My "subjective" experience with AKM vs ESS DAC chip "sound signature".

[donjoe0]

Any DAC comparisons make sense when they are only DACs or streamers and connected to our favorite one headphone amplifier with our favorite one headphones.

Well, not "favorite" but the most neutral amp you have access to. Like maybe a THX or something.

I recently got direct confirmation (that I was hoping not to get) that ESS-based DAC-amps can sound very different, i.e. the KA17 sounds veiled to me in every configuration, if I compare it to the also ESS-based Hiby FC3, with the latter being bright and crisp just like I want it - all the sound is aimed directly at me with no obstacles, like I have a front-row seat, not like the sound is lost in a large room somewhere. I consider "harsh" ESS sound to be correct and neutral, and all these other warmsauce DACs like Burr-Brown or R2R to be deviating from maximum fidelity by cutting out treble they shouldn't be cutting.

 

[theveterans]

IMHO, harsh means the equipment failed to resolve the signal correctly. And no resolving ability has nothing to do with tonality. You can have a brightly tilted system and not get harshness whatsoever if the system resolved the signal accurately. My 2cents

 

[KSB1978]

Resolving ability doesn't have anything to do with tonality, but humans do hear certain frequencies better than others. A bump in these frequencies we hear better can give an 'impression' it is more resolving.

 

[jjazzy]

Well, not "favorite" but the most neutral amp you have access to. Like maybe a THX or something.

I recently got direct confirmation (that I was hoping not to get) that ESS-based DAC-amps can sound very different, i.e. the KA17 sounds veiled to me in every configuration, if I compare it to the also ESS-based Hiby FC3, with the latter being bright and crisp just like I want it - all the sound is aimed directly at me with no obstacles, like I have a front-row seat, not like the sound is lost in a large room somewhere. I consider "harsh" ESS sound to be correct and neutral, and all these other warmsauce DACs like Burr-Brown or R2R to be deviating from maximum fidelity by cutting out treble they shouldn't be cutting.

Ess sabre out my experience is a very flat presentation, AKM does a way better job. But to be honest most single chip dacs give a flat presentation. I am more of a fan decent fpga dac’s, several high end brands do their own thing. Fpga just have more processor power.

And out of experience, a lot of people associate high treble as details, but what they really hear is distortion, which fools as more details.

The best have a deep dark background, total silence in the background, 3d and 2d depths. And I still have to hear single dacs do that in this matter.

I heared the best single dac chips, from ess to akm, cirrus. But most m flat presentation, no depth.

And r2r dacs are like tube dacs, smooth and nice sounding, but it is at the cost of some treble details yes

 

[DaveStarWalker]

Well, my experience about ESS vs AKM (portable devices and specific products : I don't make any universal advice here) :

https://www.head-fi.org/threads/qls-qa361-impression-thread.866536/page-23#post-18390763

https://www.head-fi.org/threads/qls-qa361-impression-thread.866536/page-23#post-18399193

And :

https://www.head-fi.org/threads/ran...s-that-you-heard.936707/page-62#post-18399326

Aune = ESS, QLS = AKM.

Personal conclusion :

More than the chipsets used, the whole scheme (implementation) is far more important for the final results. This is the case here...

 

[jjazzy]

Well, my experience about ESS vs AKM (portable devices and specific products : I don't make any universal advice here) :

https://www.head-fi.org/threads/qls-qa361-impression-thread.866536/page-23#post-18390763

https://www.head-fi.org/threads/qls-qa361-impression-thread.866536/page-23#post-18399193

And :

https://www.head-fi.org/threads/ran...s-that-you-heard.936707/page-62#post-18399326

Aune = ESS, QLS = AKM.

Personal conclusion :

More than the chipsets used, the whole scheme (implementation) is far more important for the final results. This is the case here...

Implementation is always a part of the equasion. But, all chips have there characteristics. For example I recognize a ess everytime, I heared good inplementations, but still don’t like ‘m. They have some glare in the high midrange and treble. Which I only hear with ess, not with other dac chips or fpga implementations. I even like cirrus more than ess

 

[DaveStarWalker]

Implementation is always a part of the equasion. But, all chips have there characteristics. For example I recognize a ess everytime, I heared good inplementations, but still don’t like ‘m. They have some glare in the high midrange and treble. Which I only hear with ess, not with other dac chips or fpga implementations. I even like cirrus more than ess

Well, I don't know how to say, really

For instance, the 2 Aune aforementioned are ESS based devices, but at the same time are super musical, "liquid", a tad "warm", by a hair... Linear Fast filter (very important precision).

So the best matching are different than the QLS...

... Which is more "transparent" and "technical", a tad "cold", by a hair... but musical, no question about that... (Sharp* and bright* settings : same).

Direct comparison, all things being equal...

As my home dac is R2R based (no PLL on : very important precision too) and super transparent...

So really, this isn't a frozen game from my point of view.

