Fostex RP discussion and mods thread: all variations of TxxRP, T-X0, and RPKIT50

[isostasy]

I've had a good look around and can't find any threads on ASR regarding this series of headphones from Fostex, using their RP planar-magnetic drivers, which are a very popular basis for modifications. I've been hacking away at a set of these for months now and have only recently struck upon a vaguely neutral, or at least listenable, frequency response. There are already various threads on other forums, articles published on audio websites, as well as plenty of commercially available mods, but I have problems with all of these, most of which boils down to a dearth of evidence for the modifications suggested, even the most popular ones. Let's make way for some more scientific based discussion.

The headphones I'm talking about, so you are clear and so we show up in search results:

• T20RP, T40RP, and T50RP, spanning three revisions (marks I to III)
• T60RP
• Massdrop T-X0
• RPKIT50

If you want a more detailed breakdown please let me know, though specs of all these are available online.

My intention in making this was that we could share measurements we have made showing before/after a variable is changed, as well as measurements of our overall final products. I understand changing a single variable isn't necessarily always helpful, however e.g. it makes little sense to compare very breathable velour pads with non-breathable leather pads without also modifying the bass response via front or rear venting, as these two changes need to be made in tandem really.

Some guidelines I think would be helpful:

• Please share details of your measurement rig!
• Unless you're using a coupler which complies with standards, it would be extremely useful to provide measurements against a well known reference, with limited unit variation, such as a Sennheiser HD6x0 series headphone. I know this doesn't at all confound the drawbacks of FPC measurements, but otherwise your measurement of what may look like a perfectly neutral T50RP mod is almost meaningless.
• Well-annotated graphs, comprehensible legends, and a reasonable y-axis resolution are welcome
• Restrained subjective description is valuable

I know I don't have the power to enforce rules, but if I could make one it would be: no claims made about any mods without evidence. Graphs or it didn't happen.

Stick around and hopefully I'll some findings of my own to share. It's tricky because once you have tweaked something to a point you're happy with you are cautious about reversing things in case you can't get back to that point again! Additionally, I'm a hypocrite because I'm lousy at making notes on what changes I've made, or even keeping past measurements at all, which is not scientific in the slightest. I might get hold of another pair just so I can make changes from scratch to share with you all.

I'll quickly share where I've got so far with my RPKIT50, as I'm happy to have finally produced something not awful. I've flouted pretty much all my above guidelines but I don't want to bog down this first post with before/after measurements etc. so just want to put this here to show one example of what is possible and to have a timestamp for where I got up to on this day!

My coupler is a cardboard box chosen to match the width of my head, with the contact area made of craft foam and acoustic damping foam, and measurements taken with a Superlux ECM-999 into a Scarlett Solo. The ~5kHz dip is not audible and needs investigating.

Thanks and look forward to seeing what others have come up with

 

[IAtaman]

My coupler is a cardboard box chosen to match the width of my head, with the contact area made of craft foam and acoustic damping foam, and measurements taken with a Superlux ECM-999 into a Scarlett Solo. The ~5kHz dip is not audible and needs investigating.

Thanks and look forward to seeing what others have come up with

Sounds like fun. I think the main issue for me, and I suspect many others as well, will be to put together a reliable measurement rig. Can you maybe tell us more about your measurement coupler and how you build it and how you control for measurement variation?

 

[solderdude]

Blue Monkey Flyer (head-fi) has done a LOT of experiments with Fostex RP drivers and mods and there is a huge thread on this.

I too have done some tinkering with mods and EQ and the Fostex'es can be made to sound good but ultimately the driver is limited in certain ways.
It was fun to experiment with for a while but in the end one can only achieve so much. More of a fun hobby thing.

Just like Dan Clark I realized that one can only take it so far and there are better sounding drivers than just trying to get the most out of the RP driver.
I found these drivers (modded quite a few) have substantial variations between them (mostly in distortion but also FR).

Nice headphone to tinker and play with. The drivers can take a beating and are not very fragile which makes them suited to 'play' with.

 

[isostasy]

Sounds like fun. I think the main issue for me, and I suspect many others as well, will be to put together a reliable measurement rig. Can you maybe tell us more about your measurement coupler and how you build it and how you control for measurement variation?

I've been doing some more tinkering about with my rig just today actually.

before I had a circle of acoustic damping foam immediately around the mic, and craft foam surrounding that. I had to push the mic out slightly further than the flat mount because the treble would roll off otherwise. Turns out this was partly the mic capsule itself, because I soldered on a different capsule and the high frequency response is much better now.

A picture would make more sense so I'll upload one tomorrow when there's more light.

In terms of controlling measurement variation, if I'm interpreting you correctly, just making sure the mic is in the same position each time is the most important thing. This isn't a problem if you build a proper rig like @solderdude, but because I've just got a measurement mic stuck through a box (and which I do sometimes want to remove to do speaker measurements), I do need to make sure it's changed position because this can impact measurements alot (again I'll share some evidence of this tomorrow).

Otherwise positioning the headphone for consistent measurements is easy. I cut the foam to the shape of the HD6XX pad so it is easy to match up most headphones in exactly the same position, and even slight variations don't introduce changes due to the nature of the coupler.

Here's my new measurement after messing about with my mic as I described above. So alot of changes, yet you can see the relative differences between the HD6XX and RPKIT50 measurement are almost the same as the earlier graph; at least in the same places. There's some definite differences in the sub-bass and treble.

Blue Monkey Flyer (head-fi) has done a LOT of experiments with Fostex RP drivers and mods and there is a huge thread on this.

I too have done some tinkering with mods and EQ and the Fostex'es can be made to sound good but ultimately the driver is limited in certain ways.
It was fun to experiment with for a while but in the end one can only achieve so much. More of a fun hobby thing.

Just like Dan Clark I realized that one can only take it so far and there are better sounding drivers than just trying to get the most out of the RP driver.
I found these drivers (modded quite a few) have substantial variations between them (mostly in distortion'closed back HD600' but also FR).

Nice headphone to tinker and play with. The drivers can take a beating and are not very fragile which makes them suited to 'play' with.

True, and I've had a good look through it. GREQ's posts are useful too. However, many of the posts are simply mayflower mods being shared over and over, along with the same old blu tac on the back of the baffles, dynamat in the cup, hyperbolic claims of how much better it sounds and no measurements to back it up. And even when there are measurements, no reference measurement such as a HD6x0 to compare. The fact no measurements show the extremely high treble peaking with pads deeper than the stock pads I measure in my RPKIT50, and the T20RP mkIII I had before (which can't be tamed with damping alone) makes me untrusting also.

The drivers in my copy do vary on measures such as distortion but nowhere near the extent yours did. Maybe due to being assembled from scratch?

You're right, I invested in these thinking I could make a good set of closed-back headphones. Reality: it's really hard, and probably not possible to make anything better than even affordable closed-backs available now. It is fun though, like you say.