Let me start by saying that I'm a pro audio guy more than an audiophile. I have no philosophical aversion to EQ. If you do, you might as well stop reading, if you believe EQ is cheating and/or that it can't possibly ever be good.
My goal with headphones has always been to get as good of a sounding rig as possible, as cheaply as possible, regardless of how that was to be achieved. I know that frequency response isn't the be all and end all to headphones. but the more I read head-fi, I realized that while there are these other concepts to a headphone's sound, they almost never get talked about. 99% of what is talked about on here is almost completely based on a headphones frequency response curve. Even when people think they're talking about something else, like soundstage depth, most of the time they are actually talking about frequency response.
This was further enhanced by the fact that 99% of people's (and my) complaints about the AD700 had to do with frequency response. The bass is too weak to be neutral. There is a noticeable drop off at around 3800 Hz. 1000 Hz is too hot. Yes, in many ways these 3 things give the AD700 its character as well. Part of the amazing depth of the soundstage comes precisely from that 4 kHz dip. the 1000 Hz bump is the main reason the mids sound so prominent. The lack of bass is a major reason they are so crystal clear. Sound, and especially headphones, are about tradeoffs, and the AD700 made its choices. That being said, I think the fundamentals of the headphone, notably its transient response, lack of THD, comfort, amazing imaging, low cost, ease to drive them and extension high and low made them great candidates to just "fix with EQ."
So first, let me talk about source. It's a hotaudio hotusb1 external DAC -> 3.5mm to XLR -> Samson D-2500 EQ -> Millett SSHM. Even though I don't think there's an audible difference between well encoded, high bitrate mp3 and lossless, just to avoid this argument in a situation I don't want to have it, everything I'm using here is FLAC. I will also try to use very popular tracks as much as possible, for the simple reason to aid people in having a clue as to what I am talking about.
My first step was to EQ it to completely flat. I used Equal loudness contours and audiometry - Test your own hearing to generate a flat equal loudness curve. What I did was simply messed with the EQ until it was totally flat at an average listening volume. That is, going straight across all tones sounded the same loudness. The Samson allows you to save settings, so I saved this as "AD700 flat". My next step was to EQ it how I like it. I like a touch more midbass than totally flat and a tiny bit more 6k-14k than flat. Nothing drastic, but the additional bass makes up for the lack of impact that headphones have naturally, a little. And the 6k-14k boost makes up for a tiny bit of high frequency drop off in my ears. that bump is VERY subtle, around .5 dB at its peak point. I saved this as "AD700". Since the difference between the AD700 Flat and AD700 settings were virtually identical, I will only use the AD700 setting for this review. Also, when I refer to "flat" in the review, I mean with no EQ, not the Flat setting. This could be confusing, so I am trying to clear that up here. So if I say AD700 flat below, I mean unEQ'd. It's "natural" sound.
Essentially with the AD700 setting on my EQ, I had 64 Hz boosted to 6dB, with a plateau there down to about 44Hz, after which I gradually rolled it off to 0 dB at 20 hZ (to avoid clipping and slowing the headphone down unnecessarily for extremely low tones, with infinite current I would have left this boosted though) From 64 Hz to 500 Hz I gradually backed it down, trying to make as smooth of a curve as possible, to 500 Hz it was at 0 dB. from 500 Hz to 700Hz I maintained a very slight taper downward and then a sharper curve down to -2 dB at 1kHz. Then slowly back up until its back at 0 dB at 2kHz. Then a pretty sharp rise to 3500 Hz where I had it at +6 dB, then a small plateau to 4200 Hz, then a sharp fall back down to +1.5 dB at 8kHz, then back to 0 dB from that point out.
Now to the meat of this, the comparisons:
1) The first thing is just a much greater sense of tonal balance with both the adjusted EQ settings. AC/DC's Back in Black now hits correctly in the opening. Whereas before it felt "neutered" and off. You can tell how poorly the hi-hat was recorded. Angus Young's guitar is in your face, just as it should be. The kick drum is both powerful from the added bass, and in your face, with the 4 kHz boost, just like it should be. This track is not supposed to sound laid back, but with no EQ on the AD700, it almost does. Things that are forward and aggressive with the drums and rhythm guitar and riffs sound laid back and flat with flat AD700's. Al Green's voice on Let's Stay together now sounds full and powerful instead of overly nasaly and weak. The thunder on Dire Straits' Brothers in Arms sounds powerful and real instead of like a cheap sound effect. But more than anything, the EQ'd version is indescribably more balanced. Everything just sounds more right and natural.
