The Border Patrol DAC - Maybe linearity in a DAC is bad ... Spitballing


Hi Everyone,
I've been thinking about a few things related to DAC's and how they behave and how we hear. Also thinking about a couple of audiophile comparisons I've heard and how we interpret what we hear.

Let's talk about this simple measurement called linearity.

In a DAC what we mean is that as the magnitude changes the output changes the same amount. That is, if the signal says "3 dB softer" you want to get exactly 3 dB softer output on the jacks.

And with modern, top tier DACs this is usually really good until around -90 dB where noise becomes the limiting factor.

For a long time I felt that a DAC which allowed me to hear the decay of a note, so that it fades instead of stops suddenly was the mark of a truly excellent sounding DAC.

I'm wondering if what I'm actually hearing is compression? Lack of linearity.

The reason I bring this up is that I was reading a long article about the complexities of reviewing a DAC from Border Patrol. One of the main failings, from measurements, is that it is really not linear at all. Sounds don't get softer fast enough. And ... low and behold, Herb Reichert actually makes many comments about how much more he can hear with this DAC than with others.

I'm going to link to a critique of the "scandal" so you all can get a better look:


https://parttimeaudiophile.com/2019/01/06/border-patrol-dac-revisited-audio-fur/


Also, take a look at the linearity charts in the original review. Honestly, awful. Not up to what we expect in state of the art DACs today, but ....


https://www.stereophile.com/content/borderpatrol-digital-analogue-converter-se-measurements

What do you all think? Do we need a compression feature in DACs so we can hear more details? That would make more sense to me than a lot of the current fad in having multiple filter types.

Best,
E

erik_squires
The issue is that even poor measuring DACs still have barely audible distortion or whathaveyou, unless it’s a total garbage product.

It also has been shown that ~10sec is the limit for when one can accurately compare the sound of one instance to another (with precise detail), so any reviewer talking about another product they reviewed 6 months ago should not and cannot be held as valid.

Also, sighted listening that’s likely not level matched should also be taken with a grain of salt. If I go to an audio show and show two DACs, one a cream of the crop $10,000 one with a beautiful exterior, and the other being the same product but in a cheap plastic casing, I bet 99% of the attending audience will pick the former as sounding better.
Thanks for pointing out that very interesting article by part-time audiophile, Eric.

I recently acquired and love the Border Patrol DAC SE. I find it to be the most transparent DAC I’ve owned ( previously owned Schiit Yggdrasil, PS Audio Perfectwave and Concert Fidelity DAC 040-BD) and Herb Reichert along with Scot Hull praised it for being so. Measurements are meaningless to me, I only care how a device sounds and the BP DAC SE is musical bliss. Both Herbs review and Scots article are spot on in my opinion. 

I’be never read such a review as Jons’ before. One where JA seemingly recruits Jon Iverson to confirm the DAC can’t be all that good because it didn’t perform on his test bench well and where Iverson admits to his preconceived bias from the start. Very interesting to say the least. 

