New DAC Chipset from AKM

[TurtlePaul]

I suspect that we will see this filterless DAC chip used in some implementations which have discrete filters. Given how mundane the inside of a perfectly measuring DAC looks, I can see DAC makers adding big inductor loops and audio grade capacitors as a discrete filter stage. then running that through a BUF or op-amp for the output. It will make the inside of the DAC look much more impressive, even if the filter sacrifices steepness.

 

[despoiler]

I think the latter. The only issue being, why would I care if it was either/or (ignorance, or a engineering trade-off). They put out their latest and most complex chip design, yet still suffer the identical "tradeoff". I just fail to understand why ESS can do it, but they aren't able.

As for the passband ending at 19.88, I don't see much of an issue there (sure it's not "ideal"), but for the nitpicky realm of discussion here, I'd rather that, than it going out to 24kHz just for the sake of equipment not having to bother trying to output such signal (which is basically the only reason I care for filters in the first place, not so much for the sake of audibility at all really, seeing as how the only difference I hear is when I use a non-oversampling filter for example, but the rest? Yeah they're basically identical to me).

ESS can do it if they don't actually output to 20k. My point is they are both making trade-offs that aren't audible. I don't think anyone's gear is negatively effected either. AKM has been using the same filters for many years. If there was a problem someone would have seen it by now. It would be interesting if Amir did an AMA with the engineers about why they designed their solutions they way they did.

 

[Tks]

ESS can do it if they don't actually output to 20k. My point is they are both making trade-offs that aren't audible. I don't think anyone's gear is negatively effected either. AKM has been using the same filters for many years. If there was a problem someone would have seen it by now. It would be interesting if Amir did an AMA with the engineers about why they designed their solutions they way they did.

You mean AKM could do it?

Anyway, if audibility were the only concern in audio devices with the level of interest people take here, I think we'd be done basically with $9 iPhone doggles and call it a day for mobile, while for desktop, using on-board audio would suffice in the latest motherboards, and for home, well... people put up with sub par stuff as a majority anyway (the performance of power amps and speakers as we're seeing isn't all that amazing), so we could just resort to powered bluetooth enabled speakers and be done with it for audibility's sake considering the opinions of some people when listening to music.

As for the notion of "if there was a problem people would have seen it by now". I highly doubt that. The only way the problem can be seen is by measurements to begin with. To say having amps and speakers trying to reproduce 25kHz or higher frequencies isn't a problem, I just find that a bit weird, seeing as how it's undue stress on the hardware for no appreciable reason.

But maybe I'm wrong. Perhaps you have some more information on this matter. At what frequency cut-off do you think a "problem" would be constituted? Like if my devices are allowing 50kHz or higher frequencies(say by the use of non over sampling filters or something of that kind), what kind of "problem" in your view do you think will manifest, and if none, then what is the point of filters to begin with?

I guess I'm just having problems conceptualizing the notion of a "problem" in your definition. With mine, it's pretty simple concerning this topic. Anything less than ideal is basically a problem to varying degrees, and if competition offers a solution to that problem compared to another, then it is a problem worth rectifying unless price differences are THAT staggering as a trade-off.

 

[A Surfer]

I am not brand loyal per say to any chipset manufacturer. Over the years I have had DACs that featured Burr Brown, Wolfson, Cirrus, ESS and AKM chips. Currently I have two products that feature AKM chips those being the 4497 and 4493 and both sound very good to me. I agree, it would be nice if the design teams behind the chips provided an explanation publicly for why their filters demonstrate the effect mentioned here. Even if the effect isn't audible some understanding of the rationale would be nice to hear.

 

[bennetng]

I think the latter. The only issue being, why would I care if it was either/or (ignorance, or a engineering trade-off). They put out their latest and most complex chip design, yet still suffer the identical "tradeoff". I just fail to understand why ESS can do it, but they aren't able.

As for the passband ending at 19.88, I don't see much of an issue there (sure it's not "ideal"), but for the nitpicky realm of discussion here, I'd rather that, than it going out to 24kHz just for the sake of equipment not having to bother trying to output such signal (which is basically the only reason I care for filters in the first place, not so much for the sake of audibility at all really, seeing as how the only difference I hear is when I use a non-oversampling filter for example, but the rest? Yeah they're basically identical to me).

ESS also have trade off:

I guess you would choose filter 7

...and filter 6, right?

They are the same one, it is called Brick Wall (Linear) in the datasheets. The datasheet plots are in digital domain, while Amir's plots involve analog conversion so the DX7s plots are limited by noise floor. On the other hand the Matrix plot does not have enough range to show the noise floor, but you can still see that filter 6 has poorer attenuation than for example, filter 1. So one may wonder whether 20dB more attenuation or 2kHz earlier attenuation is more important, if you ask me I would say I don't care.

If you compare it with Fast Roll-Off (Linear), which is probably a half-band filter, you can see that it has a deeper stopband attenuation.

