Discussion in 'Headphones' started by Pyruvate, Nov 18, 2016.
assuming that the back of the driver is open aside from the cup damping
What's the point of the new chassis exactly? A new plastic chassis with a new shape. I like the stock cups just fine.
Correct. In my limited experience changing the inner dimensions / shape of cups (even open backs) has a measured effect. In the case of T50RP, Open Alpha cups flattened midbass bloat without seemingly any drawback both measured and subjective. I also feel soundstage width increased as well as possibly adding some delicious reverb flavor. I am very much interested in hearing @zach915m wood cups.
Spoiler: HE500 vs HE6se vs Open Alpha
Yesterday I was preparing to start a/b/c comparison of HE500, HE6se, and Open Alpha (OA). But then switching from HE6se to OA I felt compelled to mod OA due to the tone and bass bloom, so comparison delayed…
While listening to Open Alpha you're always aware that the sound is colored (fun). The only time it comes off as natural is when listening to unfamiliar music. Listening to HE6se afterwards is like turning off the equalizer. Extended exuberant listening to bassy music on OA led to some fatigue from pressurization which is what really prompted the need for fixing / venting.
As I was planning to add venting (via drilling baffle, or placing a filter between baffle and pad mount ring) I realized I was turning my Open Alpha in to the T50RP Open Alpha hybrid I made earlier which was a combination of Open Alpha cups with T50RPmk3 driver baffles. I think I will just rebuild that. Im going to order another T50RP (variant) so I don't have to disassemble this one and can a/b afterwards.
Here are EARS measurements of the OA mk3 baffle hybrid VS OA. Same pads on each (Dekoni leather) but the hybrid version has little to no cup damping material applied to cups so not exactly comparable. *Remember mine is unusual because Ive removed the driver's stock damping paper which alters the frequency response a lot so ymmv! For comparison EARS measurements of HE6se w Dekoni hybrids.
T20RP mk3 arrived today.
- Sounds great for $100, Id easily put in the same category as 58X for value. From memory, 58X more neutral and cleaner bass but T20RP a little more resolving elsewheres.
- L/R balance is decent. Ill open them up in a bit and see if I find more excess glue issues.
- FPC measurements because the stock pads are too shallow for EARS
Spoiler: T20RP FPC
Spoiler: T20RP is the "open" version
Look, you can see right through the vents. From what I remember, the piece of felt over the vents on the semi-closed T50RP adds practically zero acoustic impedance value. I wonder if T50RP has any noticeable noise isolation improvement over T20RP or if its all marketing.
Interesting discovery? Ive not seen mentioned elsewhere but Im no Fostex expert like @E_Schaaf . T20RP and T50RP use different density baffle foam inserts and I suspect it's really the key difference between the two in terms of sound. I always suspected that the different Fostex (T50RP, T20RP, THX00, TH500RP, etc) used different density foams but never had a pair side by side before. Prior it wasnt clear that different foams were produced except when the Fostex RP kit (aka never publicly released Massdrop T-X0 v2) came about which came with two different sets of foam for personal tuning. The only Fostex I've seen with very obvious different baffle foam is the T60RP which is very light density and shaped differently, I believe T60RP also the brightest amongst the mk3 RP family.
Below is T20RP (marked with blue tape) and T50RP baffle foam. Not a great pic but the visual clue to key in on is how much more ragged the T20RP foam looks in comparison to the higher density T50RP foam. In your hands and with breath test the difference is obvious.
Spoiler: dim pic in the PNW sun
edit: I just remembered and must clarify that my "T50RP" baffle foam insert actually came from the Dekoni Blue. So I dont know if my baffle foam is representative of regular T50RP. Anyways, Ill switch foams and see what happens to T20RP measurements. I know modders have messed with attenuating the venting through other means (I tried micropore taping them closed) or using the Mayflower custom foam inserts to seal (they are closed-foam if I remember correctly). I didnt prefer the Mayflower inserts myself over stock T50RP (not Dekoni) baffle foam just like Im not sure I prefer the sealed front-side Open Alpha to the Open Alpha hybrid (OA cups with mk3 baffle).
