we tend to do when eq'ing a system is to over do it. Cook it until it's fried. What we wind up with is a curve that is hard to produce without artifacts. The one thing that amazed me was when I used a curve that was not overly aggressive. you can set the level at which the curve is established but I backed off of an initial curve and it sounded much better. The resultant curve was not exactly flat and I can verify that the overall response isn't flat but it sounds darned good, better than the original curve I had for those speakers that I did from my own impulse response. The best way i can describe it is that it seems to have more definition.

[Andy Hamm]

I'm extreamly skeptical and I never use these types of things but I paid the $100 to see if it does a better job than my ears. I have my doubts that this will work in a noisy room, as has been the case with every other system I've tried so far. I can't see blasting the test tones as being too practical after the opener either, but I'll try polishing some turds with it and see how it goes.

[Yogi]

Andy, do it on set up. Put the mike about 10 feet from your speaker stack (or whatever). Don't be too aggressive with the correction line. Oh, one other thing I found helps a great deal, ease up on compression (make sure compression is off when doing the testing). I was really pleased with the result.

[Andy Hamm]

Andy, do it on set up. Put the mike about 10 feet from your speaker stack (or whatever). Don't be too aggressive with the correction line. Oh, one other thing I found helps a great deal, ease up on compression (make sure compression is off when doing the testing). I was really pleased with the result.

Thanks Yogi, I'll give it a shot.

I do walk ins though so I usually don't have quiet time in the room without musicians banging drums or wanking on guitars.

I was planning on installing it on my FOH laptop to take the measurements and then load the correction file onto my mix rack. Not sure when I'll get a chance to test it, but I'll post back with results.

[Guitarkeys.com]

Well I took the plunge. I figure if Yogi was impressed then I would be too.

I haven't gotten a change to "use" it, but have played with it a little.

My first question is: Do you think the frequency responce graph is pretty accurate? In other words, is this also a tool for a quick and dirty "looK' at a speaker system?

Thanks,

Jamie

[Yogi]

Yep, I have two different kinds of speakers that we use for the mains. One is a sealed cabinet with a woofer and horn. We have these installed suspended from the ceiling. I have no idea where they came from or what brand they are (they preceded me). They work pretty well and sound decent so I haven't swapped them out. I built a pair of Jack 12s (Biamped) and mounted them on poles on each side of the stage for extra coverage and to use when we go mobile. The stage monitors (4 of them) are two different types of speakers so all told I have 4 distinct types. The response for each pair was different. I had done impulse responses on these over two years ago and had curves that I had put into EQs for each pair. The curve that auto eq came up with was different, not vastly, but enough that I could see the difference. Supposedly auto eq takes into account phase as much as freq and amplitude (making it closer to a linear phase eq than a minimal phase one). I know this, with 4 of them set up my cpu load went up 10% in SAC. The EQs I was using (reaeq from reaper) were almost 0 load. I am impressed with the way it sounds. I had done a backup of the previous setup in SAC and did some AB comparisons and had a few other people listen to it (blind testing) and every one of them picked the auto eq setup. The monitors sound the best they have ever sounded.

[Dan Fulton]

well I annoyed the neighbors this moring.... setup the mrx stacks 525's and 528's. xti 2000 on the tops xti 4000 on the subs tops cross over is done in the amps

One thing i noticed was there was a huge area that seem to be cut on the left where it wasn't on the right.

Could this be a sign of a phasing problem or is that normal. I can post the response files if needed

[Yogi]

Where was the mike set? Did you just do cabinet calibration? If so the mike should have been set up somewhere between 3 and 10 feet (1M to 3M). If you're doing a stack and you found this with the mick set up close to the cabinets then phasing was probably the issue.

[Dan Fulton]

this was a full stack doing a room calabration. I did an arch about 20 feet away from the stack

when i did a speaker single stack it was more close to the same for each stack. I stepped off ruff 10 feet from each stack.

One thing i have noticed when the autioeq is off you can here in the highs where the over lap is between the horns. then when it is that goes away.

So far I love it and will be purchasing the software soon. can't wait to try it on my moniters also.

the only thing is kinda wierd to me is there is a big cut from around 200 to 1k is that because of the double 15's and the system seeing to much mid bass?

[Yogi]

did you make sure there wasn't any EQ in between the output of the test signal and the input from the mike? One thing I did was use the auto EQ for speaker EQ rather than room EQ.

[Dan Fulton]

yeah on the first run i had the eq's on in the xti amps i turned them off for for the tests after that.

I am just wondering if my crossover points are set wrong.

