Topic suggestion: how to incorporate manual sweeps when EQing a headphone
Background: In possibly the most “deep down the rabbit hole” move possible, I have started doing manual sine sweeps with my headphones. This has proved immediately illustrative of just how variable the actual response of a given headphone on a given head is. For example: there is no 6k peak on the 800S whatsoever for me. In fact, there is a noticeable reduction of energy in the general area, which makes me think the absorber is actually “killing” a peak that doesn’t exist on my head.
My methodology so far has been to run sweeps, then smooth out obvious dips and peaks in the treble. This has seemed to be helpful to identify problem issues with headphones on my head (crazy energy at and above 10k on the 800S for me, for example).
But I am questioning whether this is what I’m supposed to do. Maybe I misinterpreted it, but on an earlier live stream I think I heard one of you say that running manual sweeps doesn’t account for your HRTF.
Put another way: is it possible that using this method of smoothing out dips could actually lead to people inadvertently trying to “fix” natural features of my own HRTF, harming sound quality in the process?
And if that is the case and what I’ve been doing is misguided, how *would you go about using manual sweeps to EQ treble?
This one will be mostly about aesthetics and comfort, so it would be neat to see how the team would design them. Keep it open, fluid and informal by doing it live with MS Paint or a similar application.
It may be better as a video but it might be fun to see it live.
We’ve had various targets come in and out of mode. The latest waves has many brands targeting meta. While I understand there’s an appetite to achieve “neutrality.” Who’s to say thats ultimately what consumers will continue to purchase? Wouldn’t they eventually get bored of all sets in the market having little to no coloration?
This would be a good topic. My opinion is that the “current” approach to “New Meta” is probably going to stay around as some flavor, but I don’t think everyone tuning their stuff with the near-exactly the same tuning like we’re seeing now is going to stick around as the “supposed ideal” or “main flavor” of this approach for much longer.
IMO the real point of the theory that underpins JM-1 as used for IEMs is that starting with a DF baseline gives you a way to get the baseline anatomical factors correct (such that IEMs are much less likely to sound outright wrong), while also giving you the freedom to make editorializations on that baseline.
The issue we have now is that everyone is choosing the exact same editorializations, either due to a lack of imagination, a lack engagement with the theory in earnest, or because manufacturers are more focused on what will sell (and the sets tuned like this seem to sell decently).
The ideal future is, I think, that we understand that having the same baseline doesn’t mean everything has to have the same end result, so we end up seeing more variety that still fits within the realm of “overall reasonable/good”. And then, of course, the people who don’t care about any of that who make wacky fun stuff can continue to make that stuff too.
As Headphones and IEMs get better and better, is it harder or somehow less rewarding being a headphone reviewer? I look back to my first set of IEMs from around 2008 or 2009 and by today’s standards they measure absolutely terribly. As everything converges closer to targets is there less joy to reviewing?
It will stay for as long as reviewers talk about it and brands can use it as a selling point. If customers are buying products recommended by reviewers, and reviewers are recommending products that cater to the widest customer base, and “meta” happens to be that, then there’s no incentive for manufacturers targeting the mass market to do anything other than “meta”. It’s an unfortunate consequence of how reviewing has exploded that reviewers are effectively the marketing arm of manufacturers now. People don’t buy IEMs without seeing a review now, so smaller brands that can’t afford to send products to everyone become pushed out of the market. Which is why we see previously more budget brands start to tackle the luxury market and position themselves as boutique. Boutique brands will continue to do whatever they want while budget IEMs will all coalesce into whatever is being talked about by reviewers.
There’s an article here I’m mulling about for why we don’t see products in the $500 - $1,000 range much anymore, and this is part of it.
I think it’s less about headphones being better, and more about the classic hobby vs. job dilemma. The more of a job it is e.g. the more you need to chase after say the YouTube algorithm, the less rewarding it is, I think. I’d say a lot of reviewers got into it from a personal satisfaction standpoint - “I have opinions and I’m going to tell it to anyone who cares to listen”. But if its about trying to hop on the next hype train to capture an audience, the less fun it is.