T20RP before / after Dekoni Blue baffle foam insert, Left-side only and averaged, FPC:
Spoiler: T20RP before after
Initial impressions subjectively confirm. Not sure which I prefer. Ill definitely try both sets with Open Alpha hybrid.
And there is glue on paper behind drivers which renders those areas impermeable. Will removing them fix our mild imbalance? Let’s find out (later).
The source of the glue is not from attaching the driver damping paper. The paper seems to be pre-adhesived like a sticker. The glue is from another part of driver… Where they connect the metal magnet bracing / scaffolding to the plastic driver capsule (?).
Spoiler: driver glue
Reinstalled original T20RP baffle foam so its stock aside from removing the driver damping paper filter.
- Sounds nice, tonally reminds me of my Open Alpha. Smooth, dark, and warm.
L/R balance improved:
Before / After paper filter:
Resonance because impermeable wall (gluey paper filter) too close to back of driver Id guess, seen similar result in vintage ortho Yamaha driver transplant to larger enclosure. Maybe acoustic impedance effect too.:
Better? But the frequency response has changed a lot so not super comparable.:
* CSDs probably deeper than necessary, for effect...
^ The ridges around 1kHz and below may be reverb trails if we're lucky. If so maybe cup filler will help but preferably(?) not too much. Ill compare against Open Alpha cups later with an overhead CSD.
Open Alpha hybrid assemble:
Spoiler: apply gasket foam tape around edges
Spoiler: cup comparison
Spoiler: cup comparison (2)
Spoiler: Always use protection
Spoiler: Open Alpha hybrid VS Open Alpha
^ The Open Alpha (blue) has Dekoni pads and ZMF strap.
KH-Open Alpha Review
Roughly a month ago, my audio friend @khbaur330162 loaned me his build of Open Alpha built for a fairly generous term. As I was highly curious about the OA project, I was excited to evaluate one myself.
(Additional disclosure: Cups were printed by @Philimon and I believe he didn’t change cup dimensions specified in the project page/github)
To my knowledge, OA can vary a lot across implementations because there are so many variables builders should decide. Major modding levers include:
Drivers .. specifically which revision of T50rp
Damping material filled in cups
Ear pads (dimension, material, perforation, etc)
Removal of paper acoustic filters attached behind drivers
Full vs partial vs zero sealing of baffle vents
As per KH, his OA pair was built with:
T50rp mk3 drivers
0.33g (per cup) angel hair -- thin synthetic material
T50-compatible sheepskin pads he purchased in Ebay: https://www.ebay.com/itm/265928328595?var=565765669995
Paper filters removed
Baffle vents sealed slightly but not fully
Paxmate used to line housings and baffles
My listening is done with HO200 (4-pin xlr out) and M500 mk3 as amp and dac respectively. And because this pair is neither for sale nor commercial products having established MSRPs, I asked him how much he would hypothetically quote if someone commissions then he estimated well below $200. So, I mainly compared the loaner to Senn HD650 due to price proximity. Other comparisons may be additionally posted upon requests.
The OA I got weighed 395g without cable. I believe this OA build largely follows the original T50rp in terms of aesthetics and comfort. Generally fine by any means. But I found the default head padding wasn’t quite enough for me so I slapped the ZMF Co-pilot head pad (much larger surface area to distribute weight + better cushion) during my evaluation.The ear pads KH used were pretty comfy.
Evaluation By Tonal Regions
KH’s OA is not very tonally accurate to my ears. Not quite tuned for neutral heads like me. But I could discover many good traits with OA. Let me elaborate.
The bass extends well down to the lowest register, but there was a strong (4-5 db) hump in mid bass, which makes both the sub bass and the upper bass a bit subdued relatively. Bassy and bloomy. I could hear some slamming qualities that are a bit different from the ‘bouncy’ feeling I found in the later DCA/MrS Ether series. Downside is, while I would not say bass is one-noted, it also lacks fine gradations, compared to HD650 or (modded) Yamaha orthos.