Not sure if DAC’s require more compression or not to reveal more detail but whatever Gary at Border Patrol did with his DAC has it revealing things in my CD’s previously unheard yet doing so in a very “analog” manner....ie, unfatiguing and natural. 
I don’t know how to reconcile the competing views among the reviewers, including those writing for Stereophile, other than to listen for yourself. You could, I suppose, write all of this off as just another example of the divide between the subjective and objective schools of audio. Doesn’t this assume that measurements showing no departure from the input signal mean ’more accurate’ and (here’s the rub in my estimation) that ’more accurate’ measurements invariably mean a more convincing reproduction of a musical event?
I’m not so sure you can always make that leap. Music is complex. Accuracy to a series of test signals tells me something but it doesn’t tell me everything. Accuracy to the master recording? Leaving aside the colorations introduced in the recording chain, what’s been done to the master to make it sound ’good’?
Accurate to the live event? Where are you sitting? How good is your sonic memory?
I think, for me, the easiest way to make judgements about reproduced sound is to ask whether it sounds real. Does a piano sound like a real piano? Very few recordings of piano sound convincing to me, convey the gravitas of the lower registers and the air around the upper registers with the bite of the attack and the gradually diminishing envelope of harmonic decay.
Subjective preference? I think some people like to hear all the detail. My preference would probably go the other way, toward sins of omission, since I often find analytical systems fatiguing. But, those are conclusions, aren’t they? One person’s "analytical" is another’s "fully detailed and revealing."
Perhaps, rather than reaching conclusions, the reviewer's job is to report what they heard, without all the gloss about hot sauce and fur. One of the most interesting things about that whole dog fight was Atkinson’s characterization of his measurements as an "opinion."[1]
Full disclosure: I bought a Border Patrol and like how it sounds. I’m basically an analog guy in a digital world. I find the biggest variable to be the difference in recording quality and mastering. That’s been a long standing issue for me in vinyl-land, and as I have only recently started to play with digital in my main system (which is far more weighted toward vinyl playback in terms of expense), I’ve been sorting through various masterings of the same recording in digital formats, mainly Redbook. There are profound differences there, which I guess should be no surprise, but I’m new at this digital thing.
Will I go all spendy on a DAC at some point? Maybe. Does the BP sound like real music to me? I’ve been pretty pleased with what I hear on some recordings, despite many years of ambivalence toward things digital....
_______
[1]"In our review, the DAC SE impressed Herb Reichert for delivering "refined, human-sounding musical pleasures," but didn’t do well on the test bench. (In particular, I criticized BorderPatrol’s use of what I felt was an "underperforming" chip.) Given this conflict, I felt a third opinion was called for...".

I agree a lot with whart. I am not any kind of a measurements guy. But I do not think we can measure what Herb R. says is 'human sounding musical pleasures'. In a live event that is what it is like,  'human sounding musical pleasures'.  And also there is something about the physical feeling to a live event, I like to have much of that in my listening experience at home. And for me that does not mean listening to my music loud or bass heavy. Just that feel of being in the presence of a real musical production. That is when it becomes more real to me. I don't think you can measure that. And I don't need the detail to hear flies bumping into each other on the recording. Only just enough detail to make things real. I certainly pay attention to measurements initially to see if components are somewhat compatible, but after that I don't make use of them. So I appreciate the measurements and specs a whole lot to get started but not for a lot of my musical pleasure, there I appreciate my ears.
But I do not think we can measure what Herb R. says is 'human sounding musical pleasures'.


In this case, I actually genuinely think we can. My hypothesis is that he is enjoying the compression at the bottom of the dynamic range.

I'm going to see if I can prove this. :)

Stay tuned.

E

That would be nice if it could be correlated to what we perceive as 'human sounding musical pleasure'. Then it could be designed into our components. But so many, by my wild approximation of 250, systems I have heard most have a very small measure of that sound. Interesting topic Erik. I think it('human sounding musical pleasure')for me is a kind of core importance.
The most common use of a compressor in music production is to reduce overall dynamic range by some combination of lowering the peaks and increasing the soft passages.  If uses judiciously it will bring out low level detail in a track.  It's a basic and widely used recording technique.
If the Border Patrol is manipulating the output signal to mimic a compressor, than it is not surprising that many listeners would like the effect.  An analogy would be upping the contrast slightly in a video signal.  There's no magic involved.
That is a a pretty natural sounding dac ,the New Audio Mirror
is very similar and uses Vacuum tubes currently $1200 Outstanding ,Audio Analog Multibit dac which yourarely see anymore the Burr Brown 1704  dac chipsetcomes to mind.
way to expensive to make ,greatchipsets of the past 

The border patrol is the best sounding dac I’ve heard. I found it very transparent and almost 
life like. I’ve not compared it to 10k plus dacs but some have said it’s the r2r arrangement. I’ve heard the denefrips dacs come close at twice the cost. Would like to hear the terminator. 
@mickeyb

I found it very transparent

Hmm, Stereophile’s measurements show it has some slightly audible distortion, and I was just in conversation with someone who owns one and says he likes that it adds a slight tube sound, and that while not as transparent as his Parasound ZDAC, he preferred it in some instances. It has piss-poor volume (bit-depth) linearity, which likely contributes to what this person described as tube like, as if we use a 0.1dB threshold like Amir from ASR likes to use (though he sometimes uses 0.5dB), the BorderPatrol technically has the resolution of 10Bit, whereas even the $80 Grace SDAC has 19Bit.