If you’re worried about “is there anything else to talk about when everything is the same”, yes that is a bit of a problem. But there’s still a large number of aspects that can be said outside of that. We’re still a long way off from the perfect headphone.
As for joy, I think the joy comes from being in the community and discussing things with other people. I have a group of friends and we meet pretty often to swap gear. They know all about meta and graphs and whatever, yet they don’t care and drop thousands on exotic IEMs. It’s actually kind of refreshing to see there’s still a wide range of preferences out in the wild.
I think this is more for a video than podcast, but here’s an idea: Everyone tries to squig unheard headphones/IEMs by hand and see how do they compare to measured results.
Should iem reviews really focus on stock tips or should they focus on whatever tips allow the reviewer the best experience with the iem? Should the measurements reflect some diversity in terms of tips?
I recognize someone new to the hobby might just go with stock tips, but I can’t remember the last time the stock tips gave me the best experience with an iem. Whiile its nice when manufacturers include more tips, its still the case I am likely to end up using some other tips that work better for me.
Where is there synergy between the headphone/iem segment of the audio hobby and the speaker hobby? Does not having access to good speakers in a well treated room constrain your understanding of what good sound or neutral sound is to you?
We live in a world where there are now different kinds of listeners with differing needs. I was doing my best to educate a dear much younger nephew of mine, on dongles and IEMs, but he was far more interested in listening in a real room. Eventually got himself a set of passive speakers and an amp. In his use case he was prone to destroying IEM’s, they just did not survive, in his hands, cos he took them everywhere with him, and probably was not very kind to them.
My use case happens to be someone who has gone the other way. I loved listening on speakers, active speakers preferably, cos that was what I used in my home studio as a mixing engineer/ singer songwriter, but now I am almost exclusively listening on IEMs. My neighbours have no idea and hear nothing. And same with members of my household, they also hear nothing, and I can listen anywhere, via a dongle connected to my computers, or IEM plugged into a mobile phone. For me cutting out the room, has helped remove the variability.
Whatever one is listenig to, I think the ear has a way of acclimatizing, so any peculiarities are tuned out over time, to an extent, be it the room acoustics, or the frequency deviations of a specific IEM.
I have a suggestion for a video in the The Headphones Show channel rather than a topic for The Noise Floor, but it is heavily inspired by Resolve’s and Listener’s endless talks about how EQ takes the headphones to the S-tier and so on. With the main argument being the frequency response at the ear drum, you guys have been consistently suggesting that any decent headphone can be improved drastically, and if it’s not working for someone that’s usually a skill issue.
So, why not take it to the next level and help the viewers fix their skill issues? There is a basic EQ video on the channel from 3 years ago, but it mostly talks about easy to fix things below the ear gain that can be done based on the graph. I suggest that you create a video where you teach us the secret art of EQing headphones by ear above 3kHz.
I personally used Owliophile.com tool with slow sweeps to detect the peaks and dips and then try translating those to the filter settings for my EQ app. But when listening to the recurring EQ praises on the channel, I can’t help getting a FOMO: I feel like I still might be doing something wrong and not getting the full potential out of my EQ process. I’ve also read about the alternating tones level matching technique, but never bothered to try it because it seems to be too time consuming.
Yeah it’s a good idea, but it often comes back to “what kinds of colorations do people actually prefer??” and this is where the evaluative side of the industry only has a crude understanding - for what we have it is thanks to Dr. Olive and Co. but it’s only wideband.
Moreover, fixing perceptual peaks with tone-gen based EQs does not necessarily yield your DFHRTF, and it’s not at all guaranteed this is going to be preferred or even sound correct. It’s a useful tool to hear where various perceptual peaks are, but does that mean people don’t prefer them to be there with music? We already know people prefer certain kinds of colorations. And, to further complicate this, humans don’t have a natural reference point for these types of pure tones.
