Schiit Jotunheim 2 Review (DAC & Headphone Amp)

[threni]

What's wrong with saying "to my eyes the [incorrect] calculations are more correct?"

The fact that they're incorrect.

If your eyes distort what they see in the opposite direction of what the calculator miscalculates, you have a correct result.

My eyes don't do that.

 

[threni]

I would guess Schiit still makes the Asgard and Jot just to placate their small but vocal hardcore fan base at SBAF and HF. I would assume that the Hearsay and Magnius units have much more sales volume then these archaic discreet designs as well as higher profit margins.

Manufacturers design and sell hardware to make a profit. The idea that they'd make something which nobody will buy just to keep a group of people who post on a forum happy is the most ridiculous thing I've read about audio for a while. And I read reddit.

 

[PeteL]

Manufacturers design and sell hardware to make a profit. The idea that they'd make something which nobody will buy just to keep a group of people who post on a forum happy is the most ridiculous thing I've read about audio for a while. And I read reddit.

That is true, but Engineers are passionate about what they do, Moffat does stuff like multibit, unison usb, and such, because he believe it adds to the state of the art, because he likes the challenge of problem solving and actually creating stuff instead of strictly implementing a IC manufacturer reference design. He may be misguided, or plain wrong with what matter to him in audio reproduction, but You don’t go to these lenghts just to fool people. stuff like that is hard, and costly. Making another ak4493 dac is easy and cheap..

 

[threni]

That is true, but Engineers are passionate about what they do, Moffat does stuff like multibit, unison usb, and such, because he believe it adds to the state of the art, because he likes the challenge of problem solving and actually creating stuff instead of strictly implementing a IC manufacturer reference design. He may be misguided, or plain wrong with what matter to him in audio reproduction, but You don’t go to these lenghts just to fool people. stuff like that is hard, and costly. Making another ak4493 dac is easy and cheap..

Schiit aren't fooling anyone. They seem to like a challenge. They are open and seem honest about what they do and why.

the creation of their multibit stuff:
https://www.head-fi.org/threads/sch...obable-start-up.701900/page-554#post-12031225

why people are suffering delays in orders recently:
https://www.head-fi.org/threads/sch...bable-start-up.701900/page-4675#post-16102543

But you'll never see them say "We know nobody's going to buy this product. But sometimes you have to waste millions of dollars making something nobody will buy to avoid negative comments on the part of the internet populated by people who know nothing about audio, electronics or business".

Also, I read elsewhere that Schiit are forced to produced stuff which measures well because of sites like ASR! Seriously - someone actually typed that!

 

[Long Audio Beard]

So. much. broscience. Like sorry you're using a lot of pretty words but is there anything to back up the things you say? Do you have an electrical engineering or signal processing background? Or can you quote any literature about the things you claim?

wow, I guess someone does not like multibit and R2R DACs and can't fathom why anyone else would, enjoy your DACs, and to answer your question: yes and yes, and perhaps you should audition some different DACs with music you know very well and see for yourself

 

[Long Audio Beard]

Sorry sir, I cannot agree with that statement, not these days! Our ears really are pretty crap in term' of 'resolution' but the mind behind them can invent all manner of flights of fancy to justify an opinion. All this fine filigree detail, air and 'atmosphere' in recorded music really isn't very far down in level (just listen to the non-speaking channel on a phono cartridge to hear what '-30dB' actually sounds like).

I had my epiphany some time back as to what reality sounds like and how many good 1950's recordings (without all the 'tech' and 'mastering' in the way) can capture this raw honesty and 'truth' quite well when played on decent quality larger monitors. Digital seems to retain this far better than 'analogue' does so any 'tech' which makes a human voice sound 'nicer' tends to make me deeply suspicious, as real voices can have an 'edge' or 'starkness' to them that we take for granted in day to day living yet we don't always like in reproduced music - usually the speakers' fault but 'we' always look to the innocent party here - the source.

Apologies if the above seems confusing. I'm done with the notion we don't measure everything properly. maybe we don't always *interpret* what the measurements tell us, but our ears are crap pretty much really I think, although they're all we have for ourselves 'Nice sounding' audio almost always has deliberate flaws or colourations built in - and then you start to find every recording tends to sound equally 'nice' and there's little to no differentiation between recordings or productions - that's when alarm bells start to ring for me if not others.