2) Soundstage. By now, some are probably wondering "the AD700's soundstage is more amazing, this guy is lying if he says its improved." Yes and no. It's not as deep when EQ'd and not as wide really either. BUt only slightly less so. I say it is improved slightly, because I believe that it's more "natural" now. Before I felt that the soundstage on the AD700 sounded artificially big. Miles Davis's So What shouldn't sound like they're playing 50 feet apart. But flat, it did. It's a small group that's supposed to sound like they're playing on a medium small stage. Before it sounded like you were about 20 rows back watching 5 guys play on the superbowl half time stage. Now it sounds like you are sitting in the 3rd table back at a medium sized jazz club. The imaging is so accurate that I can tell exactly where the microphones are placed to record the drums. I can actually hear the horn players move their horns around towards and away from the mic. The bass is there in the right amount. Before you almost had to struggle to hear it clearly, as its mastered low, like small group jazz is. The lack of bass and the forwardness of the horns previously pushed it almost out of the recording. Now its always there holding everything together. Whereas before the soundstage was wide and accurate, now its just right and hyper accurate. Classical recordings may suffer a tiny bit, but only because many classical recordings struggle with getting the soundstage wide enough in the recording. This now reproduces that trouble. If I was listening to classical, I'd be torn as to what setting to use.
3) Speed: Now let me say that the headphone's speed suffers a tiny, tiny amount EQ'd. Basically, with the added bass it seems that its now having to work harder. With an amp that delivered more current than my SSHM this might not be an issue. Muse's Hysteria sounds much more right with EQ, but unEQ'd the bass is ever slightly more crisp on the hard transients. It's very slight though and definitely not worth the tradeoff that the song otherwise sounds totally neutered without the EQ. When there is a ton of deep bass and hard transients, the headphone has a slightly hard time keeping up. But it's a $90 headphone, not an HD800. I don't think it's particularly noticeable unless you're looking for it and picking music that specifically challenges this aspect. To be clear, the unEQ'd version still has this issue, its just not as much of a problem, because it isn't "weighed down" by the demands of pumping out the bass. It kind of points out that most of what is viewed as "fast" headphones are also bass light. It takes a tremendous amount of power to adequately provide correct amounts of bass and maintain the hard transient edge.
Overall, I'd say its night and day better EQ'd. I think that everybody who isn't philosophically opposed to EQ should buy a good graphic EQ (at least 16 bands, but 32 is better, mine is 31 bands). Even better is a really, really good parametric equalizer, however they get really expensive really fast, and if you have enough bands on your EQ, it's good enough to change the overall things you want to solve with headphones. If nothing else you will learn exactly what frequencies effect sound and exactly how they do it. Also, you will want to look for one with a master gain level, so that you are free to EQ up and down without worry of clipping. If you get clipping, you can just pull the master gain down.
Anyway, this was really long, but hopefully someone found it interesting and/or helpful.
My goal with headphones has always been to get as good of a sounding rig as possible, as cheaply as possible, regardless of how that was to be achieved. I know that frequency response isn't the be all and end all to headphones. but the more I read head-fi, I realized that while there are these other concepts to a headphone's sound, they almost never get talked about. 99% of what is talked about on here is almost completely based on a headphones frequency response curve. Even when people think they're talking about something else, like soundstage depth, most of the time they are actually talking about frequency response.
This was further enhanced by the fact that 99% of people's (and my) complaints about the AD700 had to do with frequency response. The bass is too weak to be neutral. There is a noticeable drop off at around 3800 Hz. 1000 Hz is too hot. Yes, in many ways these 3 things give the AD700 its character as well. Part of the amazing depth of the soundstage comes precisely from that 4 kHz dip. the 1000 Hz bump is the main reason the mids sound so prominent. The lack of bass is a major reason they are so crystal clear. Sound, and especially headphones, are about tradeoffs, and the AD700 made its choices. That being said, I think the fundamentals of the headphone, notably its transient response, lack of THD, comfort, amazing imaging, low cost, ease to drive them and extension high and low made them great candidates to just "fix with EQ."
So first, let me talk about source. It's a hotaudio hotusb1 external DAC -> 3.5mm to XLR -> Samson D-2500 EQ -> Millett SSHM. Even though I don't think there's an audible difference between well encoded, high bitrate mp3 and lossless, just to avoid this argument in a situation I don't want to have it, everything I'm using here is FLAC. I will also try to use very popular tracks as much as possible, for the simple reason to aid people in having a clue as to what I am talking about.
My first step was to EQ it to completely flat. I used Equal loudness contours and audiometry - Test your own hearing to generate a flat equal loudness curve. What I did was simply messed with the EQ until it was totally flat at an average listening volume. That is, going straight across all tones sounded the same loudness. The Samson allows you to save settings, so I saved this as "AD700 flat". My next step was to EQ it how I like it. I like a touch more midbass than totally flat and a tiny bit more 6k-14k than flat. Nothing drastic, but the additional bass makes up for the lack of impact that headphones have naturally, a little. And the 6k-14k boost makes up for a tiny bit of high frequency drop off in my ears. that bump is VERY subtle, around .5 dB at its peak point. I saved this as "AD700". Since the difference between the AD700 Flat and AD700 settings were virtually identical, I will only use the AD700 setting for this review. Also, when I refer to "flat" in the review, I mean with no EQ, not the Flat setting. This could be confusing, so I am trying to clear that up here. So if I say AD700 flat below, I mean unEQ'd. It's "natural" sound.