So just because you can hear a perceptual peak at say, 7khz, that doesn’t guarantee it’s not supposed to be there. Perhaps it’s not supposed to be there, but that depends on a lot more than just your perception of that peak in relation to the rest of the spectrum.
At least… this is my conjecture having done this. I say this having tried to do exactly this kind of thing, and it’s sadly not as simple as “EQ it all to sound perceptually dead even with tone gen”.
In many ways it needs to be done in combination with graphs, music, test tones… and with a general recognition of what each of us prefers individually. And I don’t think there’s an easy method for achieving that, though perhaps a video on suggestions/recommendations as you indicated would be helpful there.
I get what you’re saying and it makes total sense. Still, I’d like to challenge you guys on this. The process you have just described is extremely complex and requires a lot of experience to get it “right”. Let’s put the personal preferences aside, because we are already assuming that when somebody is EQing by ear they are trying to get it “right to their ears” not “generally right”.
If we assume that only 10% of the audience (my gut feeling tells me that this is a best case scenario) has the skill or the time to make such adjustments, then you guys claiming that that’s the way to go look like wizards sitting in the ivory tower. Thus, here is the request from the rest like me: teach us the spells and how to use the magic wand properly. Otherwise all we can do is just believe you that it works. For the chosen ones.
(pun intended, I hope you’re not getting it too personally)
Yeah I think it’s a good discussion topic to add to the live stream, and eventually a dedicated video on the subject. But in large part this is also to indicate the vast majority of the people using EQ are in fact doing it wrong. That’s not for any fault of their own, merely that the tools that exists don’t have sufficient guardrails and are largely based on the existing (erroneous) paradigm of single line FR graphs.
I’m happy to go through my own process of doing EQ, and I do also use manual tone-gen if I’ve identified certain problem areas with music.
I think people may assume that since we’re the ultra-nerd wizards in the ivory tower with our EQ magic that… we’re having these incredible experiences that are locked away from the public behind endless measurements and research - that nobody else can have, when that’s not necessarily the case.
Most of the time it’s just a matter of trying things and seeing if it sounds good. The one point of demarcation that I’d personally promote is to not approach any of this with overconfidence in any one particular method, or expect that it will yield “the best sound”, given all the unknowns in this space.
^this is mostly a joke, but there is a genuine benefit in not peering too far down the rabbit hole. I don’t fault anyone for not wanting to get into it.
Sharing a personal process would be one way to do it. If you feel scientific about it, you may want to delve into the methodology a little bit. Like before filming anything, conduct a small qualitative research in a focus group of experienced EQ proclaimers. Here are some example questions you might ask them:
How do you understand that something needs to be improved in the first place?
How do you detect the frequency band that needs to be fixed?
Do you use synthesized sweeps or tones? If yes, what’s your process of using them?
How do you understand the band width/Q and level of adjustment needed?
Do you use specific test tracks to adjust the EQ parameters with music?
For each track, what exactly are you looking for and how do you translate that to the EQ adjustments?
Do you care mostly about wideband balance or do you keep chasing small artifacts?
What is your attitude to narrow, high-Q filters?
How do you combine multiple filters together? Are there any rules of thumb you use?
Where do you stop? How do you understand that the EQ preset is good enough?
A shorter version would boil down to 3 generic questions:
What do you listen for?
How do you listen?
How do you translate what you hear to the EQ settings?
If you ask several people you might notice some patterns, which can increase the probability of the method being more applicable to other people. I really liked the video you guys did on the test tracks, it already gives a clue on questions 5 and 6 above.
Social sciences are hard to do right. It’s not an excuse to give up and go with purely subjective point of view though
Strange. I thought perceptually smooth mid-to-upper treble was always desired. Like you can customize the overall amount and add a little spice here and there with low Q, low db filters, but you never want jaggies.
Here’s a suggestion: looking back over the last, say, 5 years, do you have any products you were enthusiastic about at the time that now either you’ve changed your mind about or have been completely outclassed by newer products? In a phrase: recommendations that are no longer recommended.