I didn't mean to criticize any of our measurements or standards, just that we have standard measurements and probably more measurements will be developed, the state of the art is a moving target, in my own testing some, not all, DACs are better at reproducing voices and there is less of a sense that it is a recording, when people complain about a digital sheen or grit or brittle sound they should try those same recordings with some different DACs and see if they notice a difference, in video we talk about pixel sharpening and there are audio filters that do just that and with varying results for different styles of music, that's just an example

 

[Veri]

wow, I guess someone does not like multibit and R2R DACs and can't fathom why anyone else would, enjoy your DACs, and to answer your question: yes and yes, and perhaps you should audition some different DACs with music you know very well and see for yourself

I've auditioned stuff like holo audio. I own an old Sony TDA multibit-based DAC; you assume I haven't "tried" them. There's still no magic in there for me. And the 'Yes and yes' is a pretty easy cop out from my question.. got it.

 

[don'ttrustauthority]

The fact that they're incorrect.


My eyes don't do that.

Right, but he's saying his ears do that. His ears are incorrect, so it works. That's the point eh.

 

[Long Audio Beard]

It does. It doesn't know the difference. There is no intelligence in a DAC.

My tests are a lot more than 1 kHz anyway. IMD is dual tone of 60 and 7 kHz. Jitter is dual square waves with tons of harmonics, and multitone has 32 tones. THD+N vs frequency sweeps the entire audible range and extends response to 90 kHz.

no quarrel with your testing methods at all, just wanted to note that there are various filters at play in various DACs especially where the filtering is more bespoke whether through analog electronics or DSP programming (a lot of which is proprietary), with analog filtering we can readily assess the filtering and we know there is brickwall filtering and low pass and high pass, etc., with DSP filtering there are some algorithms that are more unusual and can provide logic for certain frequencies and time coherence that is not necessarily expected and even for groups or sequences of audio that further manipulate the final product that comes out of the DAC. With the major ready-to-go DAC chipset solutions from ESS and Cirrus Logic and TI/Burr Brown and AKM etc they are typically implemented by makers with chipset default filtering including those chipsets that have filtering choices built in, but the bespoke filtering deployed by some makers such as Chord and Schitt for example is interesting and when we measure DACs we are of course getting measurements of more than just the DAC topology and filtering, we are getting the power supply and analog audio stage too, in terms of Denafrips I have not looked in detail at their filtering but it may be more ordinary, yet there is some support that R2R produces less of certain noise issues to be filtered out compared to DS and whether that makes any difference I am dubious

 

[Long Audio Beard]

I've auditioned stuff like holo audio. I own an old Sony TDA multibit-based DAC; you assume I haven't "tried" them. There's still no magic in there for me. And the 'Yes and yes' is a pretty easy cop out from my question.. got it.

tell us more about your experience with your old Sony TDA for example, I would be interested, some of the early DACs in CD players were great and some were not, you bring up a good example of a different sounding DAC, I've built outboard DACs using R2R sections from older CD players and have had some good results but not always, and yes EE and CE and yes signal processing in various settings including audio for electric guitar digital effects, I respect everyone's opinions and would like to know if you personally hear a difference with music you know well going from the older Sony to using a newer DS DAC for example, there is a strong market for some of the older R2R CD players

 

[DSJR]

I didn't mean to criticize any of our measurements or standards, just that we have standard measurements and probably more measurements will be developed, the state of the art is a moving target, in my own testing some, not all, DACs are better at reproducing voices and there is less of a sense that it is a recording, when people complain about a digital sheen or grit or brittle sound they should try those same recordings with some different DACs and see if they notice a difference, in video we talk about pixel sharpening and there are audio filters that do just that and with varying results for different styles of music, that's just an example

