O2 Build Complete: Let the objective, subjective listening tests commence! - Page 6  

Quote:

Originally Posted by DRUB 

Just realised there's an impedance mismatch with the LCD-2 and Asus actually; was reading an article (by NwAvGuy) that looked at the relation between a source's output impedance and the impedance of a headphone and its impact on the frequency response - apparently optimum response is achieved via the output impedance being 1/8th of the headphone impedance. The Asus Essence ST/X has an output impedance of 10.7 ohms, so it actually provides insufficient electrical damping for the Audeze LCD-2 (which has impedance of 60 ohms) - apparently the Asus will roll-off bass frequencies by about 3db, which is definitely audible. I mentioned on another thread that I was initially confused by the LCD-2 I was listening to on my soundcard, as I didn't hear the legendary bass extension. The impedance mismatch explains it elegantly!

Is there evidence of damping making a big difference in headphones? You don't have to worry about coloration beyond damping, because the LCD-2's impedance curve is flat. I'm not getting serious roll-off, but I didn't try any particularly bass-heavy songs when I compared to the DAC1.

Damping factor is the only thing I'm worried about, and the only reason I'll be building an O2 in the future.

Quote:

Originally Posted by Head Injury 

Is there evidence of damping making a big difference in headphones? You don't have to worry about coloration beyond damping, because the LCD-2's impedance curve is flat. I'm not getting serious roll-off, but I didn't try any particularly bass-heavy songs when I compared to the DAC1.

 

Damping factor is the only thing I'm worried about, and the only reason I'll be building an O2 in the future.

The article shows 6db changes (very audible) in FR curves on low-impedance IEMs and 1db above 20hz on the HD590 (which is just audible) - not sure about effects below 20hz though, as the LCD-2 is one of the few headphones capable of that extension. When I ran the Head-fi HDtracks Heartbeat track (bass test) on the LCD-2, I didn't notice particularly good extension out of the Asus headphone jack, but in 3 weeks I'll blind test once I have the O2 once I know for sure. If someone could measure the FR curve of the LCD-2 out of a high impedance output, versus a <2 ohm output (if they have the equipment) that would be ideal. 

Quote:

Originally Posted by DRUB 

The article shows 6db changes (very audible) in FR curves on low-impedance IEMs and 1db above 20hz on the HD590 (which is just audible) - not sure about effects below 20hz though, as the LCD-2 is one of the few headphones capable of that extension. When I ran the Head-fi HDtracks Heartbeat track (bass test) on the LCD-2, I didn't notice particularly good extension out of the Asus headphone jack, but in 3 weeks I'll blind test once I have the O2 once I know for sure. If someone could measure the FR curve of the LCD-2 out of a high impedance output, versus a <2 ohm output (if they have the equipment) that would be ideal. 

To my knowledge the article's graphs showed frequency response variations caused by impedance curve. Balanced armature IEMs in particular have wild impedance curves. High output impedance amps feed less voltage into low impedances, so the places where impedance curves are at their highest get the most relative volume. I could be wrong.

His graph for the IEMs showed no bass roll-off above 20 Hz, but I don't think he measured the headphones themselves with a dummy head, just the electrical properties. Again, I could be wrong.

The effect on damping with the LCD-2 is relatively minor compared to the IEMs, too. They're at a 6-to-1 impedance ratio, not too far from the ideal. Enough for audible effects from lack of electrical damping? Maybe, depends on how audible that is with any headphone.

Quote:

Originally Posted by Head Injury 

To my knowledge the article's graphs showed frequency response variations caused by impedance curve. Balanced armature IEMs in particular have wild impedance curves. High output impedance amps feed less voltage into low impedances, so the places where impedance curves are at their highest get the most relative volume. I could be wrong.

 

His graph for the IEMs showed no bass roll-off above 20 Hz, but I don't think he measured the headphones themselves with a dummy head, just the electrical properties. Again, I could be wrong.

