Trinity's claims vs. what actually happens in digital

Al M. · Jan 13, 2014

I simply don't buy Trinity's claims about the shape of a 10 kHz sine wave signal at 48 kHz sampling rate on their Trinity DAC website:

http://www.trinity-ed.de/typo/index.php?id=19&L=1

It seems to be pretty well debunked by this website:

http://onetwothreeaudio.wordpress.com/2013/08/07/digital-audio-basics-stair-steps-and-sample-rates/

and the video:

http://www.youtube.com/watch?v=cIQ9IXSUzuM

The video appears credible since he uses defined equipment on which his experiments should be easily reproducible -- his demonstration is convincing.

(I was alerted to the education website and video by other WBF members on another thread --thanks.)

esldude · Jan 13, 2014

If they really know as little as their claims portray, they have no business building DAC's. They are telling an untruth.

But the designer speaks about it in post #9 here:
http://www.whatsbestforum.com/showthread.php?12023-Trinity-DAC

So maybe it is just marketing department garbage.

TRINITY · Jan 14, 2014

Unfortunately you do not understand the purpose and the content of the paper. It was written by me as an engineer for engineers or for people with some technical knowledge about the sampling theorem.
The purpose of the paper was to show the limitation of the “LIANOTEC” architecture and not only the benefits.
As you can see all shown pictures are screen prints of measurement instrument and not power point sheets!! No way to cheat.
“If they really know as little as their claims portray, they have no business building DAC's. They are telling an untruth.”
I personally think that is pretty rude and if I want to be rude by myself my answer would be “two blind persons discuss about color”.:-(
Instead here the “engineering” answer:

It is very simple to understand that a sampling system with 44.1kHz sampling rate and an input signal of 10kHz can only produce roughly 4 sampling states 44100/10000=4,41states. That is what the simulation and the original article shows, but in advantage you can also see the limitation of such a sampling process. At the second peak of the sine wave you can see that the sampled value is significant lower than the original input signal. That is the point where you have to read the “A Mathematical Theory of Communication” published in 1948, Claude E. Shannon http://cm.bell-labs.com/cm/ms/what/shannonday/paper.html very carefully, since it is more important what “Nyquist Didn’t Say”, that the amplitude and phase of the signal can complete recovered, if the sampling frequency is twice as high as the highest signal frequency. Further helpful reading can be found here: . http://www.wescottdesign.com/articles/Sampling/sampling.pdf
By the way we have no marketing department at all.
Please donot expect any further conribution here in the Forum from me, since I have to spent my time for the development of the power amp.

Audiocrack · Jan 14, 2014

esldude said:
If they really know as little as their claims portray, they have no business building DAC's. They are telling an untruth.

But the designer speaks about it in post #9 here:
http://www.whatsbestforum.com/showthread.php?12023-Trinity-DAC

So maybe it is just marketing department garbage.

Don't you think it would be wise to choose your words somewhat more carefully?

esldude · Jan 14, 2014

Okay, apologies are in order from me. So I am sorry I was too quick and blunt in my disagreement with Trinity. It appeared that he was showing Nyquist is wrong because you can't get back a proper 10 khz wave with 48 khz sampling. That part isn't true except under these peculiar circumstances.

What is wrong with the claim about Nyquist? Extremely simple. Output 10 khz from your DAC and look at it on an o-scope. It will be recovered with good precision. Why did he not show both the input signal, the sampled output and the analog output above? Because a NOS DAC also lacking an output filter would indeed put out the square blocky shape he shows. But that isn't showing the Nyquist theorem is wrong. It is showing a poor design of a DAC in one sense.

Now the Trinity DAC then uses 8 times oversampling, again without output filter. Still not a great result. So then he uses multiple time staggered DACs to achieve something like 64x oversampling and still no output filter. And that result isn't bad. Still the rejection of the higher frequencies isn't all that great. Though they are by then so high it likely cannot matter. So in one sense it is a slick design. But all this, and at great expense to avoid using an output filter. It improves the impulse response. But it doesn't show a problem with Nyquist, just a problem in insisting on his way of implementing it as a filter-less design.