I hope you don't mind me butting in and 'chatting' about such stuff. The fascinating thing for me was with a modern better sorted stereo system (remember, we had Naim and Linn here back in the mid 80's), even the early CD players don't seem to sound as horrid as I remember them being. I do have a Deep Purple CD (Burn, original issue) which sounds as if the RIAA encoding was put onto the master tape used to aid record cutting but not corrected when transferred to digital making the music all but scream at you relentlessly and this for a heavier rock band where you need to hear and feel the bass as well... The remaster appears to have sorted this but discs like this are an exception. Known 'difficult' discs (Vivaldi Four Seasons, Academy of Ancient Music - Pinnock from the early 80's) used to sound absolutely horrible and 'screechy' back then (the more squeaky sounding 'original' instruments didn't help) but today, this recording sounds as it is, a nice production of original instruments in a good acoustic - in my opinion. I do appreciate you're not speaking of this kind of 'tone' but there are modern speakers that 'do' voices very well indeed with no added bloom or 'grit' and also amps with no hard or harshness in their distortion and (finally getting to the point) to me at any rate, the differences in dac tech and so on becomes fully resolved, inferior ones with higher distortion or CD players like my Micro Seiki (based on the Philips 960/Marantz CD94 derived 'universal' chassis but heavily tweaked and with additional transformer coupled balanced outs which came with their own adaptor cables for RCA use (two 'hot' pins 2 and 3 wired for signal hot and return and screen pin 1 disconnected at the RCA end) just sounding 'enhanced' or even slightly coloured, if you see what I'm getting at. I suspect my own player used from the transformer coupled outs, has a little extra bass distortion to make the sound more 'organic' and musical and I have a sneaky feeling the hf filtering in said transformers may absorb some of the extra hf 'sh*t' that some of these TDA1541 based machines could chuck out (I have no proof of this though, so it's just a hunch).

It has been said over the years that a higher noise floor (within reason) can help the perceived sound at home. I didn't care for the 'No Noise' software EMI often used and maybe still do as in excess, it sounded 'wrong' to me on backgrounds heard especially on headphones - Far better to have a master tape 'presence' on analogue recordings if not overdone. I never found UK FM radio at its best to be an issue I have to say (it's crap these days though apart from the odd BBC live recording imo) and this is basically 13 bit digital I gather and back in the days of the very first CD players from Philips, a cassette recording from the CD (metal tape on a top Nakamichi and set up for the tape before recording) sounded 'nicer' to me than the CD original.

All this 'thinking out loud' makes me wonder if the appeal of some multi-bit dacs is actually 'nice added noise' and maybe a gently rolled off top?

Forgive me if I'm thread drifting - I'm good at that.

 

[PuX]

Because not everyone agrees that only measurements matter. There's a market. Schiit offers both 'well measuring' & 'voiced' products to their customers, and they are very transparent about it. It's called choice.

is this a troll post?

you are on audio science review. measurements reflect the real sound of the product. if it's not comprehensive enough then suggest some new test that would be good enough for you. but don't bring emotion to scoring a product, it's not objective, not scientific in any way.

 

[PuX]

USB-C is absolutely more future proof, in 5 years time there probably won't be much else of older connectors still present on devices.

maybe, but probably not. keep in mind that 5 years ago type-c appeared as the only connector on a macbook and everyone said the same thing,

but in reality what connector do you have on most of your new devices? android phones - sure, but there's no consistency/interchangeability and most of the time it is type-a on the other end. ps5 controller - same. laptops - very rarely it's type-c, most use some barrel plug. DACs with type-c are rare. mouse, keyboard, printer, scanner, external ssd all come with type-a in general.
so how is it a connector of the future?
it's a bit like firewire 400 and 800, e-sata, first gen of thunderbolt (shaped as mini-displayport) etc. yet another attempt to unite a myriad of connectors, now we have myriad+1.

 

[threni]

Right, but he's saying his ears do that. His ears are incorrect, so it works. That's the point eh.

Well, brain maybe. But I don't want to have an inferior solution and have to sort of subconsciously pretend it's ok; I'd rather spend next to nothing on a measurably pretty-much-perfect device and be done with it.

 

[don'ttrustauthority]

Also, I read elsewhere that Schiit are forced to produced stuff which measures well because of sites like ASR! Seriously - someone actually typed that!

Did Schiit add to their product line to satisfy the demand they did not realize was there (and was created in part by the review)? Yes.

Did Schiit add to their line, for $100, an easier to design and manufacture, and cheaper to boot, version of the Magni 3? Yes.

Was it ready the day after the article dropped? No.

