Auto EQ
“Cool all I have to do is push a button and everything will be awesome!�
Not so fast there …First lets take a look at a few facts.
1. You CANNOT eq a room! Only a bulldozer can that.. which if you have read between the lines means if you knock down the walls….you get the picture.
2. An RTA based measurement system only takes into consideration volume based frequency specific information. Another problem is that the sampling of these frequencies is finite. This means that with a sampling of only 28 frequencies for the entire audible frequency range (20 hz to 20,000 hz), that huge numbers are completely ignored. I won’t get into to much detail about what frequencies are sampled, but they are considered “ISO centers� this means that a consortium of “audio professionals� got together and decided what frequencies a 1/3 octave EQ (28-31 bands) uses for its center frequencies. ( your center frequencies are 31.5hz, 40,50,63,80,100,125,160,200,250,315,400,500,630,800,1k,1.25k,1.5k,2k,3.15k,4k,5k,6.3k,8k,10k,12.5k, and 16k for the DRPA and 260… the 240, 400, and 4000 series have 31 bands…) this leaves HUGE gaps (for instance, 4K to 5K is 1000 hz! And EVERYTHING between 4000 and 5000hz is IGNORED completely!!!), and if you noticed the higher the frequency the larger the gaps…Instead of 1/3 octave the sampling rate should be more like 1/128th octave, or more (note…future generations will look back at our quaint sampling rates.. even 1/128 octave and snicker at the simplicity of our understanding of the technologies…) these huge blocks of ignored audio could hide comb filtering (if you don’t know what this is you need to do a Google, or other search to find out what it means) nulls, or modes that could and will adversely affect your sound reinforcement system in a room.
3. Mic placement is critical! Problem, mic placed ½ way between stacks, a symmetric room with lots of reflective surfaces and even number squared room dimensions (40 X 60, 60 x 80, 60 x 120… you get the picture.. not 55 x 73, 77 x 135…etc but that’s a topic for another diatribe.. architectural design for acoustic concerns.. ) lets look at how a measurement mic “seesâ€? audio. A measurement mic needs to be very very flat from 20 hz to 20,000 hz, and omni directional (all directional) (no an SM 58 won’t work in a pinch, neither will a $4000 Neumann U87) cheaper mics like the Behringer 8000 series suffer from inconsistencies from mic to mic (especially in the low and high end), the RTA “Mâ€? mic is the DBX offering, but I would like to suggest that the TR 40 by audix is the least expensive “decentâ€? measurement mic for professional audio measurement ( what I’m getting at is that the RTA method ISN'T a professional measurement method, this means that for our purpose the RTA “Mâ€? may be the most cost effective solution at roughly ½ the cost of the TR-40) you can spend THOUSANDS more… easily, on the mic alone. No matter where you place the mic in a room you will encounter problems…These problems are associated with the effects of reflected sound off surfaces, and will manifest themselves in the form of standing waves (these multiply and strengthen bass frequencies and usually end up resulting in the cutting of sub frequencies), nulls, the coupling and cancellations of out of phase audio waves, and reflected cancellations (usually manifested in the form of boosts of near 12db for the canceled frequencies) (this same scenario will happen at the FOH position, or any where else in the room for that matter) trouble is, move the mic a few inches and the problem changes frequencies. Was there really a cancellation where the mic was.. YES! But we seldom sit still enough to realize there is a missing frequency in a frequency rich environment, and that frequency only occurs randomly and we cannot process the change quickly enough. The mic however.. samples all frequencies at the same time, and the Auto EQ uses analytical precision to determine what is, and ISN’T there…thus the Auto EQ senses a missing band of frequencies and goes “huntingâ€? for it using that frequencies gain control! If it’s missing(more than 12 db down the Auto EQ will fully boost that frequency if it cannot get enough of it… even though it’s KILLING you in the room (read that WAY too loud). So the Bass is totally missing, and the High frequencies are tearing your head off.. sound familiar? If you’ve done an auto EQ in a room with the mic in the center of the room… that will happen MOST of the time…add to that “pre curvesâ€? and you have a formula for disaster!
4. The AUTO EQ process isn’t meant for man NOR beast… It CAN hurt you if done at the volume necessary for concert volume Auto EQ. This means if there are ANY people in the room during the “SONIC BLAST� they are likely to hunt you down and hurt you! DO NOT EVEN ATTEMPT THIS in a room with people in it! So can I Auto EQ every venue and store the curves for future use…NO. BUT...If you go in with a completely flat system response, walk the room with some pre-recorded music you know and trust, and note the sonic problems (this takes time and experience) you will eventually be able to tame the system and save it’s product for the next time you play there which will require some tweaking, especially if it’s with a different band.
