What a long, strange trip it's been -- (Robert Hunter)
post-14028321
Post #6,706 of 13,378

amirm

Member of the Trade: Madrona Digital
Joined
Sep 26, 2017
Messages
596
Reaction score
317
Location
Seattle, USA
Joined
Sep 26, 2017
Location
Seattle, USA
Posts
596
Likes
317
Website
www.audiosciencereview.com
Edit:
But the guy that dropped his measurement bomb on the Schiit happened thread didnt show up here? It saddens me...I even got some popcorn.
I didn't drop any bomb. Accusations were made that my measurements were in error/conflict with another glowing one. I chimed in to show that there was actually agreement with two sets of measurements and it was the commentary from the other dude that was wrong since it showed the same flaws.

As to not wanting to be here, I have been warned by multiple people that discussions with forum sponsors will lead to banning from the forum. I have been treated exceptionally well by the moderator in the science forum but out of abundance of caution, I am not going to post here. Now if there can be an assurance that nothing like this will happen, I am happy to participate.

As an example of that, this is the jitter measurements of Schiit BiFrost Multibit:



Everything other than the center line is distortion/jitter/Vref error. As you see, the graph in red is awful compared to the yellow/green for the cheaper DACs. The signal is full amplitude so has nothing to do with any glitch problem.

Someone wants to explain why this is acceptable in a well implemented DAC?
 
post-14028327
Post #6,707 of 13,378

amirm

Member of the Trade: Madrona Digital
Joined
Sep 26, 2017
Messages
596
Reaction score
317
Location
Seattle, USA
Joined
Sep 26, 2017
Location
Seattle, USA
Posts
596
Likes
317
Website
www.audiosciencereview.com
Disappointing to read that if it's true.
Of course it is true. Here is what I see when I go to John's thread:

upload_2018-2-7_20-0-22.png


And reply button is missing. This happened a minute after he said I should ask questions here.
 
     Share This Post       
post-14028346
Post #6,708 of 13,378

DoubleIPA

100+ Head-Fier
Joined
Apr 28, 2017
Messages
161
Reaction score
157
Location
Colorado Springs, Colorado
Joined
Apr 28, 2017
Location
Colorado Springs, Colorado
Posts
161
Likes
157
Because of what they do with their input to achieve their output. One use original samples the other resamples. Obviously that's a very very generic explanation, but so was the way your stated your question. However there are white papers that go over this in great detail.
 
     Share This Post       
post-14028349
Post #6,709 of 13,378

amirm

Member of the Trade: Madrona Digital
Joined
Sep 26, 2017
Messages
596
Reaction score
317
Location
Seattle, USA
Joined
Sep 26, 2017
Location
Seattle, USA
Posts
596
Likes
317
Website
www.audiosciencereview.com
Because of what they do with their input to achieve their output. One use original samples the other resamples. Obviously that's a very very generic explanation, but so was the way your stated your question. However there are white papers that go over this in great detail.
Here are the measured results of those two ways:



Which one do you think is more faithful to samples it is told to reproduce?
 
post-14028379
Post #6,711 of 13,378

DoubleIPA

100+ Head-Fier
Joined
Apr 28, 2017
Messages
161
Reaction score
157
Location
Colorado Springs, Colorado
Joined
Apr 28, 2017
Location
Colorado Springs, Colorado
Posts
161
Likes
157
Here are the measured results of those two ways:


Which one do you think is more faithful to samples it is told to reproduce?
I never said which on was better or more faithful, I just said they should sound different because their implementation is different. If you think all DAC's the same then that's great, I'm not hear to argue.

I'm not sure who the John is that you're referring to above?
 
     Share This Post       
post-14028403
Post #6,712 of 13,378

Don Hills

100+ Head-Fier
Joined
Jun 13, 2013
Messages
404
Reaction score
202
Joined
Jun 13, 2013
Posts
404
Likes
202
Because of what they do with their input to achieve their output. One use original samples the other resamples. Obviously that's a very very generic explanation, but so was the way your stated your question. However there are white papers that go over this in great detail.
Why should their analogue outputs differ because of the way they convert input to output? I know why they do often differ, but what rule says they have to?
 
     Share This Post       
post-14028407
Post #6,713 of 13,378

Rtg97229

Head-Fier
Joined
Oct 12, 2014
Messages
50
Reaction score
101
Location
Portland Oregon
Joined
Oct 12, 2014
Location
Portland Oregon
Posts
50
Likes
101
Here are the measured results of those two ways:



Which one do you think is more faithful to samples it is told to reproduce?
I work in test and measurement and find your lack of documentation about your setup to take away the usefulness of your presented data. Please document your setup and calibration details if you want your data to be useful. The audio stuff is outside of my field but common sense should tell you that a 16 bit DAC does not have 20 bits of range.
 
