<Renewed at 2012/03/12>The new HiFiMAN 802 & New UI named "Tai-Chi" - Page 44

Quote:

Originally Posted by LarsHP 

Maybe, or the DX100 isn't as neutral as some people think. I am thinking of the DAC section per ce as well as the unit as a whole.

How could it not be neutral? It measures ruler flat and has one of the lowest output impedances of any DAP out there.

Everyone hears differently guys. Depending on age and environmental exposure, you could easily have different levels of ability to hear certain frequencies. I wanna hear more about this HM-901.......anyone got anything more on it?

Quote:

Originally Posted by LarsHP 

Maybe, or the DX100 isn't as neutral as some people think. I am thinking of the DAC section per ce as well as the unit as a whole.

Quote:

Originally Posted by lee730 

The thing is the DX100 should actually be showing you the true colors of your IEM. So I doubt this is the case. Especially if it is on 1.1.7 or 1.2.7 firmware. Could be his miracles are defective or maybe the DX100 is defective, fit issues etc.

Quote:

Originally Posted by LarsHP 

Because of the DAC section. Many have reported a difference in sound signature between R2R and sigma-delta DAC's where the latter (including the Sabre chip in the DX100) has a brighter sound than f.i. the PCM1704UK chip in the HM-801. Implementation is of course also a possible issue.
Anyway - as Ken says - it could be different in how we perceive and hear things. It could also be because of what headphones we are used to before we get a certain IEM or CIEM, and thus relate things accordingly.

But you see the 801 does not measure flat. It has roll off in the treble, and boosted mids and bass. It never did measure as a neutral sounding DAP.

Quote:

Originally Posted by lee730 

But you see the 801 does not measure flat. It has roll off in the treble, and boosted mids and bass. It never did measure as a neutral sounding DAP.

Not entirely true. Yes, there is an intended roll-off of Butterworth filter on the treble but the mid and bass do measure flat.

Here are two fresh RMAA graph I took a few minutes ago

From top: headphone-out without load; headphone-out with 16ohm CX300 and Line-out to O2 without load. The result is consistent with roll-off at treble but definitely no boost or roll-off at other frequency. The reason for bass and mid to be slightly higher than 0dB is due to how RMAA places the line and not because it is boosted.

From top: The same headphone-out with 16ohm CX300, headphone-out with Lear LCM5 and headphone-out with LCM5 and its Studio Monitor Adapter.

The graph is to show you what impedance interact is. The reason of the odd looking FR curve on LCM5 is because it is only 28ohm and has a 3 way crossover circuit. As HM801 has a 10ohm output impedance (balanced amp module, higher with other amp module), the two impedance is too close to each other and therefore interact to change the FR curve. Although CX300 is only 16ohm, it has a very linear impedance curve and no crossover, so it doesn't interact with HM801's impedance. Last but not least, as the Studio Monitor Adapter rises LCM-5 impedance to about 180ohm, it is far away enough not to interact with HM801 any more. This tell you that any IEM with crossover (or headphone wiht non-linear impedance phase) might not be the best pairing with HM801 due to its high output impedance. However, with the right headphone or IEM, there won't be any impedance interaction and therefore no added coloration beside the treble roll-off. Also worth noting is that HM801 isn't that only source with high output impedance. Foxter HP-P1 and Sony PHA-1 are two good example that has 10ohm output impedance. The new iriiver AK100 is 22ohm.

Well I guess that is another issue to take into consideration the output impedance. So you think the 801s bass and mids appear to be boosted because of the rolled off treble?
 

Quote:

Originally Posted by lee730 

Well I guess that is another issue to take into consideration the output impedance. So you think the 801s bass and mids appear to be boosted because of the rolled off treble?
 

I don't find HM801's bass or mid range to be boosted. I can hear a more mid centric and warm sound on HM602, but not on HM801.

Quote:

Originally Posted by ClieOS 

I don't find HM801's bass or mid range to be boosted. I can hear a more mid centric and warm sound on HM602, but not on HM801.

Well I hear different from others saying the 801 still has emphasized mids and bass. A characteristic that seems to be desirable however in certain circumstances.

Quote:

Originally Posted by lee730 

Well I hear different from others saying the 801 still has emphasized mids and bass. A characteristic that seems to be desirable however in certain circumstances.

I used to own a HM-801 (still do, it's just bricked itself, silly thing). I wouldn't say it has emphasized mids and bass. That statement sounds misleading. Sounds like a long way to say treble roll off. The sound was relatively linear depending on what headphones I'd use. My DT770 250ohm sounded awfully muddy and bassy, worst I've ever heard them sound. LCD-2's sounded pretty good. The treble roll off didn't bother me too much, I didn't notice it all that often. Probably because I'm still young and have good hearing and treble sensitivity. I could see myself hate it when I get older and can't hear treble the same way.