EDIT: Oh, and the Stereophile review (done by actually listening to it)  also states it’s not as detailed as say the Benchmark (rolled off highs and softer bass), and is more for those wanting a more intimate, smooth sound, rather than every detail exposed.
If the Border Patrol is manipulating the output signal to mimic a compressor, than it is not surprising that many listeners would like the effect. An analogy would be upping the contrast slightly in a video signal.

I don't think they are manipulating it so much as going with old school designs to make it happen. Based on the Stereophile measurements, I'm not even sure they know what they did. If they had, the L and R channels would be better matched.


From the Stereophile review, a rebuttal from the first reviewer:

In order to answer that question, I have purchased and studied at least 80 genuine (analogue) “master tapes” taken directly from the archives of RCA. Many of them I played back on the exact Ampex machines they were recorded on (which I also purchased). Virtually all of them sounded VERY “pleasing good” and therefore, by your definition, must ALSO have sounded accurate good. In truth, I can’t remember hearing too many bad sounding master tapes.

For my reviews I always use high-res “master” files supplied to me by the recordist for that purpose. When I finished my BP/BM comparison, I told JA about my own “alarming audiophile episode” with Macy Gray’s HDTracks album “Stripped,” wherein the BP DAC reproduced pretty much exactly what I heard sitting behind the binaural head at the former church in Greenpoint and the Benchmark DAC3 which did not even get close. It conspicuously stripped away a huge amounts of what I and David Chesky know is on the recording.

I use Chesky recording sessions to review headphones because I can compare what I hear live to the sound coming off the so-called “mike feed.” The Border Patrol DAC reproduced the church walls, the reverb, the positions on the floor where the musicians were standing, and all the subtle breathiness of Macy Gray’s voice. With the Benchmark, the majority of that information (which is definitely on the master file and appears via David’s $100K MSB DAC and via my Holo Spring DAC) disappeared !!! Your neutral DAC “stripped” away information that is unquestionably on the master file. Not to mention the BM DAC made it sound hard cold and harmonically threadbare. I call this subtractive distortion. Did you measure any of that?  


It’s all in the ear of the beholder.

All the best,
Nonoise
Nonoise:
Yes, and this part in particular catches my eyes:
I use Chesky recording sessions to review headphones because I can compare what I hear live to the sound coming off the so-called “mike feed.” The Border Patrol DAC reproduced the church walls, the reverb, the positions on the floor where the musicians were standing, and all the subtle breathiness of Macy Gray’s voice. 

I am wondering if this is not indeed compression, and that we like compression.

Best,
E

Erik:
Don't let the objectivists hear you say that. They would consider it blasphemy and subject you to something jagged, or worse.  😄

All the best,
Nonoise

These "meters vs. ears" discussions always remind me of the film-to- video evolution of the early seventies. I owned a well established film/video studio in Manhattan. I always got a kick out of the debates between the film directors and the video engineers. The video eng. would point to his monitors and scopes and prove the image was “picture perfect” in hue and luminence. “And you can see the threads on the buttons of his shirt!” And the director would scream “no-one can walk into a dimly lit room, open a window blind to sunshine, then crawl under a bed and not have a single shadow on them! The video eng. relished the technology and the film director looked for realism and emotion. I admit I like the dancing needles on McIntosh gear but, I find my enjoyment with the listening and pay little attention to the meter readings.


@gwbeers, 
Nice analogy. My father was a film editor back when they used film. Towards the end of his career digital was coming on scene and he hated it. It's implementation lagged way behind the promises. 

Today it's great but back then, as a child, even I could see how bad it was.

All the best,
Nonoise
Numerous posts on this thread attest to the very good sound quality of the Border Patrol DAC (BP). I’ve heard the BP and Benchmark (BM) although not in the same systems. In a general overall assessment the BP was more natural sounding and realistic IMO.

Herb Reichert’s rebuttal comments are exceedingly telling comparing the BP versus the BM. The BP captured and revealed master tape information /nuance completely missed by the BM DAC. in addition the BM struck him as harmonically threadbare, hard and cold (similar to my listening impression).