5. Is the Auto EQ the best method of tuning a system to the room? NO! ANY EQ done after the system is flattened and perhaps seasoned to taste should be done by ear( this means you perhaps accentuate certain frequencies to enhance your listening experience.. we do after all hear our own way, and one mans nirvana may be another’s sonic hell! Ok, so I’ve stated that you cannot EQ a room, eq’ing the speakers is a tradeoff at best, what CAN I do to actually help the problems of playing in a room.
a. Keep the sound off the walls and floors/ceilings and on the audience.
b. Use creative speaker placement to minimize room interaction.( remember that high frequencies bounce off hard surfaces and can cause cancellations, and low bass notes reflect off walls and can couple and both cause BAD sound! Sorry gang..no “one button nirvana here� By the way�…nowhere else either. You must learn, work, and then you can effective.
c. Consider limiting the system’s volume so it does NOT inject destructive volume levels into the room.. EVERY room has a point of diminishing returns where volume is concerned, I don’t care how well your system is tuned… there is a breaking point, (you can only retune for so many possibilities, and we as sound technicians owe it to the people to “do no harm� …as much as possible. Exceeding the rooms volume threshold WILL degrade the quality of the sound. The more volume the more…DISTORTION, CANCELLATIONS, MODES, NULLS, COMB FILTERING, and more really bad things.
d. Using an ISO based non adjustable center frequency filter like a GEQ is like carpet bombing when a sniper rifle is all that is needed. This means that when you cut a frequency with one of the faders you are also cutting a bunch below AND above the center frequency. And if the frequency falls somewhere in between fader (which is all too often the case) you need two faders and all the frequencies above and below are affected!..NOT good!
Review:
1. Room auto EQ using RTA based measurement system isn’t usable.
FFT based systems take into consideration time, coherence, and frequency specific energy to make REAL sonic improvements possible
2. The Auto EQ process is not usable when people are present.
3. A measurement specific super flat response omni directional mic is required
4. The placement of the mic creates problems no matter where you place it.
5. You CANNOT EQ a room!
6. The frequency resolution of a 28-31 band GEQ is insufficient to be adequate for the purpose of system equalization. GEQ’s cover too many adjacent frequency’s when utilized. With he driverack we do have a tool available to us that is much better for this purpose, that is the PEQ. (parametric eq) and it can be used instead of the GEQ… the Auto EQ function does NOT work with the PEQ however.. There are 9 bands but consider this, one band can be used to cover the whole spectrum from 20 to 20,000 hz. Now I’m not suggesting you do this but you can. A peq is a powerful tool, but beyond the scope of this article.
For a really good example of what SMAART does, and looks like see Mike Kovachs \"white paper\":
http://www.driverack.com/drug/view_thre ... d_id=41581
It will also help you to understand why we say let the Auto EQ do only 170-700hz frequencies and tune the rest by ear.
So what CAN we do with the auto EQ?
I suspect you’re sitting there thinking what good is this thing anyway? Well I asked myself the same thing. First off Mike Kovach ran side by side tests with a SMAART rig (FFT based measurement system we talked about earlier) and found that within the limited scope of the of the Auto EQ it was fairly accurate.
If you have been reading up on this site you know that we have discussed taking the system (one stack of speakers) outdoors away from all reflective surfaces and with the mic 8-10 feet out aimed at the space between the low and high components of the top cabinets, selecting the -0- precurve from the Auto EQ menu and establishing a baseline flat preset, that we can then store and recall for a number of purposes. ( note, here would be a good time to also set up a couple of the other precurves for later recall. The “C� curve was the one most preferred but I can see a potential use for the –A- and possibly the B, and C curves as well. ) The Idea hers is to make a pass and then view the product of the Auto EQ pass and then use the post (after) crossover PEQ’s to accomplish the same thing the GEQ did and then clear the GEQ and do another pass with the Auto EQ .. again use the remaining peq’s to get closer to a flat baseline until the product Auto EQ is essentially flat. (remember that speakers react differently the closer they get to their theoretical max output so try and test at your normal concert volume… the DRPA has some limitations here so work within these parameters)This we will store to a preset location (of your choosing, I usually choose a preset I will never use but the preset you replace is always still available in a hard reset to factory default …see your manual) with three presses of the store button where you will be asked to name the preset, select a location for it, and verify that you really want to replace the existing preset. This also presumes that you have read and done the steps in the Welcome “read me first� and start here threads…
Note...If you are like me, living in a frozen white wasteland much of the year, or living in a concrete jungle and cannot use the "outdoor method" a modified procedure where we place the mic close enough to hear the speaker far louder than the reflections and yet far enough away to make sure all the elements of the speaker are represented in the product. Depending on the size of the speaker like a small monitor versus a dual 15" with horn...a distance of 4-6 feet should do nicely and sitting on something off the floor... like a speaker stand... or something like a stool or something for a bigger speaker you can't get on a stand...to minimize floor interaction...
