AKM top chip Hump?

[JJFUSA]

I think it's quite entertaining for every audiophile to compare two highest level ESS and AKM chips, especially when implemented in the same DAC model:

SMSL D400 ES - ES9039:

Measurements of SMSL D400ES ESS9039Pro DAC - L7Audiolab

台式设备 随着一个个项目评审会的结束，我忙乱的日子也开始看到了曙光……前两天双木的晶晶问我，说之前测了D400EX之后怎么没测同时送到的 SMSL D400ES 。我也只能很无奈的说了了期末+之前病休导致的忙乱。无论如何，接下来应该会逐渐拉上测量的频率……

www.l7audiolab.com

SMSL D400 EX - AK4499:

Measurements of SMSL D400EX Dual 4499EX DAC - L7Audiolab

台式设备 前阵子测了本站的第一个AKM4499EX解码，在3K多的价位还是近乎首发的情况下，表现中规中矩。当时我就在想，三小强在干嘛呢……这芯片发布那么久了还没拿出机器……结果我很快就收到了来自SMSL的4499EX机器！这台还是双4499EX！

www.l7audiolab.com

Apex of the pure sound performance, I would say.

But it's quite interesting to observe a famous hump on the AKM chip now:

https://cdn.l7audiolab.com/wp-content/uploads/2022/11/1668347120-SMPTE-Ratio.jpg

I'm not a sound expert by all means, yet after the ESS hump fuss (there is no more hump on new ESS chips like ES9038Q2M for example as far as I know) it's kinda weird to see a top AKM chip behave like this, or am I missing something here?

 

[Hemi-Demon]

It's a engineering error. The Gustard A26 does NOT have this issue.

Measurements of Gustard A26 Network DAC - L7Audiolab

台式设备 犹记得差不多两年前，和老黄聊（他当时正在做的）网播的时候，我说你只要把这玩意做好了，到时候集成到解码里会是歌诗德的一大特色。这次我收到 Gustard A26 这台解码一看，果然老黄把（之前在R26上用过的）网播模块用的到处都是了啊！

www.l7audiolab.com

SMSL and Topping will fix the issue at some point. Both companies have more pressing quality control issues to fix in my opinion.

 

[JJFUSA]

It's a engineering error. The Gustard A26 does NOT have this issue.

Measurements of Gustard A26 Network DAC - L7Audiolab

台式设备 犹记得差不多两年前，和老黄聊（他当时正在做的）网播的时候，我说你只要把这玩意做好了，到时候集成到解码里会是歌诗德的一大特色。这次我收到 Gustard A26 这台解码一看，果然老黄把（之前在R26上用过的）网播模块用的到处都是了啊！

www.l7audiolab.com

SMSL and Topping will fix the issue at some point. Both companies have more pressing quality control issues to fix in my opinion.

Strange that companies let that happened, especially in the top tier models.

Comparing ESS and AKM:

Single ES9039:

https://cdn.l7audiolab.com/wp-content/uploads/2022/11/1669821872-SMPTE-Ratio.jpg

Dual AK4499:

https://cdn.l7audiolab.com/wp-content/uploads/2022/12/1670626588-SMPTE-Ratio.jpg

Seems ESS still performs much better.

What the point to use AKM then?

 

[unpluggged]

there is no more hump on new ESS chips like ES9038Q2M

On the contrary, the Q2M was actually infamous for the hump. See, for example, Khadas Tone Board measurements.

 

[boXem]

Unless the hump comes from the I/V circuit and not from the DAC...

 

[JJFUSA]

Unless the hump comes from the I/V circuit and not from the DAC...

SMSL use identical components in both versions of D400.

D400 ES (ES9039):

D400 ES

Foshan ShuangMuSanLin electronics Co., LTD. founded in 2009 which located in Shenzhen, China. And our branch company VMV Technology(HK) Company Limited registered in HK in 2013. Our company is specialized in the audio DAC, stereo headphone amplifier, power amplifier, and with own R&D...

www.smsl-audio.com

D400 EX (AK4499):

D400 EX

Foshan ShuangMuSanLin electronics Co., LTD. founded in 2009 which located in Shenzhen, China. And our branch company VMV Technology(HK) Company Limited registered in HK in 2013. Our company is specialized in the audio DAC, stereo headphone amplifier, power amplifier, and with own R&D...

www.smsl-audio.com

I guess it means that only chip itself (AK4499) is responsible for the hump.

