- Apr 30, 2008
Thanks for this! I was also curious to try out crescent and mod it and Your end result looks pretty good. Though, I would still find it just tad bassy (mid-bass) as You do and I would miss the "sparkle" and details from up-top highs, though, I`m very sensitive to 6.2-7.5kHz (which is great You got rid of). Waiting for more measurements and tuning from You before I take jump on it (if ever). Like my KPE sooo much atm ;P
The resulting midbass was reduced quite a bit from stock. Now it’s only 3-4dB higher than the level at 1 kHz. It was 5-6 dB before, with the gap growing wider at lower frequencies. I suppose, if you want even less bass, you could decrease the rear volume even further. BTW, it reminds me of this measurement of the Moondrop Aria:
When I compared the sound with the Comply Sport Pro tips to the stock tips or the Sony hybrid tips, I found that they don’t sound that different from each other. They retain the Crescent’s overall sound. It’s not as big a change as the bass cut. Try it with a parametric EQ: Add that 7 kHz peak to a system that doesn’t have it and compare the sound with it bypassed:
The peaking filter: f = 7,017 Hz, amplitude = 7 dB, Q = 4.0
I tried it with the Crescent with Comply foam. With most music it’s not that noticeable unless you’re looking for it. At worst, it sounds like some hollow whistling. As for missing sparkle, I’ll take smooth and honest over fake sparkle resulting from the tall resonant peak at 7-8 kHz.
If you look at the response resulting from the Comply tip, you’ll see that it actually has a stronger response from 4 kHz to 12 kHz except for the spike. After 12 kHz, it’s definitely softer. So compared to stock, it’s missing air. But what’s really up there? I can still hear up to 15 kHz clearly, but 16 kHz is faint. When I EQ back that 15 kHz spike, making sure frequencies below 12 kHz aren’t affected by the peaking filter, I don’t hear that much difference with most music.
Difference: modded - stock
So close to Harman target, except for the air! I'm getting comply tips and i'll try and post a compensated measurement once they arrive
Here’s what I have for Comply foam:
Moondrop Crescent with Comply foam tips, left channel shown
Green: Sport Pro with Smart Core
Blue: TS400 from UE600vi
The S400 is more durable and washable, and it seems to damp those resonances more aggressively than the others. I also tried the T400:
Moondrop Crescent with Comply foam T400, left channel shown
Red: T400 tip not pushed all the way down
Green: T400 tip pushed all the way down the stem
Predictably, the less surface area the foam has in the ear canal, the less it absorbs sound and the less it dampens the ear canal resonance. The graph in green looks like it’s halfway between silicone and foam.
The annoying thing about these foam tips is that they start out with a shiny glaze coating which gets worn out with repeated use. It changes the acoustic properties of the foam, making it more absorbent. The S400 doesn’t have the glaze coating. It starts out rough and stays rough.
Another thing to consider is that maybe the Comply 400 series isn’t the best fit for the Moondrop Crescent. T500, S500, etc. might work better. The only one I have that slips on easily is the universal-fit Sport Pro model with Smart Core. Comply foam is relatively expensive, especially since they wear out. I should experiment with cheaper foam tips. We’ll see.
I tried a 3m micropore tape mod on the Crescent. If you hate foam tips, this might be a somewhat viable alternative:
Unfortunately it only really removes treble, doesn't actually get rid of the standing wave resonances. Any ideas about other materials?
Yep, the material right at the opening of the nozzle damps the IEM’s tube resonances instead of the ear canal resonances. If we go the other way and remove the grille mesh altogether, we get this:
Moondrop Crescent, frequency response with grille mesh removed
Red: Stock tips
Green: Comply S400 tips
Unlike the Sony MH1, there’s no foam plug in the tube in this one. It’s not hard to place the grill back in front of the tube when you’re done.
The ear canal resonances happen when the distance to the eardrum is an integer multiple of half a wavelength. This corresponds to 7 kHz and 14 kHz in the graph. The grille mesh doesn’t damp those as much, but the super absorbent Comply S400 does. What’s left are the peaks at 3 kHz and 9 kHz. Additional damping at the tube opening will reduce the one at 3k and kill the one at 9k. In fact, it’s exactly what it looks like in your graph.
I don’t have 3M micropore tape at hand, but maybe it can be used instead of the grille mesh. I wonder what it will look like with one, two, three layers, possibly with pinholes to tweak the acoustic resistance of the micropore tape filter. A combination of micropore tape and foam tips might get the Crescent a smooth response that still preserves the energy in the top octave. It’ll still counts as a reversible mod because you can undo all of the changes.
If someone hates foam tips, just stick with silicone. As I mentioned earlier, the 7-8k peak is really there, and can be heard in sine sweeps. But the one at 7 kHz at least wasn’t too bad with most of the music I tried when I EQ’ed the peak in and compared the sound with it bypassed. You wrote that you got better results with Sony hybrid tips, and my measurement does show the peak being shorter and fatter. Maybe the Sony hybrid tip retains its slant surface inside the ear, so that the exposed surface isn’t always at the same distance to the eardrum.
Maybe there are silicone sleeves out there that work similarly, like diffusers instead of absorbers. There might be one with an unusual shape that spreads out and blunts the peak even better than the Sony hybrid tip. Another alternative is to find an insertion depth that avoids the problem frequencies.