The BM has superior measurements and specifications. What does this yield? What is the benefit of the better numbers? Why does the poorer measuring BP retain critical master tape sonic and musical information that "disappears" when heard through the BM? It suggests that the ’superior measurements have little correlation with actual sound quality. Most experienced listeners already know this.
Charles
@charles1dad

@nonoise

Because all these listening impressions are done knowing the brand/price/looks.

Unless you are doing double-blind, quick-switching, level-matched comparisons, it is scientifically impossible to thoroughly compare the sound of two different products. So Michael Furmer’s story of how he heard a difference in his $18,000 speaker cables when he was in another room working while another person wired them up, cannot he taken as there truly being a difference, even if he believes so.

I could make my own speaker cables for $100, give a BS description, charge $5000, and I gaurentee you if I do a demo at an audio show, I will get positive reviews and maybe even a customer or two.

There is no way the Benchmark DAC could remove the soundfield information of being recorded in a church, the placement of the musicians, etc. Those comments easily show that it’s all in his mind. You can talk about tonal balance, distortion, noise floor, channel separation, etc., but saying it removes the church walls is just ridiculous.

And again, the Benchmark is proven to be not cold, regardless of sighted listening impressions. Being cold is a rolled off bass (or emphasized treble; kinda the same thing if volume matched), and the Benchmark is dead flat. It could only have rounded off bass if your speakers were made with tube amps in mind, very few exist today (I only recall seeing 1 brand), but maybe if your speakers are from the 60’s.

I hate to repeat this, but a speaker at an AES myth busting talk gave a story of how he tricked people into thinking a McIntosh tube amp was playing when in fact it was a solid state, and the people described it differently than the same solid state amp it was supposedly being compared against. If doctors have to give sugar cube placebos to make sure drugs work, you can bet our ears can be fooled in what we are hearing (just like our eyes were fooled with the color of the dress a while back).
@mzkmxzv,

You hydrogen based life forms are funny. Anyone can be tricked, even those who adhere strictly to measurements. It is with long term critical listening that we appreciate the differences and distinctions. 

Quick switching (A/B/X, whatever) is nothing more than a cheap parlor trick. Even your ears would be fooled into thinking you're hearing an excellently measured piece of kit about half the time when it's not.

Sighted references are just that and no amount of back seat psychobabble will account for all the situations as there are people who can overcome the stupidity of listening tests because they do have better ears. It's statistically inevitable that they can. Do the math since you're so good at it.

Also, all of your anecdotal evidence is just that. For every story you dredge up, there are those that prove the other point of view. And please, everyone, don't let someone dictate what "cold and sterile" is for you as mzkmxzc states.

For me, it's an unemotional and dry presentation. It can be full of bass and have what most would consider a normal treble but where it fails for me is a deficit of tone, body and timbre: a threadbare presentation. It simply won't pass the test of fooling anyone in an adjoining room or even off to the side  that something live is playing. It will measure well but it will not sound right.

All the best,
Nonoise
Reichert is a very experienced listener and professional reviewer who additionally had access to the live Chesky studio and master file. He had access to both DACs and could compare directly.  It doesn’t get much better than this in terms of knowing the actual sound of the recording you are listening to. 

 He is quite adamant that he heard musical information that was present on the master file  and subsequently conveyed convincingly with the BP .  The BM was notably inferior  under these listening circumstances according to him..  he simply and honestly reported what he heard. I don’t believe for a moment there are any ‘mind games’ being played here.  

One may not like the results of his listening sessions but he is reporting what he experienced.  In this scenario one DAC was clearly found to be better than the other, in this case the better sounding DAC happened to be the BP 
 Charles
He is quite adamant that he heard musical information that was present on the master file and subsequently conveyed convincingly with the BP .

Yep, and looking at the measurements, the lack of linearity is an interesting clue, one which supports a hypothesis I have suspected before I even heard of the BP DAC:
DAC’s are too linear. Very linear DAC’s cut off faster. DAC’s with compression are more revealing, without the frequency response aberrations associated with "revealing" speakers.