     Share This Post       
  • Like
Reactions: ZetsuBozu0012
post-14028410
Post #6,714 of 13,378

amirm

Member of the Trade: Madrona Digital
Joined
Sep 26, 2017
Messages
596
Reaction score
317
Location
Seattle, USA
Joined
Sep 26, 2017
Location
Seattle, USA
Posts
596
Likes
317
Website
www.audiosciencereview.com
I never said which on was better or more faithful, I just said they should sound different because their implementation is different.
High quality implementation of each eliminate sonic differences. The way we know the implementation is high quality is to look at the measurements. Not what scheme the underlying internal DAC chip uses. It is the system implementation that usually destroys the fidelity of the chip that is used.
 
     Share This Post       
post-14028418
Post #6,715 of 13,378

amirm

Member of the Trade: Madrona Digital
Joined
Sep 26, 2017
Messages
596
Reaction score
317
Location
Seattle, USA
Joined
Sep 26, 2017
Location
Seattle, USA
Posts
596
Likes
317
Website
www.audiosciencereview.com
I work in test and measurement and find your lack of documentation about your setup to take away the usefulness of your presented data. Please document your setup and calibration details if you want your data to be useful. The audio stuff is outside of my field but common sense should tell you that a 16 bit DAC does not have 20 bits of range.
What do you want documented? The devices are dead simple. A digital stream is sent to their S/PDIF and analog output is measured. This is stated in the review: https://www.audiosciencereview.com/...d-review-of-schiit-bifrost-multibit-dac.2319/

"Let's test the two units using S/PDIF digital input and see what that shows:"

As to 16 bits, did you see how the Bifrost Multibit starts to lose accuracy at just 58 db which translates into 10 bits or so as indicated on the graph?

Calibration is not material here as the two DACs are measured immediately one after the other. It is the relative difference you want to look at.

And how was your interest in test and measurement satisfied prior to my measurements being posted? You liked the product without it???
 
post-14028422
Post #6,716 of 13,378

Rtg97229

Head-Fier
Joined
Oct 12, 2014
Messages
50
Reaction score
101
Location
Portland Oregon
Joined
Oct 12, 2014
Location
Portland Oregon
Posts
50
Likes
101
What do you want documented? The devices are dead simple. A digital stream is sent to their S/PDIF and analog output is measured. This is stated in the review: https://www.audiosciencereview.com/...d-review-of-schiit-bifrost-multibit-dac.2319/

"Let's test the two units using S/PDIF digital input and see what that shows:"

As to 16 bits, did you see how the Bifrost Multibit starts to lose accuracy at just 58 db which translates into 10 bits or so as indicated on the graph?

Calibration is not material here as the two DACs are measured immediately one after the other. It is the relative difference you want to look at.

And how was your interest in test and measurement satisfied prior to my measurements being posted? You liked the product without it???
I don't own a Bifrost and never have so I can't answer if I like it or not. I work in test and measurement and it gets my attention that you do not document your setup and calibration well to go along with your presented data.
 
     Share This Post       
  • Like
Reactions: Lolito
post-14028427
Post #6,717 of 13,378

bosiemoncrieff

Headphoneus Supremus
Joined
Mar 21, 2015
Messages
2,953
Reaction score
2,220
Location
San Francisco
Joined
Mar 21, 2015
Location
San Francisco
Posts
2,953
Likes
2,220
ugh this sound science circle jerk is putting me to sleep. If you can hear flaws in Yggy, you really have too much time on your hands.

I see Parsifal a week from Saturday. What's everyone's favorite recordings?
 
Last edited:
post-14028431
Post #6,718 of 13,378

Don Hills

100+ Head-Fier
Joined
Jun 13, 2013
Messages
404
Reaction score
202
Joined
Jun 13, 2013
Posts
404
Likes
202
... Which one do you think is more faithful to samples it is told to reproduce?
Monotonicity / linearity tests like this used to be common in reviews of CD players back in the 80s before they went to D-S topologies. I don't recall ever seeing one that bad in any player with pretensions to being HiFi. It does make me wonder how a modern high performance chip could apparently be so bad.

(I think it's all a bit "angels on the head of a pin" when it comes to audibility. I'm just surprised, I thought that the state of the art in R2R DACs, even at this price point, should be past this.)
 
     Share This Post       
  • Like
Reactions: tili
post-14028436
Post #6,719 of 13,378

Rtg97229

Head-Fier
Joined
Oct 12, 2014
Messages
50
Reaction score
101
Location
Portland Oregon
Joined
Oct 12, 2014
Location
Portland Oregon
Posts
50
Likes
101
It is not like it would be hard to satisfy me on this documentation thing. Here is the associated equipment, here are the calibration dates, here is a beautiful looking noise floor test just in case, here are some pictures of the test setup, here is the data, thank you for reading.
 
     Share This Post       

Users Who Are Viewing This Thread (Users: 6, Guests: 5)

Top