Discussion in 'Measurement Setups, Systems, and Standards' started by Luckbad, Oct 5, 2015.
Sorry for pic quality. No stand yet but I can get a headphone cup next to the mic fairly easily.
Thanks so much for that. How did you poke a hole through the yellow sponge? Also, what's the length of your PVC pipe?
You could also undergo the relatively easy and painless procedure of removing the microphone capsule from the end of the stick.
There are 3 total sponges, all cut into Pacman shapes and stacked in the tube with all their mouths facing the same direction, allowing me to fit the mic in at an angle.
EDIT: PVC diameter is 4.5" and length is about 6.25"
I just set this up:
Maiden measurement (HD600):
My first measurement rig (http://www.superbestaudiofriends.or...nt-setups-post-your-rig.141/page-5#post-35732) turned out to be more of a prototype. Here's v2:
It's the same idea as before, with a bit more polish and a shorter tube to avoid stressing headphone bands so much.
Instead of a CD glued to the business end of the 4.5" ABS tube, I swung by TAP Plastics and got two 6"w x 7.25"h sheets of 1/8" ABS plastic and glued them to the ends of the 6" long 4.5" ABS tube using the cement recommended by the dude at TAP Plastics. Protip: cutting ABS plastic with a jigsaw is a disaster. TAP Plastics is the answer to my jigsaw skills.
I used three yellow poly foam car-wash sponges from the hardware store, cut into the right shape to fit snugly into the ABS tube with pacman-shaped cutouts to allow the MiniDSP measurement mic to pass through on an angle. (thanks for the tip, @Thujone).
For the coupler I used the same medium-density open cell foam (this stuff http://www.amazon.com/gp/product/B003VYAXV8) and felt as on my prototype.
Measurements seem in line with others' rigs:
Here's the measurement protocol I've been using: https://docs.google.com/document/d/1sh_js_dPETMfkEoGfC_DWb6qD_EUMr5B-9WJZqvBYpY/edit?usp=sharing
This is too awesome.
Thanks to Sorrodje, I got pointed to this thread and set up my headphone measurements rig. Primarily a speaker guy, I used to make loudspeaker measurements, but headphones was a dark area for long. As family got bigger, time grew shorter and even that moved into late nights, my headphone experiences started, mostly by tweaking all the phones I could lay my hands on. Tweaking by ear, mostly, based on my prior experience with various materials. As measurements equipment is much cheaper and yet pretty good today, even with a cheap DIY rig the results are quite usable as a tool for sanity-checking and comparing mods. Too many words, here come some pictures.
Materials: cork plates, active carbon foam stuffing, a paper tube, soft foam sheet, EPA foam sheet, leather sheet.
(I got feedback the images are not seen - there are some problems posting pictures here, so I include an external link too). https://goo.gl/photos/LhALFopgTHeQbinp8
First version, UMIK1, without the outer coupler sheet.
Leather coupler, with various wool felt disks:
A few days later I got the Creatology foam, and made the final coupler. Also, added a headphone band support from open-cell foam, but I haven't used it often (needs extra pieces of foam for most headphones).
From the other side:
Setup REW, and measurement time. Leather coupler vs Creatology foam coupler, 1/6 octave smoothing:
A visual check on Tyll's HD598 measurements confirms the Creatology coupler (violet plot) is indeed better.
I made a few in-situ (headphone stays in place) repeatability tests and it's very, very good: I was surprised to get this good results from such a cheap setup.
The marks on the foam coupler help placing the headphones to the same position. The repeatability when the same headphone is removed and replaced between measurements is also pretty good, but nowhere near the in-situ repeatability. When I start a series of measurements for a given mod, I make a base measurement for the reference, then take it off, replace, measure again, and then repeat once more - only then I measure the mod against the avg reference.
Of course measurements help me judge faster the direction of a change, but I am aware of the limitations of measurements. The fact that I measured something doesn't make it an absolute reference point, but just a relative one, in need of correlation with subjective listening tests, and other people's measurements.