I have the subs low pass at 35hz 24 then hi passed at 120hz 24
the tops are around 85 with 24 low pass

[Yogi]

That's REALLY nasty. something is really wrong there.

[Yogi]

looks like a really mean dip at 800 Hz that has a Q of 1.

[Dan Fulton]

you think of anything i can start with?
could it be a possible wiring issue of some type?

[Andy Hamm]

Do they sound like that?

[Dan Fulton]

that is the weird thing is it only thins out alittle bit when i turn the auto eq on and off but the clarity improves.

[airickess]

that is the weird thing is it only thins out alittle bit when i turn the auto eq on and off but the clarity improves.
The clarity improves when the EQ is on or off? If the clarity improves when it is on then you have some issues with your highs. Maybe even some phase cancellation going on with the stacks.
You should just measure a single speaker rather than the stack and see what type of measurement comes back.

[Dan Fulton]

it clears up on with the eq on. I consider my stack of One 525 Top and 528 sub

[airickess]

it clears up on with the eq on. I consider my stack of One 525 Top and 528 sub
Try a measurement with the 525 at full range and the sub off and see what type of measurement comes back.

[TomyN]

Hi,

This frequency-response indicates that there is something wrong in the crossover-area.
Try to invert the polarity on the subs and measure again.
Measure your cross-over to make sure that it's doing what you think it should do.

Tomy

[Dan Fulton]

think i have nailed it down the software isn't doing a full sweep of the system its' doing about 1k and up then 30hz to 80hz.

Just did a test with a powered moniter full range got the pretty much the same results

[davidss1]

dan posted ( in another thread) that a new version was out. does anyone know if it can now run instances in series?

[Dan Fulton]

I tried it in the previous version.... haven't tried in the new it seemed to work.

the only way i can see it working correctly would be if you can run the actual test thru sac which you can't do at this time. you have to use the standalone



dan posted ( in another thread) that a new version was out. does anyone know if it can now run instances in series?

[davidss1]

I tried it in the previous version.... haven't tried in the new it seemed to work.

the only way i can see it working correctly would be if you can run the actual test thru sac which you can't do at this time. you have to use the standalone

it is possible dan, if u look back to some of yogi's posts it explains how.
i have my test signal output looped back through sac in live mode so i can run on subgroups or master outs with bandpasses applied

[Dan Fulton]

it is possible dan, if u look back to some of yogi's posts it explains how.
i have my test signal output looped back through sac in live mode so i can run on subgroups or master outs with bandpasses applied
that is true Just wished it worked with correctly in the program... gonna keep emailing him to see what else i can get done

[davidss1]

hi guys,,i notice version 2.2.0 is out,,can u give a functionality update pls

[Andy Hamm]

hi guys,,i notice version 2.2.0 is out,,can u give a functionality update pls

Haven't tried the version I bought yet...

[Dan Fulton]

basically all that is new in the version is he turned on the dithering so that the behringers acted properly so you don't miss the first part of the sweep.

used it again last night sound was awesome... def helps with weird rooms. Last night the band was in a hole. and had to get the sound up to the second level. did about 8 to 10 different spots with the system and got a really even coverage with the system. better then i got with just my ears to start with. room was really hot in 200 hz range and also around 4k. It cut those down with the curve every thing just came alive. around a -6db cut to the curve.

Glad I bought it. I haven't tried it on the moniters yet but still figuring out how i want to do it. I always do the moniters in mono so I am thinking gonna have to run the eq the first time in hooked up in stereo then eq from there. Gotta play around with them again.

[Andy Hamm]

Just thought I'd report back as I've finally tested this thing out in the real world. I did a show last night that was local, and I used my mix rack with a PA that I am familiar with, so I decided to give it a shot.

I set up the VST on my mixrack, as well as on my laptop. I just used the laptop to take the measurements via my m-audio mobile pre and my Apex meansurement mic. The outputs of the mobile pre run into the mixrack via 2 channels and the mic is connected to the mobile pre.

I took several measurements, moved the mic around the room a bit, saved the correction file and then loaded it into the VST that was on the main outs of the mix rack. I would like to add that it sounds like the system does a full spectrum sweep, and then does the first part again up to about 100Hz or so on each channel, which I thought was strange, but the graphs looked reasonable based on wht I was hearing so I kept moving forward. The first time through, I lost one channel on my outs when I loaded up the correction file - even though it did appear as being there and at the same level during the measurement stage, so I had to start over again. This time I made several different files, with varying levels of correction and it loaded up on the mix rack and I had audio on both sides.

