Hey guys. Nice thread!
I´ve only read trough about 28 pages, but have some gist about things...
After not being really happy with Sennheiser PX100 II, which replaced totally awesome PX100 originals I´ve lost, my desperation about good, live, punchy sound slowly rised to a point where I lost it, and went for impulsive purchase. I know there are Sennheiser products which would satisfy me, but these are not very often on stock in brick and mortar stores, so I would never know how I end up after buying some from e-shop, the same way as it happened to me with PX100 II. These are not bad, but not satysfying...
Now for the MDR-100s. The seller recommended these highly, and after I tried these, I thought "yes, these might do".
After short listening session, I was blown away by mids and highs, especially when boosted. Awesome! Live, precise, little on the sharp side with some additional colouration or distortion at higher volumes, which I totally expected from titanium drivers. It goes the same way in PRO sound and PA with titanium diaphragms.
That´s what makes music entertaining with feeling of precision and brilance when things don´t go overboard.
So far so good. The bad thing is that youtube compressed stuff with prominent highs now sounds like absolute trash. That happens when you have good phones.
Now for the bass...
It does some bass, I hear these to go in mid 20s quite easily. stronger at 35-40, and overbearing and little bit boomy/bloated at 60-70. Nice and precise at 120Hz+ again. Very good for trance music, but not good for deeper stuff or bassy stuff. It goes close to "Porta Pro mud sound". Some EQing helps a little, but doesn´t solve the probelm. Without this one, these would be absolutely lovely. At first listening session, I was blown away with mids, highs and the soundstage, instruments separation. But the humble bassheads bass is not really there.
Too bad. I can only recommend these for balanced sound users (especially for those who would mod these)
So equipped with some knowledge about speaker drivers, I´m in the modding stage.
If you are not into deep ***, you may stop reading now.
---------------------------
Thought I might try small things here and there, step by step, measuring impedance curves and SPL at each small change to get to the point, but it totally didn´t work out this way for headphones. These have different specifics than larger speakers, but also are more similar in design issues then I ever expected. From what I learned, one can alter things quite a bit, but the driver itself does the most work on the headphones the same way as on large speakers. No mumbo jumbo around $hitty driver will do it for you. Once it doesn´t have required parameters, you cannot do much. And that sadly, is the case of these too. I though I can do a lot by altering parts around the driver, but there is not really much to do, except for clever alterations inevitably coupled with very powerfull EQ or DSP to solve other arising issues. So I know now, that the issue has no full solution. Things can only be made little better.
I´m on my way to get full results including very amateur frequency response comparisons, so I might even find out it wasn´t worth it, but, here is how it goes, and what I expect:
Impedance curve of the driver and "front baffle" without the rear cover:
Resonant frequency about 115-120Hz, medium to good motor strenght. That is our starting point. Not bad, but not good for the resonant frequency. That parameter is good indicator for direct radiating speakers, about the bass response. 115Hz is not really good for bass response. So we´re in trouble here.
What you need to do basically from this "start", is to push the resonance as low as possible, and keep general impedance curve as high as possible.
That will get you to most bass. To do that, you can basicaly do three things:
Add weight to the driver cone
Dampen things up
or Make resonant enclosure (ported) to leverage the drivers output.
...And Sony did later two here. I cannot get rid of that feel, that they (as always) again did it the worst way they could. Cheap and consumer-ish way. No serious try to make serious product. Some audio engineers might work their asses off to make a new driver, and then someone in the later stage of the engineering and design pipeline wipes a $$ with all their effort. That´s what happens again and again in Sony.
Here is the deal: On the back of the driver magnet, there is some "label" for dampening and also for keeping dust coming inside the driver. It damps the driver.
The result might look good at first glance, because resonant frequency gets way lower to 65Hz, seen here:
But the impedance level also drops by a lot. That means things get dampened, less sensitive. So more power needs to go to the speaker to play. Not even touching cooling part, where the sealed hole in the speaker stops this, making the driver more vulnerable to overheating and distortion, because air from under the membrane cannot get out effectively. Gets compressed, and creates nonlinear distortions. If I got these engineers in my hands, I would f** em up. Bustards!