Essentially with the AD700 setting on my EQ, I had 64 Hz boosted to 6dB, with a plateau there down to about 44Hz, after which I gradually rolled it off to 0 dB at 20 hZ (to avoid clipping and slowing the headphone down unnecessarily for extremely low tones, with infinite current I would have left this boosted though) From 64 Hz to 500 Hz I gradually backed it down, trying to make as smooth of a curve as possible, to 500 Hz it was at 0 dB. from 500 Hz to 700Hz I maintained a very slight taper downward and then a sharper curve down to -2 dB at 1kHz. Then slowly back up until its back at 0 dB at 2kHz. Then a pretty sharp rise to 3500 Hz where I had it at +6 dB, then a small plateau to 4200 Hz, then a sharp fall back down to +1.5 dB at 8kHz, then back to 0 dB from that point out.
Now to the meat of this, the comparisons:
1) The first thing is just a much greater sense of tonal balance with both the adjusted EQ settings. AC/DC's Back in Black now hits correctly in the opening. Whereas before it felt "neutered" and off. You can tell how poorly the hi-hat was recorded. Angus Young's guitar is in your face, just as it should be. The kick drum is both powerful from the added bass, and in your face, with the 4 kHz boost, just like it should be. This track is not supposed to sound laid back, but with no EQ on the AD700, it almost does. Things that are forward and aggressive with the drums and rhythm guitar and riffs sound laid back and flat with flat AD700's. Al Green's voice on Let's Stay together now sounds full and powerful instead of overly nasaly and weak. The thunder on Dire Straits' Brothers in Arms sounds powerful and real instead of like a cheap sound effect. But more than anything, the EQ'd version is indescribably more balanced. Everything just sounds more right and natural.
2) Soundstage. By now, some are probably wondering "the AD700's soundstage is more amazing, this guy is lying if he says its improved." Yes and no. It's not as deep when EQ'd and not as wide really either. BUt only slightly less so. I say it is improved slightly, because I believe that it's more "natural" now. Before I felt that the soundstage on the AD700 sounded artificially big. Miles Davis's So What shouldn't sound like they're playing 50 feet apart. But flat, it did. It's a small group that's supposed to sound like they're playing on a medium small stage. Before it sounded like you were about 20 rows back watching 5 guys play on the superbowl half time stage. Now it sounds like you are sitting in the 3rd table back at a medium sized jazz club. The imaging is so accurate that I can tell exactly where the microphones are placed to record the drums. I can actually hear the horn players move their horns around towards and away from the mic. The bass is there in the right amount. Before you almost had to struggle to hear it clearly, as its mastered low, like small group jazz is. The lack of bass and the forwardness of the horns previously pushed it almost out of the recording. Now its always there holding everything together. Whereas before the soundstage was wide and accurate, now its just right and hyper accurate. Classical recordings may suffer a tiny bit, but only because many classical recordings struggle with getting the soundstage wide enough in the recording. This now reproduces that trouble. If I was listening to classical, I'd be torn as to what setting to use.
3) Speed: Now let me say that the headphone's speed suffers a tiny, tiny amount EQ'd. Basically, with the added bass it seems that its now having to work harder. With an amp that delivered more current than my SSHM this might not be an issue. Muse's Hysteria sounds much more right with EQ, but unEQ'd the bass is ever slightly more crisp on the hard transients. It's very slight though and definitely not worth the tradeoff that the song otherwise sounds totally neutered without the EQ. When there is a ton of deep bass and hard transients, the headphone has a slightly hard time keeping up. But it's a $90 headphone, not an HD800. I don't think it's particularly noticeable unless you're looking for it and picking music that specifically challenges this aspect. To be clear, the unEQ'd version still has this issue, its just not as much of a problem, because it isn't "weighed down" by the demands of pumping out the bass. It kind of points out that most of what is viewed as "fast" headphones are also bass light. It takes a tremendous amount of power to adequately provide correct amounts of bass and maintain the hard transient edge.
Overall, I'd say its night and day better EQ'd. I think that everybody who isn't philosophically opposed to EQ should buy a good graphic EQ (at least 16 bands, but 32 is better, mine is 31 bands). Even better is a really, really good parametric equalizer, however they get really expensive really fast, and if you have enough bands on your EQ, it's good enough to change the overall things you want to solve with headphones. If nothing else you will learn exactly what frequencies effect sound and exactly how they do it. Also, you will want to look for one with a master gain level, so that you are free to EQ up and down without worry of clipping. If you get clipping, you can just pull the master gain down.
Anyway, this was really long, but hopefully someone found it interesting and/or helpful.
I have a new battery on order though, so maybe in 5 or so days.