I’m going to try to prove this in 2019. :)

Best,
E


I was told how superior digital was when it was introduced. It sounded terrible to me. It’s only been in the last few years that I’ve heard material played on digital systems that provided a convincing illusion of musical reality.
The numbers, aside from questioning whether they are ’testing the right thing,’ also don’t reflect what gear sounds like in actuality, playing music. I remember hearing Spectral stuff back in the day- very precise, accurate, etc. (I owned Crosby modded Quad 63s at the time that confederation of West Coast audiophiles was using and modding such gear). To me, it sounded unnaturally precise- too precise if there can be such a thing. Real instruments don’t sound like that to my ears.
I know that puts us into a relativistic universe. I don’t rely on the opinions of others, no matter how credible. I’ve got to hear it for myself, preferably on my system, with a diverse assortment of material (still, in my estimation, the biggest bugaboo in this game- the source material often varies considerably in sonics, even different iterations or masterings of the same recording). I prefer to evaluate equipment using ’regular’ recordings, not audiophile spectaculars since I don’t usually listen to audiophile records as part of my musical diet.
One other factor- not sure how much it is taken into account when people listen at shows or in similar environments- how much ambient noise is affecting what you are hearing. I don’t tend to listen at LOUD levels, preferring to get as much musical information as I can at modest dB. To do that, you not only have to work with the noise floor of the system, but the surrounding noise of the listening environment. It’s pretty instructive to take a dB meter into your room and see just how noisy it is--
I don’t think you need ’golden ears’ to hear these differences. You do need access to the equipment, though, and often, that’s not in environments that are optimized for critical listening....
I had a very similar Spectral experience (Dr. Keith Johnson controlled the room) with them  driving Avalons at CES in the late 1990s. The sound quality was threadbare,  mechanical and artificial in my opinion.  About 30 minutes after this demonstration I heard in another room the exact same Avalon model speakers driven by Jeff Rowland electronics. Substantially better, natural and very emotionally involving relative to the Spectral components. Much more instrumental realism.  Those Avalons really revealed the stark sonic differences between the two brands. I wouldn't be at all surprised if the Spectral components had the superior measurements on a test bench. Hearing both there was no comparison.
Charles 
At the end of the day, you hear what you hear, and it’s your money.

And no, I’m not having my own definition of cold. Cold is the opposite of warm, warm sound is rolled off highs (like ELAC speakers, Andrew Jones himself states this is intentional and also described it as warm). If you don’t mean that, then don’t use the incorrect descriptive word.

And also, no, even if I made my own recordings, and compared to different DACs, if they weren’t compared to with the restrictions I stated, even I wouldn’t be able to accurately state which is closer to the original. Our brains are very stupid in this regard, there’s not a human alive which can accurately compare two audio products while knowing and seeing what the products are. 
 
I would like any explanation anyone can conjour up on how the Benchmark could remove the church walls in the recording. It’s like saying one speaker wire has a wider soundstage than another, it’s simply an impossible acchievmant as the two have nothing in common. It’s like me saying using fine china over everyday ceramic plates makes the food taste better (metal spoons over plastic spoons do in fact alter taste though).
I agree, that ears matter.

I do want to kind of emphasize that when I posted this thread I was using "linearity" as a very specific measure.

I don't mean to rehash all things about all measurements.

But, what if specifically, with DACs we prefer some compression at the bottom? Why not?

What if this is in fact correcting a problem in the ADC end?
I haven’t heard the BP or BM DACs but I’m also having trouble accepting that the BM DAC is actually "removing" musical content from the original recording. Someone used the video analogy earlier and my guess is that the difference has to do with the level of "contrast" presented by these two DACs, perhaps just like the view of the same image presented by different monitors. I suspect the ear/brain just focuses on different parts of the same content depending on the relative contrast within the content. My 2 cents.
@kalali

Don't you think these effects could be explained by compression?

Best,
E


Erik, it could be but I’m sure someone familiar with the recording process is definitely more qualified to address that (valid) question. The more I think about it the more the concept of “contrast” especially measured as a ratio, makes sense to me and I feel it also applies to other audio components. It’s a metric that I don’t think is captured in the typical audio-centric measurements such as frequency response, etc. It’s sort of like putting more “emphasis” on certain areas of musical “texture” which may or may not be directly related to the amplitude at that frequency. Again, perhaps experts like Ralph, Roger, or Steve, etc.,  can comment as whether or not the concept even applies to audio. Interesting analogy nonetheless.
Subjectivist chiming in who bought the Border Patrol SE Dac and returned it....