This is also why I only think in terms of "the system" (as a chain), taken as a whole.

* = Linear Fast, ESS
* = No apodization, ESS

 

[donjoe0]

More than the chipsets used, the whole scheme (implementation) is far more important for the final results.

Yep, and "implementation" in most cases I think is about sound coloration of the amp section (not the DAC chip!), or if the manufacturer preselected a bad slow&early rolloff digital filter in "analogophile" DAC chips that offer such filters (like Cirrus Logic), where the response is several dB down from neutral before it even reaches 20 kHz, so youg people will easily hear it as being dull or veiled vs. a correctly set up DAC. Otherwise it has to be a super cheap and low quality implementation for the pure DAC to sound non-neutral.

Example: KA 17 is intentionally tuned to sound warmer in the lower treble and deviates from absolute fidelity/neutrality. This can be shown quite easily vs. the really neutral HiBy FC3 even with measurements through a normal sound card. Both are ESS but only the HiBy is neutral and truly high-fidelity, the difference is easy to hear in A/B comparison: https://forum.hifiguides.com/t/fiio-ka-17-opinions-and-usage/43721/5

And of course the other thing that greatly confuses listeners is the mismatch between their headphones' response and their personal HRTF:
- they listen to bright and harsh headphones with a warm DAC-amp (aka. warm amp or rolled-off DAC or both) => sound seems hi-fi, close to neutral, they judge the DAC-amp as being very good, "musical" etc.
- they listen to the same bright headphones with a proper neutral DAC => sound is harsh, disturbing, they blame the DAC-amp for being "too bright" when it was really the headphones mismatched for their personal HRTF.
Nobody should judge DAC-amps before they've applied EQ to match their headphones to their personal HRTF. Alternately, just measure the DAC-amp with test signals because it's easy to specify what a hi-fi DAC-amp's frequency response should look like: a flat horizontal line aka. neutral aka. transparent.

 

[Brain Damage]

But even with all of this, the following things still need to be taken into consideration:
- dynamic range
- transient speed & decay
- harmonics v SNR.

On the latter, so many people look for the lowest THD values and believe that this can help differentiate between a 'good' DAC/DAC amp and a 'bad' one. There's far more to this. For example, warm, thick, tube sounding DACs often have lower even-order harmonic distortion which can give a more natural, organic and analogue presentation v DACs that have an odd-order distortion (which can present more of a cold, bright digital glare sound).

I believe all DAC chips (and solely the chips alone) all do the same job by converting a digital signal/stream to an analogue wave. However, how they are implemented with FPGAs, filters, oscillators, harmonic distortion; not to mention how all of this works in conjunction with the whole amplification chain is what ultimately leads to a differential in sound signature.

 

[DaveStarWalker]

So agreed with these two last messages

 

[DaveStarWalker]

Latest war stories... :

https://www.head-fi.org/threads/qls-qa361-impression-thread.866536/page-23#post-18435285

 

[theveterans]

Latest war stories... :

https://www.head-fi.org/threads/qls-qa361-impression-thread.866536/page-23#post-18435285

On my SP2000, I tried every filter available and found that the apodizing filter (in between minimum phase and linear phase with slow roll-off) sounds best. I get a bit of incisiveness in the presentation while not sharpening the edges that fast roll off filter does. Also, the soundstage expands wider with this filter as well

 

[DaveStarWalker]

On my SP2000, I tried every filter available and found that the apodizing filter (in between minimum phase and linear phase with slow roll-off) sounds best. I get a bit of incisiveness in the presentation while not sharpening the edges that fast roll off filter does. Also, the soundstage expands wider with this filter as well

Same thing, but with my Aune m1p... (ESS dac based)
And with my m1s (same but another and older ESS chipset than the previous one), this is the "hybrid filter" which sound the best, I guess...

I completely changed my mind after becoming interested in the SD cards used...

 

[Jmop]

I use my HiBy RS2 (R2R) line out to various amps to play with sound coloration but kinda wanna try a more resolving source that’s still smooth. Thinking I might get that using original Shanling M6 Pro (AKM, non 21) for a smooth yet resolving and tonally linear sound foundation? Not a dedicated line out but hopefully works as intended.

For what it's worth, after I spent some solid time with my M6 Pro, I for once didn't have the craving to try something else. The tuning really fits my desires. Although I did recently purchase the Hiby RS6, essentially based on the architecture and design. I've heard enough ESS dac implementations to know AKM suits me better. On that note, the actual tuning and components used seem to play a larger role than the chips themselves.

How do like the M6 Pro’s dac vs the RS6? I expect the RS6 to be better but just curious how a high quality AKM does against R2R in general tonality.

 

[llysender]

Just curious do you folks feel that with the latest AKM and ESS chips the decay envelope for both companies have swapped personalities with 4499EX having sharp and precise sound while 3039pro has a smooth and round sound.