Now the two links above by Al. M don't really debunk it. The key thing is the Trinity DAC does not use an output filter. And you would get that stair step response if you ran your DAC at only 48 khz and no filter. Here is a review of an earlier MSB DAC which could be switched to output without filtering and 8 times oversampling. Look at figure 13 on the page. You do indeed get stair steps.

http://www.stereophile.com/content/msb-platinum-link-plus-da-processor-measurements-part-2

I have to wonder, would the sound really be harmed to use a single poll 6 db roll off along with this technique. It would improve most specs though not impulse and square wave reproduction.

XV-1 · Jan 14, 2014

esldude said:
Okay, apologies are in order from me. So I am sorry I was too quick and blunt in my disagreement with Trinity. It appeared that he was showing Nyquist is wrong because you can't get back a proper 10 khz wave with 48 khz sampling. That part isn't true except under these peculiar circumstances.

What is wrong with the claim about Nyquist? Extremely simple. Output 10 khz from your DAC and look at it on an o-scope. It will be recovered with good precision. Why did he not show both the input signal, the sampled output and the analog output above? Because a NOS DAC also lacking an output filter would indeed put out the square blocky shape he shows. But that isn't showing the Nyquist theorem is wrong. It is showing a poor design of a DAC in one sense.

Now the Trinity DAC then uses 8 times oversampling, again without output filter. Still not a great result. So then he uses multiple time staggered DACs to achieve something like 64x oversampling and still no output filter. And that result isn't bad. Still the rejection of the higher frequencies isn't all that great. Though they are by then so high it likely cannot matter. So in one sense it is a slick design. But all this, and at great expense to avoid using an output filter. It improves the impulse response. But it doesn't show a problem with Nyquist, just a problem in insisting on his way of implementing it as a filter-less design.

Now the two links above by Al. M don't really debunk it. The key thing is the Trinity DAC does not use an output filter. And you would get that stair step response if you ran your DAC at only 48 khz and no filter. Here is a review of an earlier MSB DAC which could be switched to output without filtering and 8 times oversampling. Look at figure 13 on the page. You do indeed get stair steps.

http://www.stereophile.com/content/msb-platinum-link-plus-da-processor-measurements-part-2

I have to wonder, would the sound really be harmed to use a single poll 6 db roll off along with this technique. It would improve most specs though not impulse and square wave reproduction.

Most who have heard the Trinity DAC have thought it is one of the best, if not the best.

What did you think of its sound when you heard it?

Cheers

esldude · Jan 14, 2014

XV-1 said:
Most who have heard the Trinity DAC have thought it is one of the best, if not the best.

What did you think of its sound when you heard it?

Cheers

Have not heard it, and that fact has not one thing to do with what I wrote in this thread.

Whatmore · Jan 14, 2014

Are those charts screen prints of measurement instruments or screen prints of simulations ?
And, as I don't have much knowledge in this area, does it actually matter?

microstrip · Jan 14, 2014

XV-1 said:
Most who have heard the Trinity DAC have thought it is one of the best, if not the best.

What did you think of its sound when you heard it?

Cheers

My main reservations about the Trinity DAC would apply to any top very expensive high-end equipment in similar conditions. Distribution and availability of the unit is very scarce - it was not even shown in CES2014 - so very few people have listened to it. The very few existing reports (most is a nice word to hide it) are essentially reports of the type I owned the expensive X DAC, this one was so much better that I bough it immediately, with little real value description of the evaluations.

Unfortunately, as it is not easy to listen to the Trinity DAC and the production seems to be very little, iI seems that the internet marketing on it must live on the discredit of the established top DACs and rumors, not on its real sound qualities or its real value. Anyway, I hope that as soon as more people get it and listen to it, we can have real reviews of this DAC. I should add that if the Trinity DAC was available for listening in my country, I would surely listen to it. BTW, I am grateful to all the WBF members who enthusiastically posted on the Trinity DAC - if it was not them we would not be debating it.

JackD201 · Jan 14, 2014

I believe it was shown at THE.

Al M. · Jan 14, 2014

Esldude,

if I understood correctly what you point out, the problem is not one with Nyquist. The video that I posted confirms this.