On the other hand 100dB is AKM 449x's highest possible attenuation, if they are going to provide the filter type you like, then attenuation would likely be less than 100dB.
https://www.audiosciencereview.com/...d90-balanced-usb-dac-review.10519/post-291483

So, does it mean AKM is in general, inferior to ESS? According to Archimago, his Oppo with 9038 Pro has some strange behaviour in DSD, the higher the DSD rate, the higher the noise floor.
https://www.audiosciencereview.com/...rasonic-rf-output-important.10600/post-294860

Miska's comment:
https://archimago.blogspot.com/2019...howComment=1552471342841#c5287923425357358865

Actually, if you read the DSD256 plot carefully, the noise floor is still higher than PCM by more than 10dB.

 

[Tks]

@bennetng Absolutely, filter 6/7 is what I would prefer.

 

[bennetng]

Follow up. Archimago's DSD tests on DX3 Pro with AK4493. Noise floor remains the same in PCM and DSD below 20kHz.
https://archimago.blogspot.com/2020/06/measurements-topping-dx3-pro-v2-ldac.html

On UDP-205 with ES9038 Pro:
https://archimago.blogspot.com/2019/01/measurements-oppo-udp-205-dsd-playback.html

Miska liked to show improvements in some specific products with DSD upsampling but Archimago also showed you how DSD upsampling can degrade things.

For the new AKM chipsets, if they can further improve measurements so that Miska can no longer show visible advantages in his 2MHz range measurements, and if they can improve jitter with SPDIF input like what ESS chips can do, and if the price is competitive, then it can be an attractive offer for some people.

 

[Bach]

I wonder when we will get quantum DACs.

How many qubits will be "enough"?

You need exactly pi qubits. Just a little difference and the dac maintains no longer coherence and falls back tot about 3 bit resolution at 10kHz.

 

[bigguyca]

If you asked eng. from AKG they would likely make a brief around: it is already inaudible and the current DACs are perfect for what you want to do but to justify our salaries we have improve stuff a tiny bit such that marketing can talk about something. Then it goes to marketing department that remove all of the above and rewrite it. That's life and that's reasonable from a business PoV.

AKG makes microphones:

https://www.akg.com/

Why would an AKG engineer be asked about DAC IC's? Microphones and DAC IC's are two very different things. A Google search will provide links to basic information. Before you criticize engineers at a company you should at least develop a basic understanding of the products involved.

 

[Jimbob54]

AKG makes microphones:

https://www.akg.com/

Why would an AKG engineer be asked about DAC IC's? Microphones and DAC IC's are two very different things. A Google search will provide links to basic information. Before you criticize engineers at a company you should at least develop a basic understanding of the products involved.

Means AKM

 

[bigguyca]

Follow up. Archimago's DSD tests on DX3 Pro with AK4493. Noise floor remains the same in PCM and DSD below 20kHz.
https://archimago.blogspot.com/2020/06/measurements-topping-dx3-pro-v2-ldac.html

The major change from the AK4490 to the AK4493 is the ability of the AK4493 to drive lower impedance loads. The capability to drive lower impedance loads in turn allows use of much lower value resistors in the signal path in the analog filter circuitry after the DAC IC. Low value resistors will have much less resistor (Johnson) noise. The overall DAC circuit, that is, the DAC IC and the following filter circuitry with lower value resistors, should measure better.

You referenced the link with the comparative test, which means you reviewed the overall test and approved it as complete and accurate.

Does the product that uses the AK4493 also use lower value resistors? In addition, does the revised product use TI opamps with good enough performance to reveal improved performance, if any?


- Omitted to save space -

.

 

[bennetng]

These are AKM vs ESS comparisons, not AK4490 vs AK4493. For other details about the Oppo and Topping mentioned above, please contact Archimago himself.

 

[bigguyca]

Means AKM

You can't keep the companies straight. Are you still critical of the engineers?

 

[Jimbob54]

You can't keep the companies straight. Are you still critical of the engineers?

You're talking to the wrong guy my friend. I'm explaining the other guy obviously mistyped it as akg, not akm. Learn to read a thread and back down a notch

 

[bigguyca]

You're talking to the wrong guy my friend. I'm explaining the other guy obviously mistyped it as akg, not akm. Learn to read a thread and back down a notch

Why do you try correct the posts of others? Do you know for certain that the poster is incapable responding for himself? The poster was critical of engineers he didn't even know. Evidently blame without any knowledge is OK with you.

 

[Jimbob54]

Why do you try correct the posts of others? Do you know for certain that the poster is incapable responding for himself? The poster was critical of engineers he didn't even know. Evidently blame without any knowledge is OK with you.

Are you always so obnoxious? I like you!

 

[Thomas savage]

responding for himself

'Gender assumption warning' ..

 

[Bach]

'Gender assumption warning' ..

And how to address to someone who uses Pluralis majestatis?

 

[RayDunzl]

Are you always so obnoxious? I like you!

You made me think of my idol, Gunnery Sgt. Hartman.

 

[Jimbob54]

You made me think of my idol, Gunnery Sgt. Hartman.

A legend