Discussion in 'Measurement Techniques Discussion' started by Hands, Oct 20, 2018.
rejecting audio narcissistic douchebags, yeet
These are my results with the B1E dummy head using the Verum with quicky compensation. Wanted to check for L/R matching.
I like what I am seeing. The microphones are not totally flushed mounted. They are a few mm from the ear canal opening. There is a resonance a little bit below 5kHz, but it isn't super narrow.
Instead of straight up calling it the JUDE we could call it Reaction Against Naughty Dumb Yutzes.
EDIT: or Reaction Against Notoriously Dubious Yahoos
@purr1n Did you compensate for left and right ear differences, or does yours come out looking like that without? Not sure if yours has physically different ears like mine (which is by design, since modeled off real L/R ears).
I’ve been putting this project off because compensating for that difference has been a much bigger PITA than I expected. :/
Your quick results looking good nonetheless!
Rational (or Radial) Analog Noggin for Decibel (or Dimensional) Yield
Recognizing (or Rational) Analog Natural (or Neural) Decibel Yield (RANDY) measurement system.
I'm getting there...
Rational Analog Neural Decibel Yield (RANDY) measurement system!
The ears on mine appear to be the same. At least I can't tell the difference. I'm calling mine AMIR.
Analyzer of Measured Impulse Responses.
Reasonably Accurate No-frills Dummyhead Yokel
Do they look like the one on the official site?
I was told they couldn’t offer me true symmetrical ears, so I’m a bit puzzled now...
LOL, maybe it was accidental (manufacturing, placement, microphone, etc.) Let me take measurements of other headphones to see.
(resident | reference) (aural | acoustic) noise detection yeoman
Maybe having the mics flush at canal opening exacerbate ear differences in my case? No idea. The ears look almost identical on mine until close examination. Again, makes sense, since real human ears do just that.
I ruled out the mics themselves as a variable when I replaced stock with the ones from EARs. Same deal.
Maybe some headphones are worse too and just show it off.
This is what it is. Seems like planars, having a large less directional wavefront, don't get affected by the small differences in the ears. Results from Elex and HD58X, being traditional dynamics with their smaller radiating surfaces are not consistent, probably because they do interact with the ear structures more.
Sigh. I wish they could have offered symmetrical ears.
OK, so I’m not crazy!
They said they don’t have the ability to flip the mold of each ear to make them symmetrical. Maybe it’s just cost prohibitive. No idea. I got shut down pretty quick when asked.
Comp curve to adjust won’t make it perfect, but you can get it close. Maybe close enough to make it good enough. Could also be used to show how slightly different ears might affect the response. But no matter what, it can’t be used to check for absolute channel matching...
I have browsed around for other silicone ears that I might be able to swap out. Shouldn’t be too hard. Worst case, I could use the EARs ears. At least I get the benefit of a fake head structure then?
Randy’s a Neumann Dummy’s Yokefellow
Really Anatomical Noise Decoding Yardstick
I hope you forgive the necroposting.
May I ask whether the project is still alive?
I found your thread yesterday evening, browsing for self-built headphone measurement rigs. Funnily enough, right after ordering a pair of the silicon ears from Binaural Enthusiast
Unfortunately, it is really hard to find good quality and kind of average ear replica for measurement purposes. A lot of ears are moulded very cheaply, come with unnatural anomalies in size or shape, are too stiff or simply too pricey for a hobbyist. I purchased a pair made by Inspektor Gadjet a few weeks ago. Too small and hard for my taste. Another pair from a local supplier for accupunture needs was a bit softer, altough too large...
To be honest, I don't know if those variances really play a role for making somewhat meaningful measurements at this level. As grandma said: Every subject is morphologically individual, comes with its own HRTF expectations and reverse filtering, there is no such thing as a perfect average, yada yada yada,... You know the deal.
However, working with a good quality mid-sized pair lets me sleep better at night. I hope that the Binaural Enthusiast models fit that premise to some degree.
Speaking of premises, I also thought about reproducing the ear canal to ensure that the timbre is right. But in practice, things might come out the same: no standard, just a random guess of what should be a natural treble response. Nothing gained. And if done wrong, it can even make things far worse (introducing uncontrollable resonances) than measuring just flush at the entrance of the canal. Maybe a cheap replica of a coupler like the IEC 711 could do the job? But I haven't found a somewhat confidence-inspiring source for such parts, yet (which probably lies in the fact that I am searching for cheap replicas).
Maybe you have some ideas and impressions about those aspects.
I guess I will end up with the "keep it simple" attempt, skipping those subtleties which probably only introduce further errors and contingencies.
By the way, my current base unit is a cheap but sturdy mannequin head made of plastic, bought on eBay. Shape and size are quite realistic. So it will probably also allow me to capture some usable binaural field recordings. To eliminate inner resonances and crossfeed I will fill the whole thing with absorbent foam. To ensure a good seal I plan to place some sheets of cellular rubber around the ears.
So far I have worked with matched EM172 capsules by Primo: cheap, easy to implement, great SNR and overall decent audio reproduction. They are fed via a Simple P48 circuit published by David McGriffy / Ricardo Lee, connecting to a Babyface Pro for stationary use and a Tascam DR60D for mobile use.
I might try the AOM 5024 HD capsules by PUI in the future. The DIY cracks say its a great deal. Specwise, they are on par with the Primo... Maybe I will also find a reliable procedure to calibrate / match a pair by comparing it to my Beyerdynamic MM1 response anytime soon.
So much for my approach.
I hope to hear about your progress and state of affairs, soon. Maybe we can share some experiences and further improve our concepts
Regards (and sorry for the wall of text),
Lucas / Dr. S
That the ears weren't symmetrical threw me for a loop. Huge PITA to try to make a compensation curve for both, and even then, they always measured noticeably different.
I ordered some other ears a long while back that seem a bit more symmetrical, but also not quite as realistic. Haven't yet put them in.
Just hasn't been a priority working on this for a long while.
@Dr. S , the only place I know where one can find affordable 711 couplers is Taobao. Check out this HeadFi post. It has some links to some couplers.
My pair arrived today.
The ears aren't perfectly symmetrical, indeed. Also the material is quite stiff. If I member it correctly, the humar ear comes with a hardness of 3 to 10 shore. That is much softer than the most pouring silicons are. Thus, the pinna cannot deform correctly when pressure (especially by on ears) is applied.
But on the plus side, the shape is good and the surface is really smooth. Much smoother than the other ears I've had so far.
I couldn't find any fair priced alternatives, yet. There are some options on eBay and Aliexpress. But I doubt that they outperform the pair made by Binaural Enthusiast.
Maybe we should ask a supplier like 3Dio, POP Microphones, Headrec or Soundman for a "replacement part"?
There are also some models on Shapeways. Even some with accurate ear canals. Too bad that they are being printed with plastic.
I have seen that thread before. Actually, that's where I came from.
Separate names with a comma.