Superlux HD681B (and F) - why no measurements?

[L5730]

Looking around the internet for some hours and I just cannot find anyone measuring the B 'balanced' version of the HD681, nor the F 'front row' variant.
Seeing the various mods to the 681, would it make sense to go for the -B variant in the first instance?

It would be interesting to see what changes there are between the various models, it could be that the drivers may even be the same and/or are binned based on test measurements.

The -F variant is quite widely suggested as bass-light, but perhaps offers a smoother high frequency quality after like-for-like target EQ curve.

 

[staticV3]

Here's you can find frequency response measurements of the HD681B, with three different ear pads even: https://www.reddit.com/r/oratory1990/wiki/index/list_of_presets/

 

[L5730]

@staticV3 That's interesting, thank you!
When comparing them like this, it makes sense to opt for the -B version, even though the treble spikes are higher (like the 668B), the 4k dip seems less. Hmm, I wonder how the distortion characteristics differ, probably not much when seeing 668b vs 681. Perhaps the B stands more for bass than the marketing literature seems to say :shrug:

HD681

HD681B

HD668B

 

[solderdude]

In 2009 I started tinkering with HD681. I did not have a measurement fixture then though.
I made a passive filter that removes the treble peak (by ear).
Back then I joined Rock Grotto.
Since then this hobby evolved a bit.
Later, when I made a test fixture, I found out the filter made by ear was not far off from the one based on my fixture.

In those days the HD681B and F did not exist yet.
These HD681 were € 15.- and forum members asked me for modified HD681 (filter + K240 pads).
In 2010 and 2011 I bought quite a few locally and started modifying them.
I noticed some had audible less bass (on direct comparison)
By that time I contacted Superlux and got in contact with them.
They were just entering the market with headphones and were interested in my findings and about the difference in tonal balance.
I figure, given the price, they just produced those headphones, maybe checked if sound came out and sold them.

I had been in contact with someone from development since and got quite a few prototypes over the years for testing and improvements/evaluation.
I mentioned the audible differences in bass and that it might be production spread thing in the drivers.
This was denied of course.
Low and behold a year later there were 3 versions of HD681 with the only difference in bass output.
I suspect they found the same as I did and (speculation on my part) decided not to make the HD681 a lottery but select drivers and put them in models with different amounts of bass.
Below the differences. These differences look very similar to what happens if you damp the rear of a driver a bit with extra foam.

Yet, when these models are opened up the drivers do not appear to have a different driver damping.
Drivers selected in 3 tiers or developed ?
Superlux never answered that question. They just stated they decided to make 3 versions.

So above the different models compared.
How these measurements were obtained I never got an answer to.

Below a plot made by Superlux where they compare the HD668B with DT990

And as presentation is everything below my measurements of HD668B vs DT990 and presented in a similar style.

Extrapolating from that it could be somewhat safe to say that the differences in the HD681 plot from Superlux could be representative.

Back then when I found the differences in bass levels/tonal balance I did not have a fixture yet so only anecdotal (by direct comparison with my own HD681) but the differences in bass (tonal balance) matched the spread we see in the Superlux plots. Of course, nothing scientific here, just my subjective findings back in those days.

Note I could never find the bass extension Superlux claimed in their plots below 40Hz.
Below HD681 (stock) vs HD668B on the same fixture.

There is a clear 'driver family' resemblance between these Superlux headphones.

 

[L5730]

@solderdude Dankjewel for your continued work and especially your comments on this particular 'group' of headphones.
It makes for interesting reading, your historical experience with these.

As has been mentioned numerous times, the price point is going to lead to unit variances. I do wonder, however if a -B variant might not handle some of the 3-6k a little better? It would seem that I maybe over expecting things here - driver binning may simply be based on a quick 'n' dirty test rig pass where those that fit the low-frequency parameters for each bin (-, B or F) are thus segregated accordingly. It'd be nice if the company just put cards on the table. Honesty in manufacture at this price point and willingness to try to punt for consistency would be a very appreciated thing. I know I am basically saying "nice, it's a $1, can you make them all clones at that price?". However, if Superlux are indeed binning drivers based on qualities, that is only a good thing for the buyer to get something similar to what they expect, no?

One can assume that the company might use a similar test rig, if they indeed do at all, for other drivers.
Amir's set of 668B seem to measure quite similarly albeit 2.5 dB offset from one another. I do not possess (and doubt I will) the necessary equipment to measure my specific 668B, but it's very clear I have a 1.5 dB mismatch which doesn't change audibly enough across frequency sweep (EQ bell pass filter, sweep it manually in mono mode with pink noise).

Superficially, I think the old bright red colour and the newer duller red colour accents look ridiculous and a give the impression of a child's toy rather that something remotely serious. White and grey parts look subjectively better, but black is better still. I prefer the newer style stealthy "professional monitor" matt/gloss printing, rather than the white text - as is also present as "studio monitor" on the 668B. Why not dispense with that printing rubbish entirely? It does seem odd that they decided to retail box the 668B but keep the 681 and variants in - not a blister pack - but shrink wrap! So long as a retailer dumps that in another box with sufficient packing, it's OK I guess, but first impressions and all...

 

[solderdude]

The Superluxes are all about VFM and they want them to be cheap. The HD681 are just that. Production costs are probably just a few $.
The B stands for 'Balanced' and is less bassy than the red version.
The F stands for 'Front row' and has even less bass.

Some questions I posed at Superlux are answered but some simply are not. Even when asked 3 times.

At RockGrotto there was a guy that made nice aluminium rings for it. Another member made cloth and leather headbands (made from my old torn motorcycle jacket). Replaced the pads (K240 as Superlux did not have the 'velvet' pads back then. They simply copied the K240 velours pads later on, Also replaced the cable with a supple short one.
It turned the red HD681 into this:

 

[soundwave76]

Thanks @solderdude again! So in your opinion, which Superlux model is the one to get?

 

[solderdude]

None, unless you are really on a budget and don't mind the build quality and some modifications and/or EQ.

 

[L5730]

Thanks @solderdude again! So in your opinion, which Superlux model is the one to get?

Apart from the obvious issue that they are built to a price and have lots of flaws because of that *, the general sound signatures seem to be reasonably well documented.

Both suffer from creaky sounds from the plastic when moving jaw or tilting head.
HD681 has attached cable, HD668b has removable 'extension' 3.5mm TRS cables.
HD681 metal rods resonate with an obvious 'bong' when flicked/tapped.
HD681 headband is fairy comfortable, but self adjusting via elastic bands.
HD668b 'wings' are difficult to get comfortable, many people just hate them.
HD681 clamping force is less than HD668b
Solder joints are a known source of failure - but easy enough to repair if one is handy with a soldering iron.

The entire series of HD681 and HD668b have a 7-10 kHz peak of ~8 dB or so - it's obvious that there is simply too much treble. At this frequency, to me it's not horrid but it's definitely unbalanced. Correctable by simply covering the driver with tissue paper as Solderdude has mentioned. However, it can also change the texture of the treble, where an EQ or analog filter may be preferable (even with phase changes).

I've personally found that the HD681 suffers from a 5 kHz treble peak that is rather irritating. The HD668b doesn't have this problem, probably shifted the peak higher up, and in turn is handled by the tissue paper mod. Tissue paper isn't solving this problem on the HD681. I am looking into a filter circuit to solve this.

HD681 has a lower bass extension and a more detailed textured treble compared to HD668b.
HD668b has a closer match to Harman target low mid and upper bass frequencies.