 

The effect on damping with the LCD-2 is relatively minor compared to the IEMs, too. They're at a 6-to-1 impedance ratio, not too far from the ideal. Enough for audible effects from lack of electrical damping? Maybe, depends on how audible that is with any headphone.

That is interesting, do you have a link to the impedance curve of the LCD-2? Might e-mail nwavguy and see what his hypothesis is - do you know of anyone who has the gear to test the impedance mismatch theory here on Head-fi?
 

Quote:

Originally Posted by DRUB 

That is interesting, do you have a link to the impedance curve of the LCD-2? Might e-mail nwavguy and see what his hypothesis is - do you know of anyone who has the gear to test the impedance mismatch theory here on Head-fi?

All planars (to my knowledge) have a dead flat impedance curve. LCD-2.

If Tyll could get his hands on some amps with varying output impedances, I bet he could make a very interesting article. He could check for the effects of damping factor alone using planars, and expand on NwAvGuy's graphs for everything else too.

Quote:

Originally Posted by Head Injury 

All planars (to my knowledge) have a dead flat impedance curve. LCD-2.

 

If Tyll could get his hands on some amps with varying output impedances, I bet he could make a very interesting article. He could check for the effects of damping factor alone using planars, and expand on NwAvGuy's graphs for everything else too.

I dropped a comment on the blog to ask about the LCD-2 being affected, and whether its flat impedance curve makes it immune to the sort of craziness you see with the balanced armature designs (theoretically yes as you've pointed out). I'm getting a second pair of LCD-2s in the mail on Friday to double-check that my ones aren't just bass-deficient or something (either that or I've devolved into a completely degenerate basshead - considering I listen to dubstep I may be guilty as charged).

Thanks for the link also. One thing I was wondering, and there seems to be quite a bit of debate about this, is how much power is needed to drive the LCD-2. Apparently 810 mW will drive the rev 1 to uncomfortable levels (above 120 db). As the max power output of the O2 is 640 mW @ 80 ohms, do we know how high it goes with the LCD-2, and what level it can drive it to? I'm assuming easily beyond comfortable listening levels, considering 810 can hit 120.

How do we lobby Tyll for those graphs? 

Quote:

Originally Posted by DRUB 

 

Thanks for the link also. One thing I was wondering, and there seems to be quite a bit of debate about this, is how much power is needed to drive the LCD-2. Apparently 810 mW will drive the rev 1 to uncomfortable levels (above 120 db). As the max power output of the O2 is 640 mW @ 80 ohms, do we know how high it goes with the LCD-2, and what level it can drive it to? I'm assuming easily beyond comfortable listening levels, considering 810 can hit 120.

I think I also tested it right out of an Essence STX, and it easily went loud enough with the LCD-2's. Even with low 2xish gain (I forget what I set it to) there was more than enough travel left in the pot.

Quote:

Originally Posted by deadlylover 

I think I also tested it right out of an Essence STX, and it easily went loud enough with the LCD-2's. Even with low 2xish gain (I forget what I set it to) there was more than enough travel left in the pot.

This is why I'm confused - people have been telling me minimum required wattage is not in the milliwatts, but at least 1 watt, preferably 2-4 watts, when that would drive the headphones *far* beyond comfortable listening levels - if the Asus can drive them reasonably well, then the O2 should be perfectly capable, I imagine.

Guys, this thread has gotten away from me, but I apologise for any earlier miscommunication re the Beta22

- no, I haven't heard either amp, yet

- no, I do *not* view the Beta22 as 'audio jewelry' - the complete opposite. Functionality over bling.

- yes, I realise that some folk still want to be able to hear a comparison between the two, despite the fact that the parts alone in the B22 cost over 10 times the parts cost for the O2.

I do get the 'good enough vs massively over-engineered' point, and it would be fantastic to DBT them just for the sheer hell of it.

