Astell&Kern AK380
Sep 2, 2015 at 4:21 AM Post #1,906 of 9,040
 
Sorry for that, George. If anyone has any questions, or finds the usual (egregious errors) please tell me. 

 
Hi Shizgeo,
 
I am really intrigued by the reviews opinion of the AK380's balanced amp. I have listened to SE*46's, JHA Angies and Laylas through the balanced output and found it to be sonically excellent. However, I almost never have the volume above 80 as anything else is way too loud.
 
I have also listened to Beyer dynamic T5p's through the single ended output, and again this is excellent.
 
Is the review saying that if I listen to the music at a level that is way too loud and uncomfortable for me then there will also be additional bad sonic distortions from the amp?
 
Is there anybody out there that listens to IEMs at 150 on the AK380? Or is the review not talking about IEMs at that point, but rather non-portable full-size cans, where maybe you need to go to 150 to drive them, but then it starts to get distorted?
 
Can anybody explain exactly what the parameters are - earphone sensitivity, earphone impedance, volume output, comfortable and safe listening levels. Should we be down on an amplifier section if it cannot drive earphones cleanly at an unlistenable volume? Especially if it becomes unlistenably loud long before it loses control?
 
My incredibly amazing main HiFi at home can probably produce some truly awful IMD etc if I push it to its top volume, but other things will fall apart (like the house) long before that.
 
Thanks
 
Sep 2, 2015 at 4:27 AM Post #1,907 of 9,040
I haven't yet published the RMAA results and am still compiling them. The short answer is: loud volumes through the AK380 have good effects: higher SNR, DR, and less stereo crosstalk. But they tend to induce more anomalies in both FR and THD/IMD. At normal listening levels, there is essentially no benefit for 24-bit as 24-bit relies on being louder in order to achieve better SNR, dynamic range, etc. And, at normal listening levels, the AK380 measures about as good as an iPhone 4S. 
 
If it sounds good to you (and it does to me), that should be all that matters. But in terms of actual measurable performance, it isn't super-duper by any measure. Where it truly excels is its unloaded signal, which fed to a good amp, should produce better sustained signal quality into your SE846. I hope that no one is listening at volumes of 150 or so. At volumes of 80, the AK380 outputs no better than 16-bit. 
 
Sep 2, 2015 at 4:33 AM Post #1,908 of 9,040
As a non owner yet, sitting on fence about purchase, does it mean that although a lot are agreeing it is a great sound it can be vastly improved putting through a great AMP?
 
Sep 2, 2015 at 4:39 AM Post #1,909 of 9,040
  As a non owner yet, sitting on fence about purchase, does it mean that although a lot are agreeing it is a great sound it can be vastly improved putting through a great AMP?


A really really really good amp may be able to stand up somewhat to the AK380's unloaded signal. But most probably cannot. But loaded with high-current headphones and earphones, the AK380 isn't great, especially in balanced mode. A Vorzüge PURE II or II+ will do better, but may not be able to stand up to the unloaded signal in all areas. I expect it to hold its own in balance (single ended) against the AK380 when driving headphones which overly stress the AK380.
 
Sep 2, 2015 at 4:44 AM Post #1,910 of 9,040
  I haven't yet published the RMAA results and am still compiling them. The short answer is: loud volumes through the AK380 have good effects: higher SNR, DR, and less stereo crosstalk. But they tend to induce more anomalies in both FR and THD/IMD. At normal listening levels, there is essentially no benefit for 24-bit as 24-bit relies on being louder in order to achieve better SNR, dynamic range, etc. And, at normal listening levels, the AK380 measures about as good as an iPhone 4S. 
 
If it sounds good to you (and it does to me), that should be all that matters. But in terms of actual measurable performance, it isn't super-duper by any measure. Where it truly excels is its unloaded signal, which fed to a good amp, should produce better sustained signal quality into your SE846. I hope that no one is listening at volumes of 150 or so. At volumes of 80, the AK380 outputs no better than 16-bit. 

 
OK, I'll bite - How can an analogue output be said to only map to 16 bits of volume - analogue signal space has a truly uncountable infinite space of values, and so no matter how many countable bits I decimate my signal to I can still create exactly the shape of sound I want at whatever volume I choose (provided my DAC and amp are good enough). If I perfectly output 32 bits at 0.001 decibels or 1000000 decibels the signal shape can still be completely homologous and 'identical' and simply differ in linear scale. This is because each bit change in volume can be mapped to one on the uncountably infinite real numbers that we have in analogue space. It actually has nothing to do with signal to noise ratio (except in the construction of the DAC and amp). 24 bits does NOT rely on being louder- please correct me if I am wrong.
 
