hmm i've been wondering about this: how much of a real difference is there between say, an sblive running SSRC, and an m-audio revolution through a pair of good headphones? (eg: sennheiser hd580.)
i know a lot of people say that the sblive is crap and the revo will walk all over it, but what i'm asking is -- has anyone done any real/blind tests on this, and if so, what were the results? is there really a night and day difference?

so i guess this leads on to another thing: how much of a difference does a headphone amp make for a high-impedance 'phone like the senn 580? it's said over and over again on head-fi.org that an amp is pretty much necessary to get the full potential out of the 'phones, and that the sound improves 'dramatically' by using a decent amp. i'm wondering how much i'm actually missing out on by just plugging these straight into my soundcard (sblive), and just how 'dramatic' the sound improvement really is (ie, has anyone done any blind tests?)

the reason i'm asking is, placebo effects are usually quite strong especially if you've just bought a new piece of equipment and expect it to sound much better than your old equipment, so blind tests are the most helpful.
thanks (there's a lot of threads on this topic, but no one seems to have done any real tests... is the difference so big that blind tests aren't necessary?)

[edit: bah this should probably be over in the audio hardware section. any mod care to help? ]

i don't know about blind tests but i saw a review somewhere (can't find the link) with a graphical analysis and the difference was night and day. with such a big gap i'd imagine that many people would be able to hear a clear difference. it did fall behind quite a bit on games though.

SSRC resampling introduces something similar to preecho... (very slight effect)
Of course it is much better than resampling of most cards (introducing THD).
Additionally, SB Live eats high frequencies.
There were some RightMark test results available, AFAIK.

I must have somewhere old files, one resampled with SSRC, one with the SB Live digital out, plus SoundForge q1, SoundForge q4, CoolEdit q999, and Lame --resample. I'll look when I'm home. The differences were very subtle.

I've never had a headphone amp, but I use to plug headphones in power amplifiers, that is a high impedance, but high power headphone out. There is at least an obvious difference : the volume can be pushed much louder ! And it seems that the peaks (boom boom) are more powerful, but that should be ABXed.
I'd say that turning up the volume on a cheap headphone output, the sound becomes quickly harsh, while it can stay clean and powerful for some more decibels before becoming painful to the ears, in a powerful output.
This may be a side effect of the higher impedance rather than the higher power (power amplifier headphone out = 330 ohm) : more bass, thus less painful at high volume.

I'm wondering about this headphone issue too -- I just Emailed M-Audio support about driving standard impedance (32-64 ohms) headphones from the analog line output jacks with the Audiophile 24/96, and looking forward to their response. They've been good about responding to my inquiries in the past. I didn't ask about high-impedance headphones though, but I think those should generally be driven by an amp if you want proper volume levels.

thanks a lot for the responses



ya that's exactly the kind of thing i'm looking for.

i can't readily try out any headphone amps (new zealand's not overflowing with audio shops ), so i'm really looking for quanitifiable results.

surprisingly though, the sblive with kxdrivers has no hiss through the 580s -- contrary to what everyone seems to say (although it does hiss *slightly* through easy-to-drive earphones.)



hmm well i found these on the rightmark site, and i don't see any loss of high-end~
sblive 5.1 rear out @ 48khz:


even with front-out @48khz, the cutoff at 20khz is at -1.5db. since LAME --aps cuts off at about ~19000hz, i don't think a -1.5db drop at 20khz will be noticeable:



these match up with my own rightmark tests too.

so yah, if one assumes that there's no "noticeable" dropoff in high-frequencies with the sblive, i guess THD, crosstalk, etc.. are what's improved with a better card, right? is it a night and day difference though?

I had a SB live and I bought a revo.

I use Grado SR-125 phones with a 30$ AP cable between the revo and my sennheiser amp.

The difference between the live and the revo is day and night.
The revo can't do 3d games, so I keep both and enable/disable to get the revo or the live, to shoot or music.

Thats it.

Some answers:

- Output impedance of M-Audio Audiophile is around 150 or 200 ohm, I can't remember well, I measured it some time ago. The real problem of this card should be the lack of power to drive medium and low-impedance headphones.

- SSRC resampling introduces pre-echo just at 22.050 KHz, so it should be not audible. Apart from this, the quality is just excellent.

- SBLive cards at 48 KHz, using rear outputs, have pretty good quality, as the measurement shows. Front outputs are a little bit worse, but I think not much, from an audibility point of view. Some people has reported noises or hiss at the output, so if that is your problem that could be an issue.

- Another think I've said a zillion times: high impedance headphones are not that difficult to drive, and that is just because they are high impedance. Most high impedance headphones need a low output impedance source so that the low frequencies are not given a little extra boost, but you don't need a dedicated headamp to achieve that, some soundcards achieve that. Even if not, the difference is not big.

- I don't think the difference between the Revo and a Live is day and night, I think it's much about placebo effect, but I can't say for sure since I don't own a Live.

In practice, it hasn't been a problem here (as long as the headphones aren't some weird, hard to drive model or volume falls rather lower than the Replaygain standard). In all honesty, it seems to drive ~32-64 ohm headphones rather well, if deafness isn't the goal.

I might still consider a better amp (than the Rolls HA43) sometime in the future, since I notice some transient "blurring" and a rather bland/unremarkable sound from this amp. For the moment, I've connected it directly to the phono stage output of my preamp, for listening to records when the computer is off -- it seems to sound quite a bit better than phono stage -> preamp stage -> soundcard -> headphone amp. As always, it seems "less is more."

BTW, I recently received the Denon AHD-750 headphones from the guy I bought them from on Head-Fi -- a nice improvement over the Denon AHD-550, and so much more comfortable. Definitely the best sounding closed phones I've heard so far, too bad Denon discontinued their higher end models. I guess the siren call of "5.1 surround" got to them, too.

Also of interest perhaps... after doing some rearranging of power cords nearby my phono preamp, I've gotten background noise levels (after switching on the TT motor but before dropping the needle) down to -85dB, imho a pretty decent figure for vinyl -> digital transfers.

ah thanks a lot kikeg -- that's exactly the kind of answer i'm looking for

yeah for me, the sblive does not hiss at all through the hd580s & kxdrivers (in a silent room with silent computer.) but, switch over to the creative drivers, and it hisses quite a lot (even with everything muted in the mixer.) strange...

this lack of hiss is one of the reasons that makes me wonder just how "bad" the sblive's quality is said to be. i'd love to just try out a headphone amp & a revo, but unfortunately that's just not possible.

congrats on the new 'phones, fewtch... i'm sure they'll be the first of many more if you hang around head-fi

Don't forget the "sample rate conversion quality" setting of Windows, that divided the SB Live owners into those for who SSRC plugin provided a day and night improvement (set to "good"), and those for who it made quite no difference (set to "best").