Midrange sounds a bit detached from bass possibly because there isn’t enough energy between upper bass and lower midrange. Vocal tracks sound a bit hollower than HD650 while OA has better bodies. I could also enjoy snare drum’s barking a bit better with OA. Electric guitars are less present. Acoustic guitars are tad better but still lack presence.
Treble is interesting. I immediately hear OA’s treble does not mix well with its midrange, but this actually helps OA a bit by isolating highs -- its treble doesn’t sound too withdrawn or masked by lower frequency ranges. Let me put it this way, HD650 definitely exhibits a higher energy level in highs but I hear more sparkles and crispness with OA (i.e., better ratio between lower and upper trebles). Piano and vocals are more present with OA. Sax is more biting, too. Shortly OA does not have any ambiguous parts in highs unlike HD650 that is notorious with its own laidback nature in presence region (‘veiled’). Top end extension is on par with HD650.
Please note that I don’t think OA’s treble is excellent enough to punch up. There’s some slight glareness that occasionally bothers me in some tracks. There are several things going on specifically in boundaries of presence regions that are over-excited by certain passages.
Overall, I’m inspired to say OA is tastefully tuned in its own light. And I guess KH tuned this to be complementary to his other DIY headphones collections. He said he enjoys a lot of sibilant modern tracks and OA worked great for that duty, which is fully understandable and consistent with my findings.
Evaluation By Technical Performance Criteria
I think this section sheds more light to what OA or T50rp mods are capable of, because I believe that tonal traits are heavily affected by treatments of internal reflections and interferences but drivers’ inherent potential is more weighted in many technical performances.
KH’s OA sounds moderately resolving in terms of both dynamics and details. While (as discussed in the prior section) OA’s tonality is more uneven and there are many areas that are severely recessed than HD650, I don’t feel like I lose small gradation or nuance too much almost anywhere in the audible spectrum, which indirectly suggests OA’s resolution is very competitive. Midbass is slightly more congested than I like though (and this is probably the main reason why I hear OA’s bass lacks microdynamics).
I would like to add that OA’s resolution is amplitude-dependent. In average or louder parts of songs, OA reproduces small changes totally acceptably to me. However, I often fail to monitor subtle sound developments when things are played around the ambience level of the recording venues. I am not quite able to hear ambience or environmental effects induced by the recording venues, either. HD650 does these things (at least in mids and highs) a lot better.
What I complained about in the paragraph above is also associated with macrodynamic compressions I hear. By construction, total or long-term macrodynamics are defined with the difference between lower and upper limits of a song’s amplitude distributions -- correlated with total crest factor. In contrast, short term macrodynamics is a matter of transience as it’s mostly over a very short window (often measured as short-term crest factors). During my evaluation at my usual listening volume, I didn’t experience any issues with OA’s reproduction in the loudest parts of songs. But the lower limit loudness was more limited with OA than with HD650. This harmed OA’s climax catharsis in large scale symphonies.
Soundstage was a bit inaccurate but highly enjoyable. Larger and deeper stages than HD650. A little cup reverbnation and airy feeling on top ends make stage ambience feel more lifelike even with unprocessed and closely mic’d recordings. Layering and separation are not as good as HD650 though.
My measurements were taken with the MiniDSP EARS at 95db [email protected] with the homebrew hybrid compensation target curve that mixed HPN, HEQ, and additional adjustments based on B&K Room Curve.
Please note that my frequency responses are averaged results based on 5 different positions (center/up/down/front/back) to control positional variance. Results from optimal positioning might differ even with the same measurement fixture/rig.