Rather, the problem is one with no filtering. If you have no filtering, if I understand correctly, you have digital content that is much higher in frequency. This needs to be removed, otherwise it would introduce content that is above Nyquist and thus would distort the clean wave forms shown in the video (the video does mention that higher frequencies, and only that, would distort the Nyquist sampling by adding content above the cut-off).

As the comment above graph 13 in your Stereophile link says, "(The sharply defined steps in the waveform imply the existence of content well above the signal's nominal passband of 22.05kHz.)"

But then the claim on Trinity's website is highly misleading, when they say "Figure 1 compares the output signal of a D/A converter ...". For the sake of truthfulness, the text should be modified to mention that it is a filterless D/A converter, which is emphatically not the norm in digital. Otherwise it becomes cheap marketing that preys on the ignorance of people (and yes, a few weeks ago I might have questioned Trinity's claims to a lesser extent too).

***

So it seems clear to me that there is nothing wrong whatsoever with 16/44 digital in theory, but only with its practical (filtered or non-filtered) implementations.

jsn · Jan 14, 2014

microstrip said:
My main reservations about the Trinity DAC would apply to any top very expensive high-end equipment in similar conditions. Distribution and availability of the unit is very scarce - it was not even shown in CES2014 - so very few people have listened to it. The very few existing reports (most is a nice word to hide it) are essentially reports of the type I owned the expensive X DAC, this one was so much better that I bough it immediately, with little real value description of the evaluations.

Unfortunately, as it is not easy to listen to the Trinity DAC and the production seems to be very little, iI seems that the internet marketing on it must live on the discredit of the established top DACs and rumors, not on its real sound qualities or its real value. Anyway, I hope that as soon as more people get it and listen to it, we can have real reviews of this DAC. I should add that if the Trinity DAC was available for listening in my country, I would surely listen to it. BTW, I am grateful to all the WBF members who enthusiastically posted on the Trinity DAC - if it was not them we would not be debating it.

Hi microstirp,
I'm new to this forum.
May I know example of report having 'a lot of real value description of the evaluations' in this forum so that we can have base line of rerference ?
Thank you.

redsquare · Jan 14, 2014

I will listen to the Trinity dac + media server today. Apparently it's not so
obscure as some make it out to be (at least not in the USA)
Trinity is also not a new brand so to speak so not sure why the resentment other than
jealousy from industry veterans who use the forums to perpetuate the "old
guard."
Fremer from Stereophile and Nack from Positive Feedback apparently adore the phono and
pre, so the products seem consistently superb. I will report back shortly with my
findings and judge for myself...

microstrip · Jan 14, 2014

jsn said:
Hi microstirp,
I'm new to this forum.
May I know example of report having 'a lot of real value description of the evaluations' in this forum so that we can have base line of rerference ?
Thank you.

jsn,

Welcome to the forum. We debated the value of reviews and member evaluation several times in this forum. IMHO the value we credit to a review is given mostly by our past knowledge of the member in the forum, the experience he has shown in his previous posts, and also how we correlate our and other experiences with his own. The same for professional reviewers - unless we know his preferences and full details of systems and rooms it is very difficult to learn from them. We had threads where we expressed our personal opinions about reviewers and examples of reviews we appreciated. WBF has now a large archive of threads with passionate debates - may I suggest that you look at them?

BTW, remember that in this forum subjective statements are just individual opinions, others are free to disagree and debate them.

Audiocrack · Jan 14, 2014

microstrip said:
My main reservations about the Trinity DAC would apply to any top very expensive high-end equipment in similar conditions. Distribution and availability of the unit is very scarce - it was not even shown in CES2014 - so very few people have listened to it. The very few existing reports (most is a nice word to hide it) are essentially reports of the type I owned the expensive X DAC, this one was so much better that I bough it immediately, with little real value description of the evaluations.

