.

.

.

.

Sorry for the multible posts - nothing happend when i klicked on "Submit" - except somthing did

dalgas said:

  Well - I could not forget the sweet TDA1543 sound.
 
So I bought a ready-made DIY DAC based on 8 TDA1543 chips, with reclocking and NOS (non oversampling). The thing ONLY accept 16/44 but - WOW - it blew the Gumby out of the water! Vokals in particular are much more natural and the bass (always been the weak point of the Gungnir) is more punchy!
 
The Gumby still presents more details and a bigger stereoimage - but has more digital glare than the DIY NOS. Unbelievable - since I find the Gumby very strong here.

Click to expand...

Really... I've always found the Moffatt bass to deliver. I think it's one of the strong points of both the Yggdrasil and MB Gungnir (haven't heard the MB Bifrost, so can't comment on it).
 
The TDA1543 has less dynamic range, rated at 16 bits... has anyone measured what that TDA1543 DIY DAC is actually capable of delivering for noise floor? I think the Schiit DACs delivering >16 bits of actual performance is phenomenal, and this is one of many reasons why they sound superb. 

Dalgas' post is only the 2nd I have seen that states Moffat bass is a "weak point". It seems clear Schiit DACs are not the best fit for Dalgas' listening whereas the TDA1543 DIY DAC fits very well.

As Jason and Mike have noted that building the perfect DAC (amp etc) for everyone is not possible.

For sure. This is one of the most personal, subjective hobbies around. It's also one with many stages. Listening skills, preferences, what sounds 'right'...all seem to change often.

The gear I used to think sounded 'good' or 'bad', and the flaws (and strengths!) I didn't notice in past years...

argo duck said:

Dalgas' post is only the 2nd I have seen that states Moffat bass is a "weak point". It seems clear Schiit DACs are not the best fit for Dalgas' listening whereas the TDA1543 DIY DAC fits very well.

As Jason and Mike have noted that building the perfect DAC (amp etc) for everyone is not possible.

For sure. This is one of the most personal, subjective hobbies around. It's also one with many stages. Listening skills, preferences, what sounds 'right'...all seem to change often.

The gear I used to think sounded 'good' or 'bad', and the flaws (and strengths!) I didn't notice in past years...

Click to expand...

Well said sir, well said indeed. I am sure how a person hears changes, as that person ages. Plus this hobby is very subjective. What one likes may be disliked by another person. I just hope I can try out other equipment, so I offer more subjective opinion .

mikoss said:

  Really... I've always found the Moffatt bass to deliver. I think it's one of the strong points of both the Yggdrasil and MB Gungnir (haven't heard the MB Bifrost, so can't comment on it).
 
The TDA1543 has less dynamic range, rated at 16 bits... has anyone measured what that TDA1543 DIY DAC is actually capable of delivering for noise floor? I think the Schiit DACs delivering >16 bits of actual performance is phenomenal, and this is one of many reasons why they sound superb. 

Click to expand...

I think testing revealed the Yggdrasil is really only capable of 13 bits (it looks like the MB Gungnir has an advantage @ -90dB vs -78dB). atomicbob said he would eventually retest it to make sure that was correct but never did (it's been months) and he posted on another forum that he's been withholding some of his test info on the Yggdrasil because he's afraid people will judge it by its numbers rather than its sound.
 
I have my eyes set on a Yggy but I want a black one and am willing to wait however long it takes to snag one lol (and who knows, an even better DAC may pop up before then).

maconi said:

  I think testing revealed the Yggdrasil is really only capable of 13 bits (it looks like the MB Gungnir has an advantage @ -90dB vs -78dB). atomicbob said he would eventually retest it to make sure that was correct but never did (it's been months) and he posted on another forum that he's been withholding some of his test info on the Yggdrasil because he's afraid people will judge it by its numbers rather than its sound.
 
I have my eyes set on a Yggy but I want a black one and am willing to wait however long it takes to snag one lol (and who knows, an even better DAC may pop up before then).

Click to expand...

That is counter to my measurements. If you look at the dynamic range measurement (3rd graph) in the original post 
 
http://www.head-fi.org/t/764787/yggdrasil-technical-measurements
 
you will see this graph (also included below) clearly indicating 122 dB dynamic range. That is a lot more than 78 dB.
 
The only retest necessary affected the THD measurement below 500 Hz due to a driver issue in Windows during the test. The graph indicates yggdrasil a higher THD below 500 Hz than it actually will measure when the driver issue doesn't interfere. I haven't returned to this particular graph for the yggdrasil as performing a suite of measurements typically involves several hours of setup and then 4 ~ 8 hrs of actual measurement and graphing. My corporate life has been very busy for the last several months.
 
As to the data reserved, that had to do with the Bifrost MB which does have 16 good bits, a little lower than the more expensive Gungnir MB and Yggdrasil. I published the data but voiced my reservations. People who worry over single numbers too obsessively might miss out on a great listening experience. I also did not wish to publish some graphs that take considerable explanation to non-engineering types as to their interpretation. The ones published have kept my inbox busy enough.
 
yggdrasil dynamic range:

atomicbob said:

  That is counter to my measurements. If you look at the dynamic range measurement (3rd graph) in the original post 
 
http://www.head-fi.org/t/764787/yggdrasil-technical-measurements
 
you will see this graph (also included below) clearly indicating 122 dB dynamic range. That is a lot more than 78 dB.
 
The only retest necessary affected the THD measurement below 500 Hz due to a driver issue in Windows during the test. The graph indicates yggdrasil a higher THD below 500 Hz than it actually will measure when the driver issue doesn't interfere. I haven't returned to this particular graph for the yggdrasil as performing a suite of measurements typically involves several hours of setup and then 4 ~ 8 hrs of actual measurement and graphing. My corporate life has been very busy for the last several months.
 