I then checked the level by disabling the plug in, and it was pretty close to the same level, so I left it as is.

How did it sound? I wasn't blown away by any stretch of the imagination. It pretty much robbed me of most of my subs, made the PA sound a little transparent through the low midrange and it did a reasonable job on the high mids and upper registers. I left the plug enabled and immediately ran into issues with my kick and floor tom sounding really choked and I had to insert another EQ on the mains after it to get it back to where I wanted it. I did as little as possible as I wanted to see how the VST performed.

Over-all, because of the lack of strength in the low mid region, I wasn't impressed. It sounded a little harsh to me, and it seemed to have no balls in the 200hz to 630 hz region. I also found that the system didn't respond as well to dynamics, when the PA got louder the harshness would rear it's head and a few times it actually got painful. Aside from that, I found it very fatiging to listen too.

Given the time that it takes to do the measurements, and the results that I got, I'm not likely to use this again, I feel that I can do a much better job myself and it takes me less time to get those results as well. I also compensate for changes in volume, where this system does an OK job of flattening things out at the level of the test.

If I was to make suggestions on how to make this product better they would be:

1. Make it work without being dependant on the default windows audio devices. You should be able to just insert this on the DAW channel with the measurement mic and get it to work. The VST could stop the mic signal from reaching the speakers. Having to use a seperate system to take measurements really sucks.

2. Make three bandpasses, like a crossover that the user can decide to place where they like. Have a mix option for these three bands, so as in my case where I didn't like what it did to my bottom end, I could pass the original signal for the range that I decide to pass untouched.

3. Add simple dynamic processing on each of these said bands with a threshold and reduction setting.

4. Have the VST display the correction file. Some sort of metering would be nice as well, as would a live RTA display.

[TomyN]

Hi,

this sounds like a common mistake in using measuremt gear for a PA System.
A typical PA system is not linear and should not be. In most cases there is a boost on the low-end (I've seen up to 20 dB) and a roll off in the high frequency range.
If you (or any software) tries to flatten this out, the result will sound like you describe.
As I understand the plug in one can add a target trace, so this would be the way to go.

Tomy

[Yogi]

Andy,
One thing I found was you CANNOT flatten the whole spectrum. I went back several times and the best sounding curves were the ones that had less than 10 db overall reduction below the highest peak. One of the other things you can do is modify the curve that it generates so that your bottom end isn't squashed. With most EQ setups you are plusing and minusing different freqs. Autoeq only minuses so you don't want to overdo it. Oh, one more thing, the more reverberant the room the closer to the speakers you need to be with measuring points. NOTHING can overcome a curve that is whacked out by excessive reverb. NOTHING.

[Andy Hamm]

Hey guys,

I agree with Tony that flat is not very desirable as far as a PA is concerned. I could make a curve for the Vst, but since I can just tune the PA by ear, I don't see much point in this. I've always been the guy that uses a track and a vocal mic to tune with and I'm familiar with a few different tuning systems - I purchased this just out of curiosity to see if maybe I was missing the boat of something.

I also didn't go crazy with the correction processing, I used the natural setting on a system that was fairly well balanced at the crossover, the graph showed about 4-5dB of attenuation at its peak.

I go for a sound that I describe as warm and rich and I like to have my definition articulated by the fundamental tones, not by the upper harmonics as I find that the top end ends up getting cluttered and it looses the intelligibility of the mix. I especially don't like my top end out running the rest of the rig as the volume goes up, as it makes things sound loud, brash and painful to me.

[RBIngraham]

To me it just sounds like this plugs ends up taking all the human... or thinking... part out of the job of EQing a speaker system. When a skilled operator uses an FFT or Transfer Function to tune the system, any of them worth their salt know to take any measurements on the low end with a grain of salt and ultimately to use their ears as the final judge. So if the tools say you need to boost or cut something by a ridiculous amount, your brain should kick in and say... nonsense... it'll sound better like this...

This is why all these Auto EQ things to me are just toys, or at least crutches for those that are still learning or too lazy to learn how to do it correctly.

In the end nothing beats a trained pair of ears using the source material that will eventually be pumped out of the sound system as a reference. And they can do their job a bit faster with some helpful tools, but still... Tools do not equal Talent.

[Yogi]

Understood Richard, but part of what AutoEQ does supposedly is to correct for phase too. That's why the result may not "catch" your ear quite the same.

[Andy Hamm]

To me it just sounds like this plugs ends up taking all the human... or thinking... part out of the job of EQing a speaker system. When a skilled operator uses an FFT or Transfer Function to tune the system, any of them worth their salt know to take any measurements on the low end with a grain of salt and ultimately to use their ears as the final judge. So if the tools say you need to boost or cut something by a ridiculous amount, your brain should kick in and say... nonsense... it'll sound better like this...