Quote:

Originally Posted by LizardKing1 

... and no one so far has come out of the subjectivist woodwork to accept the challenge.

 

Following the above metaphor, the step from the Beta22 to the O2 is a huge one which most don't see justified, to it would be like if to get to the top of the Everest costed you 10x more than it costs to get to the top of Kilimanjaro, even though to you the trip is pretty much the same. Most wouldn't even bother to spend more for something they can hardly tell apart. Now, if you really want the top of the Everest for bragging rights and you're ok with affording it, that's great, but most people will be happy with Kilimanjaro and will keep wondering why would you pay that much for something to similar to what they have.

 

That sounded way better in my head.

I honestly don't see the anyone accepting the challenge, the objectivist crowd believes it unnecessary and the subjectivist crowd (both terms used lightly) doesn't believe in ABX testing. Besides a proper level matched ABX is a too much a hassle to set up for many people.

Quote:

Originally Posted by estreeter 

Guys, this thread has gotten away from me, but I apologise for any earlier miscommunication re the Beta22

 

- no, I haven't heard either amp, yet

- no, I do *not* view the Beta22 as 'audio jewelry' - the complete opposite. Functionality over bling.

- yes, I realise that some folk still want to be able to hear a comparison between the two, despite the fact that the parts alone in the B22 cost over 10 times the parts cost for the O2.

 

I do get the 'good enough vs massively over-engineered' point, and it would be fantastic to DBT them just for the sheer hell of it.

Definitely - or even BT if that's easier to set-up.

Can I just say with respect to the B22; I know it has a lot of supporters here, but I think a lot of the enthusiasm for it may be misplaced.

One of the features of the B22, at least in many incarnations and in its recommended configuration, is a third channel. This is in addition to its existing 'genuine' ground. Operating the amplifier using the 3rd. channel almost certainly degrades the performance with respect to operating it with the genuine ground only. Part of the advertising for the B22 claims the 3rd. channel results in lower output impedance, but this is false. The degradation may be small but for me it throws into doubt AMB's entire design philosophy, to say nothing of his test regime.

Opamps are subjected to extensive test and redevelopment by manufacturers. Their performance exceeds that achievable with discreet designs in most cases. The B22 is overpowered unnecessarily, it is claimed it will drive 18W into 8 ohms. Many people believe that discreet designs are automatically better than integrated designs, but this is not the case. Can you imagine a Pentium processor built with discreet transistors? The only reason for using discreet components nowadays is when power levels beyond those of opamps are required. Such power levels are not required to drive headphones of any reasonable manufacture. Yes, there are many enthusiasts who build discreet opamps and who claim they produce a superior sound, but these claims are not supported by reliable measurements or blind test.

There also seems to be an inbuilt assumption operating that more expensive == better and the O2 is being excused on the basis that it costs much less than the B22.

More expensive == more expensive.

Better == better.

Things that are more expensive very often are better, and very often things are more expensive because it costs more to make them better, but things are not necessarily better because they are more expensive. It is entirely possible to make something that is expensive that actually performs worse than something that is cheaper. It could in fact be the case that, at least in terms of distortion, the B22 performs worse than the O2. Only objective testing, either instrumented or blind listening, can discover the truth. This can be a hard lesson to learn for many people, but teaching it is one of the O2's designer's goals, as I understand it.

Please take aboard some of these ideas before rushing to judgement, or merely expressing your prejudices in this matter.

w

No problem, but I'd still like to hear a 2-channel B22 from a good source into something like the LCD3. I'd also like to do that while a naughty nurse pours cocktails for me in between preparing lobster and polishing my Maserati. I'm a simple man.

I just ordered some parts for this bad boy:

Parts should come in sometime next week, but that's the easy bit. Now how am I going to pry the b22>lcd-2 rig off my friend.....I might have to give up my Stax for a weekend or two....for science.

I fully reserve the right to slap anybody who says the blind test results would be invalid because I'll be rolling with my prototype Obj2.