[JaZZ]

Do you have a link to where you got the Symphony measurements by chance?
 
-Nevermind, found some in an online Meier manual. Wonder why his designs ring for so long . . .

Yes, it's from the user manual. And it doesn't really ring longer than the Benchmark or other conventional DACs. Look at this:
 

 
The white field is the Symphony's pulse response transplanted into udauda's graph. The resolution is that the time denominator in the Symphony manual is a joke: it should mean 100 μs, not 10 ms. I have wondered that you haven't discovered it!

 
 
Nevertheless, both the Benchmark's and (probably) the Symphony's ringing borders on 0.8 ms or even (by far!) exceeds them (hard to tell from the measuring/graphic resolution), so it's not a clear case of officially approved inaudibility. Moreover those established hearing-threshold values are just guidelines, not fixed numbers; they have to be modified from time to time (like e.g. the values for harmonic distortion which are now considered audible down to 0.003% in some cases – no surprise to me, apart from the official acceptance).
 
So mental flexibility and openmindedness is still a virtue. The more so as (you know) many people swear by the time-optimized filters from their listening experiences.
.

 

[Shike]

     Quote:

Yes, it's from the user manual. And it doesn't really ring longer than the Benchmark or other conventional DACs. Look at this:
 

 
The white field is the Symphony's pulse response transplanted into udauda's graph. The resolution is that the time denominator in the Symphony manual is a joke: it should mean 100 μs, not 10 ms. I have wondered that you haven't discovered it!

I was starting to wonder honestly, but I questioned if Meier would let such a large oversight go into his documentation like that.
 
Needless to say, I rather see a documentation error rather than such a large on in his circuits . . . still, lack of oversight ftl.
 
 
Quote:

Nevertheless, both the Benchmark's and (probably) the Symphony's ringing borders on 0.8 ms or even exceeds them (hard to tell from the measuring/graphic resolution), so it's not a clear case of officially approved inaudibility. Moreover those established hearing-threshold values are just guidelines, not fixed numbers; they have to be modified from time to time (like e.g. the values for harmonic distortion which are now considered audible down to 0.003% in some cases – no surprise to me, apart from the official acceptance).

 
To me it looks like the Benchmark would be .7 at worst, but I agree the resolution on the time domain sucks quite badly.
 
Quote:

So mental flexibility and openmindedness is still a virtue. The more so as (you know) many people swear by the time-optimized filters from their listening experiences.

 
Of course - if it was found that any ringing or less than .8 ms were found audible I would correct my stance and be more inclined to agree with you.  However we can't abandon scientific findings just because it's is more agreeable for us - by doing so humanity would be stunted in progress on many fronts.

 

[JaZZ]

 

However we can't abandon scientific findings just because it's is more agreeable for us - by doing so humanity would be stunted in progress on many fronts.

Now that's too big words and too much dramatization for me...

.

 

[immtbiker]

Good job getting the thread back on track!

 

[FrederikS|TPU]

Thank you JaZZ for taking the time to explain the impact of different filter implementation techniques, very good read!
 
 

 

[udauda]

I just found out Mr.ソノベ measured HM-801's IR:

 
and the results seem to correspond to my simulation quite well. For its timid effectiveness on the transient response, I still think the filter was too much trade-off.

 

[b0dhi]

Those results show that the HM801 is using a sharp roll-off output filter. The slower roll-off is either due to an additional filter, possibly internal EQ, or due to NOS.
 
Gibbs phenomenon is roughly as bad as other DACs, as expected from the sharp roll-off. There's nothing special about the ringing in this DAC.

 

[JaZZ]

 

 
For its timid effectiveness on the transient response, I still think the filter was too much trade-off.

I agree.
.

 

[Sonic 748i]

Hey, don't tell me this thread is dying. I love reading this thread while listening to my setup.

 

[HeadphoneAddict]

Don't know whether to call you a masochist or sadist...

 

[FrederikS|TPU]

I think anything audio has to be heard before you pass judgment on it, especially when it is a case of whether a filter is too much of a trade off or not. Personally I notice the roll off on some recordings, but it only if I listen for it and compared it to my other sources. Compared to my Marantz CD6002 the HM-801 generally sounds better, except for the roll off. Imaging, bass, detail is just in a different league (tested using the same CD and either using the analogue outs from the CD6002 to my EF1 or via the coax out to the HM-801).
 
I do not like the idea of a non linear frequency response, but given that that is the premise for having a DAC that sounds like that of the HM-801 then it is something I am willing to accept.
 
Of topic'ish: Anyone know a RMAA equivalent that works with Mac? was thinking that I might run by a friends recording studio and do some measurements with their pro tools setup.
 

 

[Sonic 748i]

 
Quote:

I think anything audio has to be heard before you pass judgment on it

You speak the truth.

 

[Head Injury]

Quote:

You speak the truth.

He speaks opinion.

 

[ajkda]

I don't get it.
 
Hifi man haters posted alot of graphs and charts to prove their point.
 
Hifi man lovers posted alot of nonsensical hear it comments and attacked other users in the process ... etc to prove theirs.
 
Who will you believe?
 

 

[Sonic 748i]

This thread is slowly dying.