AK4395 vs. AK4396 listening comparisons

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The AK4396 has been around for a couple of years now and I have had a lot of people asking me about it as a possible upgrade to the AK4393 that comes stock in the Behringer 2496 audio gear but I have just now been able to get my US distributor to supply them. I installed an AK4396 dac chip in a direct out modified DEQ in order to compare it with the AK4395 that I have been recommending. The 4396 drops in to the board space vacated by the AK4393 and works fine on 3.3v with no other mods whereas the 4395 needs an extra regulator to provide 5v to pin 2, assuming that you don’t want to share the analog 5v with the digital pin. It’s interesting that the two chips do sound quite different. Although I usually have little trouble picking one component I like best from listening trials, it has been very difficult for me to choose a clear winner this time. These chips both offer stratospheric performance. It’s funny that I actually started to feel some pressure from my indecision. I also swapped the boards into the opposite chassis to make sure that I was minimizing the variables and tried both 4k and 20k stepped attenuators even though the 20k attenuators don’t work as well with my 22k amps. The 4396 has a more powerful sound even working into my 4k stepped attenuators despite it’s lower stated power consumption. It throws it’s soundstage closer to the listener, more toward the front line of the speakers and actually plays about 1db louder depending on the program material. The 4395’s bass was heard to extend much further than the stock 4393’s, along with a big improvement in resolving ability, and the 4396 has just as much extension, with a higher level, up into the mid bass. This makes the 4395 sound a bit lean in comparison. On the other hand, the 4395 throws it’s sound stage much deeper, starting just behind the plane of the speakers and going back beyond the front wall of the room. The 4396’s stage is pleasantly a bit taller. The 96 lights the stage more brightly, making each instrument stand apart from the others but lacks the ultimate resolution of the 95’s ability to follow the sounds right to the fine end. Some tracks favor the 4396’s closer presentation as feeling more involving and easier to follow. On other cuts I preferred the 4395’s extra ability to resolve reverberation tails and ambient information, making the 96 feel like it is leaving something behind. So the trade offs went back and forth causing one of those listening binges where you just keep pulling one cd after another off of the shelf, and can’t wait to get home the next day to do it all over again. At this point, with my current associated equipment, I will have to choose the AK4395 for it’s extra resolving ability even though I was hoping the 4396’s more focused and powerful sound would win as it would be easier to install. With different equipment I can see where this might go the other way so I will check back on the 4396 as things in my system change.
 
Scott, thanks for the comparison. I have one question about the test, though. Did you use your external regulator for the AK4396 as well? It seems to me from the descrition that you used the onboard regulator on the '96 while using an external regulator on the '95. Considering the '96 can run on 5V as well, it would seem to eliminate one more variable. Do you think that had any effect on the results? That is, of course, assuming you only used the external regulator on the '95.
 
Additional regulator for 4396 makes little change

I did consider the additional regulator as a variable so decided to revisit the comparison and was able to lift pin 2 of the 4396 to add a 5v digital supply to make the circuit identical to that which the 4395 was running . Without having another 4396 equipped unit to compare, it's hard to say whether there were any slight differences from the change. The stock power supply implementations in the Behringer 2496s are actually pretty good. The comparative attributes of either chip remained about the same. Another epic listening session though! XRCD Sampler 2, Mapleshade's Music Festival, Old VanHalen. A modified DEQ makes an amazing source for the money!
 
fine surface mount

The AK4396 is a fine pitch surface mount chip. I cut the old chip out with a razor knife and desolder the old legs/ solder the new chip in place with a fine tip iron and lots of paste flux wearing a 10x scope strapped to my eye. It is very easy to damage the board or new chip with too much heat. The flux prevents bridging. There will be enough solder left on the pads after removing the cut legs.
 
AK4395 update

Here is an update on my preference of the AKM AK4395 versus the AK4396 "Miracle Dac". I have changed amps to the even more revealing modified Sure 2X100 amps and guess what they are revealing? I now feel that the AK4395 dac chip is clearly superior across the board. It sounds deeper and taller, with a quieter, blacker quality and more reverberant information. It makes the 4396 sound a little congested at times. With the AK4395, the bass is also better defined and quicker but it is not leaner by comparison, making the instruments beautifully more fleshed out, fuller and more musical. So now I need help in coining a new name for the AK4395. If the 96 is the miracle dac, what could the 4395 moniker be?
 
Can you post response, distortion graphs for these components? I can't imagine you would upgrade electronics based only on listening.

Please explain to me, the purpose of the measurements. Thanks.

I hope that we can agree, that what we are looking for is the best sonic performance = the best sound.

Then: What will you do, if a change results in:
- Better sound, and worse measurements? - Go
- Better sound, and better measurements? - Go
- Worse sound, and worse measurements? - No Go
- Worse sound, and better measurements? - No Go

As you see, if you really are concerned about achieving the best sound performance, the measurements is a "don't care".

People often lean towards measurements, because they don't believe they can point out the best performer in a listening test. And in that case, it really do not matter what you choose, since you can't hear the difference ;)

My experience is, that many people are searching for what they believe should be the best performance based on reviews and storytelling, rather listening to it themself. And that makes these people an easy target for the growing "snake-oil business" selling cheap chinese cables, covered in nice looking outer jackets and a fantasy high prices :D
 
- Better sound, and worse measurements? - Go Actually this is "check your measurement tools, they have not been calibrated in ages". Or you have a strong imagination and hear what you want to hear not what is really playing.
- Better sound, and better measurements? - Go That's the only way to go.
- Worse sound, and worse measurements? - No Go Agree with this.
- Worse sound, and better measurements? - No Go Actually this is "check your brain". It is compensating for the bad sound that you are listening every day and the good sound sounds bad now
 
Last edited:
Now about the 4395 VS 4396...
4396 has extra the DSD capability. 4395 has better filter attenuation (110dB VS 75dB for 4396). THD and S/n are identical... but the software attenuation of 4396 shall be not used for this results.
So... depends of the implementation. Filtering can be improved with better OpAmps (filers) following the DAC's. Just swaping the DAC's, with no change in the circuit that follows can be misleading.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.