 

[boXem]

SMSL use identical components in both versions of D400.

D400 ES (ES9039):

D400 ES

Foshan ShuangMuSanLin electronics Co., LTD. founded in 2009 which located in Shenzhen, China. And our branch company VMV Technology(HK) Company Limited registered in HK in 2013. Our company is specialized in the audio DAC, stereo headphone amplifier, power amplifier, and with own R&D...

www.smsl-audio.com

D400 EX (AK4499):

D400 EX

Foshan ShuangMuSanLin electronics Co., LTD. founded in 2009 which located in Shenzhen, China. And our branch company VMV Technology(HK) Company Limited registered in HK in 2013. Our company is specialized in the audio DAC, stereo headphone amplifier, power amplifier, and with own R&D...

www.smsl-audio.com

I guess it means that only chip itself (AK4499) is responsible for the hump.

It is not the components that make the performance, it is how you use them. We do not know the I/V topology used, the different amount of them makes me wonder if the topology is the same.

 

[BR52]

Actually, the Gustard appears to hide the hump. Look at the graphs of the Gustard (top) and D400EX . I can't overlay the lines but it can be seen. Look at the x vs y values at each axis intersect. Almost like a parlor trick .


View attachment 261281
View attachment 261282

You are right. It’s easy to do. Ad dither ( broad band noise) to the signal. Smsl did it with the 200 as well. See @armirm last review. This can be done in the configuration of the chip.

 

[BR52]

It shows how important it is not only to see the sinad numbers all other measurements are important as well. Here the thd/ noise measurements.

 

[BR52]

Here a more extreme example:

 

[mocenigo]

Seems ESS still performs much better.

What the point to use AKM then?

I would say marginally better, actually.

 

[djgodlike]

Someone from SMSL have send me a new firmware to fix the IMD problem of our D400EX. I have no equipment to measure but it seems to sound fine and a little less harsch. Maybe someone here can do another measurement? Flashing worked fine (different method than usb firmware). Use at own risk!

 

[Snoopy]

Is it possible to flash the firmware without a windows computer?

 

[djgodlike]

Is it possible to flash the firmware without a windows computer?

Nope. But (to me) it was worth finding an old windows pc for this.

 

[Snoopy]

Nope. But (to me) it was worth finding an old windows pc for this.

How about parallels on Mac with a Windows VM?

 

[BR52]

Someone from SMSL have send me a new firmware to fix the IMD problem of our D400EX. I have no equipment to measure but it seems to sound fine and a little less harsch. Maybe someone here can do another measurement? Flashing worked fine (different method than usb firmware). Use at own risk!

I'm not sure a fix via software will work. I guess via SW it can be only hidden in higher broadband noise (dither). And this will make measurement charts at a not so critical look nicer. See some posts here before.
I guess it's a I/V stage problem and can only be solved via HW.
Would be great to have some measurements after the SW change, to see what's really happen.

 

[Snoopy]

Looks like with these DACs it's better to wait a couple of months till they go through a few software Updates and maybe hardware revisions.

 

[BR52]

Looks like with these DACs it's better to wait a couple of months till they go through a few software Updates and maybe hardware revisions.

Normally they are coming with a new model.
Model maintenance………

 

[Dawsoo]

AKM4499EX really has a IMD hump problem in all implementations here you can see a picture of E70V vs E70. No noise masking really happens with the ES9028pro.

 

[Sokel]

ES9038Q2M in Khadas tone1 is definitely responsible for the hump,output stage has little to do as seen here:

PCM vs DSD

Did a little comparison and I was wondering if any friend with better gear here has tested this. Results you're about to see is probably narrowed to the specific DAC but it's interesting to see how it handles things differently. I will not try to decipher the results other that DSD seems much...

www.audiosciencereview.com

Unless there's a different output stage for DSD (which I don't think so) .
IMD vs level seems to be the most important measurement in DACs,along with SINAD vs level when used as an attenuator too.
Mind that music levels are about at -15db normalized,so about there is a nice point to look when played at full volume or subtract accordingly when used as an attenuator too.