Thanks for the offer! I will just order it from them. I think it will be good to make this comparison going forward as well.
Nobody is debating on whether an amp can play a headphone to loud enough volumes, no one ever is. My cheap dongle can power a susvara to ear bleeding levels.
What is important when picking an amp for a headphone is determining not the power output, but the amplifiers capability of moving the transducer to the desired location, at the desired time. The error in this controllability is what separates a good amp for a certain headphone from a bad amp.
Dynamic voice coil headphones are really good at taking a power signal, and converting it to electromagnetic force that interacts with the magnet, moving the diaphragm. This is due to the high number of turns that the wire makes for the voice coil.
Planar magnetic voice coils essentially have one “turn”, and therefore react differently than a standard voice coil when amplified.
When any amp receives a load, it tries to send an exact voltage signal to it. Typically, the signal is driven, and controlled, by the error in voltage between the amplifiers input and output. When certain signals are sent to the amp, it will sacrifice current in favor of voltage in order to keep outputting the desired voltage waveform.
For a dynamic voice coil headphone, this is not as much of a problem, with the transducer meeting the desired location with little error even if the current starts to drop on a demanding input signal waveform.
Certain planar headphones, however, are designed to require every bit of current available in order to drive the diaphragm to the desired location. When an amp drives a demanding waveform, such as a kick drum hit, to the Susvara, the amplifier will control the voltage to match the input signal. However, the current will start to drop as the amp pushes more voltage into lowering the output error. This lowering of current to the output will in turn create a weaker magnetic field in the instance of the kick drum hit, making it sound weak and un-dynamic.
This is not measurable in the time domain. If measuring a 30hz signal at a certain SPL, it’s response will be solely based off of the technical capabilities of the headphone, and not the amp. That is because the measurement of the sine wave at that frequency takes place when the amplifier has reached a steady state. In reality, there could be a large error being produced. This error is not seen in the amplitude, it is seen in the phase.
So the end solution at the moment is to use an amp that brute-forces the power the headphone by having a huge power well that the headphone would be incapable of draining. The exception would be certain amplifier topologies that are current driven, or have circuits that naturally deliver even, fast power.
Ok, that is the end of my rant. Thanks / Sorry!
Holy cow - that’s the best explanation (or at least the easiest for me to understand) I’ve ever seen on this topic - thank you!!
There’s a lot to unpack in this comment.
It’s true that planar designs are - all else equal - typically less sensitive than moving coil designs, although it’s not per se about turn count (you could theoretically make a flat trace with an equal number of “gangs” to the windings of a normal headphone coil). With a moving coil, the magnetic field gap is lateral to the direction of movement, so it can be very thin (assuming that your driver is very well centered and doesn’t rock), allowing for high sensitivity without limiting maximum travel.
With a planar magnetic design, the direction of motion is towards or away from the magnet array (or, with a double-sided design, “pushed” away from one and “pulled” towards the other), meaning that the distance between the magnet array and the diaphragm impacts both the density of magnetic flux, and the maximum travel of the diaphragm. Obviously, we don’t want our diaphragm hitting the magnets (even if it’s harmless, it sounds bad), so planar designs need larger magnet gaps to allow for space for the diaphragm to move.
This makes very little sense. A low-output-impedance amplifier is approximately a voltage source - that is, it maintains a constant output voltage relative to load impedance, varying output current, while it can source and sink sufficient current. But think about it: that’s dependent on load impedance. In order to maintain the same output voltage with lower current, Ohm’s law would require the impedance of the load to rise. The amplifier can’t control the impedance of the headphone, and it definitely can’t transiently increase the impedance.
Instead, what will happen is that output voltage will be limited when the limit of output current is reached - this will result in a “clipping” of the peaks of the waveform where more current than the amplifier is capable of delivering would be required to reach the maximum voltage.
If this occurred, you would see something resembling the power compression we see in loudspeakers
That is, frequency response would change based on the input level. In practice, we do see that phenomenon with headphones played at extremely high levels, and with amplifiers driven to the limits of their power delivery, but in either case, you’re going to hear the horrifyingly large amount of distortion that comes with that nonlinearity more than you’ll hear the level change.
If this is true, it could be measured with a toneburst, impulse, or other non-continuous signal and an oscilloscope or audio analyzer set to capture at a given threshold, and unlike with most headphone stuff you wouldn’t need a fancy ear simulator to observe the behavior (it would be visible at the electrical output of the amplifier, with probes parallel with the headphone load). That being the case, it should be pretty trivial to show a transient phase shifting effect, if such a thing exists.
Of course, that isn’t what happens when an amplifier is short of current, so it’d be pretty surprising to see that. And, of course, if there is a phase shift, we can also see that in conventional measurements (frequency response isn’t just magnitude vs frequency, after all - it’s phase vs. frequency as well), and I’ve yet to encounter that, but if you have an example of a headphone-amplifier combination that you think exhibits this, I’m amenable to testing it, or perhaps @Resolve can if he has them.
(In reasonably high volume) Regarding the current delivery reduction in a high voltage amplitude wave ( drum kick ). Is this phenomenon a general statements regarding V=RI ? or do high current delivery capable amps ( like the FLA Volot) still delivery more current during those specific sequences ?
In any case it does look to match the general impression that the low sensitivity planars may suffer from lack of bass energy depending on the amp.
I mean, I = V/R tautologically. In a scenario where a weaker headphone amplifier would be limited in its voltage delivery by lack of current, then yes, you’ll see more current with a stronger amp. You will also see proportionately more voltage, and in the cases where the current limits the output voltage, you’ll see some form of clipping of the peak of the waveform.
