Nothing beats trusting your own ears. 
5 Likes
1 Like
Very fitting:
1 Like
Growing up, EQ Hi-Fi components were so common. As a kid, I loved watching the FR changes with the lights out, in the dark.
14 Likes
Thanks very much its amazing information about great and best headphones . i really appreciate you to share great information than help me to chose the headphone . i like it very much. Stay happy.i need also some suggestion about which company headphones carry high time.
2 Likes
Back in the early 70’s I had a SAE EQ.2 chl. Always enjoy adjusting or equalize (alter the frequency-specific amplitude of) the left and right channels of a stereo audio track/file independently. I was pretty cool.
5 Likes
I guess we like the show “Vikings.” Welcome to the Community.
Vikings is a fantastic show!
1 Like
Also posting this here so that people can understand how EQ factors into my reviews and also the reasons why you shouldn’t just throw on profiles in a fire and forget manner:
13 Likes
Nice job in listing some counter points to using other people’s EQ profiles.
Off-topic, what is the headphone stand that you’re using for the LCD-4 in the video? It almost appears to be custom tailored for Audeze headphones.
That’s one that a friend gave me. He 3D prints them. I’ll send you a link to it when I find it haha.
2 Likes
@Resolve, can you measure the LCD-4 on the new GRAS rig? I don’t see any measurements posted from you on the LCD-4 (unless I’ve missed it).
Yeah I did, I just haven’t posted it yet. I’ll dig that out and post it in the thread.
3 Likes
Hmm. Lots to unpack here. While you covered a lot of limitations of using EQ profiles well, I can’t agree with the overall message of this video. It’s a matter of degree. The limitations exist, but you seem to be exaggerating their effects. I think the best example of this is in the conclusion when you call applying an EQ preset as “a crapshoot”. While it’s true that applying a preset does not correct the headphone’s FR “dead on” to the target when it’s on your head, it should be more likely to get closer than further away.
Of course much of that depends on the quality of the preset. If you are using the oratory1990 presets that use the Harman target, then many of the limitations you bring up are mitigated to some degree.
Oratory is using the same Gras equipment to measure FR that was used to develop the Harman target. The Gras uses a pinnae that is based on an average of a selected population. The Harman target is based on what a group of listeners considers correct.
Oratory listens to the corrected headphones against a series of test tracks, some of which he recorded himself so he knows what they should sound like. Actually he stopped doing this after awhile because the presets never needed it.
Oratory measures a number of the same model headphone and averages them to account for unit variation. This is actually an area that greatly interests me as I can’t find a lot of data on how much unit variation there is or what it looks like. But we do know that the measurements on the same rig are visually comparable. It’s the silent revisions you really have to watch out for.
Oratory doesn’t apply corrections above 10k since the Gras unit is not accurate there and knows not to correct the “9k dip”. You bring up the 1/3 octave resolution of the Harman target as a limitation, but I haven’t seen any evidence that it needs higher resolution below 8k. If you have, please let me know. I thought it was generally accepted that the FR of a good sounding headphone was smooth, even before the Harman target was developed.
I get why using a fire and forget preset may not guarantee the subjectively best sound to you. But if you are looking to get closer to objectively neutral or an EQ baseline to start from, I don’t see why applying a quality preset like oratory1990’s wouldn’t be a better place to start from than the factory tuning with all of its own limitations.
Having said all of that, nice work! I always enjoy your videos and this one really made me think. 
7 Likes
Pro / anti EQ is an interesting topic. In a nutshell, I think that quality presets that can bring a headphone (or IEM - can’t forget our siblings) to some standard version of neutral are valuable. Beyond that, it becomes personal preference. For me, often depending on the song I am listening to. Crank up the bass. Or the vocals. What resonates? Beyond that, you have to know the headphone’s limitations. Not every headphone can produce head-thumping bass even with EQ. So I see this as a 2 step process: (1) dial in “neutral” and (2) work out to what you want to hear. Both are restricted by the physical limitations of the headphone. I would suggest that “neutral” is music as stored and transmitted - which might even be as recorded and/or intended - and “what I want to hear” is my artistic (! / personal) modulation on that.
