WTF Revisited: Sennheiser HD700 Stories, Review, and Measurements

A boosted 6kHz transfer function will accentuate certain overtones at the expense tonal masking of the adjacent frequencies. Not to mention the phase distortion such a boost will incur which won't do any favors to soundstage.

 

haha yes it does. Shitty features.

 

In my experience playing with EQ, boosting 6khz or 8khz will outline the borders of your stage, helping you notice where the stage ends. This can perceived as increased stage size but in reality it's not. It's just you're more aware where the borders end. This trick is useful if you want a help to know the borders. It's also useful for imaging if the headphones struggle locating things.

Darker presentations will do the opposite, less clear border outlines > more sense of a bigger stage size.
Also if you listen to Jazz or orchestral pieces (not sure if it's applicable to all genres), a lot of intimate (forward for speaker guys) sounds are located from 1khz and up. Having a headphones or even speakers with a downward slop starting just before 1khz will place things further back, again giving you a bigger sense of the stage size.

But darker presentations will demand a superb imaging from the headphones to sound pleasing... or at least to me.


@Ruined You can stop the theoretical talk and start experimenting with all of this for free using an EQ!.

 

It does somewhat show up in the measurements. The HD800 has a lot more pinna gain than the HD600, for example. See here: https://www.superbestaudiofriends.o...easurements-by-serious.2518/page-2#post-72717

The HD600 has almost 5dB less pinna gain in the whole 2-5kHz region than the HD800 when comparing my own flat plate coupler to my own in-ear measurement method. However planars have even more pinna gain and (to my ears) don't have such a large soundatage. So maybe part of it is also the shape of the pinna gain response and not just the amount of pinna gain. Note that that HD800 has a lot more pinna gain at 4kHz than it does at 3kHz, at least in this comparison.
That's the measurement I'd point at for soundstage size. It's not perfect, but with more data I bet we could find some correlations.

 

Besides moving diaphragms further from listening point e.g. HD800, K1000, etc wasn't there something about how faffing about with phase, group delay, etc. "expands" headstage at expense of image definition?

Also I can't find where this is on forum precisely but I recall reading on here about how it's primarily midrange response level that's influenced by driver distance from ear and angling to front (which coincides with the like of HD800 and HP-3 having troughs between ~1kHz to ~5kHz and their respective headstage depth). There's also something a person far more informed than I said about how reflections between ear and baffle front contribute to a sense of staging.

Just to be clear (tendency to misunderstand and all) is the implication then that headphones that have greater variability between different pinnae positively correlated with headstaging? Wondering if anything that can be sussed out of flat coupler measurements can speak to headstage.

 

Yup, glad I read that right haha. Also now saving up for 3DIO ears I guess, or whatever other comparatively cheap crud might be convinced to stick to the MiniDSP. Having everyone use the same EARS rig would be boring

Requests I guess would be for ADX5000 (large stage but tight, more cohesive images than HD800 even out of weaker upstream) if it's not a pain! Love em save for the upper mid stabbiness and their price.

 

A bit OT but maybe relevant for this part of your question: closed-back phone (T50RP3) with different thickness pleather pads, FRs roughly aligned at 600-900 Hz to show effects on upper mids; flat-plate coupler, no pinna.



Red: stock pads (thinnest)
Green: Shure 840 pads (thicker)
Blue: Dekoni pleather pads (thickest)
Orange: slant pads 1 (softer)
Tan: slant pads 2

Both slant pairs are a little thicker than the Dekonis at the back, thinner at the front. Small dataset, but consistent with thicker pads elevating the upper mids.

Both slant pairs seemed to me to stage wider than the Dekonis.

 

Nah, I don't buy this 1-ish kHz to 5-ish kHz depression being the root all that's holy with regards to soundstage magiks in the HD700. just tried some filters through Roon, low shelf plus high shelf high Q, and although it does seem to expand things a bit the timbre of things like piano is completely off.