Unfortunately, as it is not easy to listen to the Trinity DAC and the production seems to be very little, iI seems that the internet marketing on it must live on the discredit of the established top DACs and rumors, not on its real sound qualities or its real value. Anyway, I hope that as soon as more people get it and listen to it, we can have real reviews of this DAC. I should add that if the Trinity DAC was available for listening in my country, I would surely listen to it. BTW, I am grateful to all the WBF members who enthusiastically posted on the Trinity DAC - if it was not them we would not be debating it.

Actually - as mentioned earlier - the Trinity dac was reviewed by Audio Technique and got a very positive review.

Audiocrack · Jan 14, 2014

redsquare said:
I will listen to the Trinity dac + media server today. Apparently it's not so
obscure as some make it out to be (at least not in the USA)
Trinity is also not a new brand so to speak so not sure why the resentment other than
jealousy from industry veterans who use the forums to perpetuate the "old
guard."
Fremer from Stereophile and Nack from Positive Feedback apparently adore the phono and
pre, so the products seem consistently superb. I will report back shortly with my
findings and judge for myself...

Please do. However, if you would really like the Trinity dac and post your honest feedback some guys on this Forum for whatever reason (and yes sometimes I do think it is indeed just jealousy and hopefully not stupidity) are not going to accept it.

microstrip · Jan 14, 2014

Audiocrack said:
Actually - as mentioned earlier - the Trinity dac was reviewed by Audio Technique and got a very positive review.

Thanks Audiocrack - I would love to read the review. Is it accessible?

Audiocrack · Jan 14, 2014

microstrip said:
Thanks Audiocrack - I would love to read the review. Is it accessible?

As far as I know not yet in English. I hope an English translation will come available because apparently it is rave review and I would like to know why and in what regard the reviewer prefers the Trinity dac to the Dcs Vivaldi the reviewer also uses. But I am not sure whether such a translation will be published.

microstrip · Jan 14, 2014

Audiocrack said:
As far as I know not yet in English. I hope an English translation will come available because apparently it is rave review and I would like to know why and in what regard the reviewer prefers the Trinity dac to the Dcs Vivaldi the reviewer also uses. But I am not sure whether such a translation will be published.

Yes, the why's and the what's are the interesting part of the game ... Adam referred to the Scarlatti in these threads - but IMHO the Vivaldi should be the valuable comparison now-a-days, as many people know how it sounds. BTW, my interests are biased towards high quality reproduction of Redbook - the music I listen and really enjoy exist mostly in CD and LP.

Al M. · Jan 14, 2014

Audiocrack said:
Please do. However, if you would really like the Trinity dac and post your honest feedback some guys on this Forum for whatever reason (and yes sometimes I do think it is indeed just jealousy and hopefully not stupidity) are not going to accept it.

Why would you think that? I certainly would have no problem with the idea that it sounds great or even may be the best (while I could never afford a $50,000 DAC and am enormously happy with my Berkeley Alpha DAC 2 at 1/10th of the price). On the other thread I only had quibbles about comparing a barely warmed up dCS Vivaldi with a continuously powered Trinity DAC -- while keeping the option open that the Trinity DAC indeed might be better.

And my problems with Trinity as posted on this thread are not with the Trinity DAC itself, but with the gross misrepresentation on their website of what digital really does. As Esldude pointed out, as written the suggestion appears to be that Nyquist sampling at 48 kHz simply isn't good enough, which is debunked in the video. Rather, the issue appears to be one of filtering. I hate misrepresentations of 16/44 digital since the medium has been, as it turns out with recent sonic developments, so often unjustly maligned over the decades -- and many have pretended that hi-rez is the only solution (really, with the measly music selection available there?). I can only imagine biased vinyl fans (nothing against vinyl per se) pointing at the graphs on the Trinity website and gleefully proclaiming, "see I told you so" about 16/44 digital -- while the claims made there are not revealing the entire truth.

Trinity's claims vs. what actually happens in digital

VIP/Donor

New Member

Well-Known Member

Attachments

Well-Known Member

New Member

Well-Known Member

New Member

Well-Known Member

VIP/Donor

WBF Founding Member

VIP/Donor

VIP/Donor

New Member

VIP/Donor

Well-Known Member

Well-Known Member

VIP/Donor

Well-Known Member

VIP/Donor

VIP/Donor