What we have now is a system that outputs all frequencies equally (hopefully.. and of course within the limitations of the Auto EQ system). No hype to any frequencies, no lost frequencies that need to be boosted in order to correct for some system deficiency. Now, IF all frequencies are equal any vocal or instrument that sounds good in and of itself should sound essentially the same (or at least reasonably close if the system has quality components) when reinforced. You may want to boost the 80 hz (first harmonic of the kick drum, the actual center frequency is closer to 300-400hz) , and guitars are @ about 2k, vocal clarity 3.5K and if cymbals are piercing, lower the 7-12 k region. I have even sweetened a preset or two, set up a heavy metal, blues and light rock presets and stored those for recall since the more options you have when you’re under the gun the better chance you have of having a quick preset that is going to work for your situation.
The stored flat preset can also be used to evaluate what a room is doing to your sound… when we can actually DO an Auto EQ that is. I like to take the measurement mic and aim it at the floor or ceiling in the room. This eliminates a lot of potential reflections that will color the results, The mic should be no farther away from a surface than about ¼ of an inch, and aimed directly at the surface at 90 degrees to the surface. This then becomes a pressure zone microphone (PZM). Where do you put the mic? Well.. that’s a loaded question. The best thing is to take multiple measurements in varying locations and average the results. This can be somewhat challenging and so will interpreting the results of the Auto EQ passes. If your THAT interested in the measurement of your system may I again suggest you get into FFT measurement systems like SMAART, which, once you get up to speed can REALLY allow you to tune your system to it’s full potential. An adder bonus is the that SMAART is capable of working in conjunction with the Driverack system (not the DRPA though.. sorry)
Please see \"the new indoor Auto EQ method\" for some further thoughts and techniques that work.
Not so fast there …First lets take a look at a few facts.
1. You CANNOT eq a room! Only a bulldozer can that.. which if you have read between the lines means if you knock down the walls….you get the picture.
2. An RTA based measurement system only takes into consideration volume based frequency specific information. Another problem is that the sampling of these frequencies is finite. This means that with a sampling of only 28 frequencies for the entire audible frequency range (20 hz to 20,000 hz), that huge numbers are completely ignored. I won’t get into to much detail about what frequencies are sampled, but they are considered “ISO centers� this means that a consortium of “audio professionals� got together and decided what frequencies a 1/3 octave EQ (28-31 bands) uses for its center frequencies. ( your center frequencies are 31.5hz, 40,50,63,80,100,125,160,200,250,315,400,500,630,800,1k,1.25k,1.5k,2k,3.15k,4k,5k,6.3k,8k,10k,12.5k, and 16k for the DRPA and 260… the 240, 400, and 4000 series have 31 bands…) this leaves HUGE gaps (for instance, 4K to 5K is 1000 hz! And EVERYTHING between 4000 and 5000hz is IGNORED completely!!!), and if you noticed the higher the frequency the larger the gaps…Instead of 1/3 octave the sampling rate should be more like 1/128th octave, or more (note…future generations will look back at our quaint sampling rates.. even 1/128 octave and snicker at the simplicity of our understanding of the technologies…) these huge blocks of ignored audio could hide comb filtering (if you don’t know what this is you need to do a Google, or other search to find out what it means) nulls, or modes that could and will adversely affect your sound reinforcement system in a room.
