Looking for a relatively inexpensive (<=£50) USB DAC (preferred) or headphone amp to power my Sennheiser Momentum cans at work - they seem to need slightly too much juice for my onboard card (18 Ohm impedance). Can anyone recommend me one? Ideally it should be either mains or USB powered, but can be used at the same time as receiving power.
If you can wait try one of these http://www.ebay.co.uk/itm/PCM2704-U...236?pt=LH_DefaultDomain_3&hash=item3f44485724 I have had one for about a year, its utterly fantastic for ten quid. Also can be modded, I have changed the caps in mine for better ones and it improved it even more. I run a pair of Sen HD429's off them no problems
DAC: UCA 202 or 222 (WARNING: The headphone-out is dire, having a monstrously high 50ohm impedance. Only get it if you intend to use it with the line-out and a separate amp). There's also the ODAC if you want to integrate it with the O2. Amp: O2 (Objective 2). Unless you intend to use it as the input for some crazy audio processing chain, there's no reason to get anything more expensive other than looks or branding. A budget option would be the Fiio E10, an integrated DAC + amp combo that, while not perfect, is pretty solid for the price. In all honesty, unless you motherboard line-out is REALLY terrible, you can probably get away with just a headphone amp and forego the DAC.
It's the output impedance of the DAC's headphone amp that's the issue. At 50ohm, it will cause distortion (due to impedance mismatch) if your headphone's impedance is anything below around 400ohm (8x rule to minimise distortion to below audible levels). Which is pretty much any headphone. The HD 445 has an impedance of 52ohm, so will have noticeable distortion due to changing circuit impedance at different frequencies. It will work in that sound will come out of the headphones at a reasonable volume, but it will be distorted. The lowest headphone impedance you will generally encounter is around 16ohm (ipod buds, IEMs/canalphones, and the like), so the highest acceptable output for a headphone amp, unless you have a set of old-style ultra-high-impedance studio headphones, is 2ohm. 'Modern' (i.e. last few decades) studio headphones are also low impedance; the ubiquitous MDR7506 is 63ohm, so would need an amp output impedance below 8 ohm. Notably, there is no concept of 'impedance matching' for audio signals. For RF (or raw unencoded or unmodulated audio over many hundreds of km), impedance matching is important to prevent signal reflections between source and sink. This is NOT the case for audio signals going from an amp to speakers/headphones.
Onboard audio isn't the best thing to compare to when testing for distortion, and using your ears alone (i.e. without measuring the frequency response) is absolutely not going to be effective in identifying distortion. Your ears are really bad at identifying distortion without a direct and immediate comparison*, because your brain's auditory system is very fast at adjusting what you hear into what you expect to hear. this is why soundmeters and spectrum analysers exist, and why even professional mixing engineers will double-check with the meter no matter what they think it sounds like. * And whatever you;re comparing to needs to be measured to confirm you;re comparing to something correct anyway!
If you are not a sound mixing engineer, all that matters is what you hear. A tiny amount of distortion showing-up on a sensitive meter matters not a jot, if you cannot actually hear it.
Because it's excellent value as a quality DAC. I use one myself. It's just the build in headphone output is no good for the vast majority of headphones. The distortion from a 50ohm output impedance is not tiny, particularly with low-impedance headphones. OP's momentum's are 18ohm, so a 50ohm output-impedance device will result in roughly a 12dB variation in frequency response across the audible range. a 1dB variation is below the audible level, so when compared to a correctly set up output, the high output impedance from the UCA202/UCA222 will very much give an audibly distorted output. The problem comes from trying to 'hear' this distortion in isolation. I recommending watching the Audio Myths Workshop video I linked earlier for why relying on just listening without measurements will make it very difficult to identify distortion. But does this distortion have a bearing on music? Very much yes! Such a massive variation in frequency response will completely change the sound you hear. This is easy to test: - Get a Fiio E5 (under £15), a tiny battery-operated headphone amplifier with pretty low distortion and a very low output impedance. - Attach it to the line-out of your UCA202/222 - Play a test tone through your UCA202/222 - Adjust the volume of the E5 and the UCA202/222 so they are the same volume (IMPORTANT. Comparing two sources with different volumes is effectively worthless) - Swap your headphones between the uCA202/222's output and the E5's while playing music of your choice. The difference will be very clearly audible.
Cheers for the recommendations, managed to get a Little Bear P2-1 for cheaps on eBay (£17 incl. postage from China) - will see how that fares
I know I'm late to the party, but if it's just additional volume you need, grab a Xonar U3. It can't hold a candle to an expensive DAC or amp, but it will shove out plenty of volume at decent quality for next to no money. I can't tell much difference between it an my old Fiio. Disclaimer: I am not an audiophile
Whether sound is pleasing to the ear comes down to a very personal, subjective judgement of what is pleasing to the ear. Objective measurements can only tell you so much, they cannot tell you whether the character of the sound reproduced is going to be pleasing or not. For example, testing a valve amp may show an apparently significant distortion which would horrify someone who wants their sound to be clinically accurate but, be pleasing warm and mellow to someone who likes the valve amp sound.
They can, however, tell you how the sound is being produced, and can also tell you if it is being reproduced as intended by the audio engineer. If you prefer some distortion out of personable preference, that's fine. That distortion can be implemented in software though, giving you the best of both worlds (an accurate output switchable to your preferable distortion), whereas attempting to correct a distorted amplifier is rather more difficult. Calculating and replicating a non-linear transfer function is easier than trying to calculate and apply it's perfect inverse. Bob Carver famously did this back in the 80s in hardware (by tweaking transfer function coefficients manually using posts) but with modern computers it can be done entirely within software.
50ohm output impedance to 52ohm headphones is pretty good matching, even with RF applications that is not bad.
For RF applications it's an OK match. For audio applications it's not good at all. Signal reflection (what impedance matching minimises) is a non-issue for audio frequencies at short (sub-kilometre) distances, but frequency response is. Having a high-impedance output will result in an uneven frequency response. It will also very depending on what headphones are attached (because headphones have different effective impedances at different frequencies), so an amp with a high output impedance will sound one way with one set of low impedance headphones, and completely different with another!
