Discussion in 'Modifications and Tweaks' started by Bill-P, Sep 27, 2015.
Yep, stupid peak >.< That's a fucking *great* pair.
Yeah, I agree.
Oh, and I have measurement of the whole mod that you had in there as well. It did work to reduce something, but unfortunately, that 6KHz peak stood.
Seriously, it's like an immovable fortress. I'll see what I can do...
Since you have a stock one, it'd be interesting to try the Sonarworks plugin that EQs to neutral (there is a HD800 profile with a complex EQ curve). On OSX, I run the Sonarworks plugin through Audio Hijack Pro. There's trial of both software. I actually took my mods off and am just running Sonarworks to test it out and may prefer this as I can click a switch to disable it or tweak it.
If you ran the stock HD800 through Sonarworks and measured it, it'd be interesting to see how flat it'd come out on your measurement system.
You mean save for 6KHz peak AND 4KHz dip, right?
In all honesty the dip is probably not that strong when listening.
Okay... time to take measurements. I will make a separate post detailing my measurement rig at a later time and link it to the first post.
It's all about relative differences here. I measured each state at least 3-4 times before deciding that the results are consistent enough to make sure there was no problem with fit/placement/seal on my coupler. Volume is digital and is fixed at max, while I control attenuation via REW, which is fixed, so I don't "normalize" to 90dB or anything like that. This is intentional because I want to show the effect of the mod at the same volume.
First off, the effect of removing the dust cover:
While it is easy to say that since the peaks stand out more when the dust cover is removed, and thus the headphone is brighter-sounding, I think the dust cover should be removed because it's adding high frequency contents that shouldn't be there. Or otherwise it's diffusing them so it's more confused-sounding. I actually prefer the dust cover off either way, and not just due to measurements either.
But just as a teaser, here's what happens when I push the HD800 backwards on my coupler (as in... I make it so that the microphone is further away from the trapezoid area):
Look, ma! 4KHz dip not as strong, 6KHz peak not as pronounced!
So... I highly suspect some insane reflection trickery is going on the closer the mic is to the trapezoid piece. Subjectively, I share the same sentiment as Hands... that the 4KHz dip really isn't that severe in actual listening. Eric Clapton's "Tears in Heaven" in his Unplugged album still sounds hella sibilant, and I know the sibilance in that song is related to the 4KHz region.
Anyway, so what happens with a bit of front damping? A la what Tak was doing with the HD700? The middle section of the HD800's ring driver is practically just free air being let in, so blocking that should be fine, right? Um...
Yeah, no. Blocking that causes internal driver reflections that push treble up even more. Granted, you can say that this subjectively "flattens" the treble region so it sounds smoother, but honestly, I didn't like this one bit. This is torture to listen to. It is HIGHLY not advised that any front-damping be done to the HD800's driver, and based on the results of the dust cover above, I must also conclude that having the dust filter (which is front damping) is also a bad idea.
More to come... (including SonarWork's EQ results) <- I will be editing this post soon, and when I reach the maximum post length, I'll post a new one with more measurements.
Reserved to keep things linear...
Reserved again because I know I will absolutely run out of space.
I noticed the same effect of lowered bass with the dust cover off as well, although I couldn't conclude anything in the treble area like you did since my crude measurement method produces too much measurement variance after 6k. Nice findings so far!
I settled on my ideal Sonarworks setting, it sounds VERY nice. I'll post details in the other HD800 thread to not pollute yours, but I'm looking to hear how Sonarworks measures out.
Yeah. For some reason, though, the lowered bass level is not shown, which leads me to believe something else is at work here other than just sealed/not-sealed pressure.
The last coupler I had was also somewhat unreliable past 8KHz, but this one has been more consistent all the way up to 11-12KHz, so... that's good. >12KHz is still rolling dice so please take results above there with a grain of healthy salt.
With that said, since I'm not able to reliably measure a bass increase/decrease on the HD800, the above measurements should be taken with salt as well. The main thing to look for here is relative differences in the mid and treble, me thinks.
Ah, so that's why I heard the HD800 is slightly harsher without the dust-covers. It makes the transition from peaks to nulls in the treble more hectic. Granted, this is a subtle difference in sound.
Well no actually, the transitions might look worse but it is less bright like I said. Just want to add that my HD800 measurement rig shows a much less harsher 6k peak than what Bill's show, but the same effect of the dustcover making it brighter.
As far as I'm concerned the dust covers do not improve anything whatsoever. Nada. Zip. Zilch. 6k is simply too low for any relatively transparent material to push down. Of course you can stick your HD800 full of cotton balls and kill the 6k peak along with everything above 1k. My hypothesis at this point is that the dustcover makes the treble worse by adding another layer of material for treble energy to be stored, hence that extra brightness. Causing a sort of turbulant flow rather than letting the energy move naturally. It's similar to the effect of putting a grill over a tweeter which any speaker guys should be familiar with. You lose air and articulation, while simultaneously making things harsher and more confusing. I would advise anyone serious about improving their HD800 to float their dustcovers down the LA river.
Yeah, I'm of the same sentiment. The transition looks worse on measurement, but the headphone does seem less bright.
It's kinda more "in my face" though, possibly because of heightened upper mid or something.
But anyway, will post more measurements later. I've found that this process (measuring each small step 3-4 times) is taking quite a bit of time, and I'd love to enjoy my HD800 now!!!
On that note, though, will post my 1.0 mod now, too. I applied the 1.0 mod to CEE TEE's HD800, and he liked the results so far. No measurement on this, as my last coupler is done for now, and I'm not confident the results translate well. You'll just have to try it and see.
Oh, and version 1.5 is already currently being tested. Will post that one a day or two later, and there will be measurements for it, of course. It's done mainly because both CEE TEE and I wanted a version of 1.0 that's even darker-sounding. Some of you may not like this, so... read on below.