2 Likes
This isn’t technically correct. I called the treble range (above 5khz) a crapshoot for individual ears, and it really is when you look at individual people’s HRTF - and I’ve shamelessly stolen that description from Blaine who knows much more than I do about this topic haha.
You mentioned Oratory, I expect he and I actually agree on this topic for the reasons I expressed in the video. That’s not to say… you shouldn’t use his profiles, but rather also treat his profiles as a starting point. In my view, his are probably the best because he’s taking individual HRTF into consideration, and because he does the treble by ear. He even talked about it on the livestream we did with him a long time ago. It’s not just the concha dip at 9khz, but the lower and mid-treble as well.
This is found in the three papers from Hammershoi and Moller that I showed in the 10 stages video.
This is the key paper for this stuff.
Here’s the free field result of 40 human subjects:
And here’s the diffuse field result:
Note that these are ‘mean’ averages, so if you look at the distribution on the left you can see that it varies quite a bit across individuals.
Even if you look at the GRAS KEMAR FF and DF HRTF, which is what Harman was developed for, it looks like this:
Harman’s in-room target (the important starting point of the research) is based on a sum of these, and the target we use today is effectively that with the preference tilt applied.
Now, when we did that 10 stages video, Sean Olive mentioned that he thinks it’s okay to have a highly smoothed target given that these non-smooth features in the high res DFHRTF are direction specific. As you approximate that, you end up with a smooth result (at least that’s how I understand it).
The problem with this, as has been demonstrated to me recently by Blaine, is that when you do change the direction with an individual HRTF, it’s not like those features go away. They just shift to a different spot, and will also change differently depending on the individual. Moreover, it’s not just the features above 5khz, you see this kind of change even at 3khz - and we still use that feature. The point being, just because the treble feature shifts, that doesn’t mean we should ignore that it exists.
In some ways the critique in stage 8 of that video was related to it being a mean average and not a median or mode - or something more specific to an ‘average ear’. But at the same time, there’s also a sense in which a mean average and a highly smoothed target like the one we have is going to be more useful for evaluative purposes given the variation that exists among individuals. We just have to know that going into it. Unfortunately, this also means we can’t use it as a guide for EQ in quite so straightforward a manner.
15 Likes
Man, that is a great clarifying post. Thanks for always sharing valuable headphone equalization knowledge.
3 Likes
I’ve never been one to unquestioningly use someone else’s eq settings but they can be useful starting points. Sometimes I find Resolves’ useful, other times I prefer Chrono’s or Oratory’s. Since I mostly listen to acoustic music from large orchestral pieces to string quartets to live Jazz and have been to countless concerts esp for classical music, I know what real instruments sound like and adjust the eq profiles i start with accordingly and then save my variation. I’m not one who demands the headphone manufacturers get it right, especially when I know they have their own frame of reference like Audeze consulting mastering engineers for feedback. I’m not bothered (as some are) that eq means the music is no longer bit perfect. I’m not sure I can hear the alleged distortion that eq-ing creates. Mostly I want to forget about all the audiophile analysis and enjoy my listening sessions so once I get something enjoyable I stop tinkering. I do recognise that different albums may require slight modifications but I try to not be hypercritical.
6 Likes
Yeah I also find them valuable starting points. In some ways it’s also kind of fun to throw on a profile and understand what sounds good to someone else.
At the risk of being pelted with rocks and garbage, that’s why I’m a fan of analog EQ.
Yep, you can’t get anywhere near the same precision as you can with 10-band parametric EQ software, but it’s easier and faster for me to turn one, two or three dials on my JDS Labs Subjective 3 and get close enough to where I need to be instead of messing with different software-based EQ profiles.
Different strokes … 
4 Likes