3. Mic placement is critical! Problem, mic placed ½ way between stacks, a symmetric room with lots of reflective surfaces and even number squared room dimensions (40 X 60, 60 x 80, 60 x 120… you get the picture.. not 55 x 73, 77 x 135…etc but that’s a topic for another diatribe.. architectural design for acoustic concerns.. ) lets look at how a measurement mic “seesâ€? audio. A measurement mic needs to be very very flat from 20 hz to 20,000 hz, and omni directional (all directional) (no an SM 58 won’t work in a pinch, neither will a $4000 Neumann U87) cheaper mics like the Behringer 8000 series suffer from inconsistencies from mic to mic (especially in the low and high end), the RTA “Mâ€? mic is the DBX offering, but I would like to suggest that the TR 40 by audix is the least expensive “decentâ€? measurement mic for professional audio measurement ( what I’m getting at is that the RTA method ISN'T a professional measurement method, this means that for our purpose the RTA “Mâ€? may be the most cost effective solution at roughly ½ the cost of the TR-40) you can spend THOUSANDS more… easily, on the mic alone. No matter where you place the mic in a room you will encounter problems…These problems are associated with the effects of reflected sound off surfaces, and will manifest themselves in the form of standing waves (these multiply and strengthen bass frequencies and usually end up resulting in the cutting of sub frequencies), nulls, the coupling and cancellations of out of phase audio waves, and reflected cancellations (usually manifested in the form of boosts of near 12db for the canceled frequencies) (this same scenario will happen at the FOH position, or any where else in the room for that matter) trouble is, move the mic a few inches and the problem changes frequencies. Was there really a cancellation where the mic was.. YES! But we seldom sit still enough to realize there is a missing frequency in a frequency rich environment, and that frequency only occurs randomly and we cannot process the change quickly enough. The mic however.. samples all frequencies at the same time, and the Auto EQ uses analytical precision to determine what is, and ISN’T there…thus the Auto EQ senses a missing band of frequencies and goes “huntingâ€? for it using that frequencies gain control! If it’s missing(more than 12 db down the Auto EQ will fully boost that frequency if it cannot get enough of it… even though it’s KILLING you in the room (read that WAY too loud). So the Bass is totally missing, and the High frequencies are tearing your head off.. sound familiar? If you’ve done an auto EQ in a room with the mic in the center of the room… that will happen MOST of the time…add to that “pre curvesâ€? and you have a formula for disaster!
4. The AUTO EQ process isn’t meant for man NOR beast… It CAN hurt you if done at the volume necessary for concert volume Auto EQ. This means if there are ANY people in the room during the “SONIC BLAST� they are likely to hunt you down and hurt you! DO NOT EVEN ATTEMPT THIS in a room with people in it! So can I Auto EQ every venue and store the curves for future use…NO. BUT...If you go in with a completely flat system response, walk the room with some pre-recorded music you know and trust, and note the sonic problems (this takes time and experience) you will eventually be able to tame the system and save it’s product for the next time you play there which will require some tweaking, especially if it’s with a different band.
5. Is the Auto EQ the best method of tuning a system to the room? NO! ANY EQ done after the system is flattened and perhaps seasoned to taste should be done by ear( this means you perhaps accentuate certain frequencies to enhance your listening experience.. we do after all hear our own way, and one mans nirvana may be another’s sonic hell! Ok, so I’ve stated that you cannot EQ a room, eq’ing the speakers is a tradeoff at best, what CAN I do to actually help the problems of playing in a room.
a. Keep the sound off the walls and floors/ceilings and on the audience.
b. Use creative speaker placement to minimize room interaction.( remember that high frequencies bounce off hard surfaces and can cause cancellations, and low bass notes reflect off walls and can couple and both cause BAD sound! Sorry gang..no “one button nirvana here� By the way�…nowhere else either. You must learn, work, and then you can effective.
c. Consider limiting the system’s volume so it does NOT inject destructive volume levels into the room.. EVERY room has a point of diminishing returns where volume is concerned, I don’t care how well your system is tuned… there is a breaking point, (you can only retune for so many possibilities, and we as sound technicians owe it to the people to “do no harm� …as much as possible. Exceeding the rooms volume threshold WILL degrade the quality of the sound. The more volume the more…DISTORTION, CANCELLATIONS, MODES, NULLS, COMB FILTERING, and more really bad things.
d. Using an ISO based non adjustable center frequency filter like a GEQ is like carpet bombing when a sniper rifle is all that is needed. This means that when you cut a frequency with one of the faders you are also cutting a bunch below AND above the center frequency. And if the frequency falls somewhere in between fader (which is all too often the case) you need two faders and all the frequencies above and below are affected!..NOT good!
Review:
1. Room auto EQ using RTA based measurement system isn’t usable.
FFT based systems take into consideration time, coherence, and frequency specific energy to make REAL sonic improvements possible
2. The Auto EQ process is not usable when people are present.
3. A measurement specific super flat response omni directional mic is required
4. The placement of the mic creates problems no matter where you place it.
5. You CANNOT EQ a room!
6. The frequency resolution of a 28-31 band GEQ is insufficient to be adequate for the purpose of system equalization. GEQ’s cover too many adjacent frequency’s when utilized. With he driverack we do have a tool available to us that is much better for this purpose, that is the PEQ. (parametric eq) and it can be used instead of the GEQ… the Auto EQ function does NOT work with the PEQ however.. There are 9 bands but consider this, one band can be used to cover the whole spectrum from 20 to 20,000 hz. Now I’m not suggesting you do this but you can. A peq is a powerful tool, but beyond the scope of this article.