However, especially with mods, measurements help a great deal to achieve left-right FR alignment, check distortions, resonances, and any strange behavior. Most of that is also hearable, but measurements save me so much time. Thanks for all the tips on this forum!
Question for Experimented measurbators
Just measured a mistery headphone I've on loan.
FR - R ( all raw takes)
I'm afraid the dip at 7khz is an measurements artifact ? Did someone encoutered such a problem with orthos ? I did see some problem with Stats but not for orthos thus far.
Same kind of dip but less weird.... Bass roll off and Driver matching are horribad though.
Before any more publication, I'd prefer to ask if you think that's somewhat believable.
Use a sine sweeper to see if you can hear the dip. Do you notice anything missing in music? Even if there is some sort of sharp dip, they're usually hard to hear.
Does the dip translate into a nearby ridge on CSDs? If not, probably not audible and a measurement interaction thing (always almost related to the size of cup - it's a null / cancellation artifact). Especially seen in orthos and stats, since they effectively have sealed internal volumes more than other types of designs.
Holy shit... just listened to a sweep tone... I hear the 6,5 k dip and variations of the left right balance accordingly to frequencies.
@Hands : got the same idea. huge dip at 6,5, Sub bass and overally bad driver matching is easily noticeable.
@purrin : Will post CSD later or tomorrow... stay tuned.
So I ordered a UMIK-1 to dip my toe into the measurement world.
Would Yoga blocks with holes cut in for the mic and creatology foam stuck on be no good for a coupler?
I see others getting reasonable results with even more humble rigs, so the yoga block route seems worth trying. If nothing else it'd made a good prototype.
Made a rig of my own to help with modding. It's been great fun testing what works and what doesn't in REW!
Gift box from a dollar store filled with car wash sponge, hole dug straight through for the mic with very little finesse. The measuring surface uses adhesive EVA foam on top of a CD glued to the box. I used Paxmate (EPDM foam?) previously, but measured highs from 9khz onward seemed to roll off too much with that - the EVA foam seemed to be more accurate to what I was hearing.
For some reason, using the felt cutout ensured better seal around the UMIK's tip and prevented measurements from being too rolled off in the bass.
That box does make me feel happy every time I look at it. It's a party every time I measure!
Unfortunately, the only standard comparable thing I had around to measure was a Portapro (one take, 1/24 smoothing):
Looks quite similar to Solderdude's measurements on the DIY Audio Heaven site, although I'm still uncertain of the accuracy of my rig (which is why I haven't posted measurements of my T50RPmk3 mod yet - quite like the results though). Would really like to get a HD650 and maybe one or two other oft-measured headphones to make sure.
After having finally gotten around to making this, I'd say it's one of the best bang-for-your-buck things one in the hobby could do; a UMIK-1 is cheaper than many gear purchases, can be used for speaker measurements too, and the remaining materials can be gotten for spare change or stuff lying around the house. It's a great time to be in this hobby indeed.
Many thanks to the people who provided examples and advice to learn from!
Try to experiment with different felt (and no felt), and also the height of the tip of the microphone. The deep narrow dip at 4 kHz is likely cancellation for (partly) these factors - and of course the internal volume will always have null points, just like rooms. If it was possible I'd use 1/6 octave smoothing on the treble and 1/24 up to about 4 kHz (or ideally in-ear measurements for the treble). I don't pay much attention to the treble responses, I only care about consistency between measurements in order to be able to evaluate mods.
My rig looks worse than a 3rd grader's art project. I really don't have the tools on hand to make something nicer at the moment. The box is stuffed with that foam you find in cushions and I just cut up a box I had lying around.
Here are some quick measurements I did:
Derp those controls menus weren't meant to be in there.
Also DT880 250 since I had them on hand.
Can I just say, thanks, @purrin, for sharing your setup so that we could all go balls deep.
Alrighty new coupler ready, will edit with some measurements soon. Much more portable little unit. Acronym pending*.
Getting relatively consistent results now, I'm still not sure why my distortion is as high is it is but I imagine it's just a limitation of the mic? @sorrodje distortion on his HD800S was much lower in the midrange/treble regions.
@chakku sweet bro
Separate names with a comma.