Sep 2, 2015 at 4:59 AM Post #1,911 of 9,040
You can amp any signal up to any volume, but within that range, a signal is limited to a finite amount of SNR, DR, etc. And 16-bit tends to top out at around 98dB in any metric. 24-bit, even at lowly 44kHz, jumps up to past 120dB in the same metrics, and takes a massive diet on distortion. Even if a 16-bit signal exceeds 98dB, it is limited to the format. Ditto 24-bit. 
 
So... unloaded (no headphones plugged in), the AK380, at high enough volumes to make use of 24-bit benefits, is amazing. Truly stunning. But for most people, it will be too loud. By lowering the volume to usual listening levels, you also limit the dynamic range, which is dependent on the most silent part of the signal and the loudest. The contrast between them is the measurement. 
 
You can play 24-bit files at 0dB if you want. The file doesn't change. But the degree to which the signal reflects the fidelity of the file does. And at mean volumes of, say, 90dB, you are limited to at most, 90dB; thus, the benefit wrought by higher bit and sampling rates is negligible. 
 
Sep 2, 2015 at 5:22 AM Post #1,912 of 9,040
I presume the better source you are using connected to Hugo dac will convert from there.

Tried it before with fiio x5, ak240ss and ak380, both sounds different with using Hugo.

I find that with ak380 should use a direct amp instead of connect to another dac.

now i am hoping the PC as a source will sound better than AK380 for my Hugo :p 
 
 
Sep 2, 2015 at 5:35 AM Post #1,913 of 9,040
  now i am hoping the PC as a source will sound better than AK380 for my Hugo :p 
 


Pray for voodoo. That said, not all optical outputs are created the same. I've tested a few that had some errors. I assume that the AK380's optical output is good. 
 
Sep 2, 2015 at 7:01 AM Post #1,914 of 9,040
  You can amp any signal up to any volume, but within that range, a signal is limited to a finite amount of SNR, DR, etc. And 16-bit tends to top out at around 98dB in any metric. 24-bit, even at lowly 44kHz, jumps up to past 120dB in the same metrics, and takes a massive diet on distortion. Even if a 16-bit signal exceeds 98dB, it is limited to the format. Ditto 24-bit. 
 
So... unloaded (no headphones plugged in), the AK380, at high enough volumes to make use of 24-bit benefits, is amazing. Truly stunning. But for most people, it will be too loud. By lowering the volume to usual listening levels, you also limit the dynamic range, which is dependent on the most silent part of the signal and the loudest. The contrast between them is the measurement. 
 
You can play 24-bit files at 0dB if you want. The file doesn't change. But the degree to which the signal reflects the fidelity of the file does. And at mean volumes of, say, 90dB, you are limited to at most, 90dB; thus, the benefit wrought by higher bit and sampling rates is negligible. 

 
I have to say I totally disagree.
 
The difference between a 16bit signal and a 24bit signal is resolution not volume. 
 
If we start with an analogue signal A. We then sample this signal at 16bit and 24bit resolutions, at the same frequency, making sure that the volume is normalised so that the peaks of both are at the same volume V.
 
For the 16 bit signal the individual steps of volume are
 
V/65533
 
For the 24bit signal the individual steps of volume are
 
V/16777215
 
The 24 bit steps are smaller than the 16 bits steps and so the 24 bit digital file D24 is a better approximation than the 16 bit file D16.
 
If we assume a perfect DAC and amp without noise and reproduce the 16 bit signal A16 and also the 24 bit signal A24 we will find that A24 is a better match to A then A16 is, at whatever volume we choose to play them back at. In fact,  the only difference between the files is that one is a more accurate representation than the other.
 
Please tell me where volume comes into this, and why playing D24 quieter is the same as playing D16 louder, which is what you seem to be saying?
 
Sep 2, 2015 at 7:33 AM Post #1,915 of 9,040
Piercer, it is totally fine to disagree. And I've been warned already about talking about 16 and 24 bit audio. So I'll stop. If you want to read up on it, headfi has a great thread about it here. Myriad others exist. 
 
Anyway, I've found that with most earphones, the sweet spot for the AK380's balanced output is roughly at a volume setting of 110, beyond which, things go wonky. Below that, however, and basically all of the theoretical benefits of 24-bit audio are all but gone. 
 
Sep 2, 2015 at 11:51 AM Post #1,917 of 9,040
Any other AK380 owners have light bleed on their screens?

This is my 2nd unit with the same issue in the same location (bottom left).

If it wasn't so darn expensive, I would just live with it, but at this price, i'm expecting perfection.

 
Sep 2, 2015 at 12:14 PM Post #1,920 of 9,040
Any other AK380 owners have light bleed on their screens?

This is my 2nd unit with the same issue in the same location (bottom left).

If it wasn't so darn expensive, I would just live with it, but at this price, i'm expecting perfection.


Maybe they just sent you the same unit you returned?  That would bug me a bit . . . seems like the kind of thing that could get worse with time.  Mine arrives tomorrow . . .  I'll be looking for light bleed !!!  
angry_face.gif

 

Users who are viewing this thread

Back
Top