Are op amps so cheap and easy to manufacture with stupid low distortion numbers that they also came out with a $200 balanced version (that easily could've been a $400 Jot 2 and made everyone in this forum happier since it measures better) just to give the finger to all of us? Yes.

And I don't know if they thanked Amir personally, but they do acknowledge the role this community played in these changes.

Personally I would not be suprised if this entire episode wasn't some kind of get rich quick marketing scheme dreamed up by Amir and Jason to sell cheap, low distortion opamps by ton but ...

FOILED BY COVID. Really kind of unfair during their best year Trump Tariffs, Covid disruptions and poor logist... er, suffice it to say the year was good Schiit, but not the great Schiit it could've been. I'm not in agreement with the entire made in America thing, but I do support it. DCA headphones ftw!

 

[PeteL]

is this a troll post?

you are on audio science review. measurements reflect the real sound of the product. if it's not comprehensive enough then suggest some new test that would be good enough for you. but don't bring emotion to scoring a product, it's not objective, not scientific in any way.

Well, sorry, but the engineers at Schiit, so many other electronic designers know a bit or two about science, and visibly don’t think that only measurments matter, or the ones performed here. If they did they wouldn’t realease these. Audio science, if there is such a thing is certainly not a field of interest that can all be enclosed in a apx555. it is much more vast than that, fortunately. I don’t see emotions, nor ”scoring” a product, nor throlling there, just a fact, Schitt offer a choice of highly “scoring” in term of specs, product, and other types that don’t perform as well on ASR test bench, that arguably, may sound different and also may “score” better on other metrics not necessarily tested here (and it does) limiting the audio science to this set of benchmarks is quite condescending to the engineers that designs these products.

 

[Jimbob54]

Schitt offer a choice of highly “scoring” in term of specs, product, and other types that don’t perform as well on ASR test bench, that arguably, may sound different and also may “score” better on other metrics not necessarily tested here (and it does) limiting the audio science to this set of benchmarks is quite condescending to the engineers that designs these products.

Genuinely interested. Which metrics and compared to what? Got a link?

 

[PeteL]

Genuinely interested. Which metrics and compared to what? Got a link?

Metrics, don’t mean measurments, for example, multibit dacs use less noise shaping calculations that’s a “metric” less interpolatin in their algorythm, that is a metric. I am not saying it offers a benefit in term of reproduction, if that’s what you tought I meant, but yes, some of the calculations involved in the conversion are more mathematically exact.

 

[Jimbob54]

Metrics, don’t mean measurments, for example, multibit dacs use less noise shaping calculations that’s a “metric” less interpolatin in their algorythm, that is a metric. I am not saying it offers a benefit in term of reproduction, if that’s what you tought I meant, but yes, some of the calculations involved in the conversion are more mathematically exact.

I didnt say measurements . I am asking which other metrics might a review site employ to compare products? Why, as a consumer, should I be worried about how the device does what it does? Should I not be more interested in what comes out of it and how I interact with it?

 

[PeteL]

I didnt say measurements . I am asking which other metrics might a review site employ to compare products? Why, as a consumer, should I be worried about how the device does what it does? Should I not be more interested in what comes out of it and how I interact with it?

I did not say that neither, but for a start, in this particular one, we have one metric, Sinad at 1kHz. I can assume that the limitation would be reflected to other measurments, but what I don't like is that it suggest, along with the SINAD chart, that it is enough to say it's a product that perform poorly. Now sure, I would like to see thd +N over the entire band, I would like to see impulse response to visualise pre ringing phenomenons,etc. Does it means that it would change the score, not necessarily, but assuming that this dashboard tells us 100% of how a product behave, for 100% of the content, in 100% of the listening conditions, would not be scientifically accurate neither. i was reading an interview with Bruno Putzeys, saying that he used measurments, a lot, but vert targetted measurments, specifically looking for very precise issues, that are not necessarilly the common ones that spits out a spec. He was also saying that generally, enthusiasth are good listeners, that they don't have the knowledge to pinpoint what is truly wrong or limitative, but that he allow great value to subjective assessments of listeners, that they are more often than not right, but that's his job to find the right measurment to find why, and the right engineering solution. When I have more time I'll try to find it. I tend to agree with this.