This is why all these Auto EQ things to me are just toys, or at least crutches for those that are still learning or too lazy to learn how to do it correctly.

In the end nothing beats a trained pair of ears using the source material that will eventually be pumped out of the sound system as a reference. And they can do their job a bit faster with some helpful tools, but still... Tools do not equal Talent.

Yep I agree 100%

[RBIngraham]

Understood Richard, but part of what AutoEQ does supposedly is to correct for phase too. That's why the result may not "catch" your ear quite the same.

Yes I understand. But even though a transfer function is a better representation than say an RTA or FFT, I'll still trust my own ears any day over what some screen tells me to do.

They are just tools. And to me this tool tries to take too much of the human interaction out of the process. At least from the brief looks I've taken.

I use tools like SMAART all the time. They save me a lot of time and trial and error. But I've watched many a so called "expert" show up to tune a system with TEF systems, SMAART, etc.. and then when they are done the system sounds like utter ****. Some didn't even take the time to bring up a mic into the system and hear what something akin to the actual source material going into the system would sound like.

On the other hand, at the other extreme, I had the pleasure of working with a guy that mixed Sinatra and many of the other Rat Pack types for many years on a one off benefit concert with Joel Gray. He only used a SM58, Joel's Wireless handheld and his own voice to tune the system. And you know what... that system sounded fantastic. Of course that was many years ago before a lot of the tools we have now were affordable. But even so... it's how I used to do it for years as well. Bring up a lav mic, make the system sound good with that.. because that's what going to go through it anyway. We're not doing movie sound here... or at least most of us are not. :)

Just saying that any of these tools are only as useful as the people the operate them and I would never give up the control to actually turn the dials on the parametric myself. Because I know I can still make better choices than some computer or a plug in can.

For those that need a cheap way to tune their systems, I recommend looking at Room EQ Wizard and learning how to use a parametric EQ.

[Brent Evans]

For those that need a cheap way to tune their systems, I recommend looking at Room EQ Wizard and learning how to use a parametric EQ.

+1. REW is a great tool when used as a measurement tool. There's lots of good info there to help locate problem spots.

The autoEQ funtion in REW is about as poor as any other i've seen.

[RBIngraham]

The autoEQ function in REW is about as poor as any other i've seen.

Needless to say, I never even looked at that part of the software. :D

About the only thing I trust software for more than my ears is determining delay times.

[airickess]

Needless to say, I never even looked at that part of the software. :D

About the only thing I trust software for more than my ears is determining delay times.Neither have I. I only use REW for speaker and room measurements. Almost all of my stuff is one-offs these days so I'm rarely using delays anyway, and if I do have some and am using SAC I'll rough in the delay times by way of a tape measure and picking the delay time from SAC's delay plug-in by reading the ft. measurement next to the delay time preset. It gets me close enough to perfect that I can't notice if it is off by a millisecond or two.

[RBIngraham]

Neither have I. I only use REW for speaker and room measurements. Almost all of my stuff is one-offs these days so I'm rarely using delays anyway, and if I do have some and am using SAC I'll rough in the delay times by way of a tape measure and picking the delay time from SAC's delay plug-in by reading the ft. measurement next to the delay time preset. It gets me close enough to perfect that I can't notice if it is off by a millisecond or two.

I've done plenty of it that way as well. Although I prefer to use one of those laser tape measure toys. Very cool stuff. :)

I almost always end up adding on a few ms to most delay systems to shift the image back to the stage/mains. I also, when I have time put a speaker out on stage and use that as a timing reference for all of the main system.

[Andy Hamm]

Just an update:

The software authors sent me a survey, and as per my request made me a copy of the measurement software that allows you to choose an asio source rather than the default audio interface.

I haven't tested it yet, but at least it's very promising behavior. :)

[Yogi]

do they plan on posting that one?

[PhaseShifter]

Just an update:

The software authors sent me a survey, and as per my request made me a copy of the measurement software that allows you to choose an asio source rather than the default audio interface.

I haven't tested it yet, but at least it's very promising behavior. :)

What kind of software company would care about how you, the customer wants to use the software?

Sounds pretty suspect if you ask me!

[Andy Hamm]

do they plan on posting that one?

I believe so, after it's finished.

It allows you to select inputs and outputs, other than that it looks the same. A great change though, because otherwise I needed a laptop with an interface to take measurements.