But what else can one do, right? Nothing much at all. Only what you can try is to damp what you want to damp, and not damp what you want to boost in relative comparison to the rest of the signal you wanted to damp, or actually damped. So they did it in their way.
Now for the resonant enclosure - the port. You already know about that one on top of each shell.
As the headphone generaly has two cavities (rear to the speaker, and front to the speaker when put on your head, connected with the "wall" with grilles inside and the padding is not absolutely stiff and sealing, this portage won´t have very strong effect on the sound. I can clearly hear the difference when the port is stuffed, but it might only approach 1-2db or so. Good if it adds where you need it, but not good where it adds where you don´t. Also the port adds to the group delay and distortions, so you need it to be effective where ear is not sensitive to this. That is rather in the sub 40Hz range. Didn´t post the impedance curves in stock port situation and "sealed" port situation, as these differences are very small, but it was obvious that it only added some SPL at the resonance of the driver at 65Hz, which is what we really don´t want. That way it´s better to seal these ports off. I recommend it. You won´t lose any bass under 50-40Hz that way.
After playing here and there with dampening, I managed only very sour victory in chase for lower than 50Hz bass.
Here is impedance comparsion between unmodded (green) and modded (orange) speakers:
As can be seen, the general impedance value in upper range is higher with modded speaker. That means more sensitive speaker overall in our case. I´ve seen some measurements elsewhere, and from 200Hz down, these headphones have a hump to add some bass. With this mod, we basically up the mids and highs (less so for these), so the hump in the frequency response should be lower. that´s good. But more mids and highs means less sensation from the bass right? So that one is only good addition, but does nothing for us about the bass.
Now for the bass range:
The resonance of the driver actually gets about 12Hz higher. Not really good, but as other things (to describe) compensated for it, I don´t mind.
The hump is milder (a tick lower) and also wider. That should mean more balanced dispersion/distribution of the bass. So It should not sound that boomy anymore. And it indeed, doesn´t.
For the 50-70Hz range, we have slightly lower impedance. Due to the modding, I´d say that sealing the port and damping the rear part of the shell did this. And it is good. It is exactly that point, which gets us rid of the boomy sound.
Now the main part. From 47 to 20Hz, the impedance rised. Means more sensitivity. But not by much. Again, we´re probably talking here about 1-2db.
But in addition to toning down the boominess, yes, we did 2-4db change overall all other things being equal. (little less in reality, if we speak rather about efficiency, and not sensitivity).
I have one shell modded, and one shell unmodded.
Here is amateurish frequency response measurement of the differences:
Top one is stock.
Middle one is modded - see, gained some sensitivity, toned down the 65Hz bump a little bit, and the rolloff is gentle, compared to the stock.
Third (bottom) measurement is after humble bass heads DSPing of the bass on the computer.
Don´t mind the upper range, the measurement wasn´t set up in a controlled manner for the full range. Everything went quick.
The bad thing is that for mobile phone (where I listen the most trough headphones), there is no serious DSPing available, so one ends up with pretty stock sound.
The mod (work instructions):
Disassemble the shell.
Take out the label from back of the driver, make a hole in it, and put it back on the driver. The hole size in the label should be 1/3 to 1/2 of the hole size in the speaker).
Put some dampening "wool" material in the shell. Should be quite thin. Normal cotton wool is too thick. For the amount, that one is tricky. You need to put as much as you can the natural way to the shell, so when you close the shell, it closes fairly easily, and it doesn´t push parts of the shell out (even with screws mounted). Also no wool should stick out of the shell. If you add too much, those screws won´t hold the shell evenly around the whole "circumference". In that case try harder with litte less of the wool material.
Add some more to the top grille of the "rear side - front baffle" to block the sound some more here...
Seal off the port on the top of the shell completely.
Done.
I´ve been playing long hours with other possible mods, with mods of the shell part where your ear fits, with no good outcome.Once you´re finished, you have slightly better headphones now. But don´t be too excited, it doesn´t save the world.