Bass hump at 100 Hz is exactly what I heard. Dips/Scoops at 300 Hz and 3.5 kHz are possibly associated with what I perceived as detached/isolated tonal regions
Upper mids and lower treble sound pressures are measured much lower than the midrange peak by 10-17 db. I’m not sure I hear attenuation to that extent
Noise reduction is fairly good, particularly for 3 to 5 kHz signals
One-off builds are often difficult to make any meaningful or informative conclusions for multiple reasons. First, you cannot buy the same thing I review at all unless the builder put his/hers on sale or they run their own businesses. Second, it is precisely tailored to the builder’s need and taste, not necessarily toward general markets. Finally, DIY building is an endlessly and iteratively ongoing process -- this can be a big hassle for commercialized products (what MrSpeakers used to do previously) but rather a great beauty for DIYs. So any evaluation should not be over-interpreted beyond a snapshot at the moment.
As in, I liked and enjoyed KH’s take on the OA -- maybe a lot better than I stated in this review. Most shortcomings or drawbacks I listed are as a primary/main pair of headphones to evaluate audio and music. As a secondary pair to enjoy different presentations/flavors, it almost gets jobs done.
That said, if I build my own OA or commission him to build a pair, I may propose the following adjustments:
Fill cups with more damping material
Unremove paper filters or try out different backwave damping acoustic filters
Try out whatever to make response more neutral + Equate tonal levels across the spectrum
Aside: There are certain things I deem attributable to the T50rp driver’s limitations, which I’m totally fine to live with for this price class.
All things considered, I maintain that the OA is a wonderful open-source project (perhaps Dan Clark’s biggest contribution to the hobby). I’m looking forward to seeing more people coming up with their own pairs and respective magic ingredients.
Fantastic review @Vtory .
You are correct, I didn't alter the 3DP cups from original dimensions however these are a community edited version that allowed use of screw inserts for extra durability if one is to open+close the cups a lot for tuning reasons. Requires a very small change to internal dimensions.
As you explained Vtory, comparing diy headphones against each other is tricky. @khbaur330162 and I both veered off the original plans in our own way to suit our preferences which is the beauty of diy as well as the OCD curse.
I wanted to write a comparison to HE6se and declare Open Alpha (OA) winner in terms of overall preference. Open Alpha is reasonably technically quite adept in most ways for a cheap closed back as you found when compared to comparably priced HD650 which makes it completely adequate for this mid-fi budgeteer. When I started my a/b notes I quickly realized that OA needed some adjustments which is why I went back to Open Alpha - hybrid. I know OA could not beat HE6se in terms neutrality or technicality (Id argue that OA could trade back some blows), but I find the OA tonal signature really quite pleasing even if obviously colored as I have it. I've chosen to listen to OA on more occasions than HE6se even when environmental noise is not a factor. Im excited to finish my OA mod so I can get back to writing an a/b comparison and argue in OA's favor (hopefully).
Midbass is not the most nuanced but well within acceptable margins of enjoy ability unlike HD650. My OA is still a work in progress as I find stock baffle under pad venting a compromise of bass microdynamics vs bass bloom.
Id like to have finished modding OA already but Ive been sidetracked by other projects. Like right now Im without my laptop because it's tethered to a 3DPrinter which means I now dont have source for my desktop music rig so am relegated to portable rig this week. A situation like this is where something like an ETA Mini-s or Grado RS2x would come in handy since they work well off an iPhone... ya! more headphones (serious/sarcasm)
In terms of sibilance Ive always found that a fine thread to walk in past T50RP mods ime. Even a small measured change (for example) didnt very obviously explain how sibilant they could sound which is why Ive been lazy thus far to try altering course from the Dekoni leather pads.
That's a strong statement. I've not heard any of DCA's headphones so I can't say but he has made so many other headphone so that's an interesting observation you make. From what I've read the DCA Expanse looks very interesting but way over my budget. I asked Dan Clark for an SBAF loaner but he said he only lends to professional reviewers with websites and not forums. Reasonable considering the cost of his wares and the investment Im sure he has tied up in all of it to tolerate possible forum randos... though I was very clear that the loaner could be limited to the most prominent and credible members on the forum. Otherwise my conversation with Dan was very amiable, informative, and was open to discussing anything I asked regarding the product which is rare for a MOT ime. The only other ones that transparent and friendly that come to mind are Schitt, ZMF, and Nectar. Another thing, Dan shared an interesting hypothesis about bass impact perception which Ive not taken his advice yet to explore... but I will eventually - its in the que and Ill share my trial results. I hope Dan changes his mind to lend us an Expanse, but I can imagine he is indebted to Jude (reasonable business decision but hopefully that changes) which makes SBAF off-limits.
edit: I don't want to clutter the thread with too much incremental mods and half-assessments. Ill update with more concise and conclusive reporting later. Original post contained in spoiler.:
Spoiler: caution! rambling, nothing to see here...
