Jump to content
Search In
  • More options...
Find results that contain...
Find results in...


  • Content Count

  • Joined

  • Last visited


This user doesn't have any awards


About Nimrodor

  • Title

Profile Information

  • Location

Recent Profile Visitors

1,549 profile views
  1. What makes the Superflower Golden Greens in tiers A (Rosewill Capstone-, Lightning), B, and C (Kingwin Lazer Gold, EVGA B2) meaningfully different?
  2. Cheap potentiometers, like those often used in headphones, tend to have large channel imbalance at low levels since mechanical inaccuracies in the construction have a magnified effect in those sections. The recommendation that you not use digital volume control generally only applies if you have a good powered analog volume control like the potentiometer on a heapdhone amplifier (and even then, isn't always valid). The impedance mismatch and reduced damping from lowering the volume using an unbuffered series potentiometer like those used on passive headphones are massive compared t
  3. The MSR7 has a known fake model floating around, which has even gotten reviews of its own. That eBay listing is especially suspicious because the description is copied from the eBay description used by an actual ATH distributor, complete with links to their EU webstore.
  4. Sennheiser sells the original cable for $11 with free shipping in the US.
  5. Aliexpress reviews are for the transaction, not the product. Highly reviewed on Ali means that the seller ships the product as described in a reasonable amount of time, not that the product itself is good.
  6. They also have the Airdots 2 for the same price: https://www.pbtech.co.nz/product/HSTMIX28592/Xiaomi-Mi-AirDots-2-True-Wireless-Earphones---Blac If you're in no hurry, most Aliexpress sellers ship to NZ for free in ~3 weeks. You can get the Airdots for about half the price there.
  7. The flac is 7db louder than the mp3. The peak on the mp3 is at -5dBfs. Something is wrong with the levels in both files.
  8. Rtings has great test data. The rating numbers can be misleading; they're just a quick approximate ranking tool (in the case of comfort, based on subjective impressions, clamping force, and weight) that exists so you can filter and quick-compare things more easily. From past recollection (I don't own either of these anymore) the QC25/35 and HD598 are fairly comfortable with glasses due to their low clamp force. The Senns can be a bit iffy at the smaller settings because of the deceptively stiff pads which do apply pressure to glasses arms. Velour pads also have a tendency to squeak
  9. If you have the app, you can permanently disable adaptive sound control and turn off noise cancelling. Unfortunately turning off the ANC significantly reduces the sound quality, since the low and mid frequency response rely heavily on the ANC feedback loop.
  10. I'm not really sure. One idea is that the bias voltage on the microphone jack is leaking into other channels somehow – perhaps the impedance of the microphone is too low or the biased conductor is being shorted to another conductor. Depending on the wiring this could inject current into other signal lines.
  11. This is true for the resistor ladder DACs that are usually taught in most EE classes, but not in practice. Virtually all modern audio frequency sampling uses a different process which samples at an incredibly high rate then trades time resolution for level resolution. Although switching does occur, it usually happens at a frequency so far above the cutoff of the analog output filter that steps do not appear in the output signal. EMI is the most well-known source of noise, but in practice is rarely an issue; it only tends to be significant at high (non-audible) frequencies. The most com
  12. Notebookcheck includes frequency measurements for the speakers on the phones and laptops they measure. Example: from Galaxy S10 article: https://www.notebookcheck.net/Samsung-Galaxy-S10-Smartphone-Review.414549.0.html#toc-emissions. In general even good phones don't become noticeably loud until above 100Hz.
  13. It doesn't. Electrically it's still wired in series with both headphone channels.
  14. He's planning on showing the quantization error caused by the conversion from floating point back to integer bits. It'll definitely be measurable at 16 and maybe measurable at 24 bits, but at a true 32 bit depth it'll be well below the noise of the best electronics currently available. Unless Microsoft managed to spectacularly screw up simple rounding. FWIW LSB errors at 32 bits are comparable to the thermal noise of a 1mΩ resistor. If your signal chain contains a series resistance of greater than 1mΩ, it likely has a larger contribution to noise than the digital volume control – e
  15. This was true 10-20 years ago, before processing got a lot better and before the proliferation of cheap 32 bit delta-sigma ICs. This is true. However, the processing bit depth is virtually always 32 bits. Even if you attenuate by 16 bits (which is absurd), there's still a full 16 bits of dynamic range remaining. Dynamic range limitations will NEVER be an issue on a properly configured modern computer. Quantization error might be under some circumstances, depending on the thermal noise contribution of the potentiometer. See Benchmark's application notes on the lengths they have to go to