For a really good example of what SMAART does, and looks like see Mike Kovachs \"white paper\":
http://www.driverack.com/drug/view_thre ... d_id=41581
It will also help you to understand why we say let the Auto EQ do only 170-700hz frequencies and tune the rest by ear.
So what CAN we do with the auto EQ?
I suspect you’re sitting there thinking what good is this thing anyway? Well I asked myself the same thing. First off Mike Kovach ran side by side tests with a SMAART rig (FFT based measurement system we talked about earlier) and found that within the limited scope of the of the Auto EQ it was fairly accurate.
If you have been reading up on this site you know that we have discussed taking the system (one stack of speakers) outdoors away from all reflective surfaces and with the mic 8-10 feet out aimed at the space between the low and high components of the top cabinets, selecting the -0- precurve from the Auto EQ menu and establishing a baseline flat preset, that we can then store and recall for a number of purposes. ( note, here would be a good time to also set up a couple of the other precurves for later recall. The “C� curve was the one most preferred but I can see a potential use for the –A- and possibly the B, and C curves as well. ) The Idea hers is to make a pass and then view the product of the Auto EQ pass and then use the post (after) crossover PEQ’s to accomplish the same thing the GEQ did and then clear the GEQ and do another pass with the Auto EQ .. again use the remaining peq’s to get closer to a flat baseline until the product Auto EQ is essentially flat. (remember that speakers react differently the closer they get to their theoretical max output so try and test at your normal concert volume… the DRPA has some limitations here so work within these parameters)This we will store to a preset location (of your choosing, I usually choose a preset I will never use but the preset you replace is always still available in a hard reset to factory default …see your manual) with three presses of the store button where you will be asked to name the preset, select a location for it, and verify that you really want to replace the existing preset. This also presumes that you have read and done the steps in the Welcome “read me first� and start here threads…
Note...If you are like me, living in a frozen white wasteland much of the year, or living in a concrete jungle and cannot use the "outdoor method" a modified procedure where we place the mic close enough to hear the speaker far louder than the reflections and yet far enough away to make sure all the elements of the speaker are represented in the product. Depending on the size of the speaker like a small monitor versus a dual 15" with horn...a distance of 4-6 feet should do nicely and sitting on something off the floor... like a speaker stand... or something like a stool or something for a bigger speaker you can't get on a stand...to minimize floor interaction...
What we have now is a system that outputs all frequencies equally (hopefully.. and of course within the limitations of the Auto EQ system). No hype to any frequencies, no lost frequencies that need to be boosted in order to correct for some system deficiency. Now, IF all frequencies are equal any vocal or instrument that sounds good in and of itself should sound essentially the same (or at least reasonably close if the system has quality components) when reinforced. You may want to boost the 80 hz (first harmonic of the kick drum, the actual center frequency is closer to 300-400hz) , and guitars are @ about 2k, vocal clarity 3.5K and if cymbals are piercing, lower the 7-12 k region. I have even sweetened a preset or two, set up a heavy metal, blues and light rock presets and stored those for recall since the more options you have when you’re under the gun the better chance you have of having a quick preset that is going to work for your situation.
The stored flat preset can also be used to evaluate what a room is doing to your sound… when we can actually DO an Auto EQ that is. I like to take the measurement mic and aim it at the floor or ceiling in the room. This eliminates a lot of potential reflections that will color the results, The mic should be no farther away from a surface than about ¼ of an inch, and aimed directly at the surface at 90 degrees to the surface. This then becomes a pressure zone microphone (PZM). Where do you put the mic? Well.. that’s a loaded question. The best thing is to take multiple measurements in varying locations and average the results. This can be somewhat challenging and so will interpreting the results of the Auto EQ passes. If your THAT interested in the measurement of your system may I again suggest you get into FFT measurement systems like SMAART, which, once you get up to speed can REALLY allow you to tune your system to it’s full potential. An adder bonus is the that SMAART is capable of working in conjunction with the Driverack system (not the DRPA though.. sorry)
Please see \"the new indoor Auto EQ method\" for some further thoughts and techniques that work.
0