As to the data reserved, that had to do with the Bifrost MB which does have 16 good bits, a little lower than the more expensive Gungnir MB and Yggdrasil. I published the data but voiced my reservations. People who worry over single numbers too obsessively might miss out on a great listening experience. I also did not wish to publish some graphs that take considerable explanation to non-engineering types as to their interpretation. The ones published have kept my inbox busy enough.
 
yggdrasil dynamic range:

Click to expand...

 
I was referring to the Yggdrasil's THD measurement, 0.01169% which is 78dB or 13 bits right? I'll admit I don't even fully understand the measurement (how 13 bits is derived from it). A lot of people like to bring it up when talking about the Yggy because it seems to be one of the only marks against it (for those who are obsessed with the numbers).
 
Do you think the technical difficulties you had with the measurement is why the THD measured higher on the Yggy than the GMB, or is that just a trait of the GMB?
 

 
Personally it doesn't bother me, especially since so many (trustworthy) people still seem to gravitate towards the Yggy. Which is why I'm still holding out for a black one (to go with my Cavalli Liquid Carbon) soon. 
 
Thanks for all you do BTW. Obviously someone with your knowledge/tools is going to stay busy.  

maconi said:

   
I was referring to the Yggdrasil's THD measurement, 0.01169% which is 78dB or 13 bits right? I'll admit I don't even fully understand the measurement (how 13 bits is derived from it). A lot of people like to bring it up when talking about the Yggy because it seems to be one of the only marks against it (for those who are obsessed with the numbers).
 
Do you think the technical difficulties you had with the measurement is why the THD measured higher on the Yggy than the GMB, or is that just a trait of the GMB?
 

 
Personally it doesn't bother me, especially since so many (trustworthy) people still seem to gravitate towards the Yggy. Which is why I'm still holding out for a black one (to go with my Cavalli Liquid Carbon) soon. 

 
Thanks for all you do BTW. Obviously someone with your knowledge/tools is going to stay busy. 

 

Click to expand...

 
From the title of that graph:
THD and THD+N - unweighted - update 20150603 - this measurement is not correct - update coming soon
 
From the commentary I wrote at the end of that post:
Commentary:
 
By now, some of you probably have noted the THD+N graph may not be the lowest ever seen.
 
The THD+N graph measures artificially high due to a driver / OS interaction on my computer. An update will be forthcoming.
 
I haven't made the updated yggdrasil THD measurement. Each set of measurements that I have published on these forums have been made on borrowed time. They consume a fair amount of time and I really just haven't revisited the yggdrasil. By the time I measured the Gungnir MB I had worked out the os driver interaction issues and the Gungnir THD measurement is correct. The yggdrasil THD measurement is not correct. Those numbers reported in the graph should be IGNORED. They are due to measurement error. When I find time to work in the yggdrasil measurements again I am confident they will measure as good as or better then the Gungner MB.
 
One last time, my THD measurement of the yggdrasil is in error. Would everybody please ignore the numbers associated with that graph such as the 0.01169%.
 
When I have re-measured I will post the updated graph and numbers and also note on the post that the graph has been corrected.

I'm a long time audio hobbyist who lacks a degree in EE. When I look at a discussion like the above I have a lot of trouble getting my mind around why 0.01169%THD would be of any concern at all since the transducers that the sound comes out of are doing well to have 0.1% THD if we're lucky and often are a lot worse. I've also read that when masked by music we can't hear much higher levels of distortion.  I admit I may be missing something here but it's always seemed to me that most modern DACs test out on paper way better than differences we actually hear between them.  I've come to believe in my blissful ignorance that there are other differences between them than what is measured that makes one DAC sound distinctly good like Yggy and others with great specs which sound, well, "meh."  My suspicion is it has to do with timing, jitter, and phase errors, and a bunch of stuff like that--but what the heck do I know?  I just read forums..  As a research scientist I am into data and objectivism so I have no problem with measurements as long as they are meaningful.  I even believe in double-blind testing.  Like many of you I have concluded that while the measurements need to be good and that's why we ask for them, they are simply not measuring whatever it is that makes a difference between DACs so I use my ears.  I think someone has already said this.
 
Please look on this post as more of a plea to have someone explain this all to me rather than a criticism of anything anybody has said above. I have read a lot about how to interpret measurements  and still come up empty.  

My own experimentation and the work of others has suggested a high level of 2nd order harmonic distortion may be tolerated and possibly viewed as a positive enhancement to our auditory experiences, but even slight amounts of 3rd harmonic distortion bother us, detracting from the experience. Unfortunately THD combines both and worse rolls it all up into a single number. Graphs are far more informative, especially when 2nd and 3rd harmonic are independently reported. In the future I may write scripts for my measurement systems to produce these graphs a bit more automatically rather than the very manual method I employ suffer.
 
Here is a good presentation on the topic from Rocky Mountain Audio Fest by a very knowledgeable fellow from Audio Precision Jonathan Novick:
 
https://www.youtube.com/watch?v=2V6YN-mshmY
 
It is a one hour presentation and will tell you much more than I can type in a short post.

Another approach in learning about what we hear is to read about Perceptual Audio Codec developments and patents. Many years ago I read some of the patents in the area, but sadly its been so long that I've forgotten the details. There was a lot of research also done by the phone company about what we hear.
 
I'm only putting this forward as a method of learning about accoustics, and I'm not advocating for Perceptual Audio Codec's (I listen to FLAC). But there is a lot of information about how we hear and how sound interacts from that research.