To raise the output current without raising the output voltage, your amplifier would need to be reducing the load impedance, which it really can’t do (unless, I suppose, it has some control over a very powerful cooling system).
It seems much more likely that this is due to a lower overall level changing the perceived level of bass, unless the levels were very carefully matched in those comparisons. More powerful amps will likely tend to have higher gain (not much use to be able to throw around 16W/32Ω if you can only reach 2W/32Ω from a normal DAC), and even a small difference in level can make a substantial impact on perceived low-end.
I didn’t quite your amps gain reference.(for the low sensitivity planars) you mean, Is it better to have an amp with low gain and high current/voltage capability. Or does high gain somehow remedy the potential current shortcoming ?
There’s really no “good” or “bad” to gain - just “enough”, “too little”, or “too much” from a convenience standpoint.
Gain has nothing to do with output current, but when gain is higher, the odds of you ending up with a louder source (if you don’t match them precisely) are higher - and louder stuff sounds more “impactful”, typically.
I am really enjoying your latest EQ; for my personal tastes in EDM I took the bass frequencies down a 0.5 db notch but otherwise it’s perfect.
What did you mean by “for the unit I evaluated”? Did you end up buying an LCD-5 outright, and are EQing it a little differently?
Oh, that’s just because I know our unit measures differently from the unit Crin measured - and also Metal571’s seems to be closer to Crins than ours as well, at least based on what he EQ’d. Chrono and I evaluated the same unit and seemed to EQ to a similar result. I’ll post those results here soon.
Are you hoping to achieve an EQ that works equally well across the different units? Is it normal to have this much variance?
I’m new to the EQ game, so I have no idea how much different units tend to vary.
Depends on how the headphone’s tuning is achieved, and what the manufacturer considers acceptable for unit variation tolerances. It’s actually far more common than people think, and is likely responsible when you see differences in impressions online for the same product.
Some designs, like the HD 800 S, will generally have less unit variation, largely because the pads aren’t as influential as say an Audeze or DCA, or other headphones with thick pads. The materials for this stuff are significantly more… ‘imperfect’ than what you can do in transducer development, although there too I’ve come across some differences, so again it depends on what the manufacturer deems acceptable. This is why I always caution users with EQ - that a profile isn’t necessarily fire and forget. Use the profiles as a starting point at best.
So, after a week or so getting used to my new CRBNs, I went back to the LCD-5s for about an hour on Saturday night, a short period on Sunday (trying to A/B test between the two back to back), and then did zero listening on Monday.
Last night, I put them on to listen to light music while playing a computer game next to my daughter (also playing the same game on another computer sitting next to me). The volume was low so I could hear if she said something and could then quickly mute the volume. I took them off briefly, then put them back on only to hear a crinkling noise that wasn’t there before. So, through the process of troubleshooting the new (concerning) noise, I took them off and tried to make sure everything looked okay, ensure there was nothing in them, etc. After putting them back on again, the right side no longer played anything.
Upon further inspection, I could see wrinkles in the material behind the grill of the headphone on the right side of the image below. Also, I did try swapping the headphone cable to double check it wasn’t the amp or cable in any way. So, the right side is now dead.
I’ve emailed Audeze’s support and have been escalated to someone higher up for help here. No idea what will happen next however.
Unfortunately, I won’t get a fair chance to compare the LCD-5s to the CRBNs before my time runs out to return either the LCD-5 amp, or the CRBNs themselves. My guess is that this repair process will take weeks to resolve. It took a week just to get my CRBNs due to FedEx delaying shipping 3 days. Why Audeze doesn’t ship with a guaranteed timeframe on $4500 headphones is another topic however.
Oof. I was about to say crinkling on totl planars is normal until I saw your pic. Chances are low, but it may be worth an ask to see if you can get a longer return window in the CRBNs so that you can finish compairing.
Sorry to hear about your headphones. I will say that I have communicated with Audeze a few times and that the service and communication that I received every single time has been exceptional.
Ahh man that sucks.
I just put a pre-order in for the January batch. I hope this is just a one-off failure that is quickly resolved for you.
I also hope that this doesn’t become a widespread issue and that these new drivers are found to be very durable.
Keep us posted on what you learn!
Here’s the latest update from Audeze (after asking about timelines and the new headband),
Hey Adam
We’re currently working on a new run of LCD-5s, and don’t currently have any in stock, but we’ll be shipping those out as soon as more become available. We’re looking at probably the second week of January on that new run.
With regards to the clamping force of the headband, with our latest production run we’re officially switching over to the lower clamping force version you may have read about, so your new unit will come with that alteration. I hope you find this change a little more comfortable.
So, for now, we wait. I’ll just have slum it with the CRBNs.
And, no complaints with their support. It’s mostly on par with what I’d expect for a high end product. The only recommendation I’d have for them to make this a better experience would be to keep a stock of warranty units on hand to immediately overnight ship when failures happen.
They also asked if I added excess pressure taking them on / off, or setting them down on their pads. I keep them off their pads 100% of the time (and hung when not in use), and I’ve never thought I was putting them on with excessive force. I’ll pay more attention to the new set, but I VERY MUCH doubt there was anything I did that lead to these failing. I’m pretty careful with my equipment, and seldom have issues. So I’m chalking this up to a defect. And, regardless, Audeze is taking care of me, so we’re good here.
I have found over the years that audeze has the best customer service of all of the TOTL companies
Honestly, I quite like hifimans support as well. They suck to cominicate with because of the time difference for me (they respond to my emails at like 2am my time) but their propensity to just go “just ship us the can and we’ll give you a new one” is pretyy damn nice tbh. I think they rank on a similar tier to audeze in this aspect.
@jlbrach - I agree completely.