Spoiler: Open Alpha - hybrid #2
Spoiler: OA-h m2 (EARS).jpg
OA-hybrid VS Open Alpha:
- The bass "bloom" issue that I am sensitive to is gone with the hybrid since it has the mk3 baffle / under pad vents. fyi, these are using the Dekoni Blue driver baffle vent foam. I've not done an a/b yet so can't comment on other changes atm.
OA-h VS T20RP (mod):
From memory I cant say OA-h is definitely preferable to T20RP (with stock flat pads and stock driver damping paper removed). I spent only a little time with T20RP, but I might buy another for comparison and to maybe use. Im a fan of the flat stock pads. Stock flat pads are so shallow that the driver lays pretty much flat on your ear… Unfortunately the stock pads are not a good match for the OA cups.
edit: organization and addendum
I forgot to mention that I added a variable to newest OA-h by adding larger cup venting. Ill update to share effects of closing those up. But the basic reasoning for why adding larger vents is the same as why @khbaur330162 modified the seal on his cup-baffle gasket. Ive not experimented much with this so could be better optimized. Adding larger vents may not even be necessary or possibly detrimental in a trade-off sorta way. I dont know yet. Will update in next post.
My next mod experiments will be attempting the much more tedious driver damping mods that were shared in the sbaf mk3 thread. Im not looking forward to this part because takes much more time to install these mods and Im not displeased with how these sound currently as-is. But I know there is room for major improvement as seen in sbaf mk3 thread. Big thank you guys for sharing your mods.
Ill definitely add some a/b comparisons when mod completed, particularly against HE6se and HE500. Id like to pickup another T20RP (or T60RP) and another 6XX so those might get some a/b comparisons later.
The headband support in pic is this (link to Amazon). I originally purchased for HE6se but it didnt work out because the added padding basically reduced the amount of headband length adjustment so became not enough to accomodate my skull plus headband support.
[Continuing adventures from here.]
Previously left off with stock T20RP cups and stock driver damping removed. Next let's add Open Alpha cups.
Spoiler: OA-h T20f Dekoni pads (FPC)
* mod1 = Dekoni leather pads, T20RP under pad vent foam, and some basic schitt in cups (1x layer felt, arctic cotton)
Spoiler: OA-h VS T20RP mod (stock driver damping removed, stock flat pads)
Spoiler: OA-h mod1 CSD
^ CSD less nice than imagined. Perhaps extra cup/driver damping will improve.
I realized what I liked most about T20RP is the lighter density foam used in the under pad venting which makes for the least bass bloom. I imagine like open baffle speakers vs acoustic suspension speakers and how their respective bass cancellation attributes effect room modes and decay. CSDs with extended limits could shed some light. fyi Id guess T20RP foam approximately density of Grado f-pads.
- Felt is SAE certified with much better consistency in density than typical stuff. The particular one I used is of very light density but can be purchased in other dimensions and densities.
- Arctic Cotton is some out of production product but I was kindly donated some extra surplus by @khbaur330162 . I use as a high density felt basically.
I hate to admit but I dont think OA cups are a complete win over stock cups. While OA have larger overall volume, they may actually be shallower in center area than stocks cups. The OA cups have an unecessarily large internal hump for where the headband joint attaches. This hump comes nearer to the back of the driver than stock cups. I think it may be the reason for OA’s stronger 1kHz resonance over stock cups.