When looking at the "worst" of the frequency responses, it shows a boost in the bass and upper mids, but even at the highest point, it is not even 2 dB. In other words (with this particular IEM) the alteration will be audible, but not more than that. Headphones and IEMs unfortunately have much, much higher boosts and dips than this. Anyway - I hear a little bass boost from the LO of the HM-801 compared to my Ref 7.1 DAC, which I have no good explanation for.

The treble roll off (caused intentionally by the Butterworth filter) is about 2,5 dB's down at the hearing threshold of even young people. This is undesired from a frequency response viewpoint, especially if your IEM already has a treble roll off, but if you have a slightly bright sounding IEM, it would be a perfect match.

All this being said, I do think the UM Miracle sounds ruler flat from the Burson Soloist except for a little bass boost, while from the HM-801 it has also a little boost in the uppermost mids and treble. This difference may well be caused by a higher output impedance from the HO of the HiFiMan.

Quote:

Originally Posted by ClieOS 

 

Not entirely true. Yes, there is an intended roll-off of Butterworth filter on the treble but the mid and bass do measure flat.

 

Here are two fresh RMAA graph I took a few minutes ago

 

 

From top: headphone-out without load; headphone-out with 16ohm CX300 and Line-out to O2 without load. The result is consistent with roll-off at treble but definitely no boost or roll-off at other frequency. The reason for bass and mid to be slightly higher than 0dB is due to how RMAA places the line and not because it is boosted.

 

 

From top: The same headphone-out with 16ohm CX300, headphone-out with Lear LCM5 and headphone-out with LCM5 and its Studio Monitor Adapter.

 

The graph is to show you what impedance interact is. The reason of the odd looking FR curve on LCM5 is because it is only 28ohm and has a 3 way crossover circuit. As HM801 has a 10ohm output impedance (balanced amp module, higher with other amp module), the two impedance is too close to each other and therefore interact to change the FR curve. Although CX300 is only 16ohm, it has a very linear impedance curve and no crossover, so it doesn't interact with HM801's impedance. Last but not least, as the Studio Monitor Adapter rises LCM-5 impedance to about 180ohm, it is far away enough not to interact with HM801 any more. This tell you that any IEM with crossover (or headphone wiht non-linear impedance phase) might not be the best pairing with HM801 due to its high output impedance. However, with the right headphone or IEM, there won't be any impedance interaction and therefore no added coloration beside the treble roll-off. Also worth noting is that HM801 isn't that only source with high output impedance. Foxter HP-P1 and Sony PHA-1 are two good example that has 10ohm output impedance. The new iriiver AK100 is 22ohm.

There are other parameters that affects the perceived sound of a DAC, such as THD, SNR, etc. These are not something that can be explained by a simple RMAA FR curve and really what differentiate a good DAC from a sub-par DAC (as even sub-par DAC can have very flat FR curve). HP-P1 actually has two filter settings - one is the Buttleworth filter (slow roll-off) and another is the Chebyshev Type 1 (fast roll-off) filter. Without actually seeing the roll-off, you can read how people prefer one to another in the HP-P1 discussion purely by its sound. Anyway, HM901 will have two filters setting too, so the main focus is whether they will lower the output impedance enough to make it multi-way IEM friendly.

That makes sense but then again the graph worshipers won't agree with you on those points (sound is sound and if it ain't measurable it's all in our heads then lol ;).
 

It doesn't take much to make HM801 ruler flat even with the Butterworth filter, all you really need is able to calculate the value of the resistors and caps needed (the formula can be found online easily) and able to solder SMT parts so you can make some mod to the Butterworth filter to push the corner frequency higher, around 24kHz~26kHz probably. I would think the only reason why HM801 isn't flat is not because it can't be done, but it has been done intentionally to achieve a sound that Fang wants, whether others agree to his philosophy or not is of course a discussion on its own. That's probably the reason why he still keeps the Butterworth filter on HM901 as well.

As I have said earlier, the Miracle has a little more treble with the HM-801 (GAME amp) than from the Burson Soloist. This is most likely caused by high output impedance, I now understand.

ClieOS helped me find a page where the output impedance of the GAME and Standard modules are stated. It is far from less than 1 Ohm.

GAME: almost 33,7 Ohm.

Standard: 18 Ohm.

http://translate.google.no/translate?sl=auto&tl=en&js=n&prev=_t&hl=no&ie=UTF-8&eotf=1&u=http%3A%2F%2Fmonoadc.blog64.fc2.com%2Fblog-entry-99.html

(scroll far down to see the graph)

Posting it here also for conveinience:

I asked HiFiMan to comment on this here:

http://www.head-fi.org/t/629253/head-direct-hifiman-customer-service-team/300#post_8930015

Wow that is horrible.... I thought the game amp was suppose to fix the output impedance issues. They actually made it worse?