Apparently I like the decay and transparency of a typical DAC because I felt the Border Patrol sucked the life right out of my system.  It was coloring the sound but definitely not to my tastes and it made me want to leave my system turned off.  Anyone who thinks DACS all sound the same just needs to plug this one in!  

I honestly don’t get the rave reviews for it but hey, everyone’s tastes are different.  I certainly would never recommend this as a blind buy for someone though.

It’s very simple to explain how a DAC can "remove" musical content. It’s not that it’s literally removed, but rather that the resolution of one is superior, allowing to hear deeper and wider into the soundstage, thus capturing the extremely minute character of the reflections off the walls of the venue. (I will add that this is simply discussion of the phenomenon, not my conclusion re: these two products.) Without referencing the article I believe that is what is intended, and perhaps was not stated so eloquently. It is a very regular occurrence when using components of different manufacturers. All components vary in not just that respect, but many others.

See what I said about the comparison of the Benchmark DAC3 DX to the BP DAC SE at Dagogo.com

They are substantially different builds, and have substantially different sound characteristics.

Also, feel free to reference the discussion in the Cable forum regarding Doug Schroeder Method Double Interconnects.

Note, please that Border Patrol has indicated that this is NOT a suitable method for its NOS DAC. I have been using it very effectively with the Benchmark DAC3 DX. Note that Benchmark has not officially endorsed Schroeder Method; this is my exploration/research. This is an entirely do at your own risk activity. But, the results are stunning.


If you are tempted to rebut me using theory, know I am not interested in arguing or debating my viewpoint or methods. :)


@douglas_schroeder

It’s not that it’s literally removed, but rather that the resolution of one is superior, allowing to hear deeper and wider into the soundstage, thus capturing the extremely minute character of the reflections off the walls of the venue.


That would make sense if we weren’t talking about the Benchmark losing detail to the BP, as the former is vastly more transparent.

It’s more a simple fact that no one on the planet can accurately compare DACs unless it’s double-blind, level-matched, and quick-switching.
Apparently I like the decay and transparency of a typical DAC because I felt the Border Patrol sucked the life right out of my system.

If I interpret this as a loss of dynamics, it would be too much compression.

It’s not that it’s literally removed, but rather that the resolution of one is superior, allowing to hear deeper and wider into the soundstage, thus capturing the extremely minute character of the reflections off the walls of the venue. 


Which would be easier to hear with compression, not expansion.


Anyway, I don't know anythiung for sure, I'm just spit balling about compresion, and air and dynamics.

I used to think more air = more linear. I'm not sure now. :)

One album I remember distinctly with the Border Patrol was the Daft Punk RAM album.  It has several tracks with beautifully recorded drums and cymbals.

The cymbals with the BP Dac sounded dark and with very little decay, no sparkle and air.  This is not the cymbal sound I know from a live drum kit.  “Live” sounds wide open with plenty of HF energy and of course natural decay with a cymbal.  

Other DACs I’ve had in my system give goosebump moments on this album but it was a strange experience with the BP.
In the Stereophile 2nd opinion review the reviewer specifically mentions that the Border Patrol sounds like the Benchmark, but with a vintage tube compressor inserted.  It's not surprising that listeners would like the resulting sound.  It's a fleshed out, pleasurable sound.  Over the past years I've used two different devices to introduce slight amounts of compression into my primary playback system.  One was all analog, the SPL Tube Vitalizer and the other was digital, a Drawmer 2476.  Both worked well, but ultimately I stopped using them.  When taken out of the system I realized how they "homogenized" the sound.  I was a likable effect, but it was an effect.
There's no right or wrong with the Border Patrol vs. Benchmark.  I just wish people would approach it rationally with some understanding of the engineering involved.  Tube compression sounds really nice and recording engineers have known that since the 1950s.
In the Stereophile 2nd opinion review the reviewer specifically mentions that the Border Patrol sounds like the Benchmark, but with a vintage tube compressor inserted.

Good find, I missed it.

Also, given the difference in non-linearity and difference in listening experiences, I wonder if a major problem is manufacturing consistency in the R2R DAC?

Best,
E



Here is Michael Lavorgna's Twittering Machines review of the Border Patrol SE DAC:

https://twitteringmachines.com/review-border-patrol-dac-se/