The official Bill-P's HD800 mod version 1.0:
List of materials:
Bird cage skirt:
Better and recommended alternative to 2mm foam = 1/10" sorbothane:
Optional: dynamat <- not shown because it's the most expensive of all of the components, and I'd suggest sorbothane over this anyway, but you're welcome to use it if you can't find sorbothane.
Step 0: you must have an HD800. I think you can probably steal or borrow one somehow.
Step 1: throw that dust cover away! You can either store it somewhere dark and not humid, or you can just burn it. I think burning it is potentially messy, so make sure you do that away from the hay, ya? Oh, and yank off the ear pad here. It'll make your life easier. Just insert your fingers along the edges and the ear pad will pop right off.
Step 2: cut out a ring from the adhesive felt and stick it to the metallic/aluminum ring around the driver. Make absolutely sure you are not covering the driver in any way, and you are not overshooting the edges of the ring. And yeah, you can leave a bit of the edges open, but make sure you don't cover the driver or the edges.
Step 3: small felt piece here to cover the ridge thing. I think the cut-out from Rabid Dog 1.0 somewhat omitted the fine contour of this place, so a separate piece is needed this time. Be careful that you don't obstruct the mesh while applying this piece.
Step 4: cut out a foam piece or a piece of the sorbothane to cover the trapezoid area. IMO, sorbothane gives better sound due to its vibration damping properties, but... maybe I'm just crazy. Also, if you'd like to apply dynamat anywhere, here is the best place. Put the foam on top of the dynamat piece if you choose to do so. But... it is highly highly HIGHLY recommended that you use sorbothane.
Step 5: cut out a strip of adhesive felt to cover the edges of the metal mesh like this.... The goal is to cover the places where there are plastic pieces on the other sides of the mesh. You'll have to shine a light through the back of the mesh to see this.
Step 6: cut out pieces of the bird cage cover thing in order to cover the rest of the mesh. Note that the area I highlighted needs an extra piece (so 2 layers).
Step 7: cut the bath mat thing in this Omega shape. Note the small square cut-out near the top. The goal is to make sure this piece fits just right into the cavity, and cover just the mesh (not the plastic parts). Also, the small cutout can be stuffed with a piece of open-cell foam to reduce a bit of treble, or you can leave it as is for a bit extra stage depth.
Step 8: make a ring and trapezoid pieces with the open-cell foam, then... insert them into their respective places. You may want to use some small pieces of double-sided tape to keep them adhered to the other pieces here.
Step 9: you're done. Put the ear pad back on and enjoy.
Yep, no dust cover, nothing over the driver. Pure bliss! This also has the added bonus of improving bass and mid response, and it doesn't look like someone else just did a science project to your HD800. Is there a loss of air this way? Well, one way to know: hold your HD800 up to the light and see if you'll still be able to see through? Looks like the driver is still transparent to me, and in fact, may be even more transparent now that I don't have the dust cover on anymore.
And... that's that. If you do that and find it's still too bright, there's always version 1.5.
If this is just way too many materials, and way too complicated for ya, then look out for what Marv has in store for the BTAmod. I don't intend to simplify the process any time soon because I'd absolutely like to make the HD800 more listenable first before anything else. If it can't be listenable to me, then I might as well go for a different headphone altogether, and even then, nothing else on the market currently is as resolving as the HD800.
Just wow. Awesome guide! I am still getting to know my vanilla HD800 but I will be watching this mod with interest, thanks.
Great work Bill. I'm going to try it and see if I can talk my friends in CAD to make some templates if I like the sound. Now on to Amazon.com
At this rate, you'll have mod v23.7 out by the end of next month.
Though I think I'll stop eventually when it's too tedious to make improvements. Just like with the HD600.
Something interesting to think about... as I looked again at the Spectrograms of my past data...
I think spectrograms are much more telling of the characteristics of the HD800 than either FR or CSD. I'm tempted to redo all of the above measurements but only looking at Spectrogram now instead of at CSD or FR.
Failure to "describe" the characteristics of the HD800 contributed to my hesitation to continue with the above measurement quest (laziness still dominates), but now... spectrograms are making it more interesting again. This is the only thing I've seen so far that reliably shows the "forwardness" or "ringy" characteristics when dust covers are off (notice 6KHz ringing and 12-13KHz ringing are worse than stock, plus 2-3KHz is more present).
Also that weird decay characteristic at 1KHz is interesting, too. I think that is what causes the weird "thinness" of the sound, rather than the treble regions up top. Somehow taking dust cover off kinda makes it less thin. Maybe this is why both OJ and I perceive the HD800 as being slightly less bright when the dust cover is off, though it does get quite a bit more "in yo face" that way.
CEE TEE's older HD800 (s/n 08xxx) versus my newer HD800 (s/n 30xxx). Ear pads only for now... will take super high res photos of the drivers soon.
So... what the eff is going on, Sennheiser?
Measurements incoming. I'm putting both of these pairs to stock and measuring them.
This came up when we were comparing my pair to CEE TEE's, and CEE TEE remarked on there being more bass on his pair. I'd... have to agree. And here I was thinking something was off about my pair.
But my preliminary conclusion about variations in HD800? Well... it's like the differences between those pads.
And suddenly, this makes a universal mod that applies to all HD800s kind of an exercise in futility IMO.
Oh, and high resolution photos of drivers incoming...
What are we supposed to be looking for? Post very vague.
Yeah, post very vague.
Here's an even more vague post.
CEE TEE's older HD800 driver: finer mesh
My newer HD800 driver: mesh is less fine
CEE TEE's older HD800 ear pad:
My newer HD800 ear pad:
Separate names with a comma.