If true, then I can modify the OA 3DP file to reduce this hump and/or make the cups deeper. Another option is to shave down my already printed cups with some sanding tool. But before going to that trouble, let me try measuring the drivers without any cups to see if this hypothesis has merit.
Not sure that cup volume/fill on its own will suppress that 1 kHz bump; I tried different amounts of cotton wool in the T50s without it making too much difference there. Instead (or at least, before doing anything drastic with cup structure), try more damping directly on the driver backs. See the before & after curves here (using gauze bandage) and @spoony 's result here using cosmetic foams.
FYI, in the open alpha thread on the other headphone related forum, in post #642, page 43 a contributor shared files for printing half sphere cups.
My first attempt to apply custom back of driver damping didn't go very well. I used magic eraser which is like a very high density open cell foam while still having good airflow. Its what @E_Schaaf used in his T60RP mod. The result was decreased bass and increased treble. To balance for the bass loss I switched from T20RP foam for the front-side under-pad vents to the denser Dekoni Blue foam. The tradeoff however was decreased sound stage and increased bass bloom. I don't think the trade offs are worthwhile for my particular preferences. Bass bloom is the reason why I've gone hybrid. I know I'm particularly sensitive to this issue so I don't imagine others would care. I could try like you @Biodegraded to custom seal most of the front-side vents while leaving 1-2 completely open which could fix my issues with bass bloom while still providing plenty of bass. The work you guys shared in the mk3 thread is awesome!
Spoiler: OA-h mod4 (1st driver damp attempt), FPC
^ Custom sealing some of the vents would yield more bass which is what Open Alpha regular provides.
I know there is a lot more nuance to the mods you guys did... Im lazy and just slap shit together. That's why pro modders get paid the big bucks.
Im still going to measure the naked driver out of curiosity. I did this with vintage Yamaha orthos and in most of the cases I found 2kHz resonance would disappear after removing drivers from cups. I would then seal back of drivers with micropore tape for damping to produce an even response and speed up mid-high transients, but 2kHz resonance would resurface as soon as I put the drivers back into the cups. This is why I guessed the T20RP's stock damping paper that was covered in glue was causing the 2kHz resonance. And the shallow cups are my suspect for the 1kHz resonance. If cups are not the issue then I will very much appreciate copying your mods in the mk3 thread because they look perfect.
An aside but something I didnt really discuss because I dont trust my distortion measurements due to environmental noise, but whenever I measured vintage Yamaha ortho drivers with damping materials close to rear of driver it always had the trend of increased broad distortion. I worried this was an effect of possibly reflections caused by possibly too close and dense damping materials too near rear of driver. Subjectively, I felt soundstage and blackground decreased.
FYI, there’s no rules against linking to other sites here. purr1n isn’t petty like Jude.
FWIW, here's where I left things at with my T50RP mk II mods:
Dense open-cell foam lining the back of the cups to provide support for the cosmetic pad.
Pre-perforated open-cell polyurethane cosmetic pad (can't remember where I sourced it, probably some random beauty shop).
Hole cut from stock damping paper to align with cup vent.
Blue fiber taken from Scotch Brite Non-Scratch scrubber, covering the whole back side of the driver, providing an airflow-increasing buffer to the cosmetic pad. This can be replaced with various felts and foams to tune the overall downslope to taste.
Plasticine or maybe tungsten putty on the back of the baffle (not the most impactful to sound IMO).
Adhesive window caulk foam around the baffle where it meets the cup to ensure tight seal.
Small-cavity earpads are a must for me for treble balance, these drivers can sound scratchy in the treble super easily, especially using larger or deeper pads. These pads are ETA OG G pads but I also had good results with a variety of ATH-MX0x replacement pads too, you can source sheepskin pads from Aliexpress for cheap. They can't be mounted on the cups so I taped them to the baffle.
Here's a FR capture from my in-ear rig:
Sorry for no comparison plots, I don't have time to do measurements right now.
My personal target is a smooth, ~6-10 dB downslope fom ~100 Hz to 20 KHz and this was close to that.
Separate names with a comma.