 

RTings.com points to the depth and placement of the 10khz "notch". That is, a 10khz notch (trough) contributes to a sense of distance from the stage. The deeper the notch and the better that it is centered at 10k, the farther the perceived soundstage distance.

 

There are certainly FR tricks which make the soundstage seem wider than it is, Hifiman being the absolute best example of this. The whole big dip at 2k then spike at 4k thing is a big reason why so many people rave about how open and wide Hifiman is. THe dip from 1-2k drops the presence region making thing seem more distant, then the 4-5k spike causes the attack of some instruments to be very forward which gives the impression of great depth. Its not "real soundstage" in that the drivers and housing of say a HE560 for instance arent actually capable of projecting a huge stage the way a HD800 can but it certainly can "trick" one into thinking it does.

This is why I personally think you should try to EQ a headphone to some type of normalized FR to find out what the "real soundstage" is as that removes any of the tricks from the equation. For instance Grados have a wider soundstage than one would realize once the 2k spike is eqed down (though its still not anything close to a HD800), as that spike makes everything sound super in your face. The HD800 actually does have a huge soundstage as even when you EQ down the 5-6k spike it still sounds very wide.

I do think the 5-6k region being boosted can trick one into thinking the stage is larger than it is as that region tends to bring out "room noise" and reverb effects. If you are listening to something like Pearl Jam's Ten which has a load of reverb it does sound "larger" with extra 5-6k energy. There are actually quite a few places in the FR that affect the perceived soundstage though it is difficult to pin down one and much of it would depend on what you are actually listening to as boosting or cutting different areas would make specific tracks sound larger or smaller.

In the case of the HD700 I do think the dip at 3k and spike at 6k contributes to its perception of width on many tracks but its definitely not the only thing, it really does have a truly wide stage. When I had one I EQed it to a mostly normal FR and it was still well wider and more open sounding than most other open backs out there.

 

Do you mean that because you don't agree with their compensation curve, they must be wrong about how they view other measurements? That seems a little dismissive of the measurement effort. Srsly, don't get your comment.

 

Certainly not the only contributing factor but it does go a ways towards increasing perceived sense of distance from "listening point". The fact that EQ doesn't immediately help with it and that headphone designs that inherently measure with a midrange trough appear to all have a larger headstage than baseline is part of why I think EQ isn't ideal, much as it can help with troublesome spikes and all

From what I can tell lower SPLs= sense of distance in fine gradations, e.g. when someone speaks to you from across a room v. speaking by your side. That said the former will have more sound added to it from the room, and that's not necessarily something EQing can address. Advanced DSP though (hi @Magnetostatic_Tubephile )? The notion is piquant.

As much as I hate being a lemming I do share the belief that RTINGS's investigations into how certain measurements may coincide with aspects of subjective perception don't necessarily pan out. At least with regard to limited subjective experiences with some of their headphones evaluated, I recall some models they asserted have great stage size sounded as if they didn't, though their metric for imaging "tightness" does seem to align with my impressions, yes. Need to get ears on more headphones I guess, when able to.

At the end of the day given this isn't life-critical stuff headphone measurement interpretation appears very much nascent relative to other fields of investigation and I do think we're all still learning. This is why I'm a fan of the purr1ns and Keith Howards of the world with how they look into novel ideas to better align subjectives and measured like with how burst decay could serve to represent transients or how poorly-designed headbands could have negative contributions to reproduced sound.

Not so much the heavy reductionism going on by others, no.

 

And people who have to use headphones by necessity are more naturally inclined to be picky about stage, I guess. No speakers to fall back on!

This isn't even accounting for how distortion characteristics might contribute to more of a "room-feel" either, as in the case of dBel's distortion machine. I don't fancy myself an intelligent person but I've been messing around with random DSP trying to get the HD600 to sound more out of my head but yeah, not much dice.

The software + head tracker @tommytakis mentioned is genuinely interesting to me, pricey as it is. Hoping more people here get ears on it so people can learn that DSP isn't the end of the world )