Updates - Bravo amp, Laptop settings, Senn HD650
Might be speaking to myself, and if so then maybe this'll just help me choose things in my own head. Replies would be most welcome.
Amp
I put a bunch of snipes on eBay until I won/bought a Bravo amp at $10. Ended up with the V3. Would've preferred a V2, but who cares for ten bucks? My understanding of the tone controls is they are all passive attenuators, so I am just leaving them all turned up full, which should be close to flat, or bypassing them. They do have a mid-point notch that visually and tangibly feels like it should be the mid-point, but from the circuit layout, this seems misleading. As the tone controls are turned down (and the volume up to compensate), the sound feels claustraphobic and constrained. When up full (and less volume to compensate) the sound is open and relaxed.
The Bravo amp is a better than the HP/out on my laptop with my Senn HD435, but nothing to shout about. Of course, one can argue that the laptop HP/out is the limiting thing in this setup, and the amp is just reproducing a mediocre source perfectly. That's a story for another day. When the wife's not looking, I'll pull out the hifi find a proper line out from the Arcam AV8.
There's a nasty spike as the amp is turned off, so I have learned to unplug the 'phones before turning it off. FWIW, the connection is a Cambridge Audio Pacific 3.5 to RCA cable that probably cost more than the amp.
Headphones
The HD435 pads arrived and were fitted, which made them seem like new. Took them and the PXC450 on a 5 hour train journey (10 hour return) with my Sansa Fuze. Whilst the HD435 are better in my quiet study, the PXC450 are much, much, much better on a noisy train (in the same way that I loved them transAtlantic last month).
I had the opportunity of new HD650s at £199. So I took them. The Sansa Fuze sounds muddy through them, which is a disappointment, but perhaps not too surprising. The laptop to Bravo to HD650 sounds absolutely amazing. Surprisingly, the laptop does a reasonable job with them, but the bass gets loose and the mid- and upper-presence and definition is lacking.
Source
Massive improvement here! Separate from the Windows audio mixer, there is an IDT Control in Windows Control Panel. It is colouring the audio. I can turn it off which gives a big improvement, but every time the computer is restarted and/or every time I plug headphones in and out, it is automatically turned back on and I have to go into Windows Control Panel to turn it off again. No updates are available, and there is no obvious option to disable it. I need to experiment with deleting drivers and seeing how to keep the audio ports working without this annoying "feature."
Of course, you should all be screaming at me for using a laptop headphone-out as my source, and I am digging through reviews and pros/cons to move forwards. My only digital-outs are HDMI & USB. Unfortunately this laptop does not have an optical SPDIF buried at the end of a 3.5 TRL socket.
My shortlist based only on reviews is as follows, each with pros & cons (bearing in-mind that my main source is likely to be USB, both the HD435 and HD650 can become balanced, and I may one day get a balanced amp)
Matrix Cube
+ve ASRC as 2/4x multiples; small; cheap; line in
-ve max USB 16/48; no XLR out
Matrix mini-i
+ve small; great reviews, XLR out; track names show on screen
-ve noASRC; max USB 16/48; no line-in
Yulong D100
+ve ASRC; USB to 24/96; XLR out; great HP amp; best reviews.
-ve most expensive; no line-in; ASRC is not to complete multiples of source, which bugs me - taking a 44KHz signal and upsampling to 110KHz must introduce temporal offset similar to jitter, but that is not random so cannot be called jitter. Moire would probably be a better term. It must impart a delay of 1s/220K, or 4us, to every second pulse. I guess this is adding a 44KHz frequency modulation (FM) signal to a 110K carrier signal. This artefact is absent from either Matrix implementation. There is a more complicated artefact added with input signals of 48, 88 or 96KHz that I cannot be bothered to work out. In all cases the offset should not be more than 4us between pulses and reviews seem to show that the end result works, but knowing there is a hidden cludge is always going to bug me. Hmm, I just scribbled it out on paper for both 44 & 88. Looks bad. For 20 pulses of 110KHz signal, the actual music signal is as follows
Carrier 110KHz 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20
Signal 44KHz 00 02 05 07 10 12 15 17 20
Signal 88KHz 00 01 02 04 05 06 07 09 10 11 12 14 15 16 17 19 20
See how the pulses that start at 44 or 88 get upsampled into an uneven spacing? That wouldn't happen either if not upsampled or if upsampled to a multiple of the original frequency. Still, the reviews are great and this is the model I am currently leaning towards. Maybe another bit of electronics in the signal path compensates for this patterning. Or maybe points are interpolated rather than just resynchronised. I don't know. I'm speculating. More reading required.
Edited by DefLugs - 6/18/11 at 3:45am