[jamato8]

The "higher" end cards get mentioned here so I am posting this here. I have the Micron card non CS, 256GB and I have it in 512GB in the CS. They are very different. The CS more in line with most cards that are decent. It does have an excellent transfer speed but the sound is not near as good as the non CS. I found the non CS to be good out of the box, then the mids got hard, pushed forward and not very enjoyable. The sound was very spacious but to me, not very musical. I continued to let it burn in and it mellowed, but too much so. I figured the spacial quality and extreme openness, was a passing phase and it was no better than anything else. I just let it continue to burn in and filled it with music. At 200 hours it has evened out, the spacial quality is back, the extremely good impact is there as is the black background. It does not cover anything but if you have a good recording, it shines through with excellent timbre, impact and bass quality. How a 1's and 0's would be changed, well I can only see it as in a more pure delivery of the digital signal, with less corrections having to be made by the DAP. If I had not heard these changes I would call it BS. But I have and am hearing the changes and the current excellent sound. Great recordings that an extra layer of quality added to them, a naturalness and analog nature. But it takes time so the components on the Micro Card, form and reach their maturity in a positive way. I will be buying more. This is on the DX300 with AMP12 and using the Utopia, IT07 and the SR2.

 

[Whitigir]

The "higher" end cards get mentioned here so I am posting this here. I have the Micron card non CS, 256GB and I have it in 512GB in the CS. They are very different. The CS more in line with most cards that are decent. It does have an excellent transfer speed but the sound is not near as good as the non CS. I found the non CS to be good out of the box, then the mids got hard, pushed forward and not very enjoyable. The sound was very spacious but to me, not very musical. I continued to let it burn in and it mellowed, but too much so. I figured the spacial quality and extreme openness, was a passing phase and it was no better than anything else. I just let it continue to burn in and filled it with music. At 200 hours it has evened out, the spacial quality is back, the extremely good impact is there as is the black background. It does not cover anything but if you have a good recording, it shines through with excellent timbre, impact and bass quality. How a 1's and 0's would be changed, well I can only see it as in a more pure delivery of the digital signal, with less corrections having to be made by the DAP. If I had not heard these changes I would call it BS. But I have and am hearing the changes and the current excellent sound. Great recordings that an extra layer of quality added to them, a naturalness and analog nature. But it takes time so the components on the Micro Card, form and reach their maturity in a positive way. I will be buying more. This is on the DX300 with AMP12 and using the Utopia, IT07 and the SR2.

Excellently said and great observations John ! I actually gave up after 75 hours mark LoL!!! Because it is enjoyable by then. Thanks for keeping notes and check in on the burn-in. Now we have a pretty good mark for burning in for MicroSD

I am glad we can find better ways to squeeze that extra performances for our hobby!! I am going to grab my 2nd 1Tb soon!!!

 

[Maxx134]

Edit**
Actually I am wrong. I have been convinced (im PMs) that the audio boards are actually tunned to the strengths/traits of the chassis, because that the density and other aspects of the metal change the outcome..

So far, in general I myself have observed these traits on the types of metal chassis influences:

Copper~ thicker sounding.
Aluminum~ brighter sounding.
Stainless Steel~ neutral sounding.
Titanium~???? But said to be more "airy" sounding.

I added further comments which I think is pertinent.

 

[Lappy27]

That would be awesome if it does accordingly and automatically that way. I recalled someone had that problem previously ?

It was me. I got 705.6 with 192 file with the OS set at 16 x in NMP.

with your advice, I retro back to 4 x and it showed the correct 768 frequency.

doesn’t explain why i was getting 768 with a 96 file at 16 x though.

 

[efftee]

Question. How does one burn in SD cards? Does the card have to be filled to the max with files and then play back every file? All of which for those hundreds of hours?

 

[Whitigir]

Just fill it up with files, then play it back. I don’t know why it needs burn-in. But here is from Datasheets

Intelligent controller

errors corrections Algorithms and so on

how intelligent are these stuff on the little chip ? No one know ..... could it be that the more you use it, the more efficient the card controller is at handling it tasks ? Like error corrections correct better ? Etc... ?

 

[Maxx134]

how intelligent are these stuff on the little chip ? No one know ..... could it be that the more you use it, the more efficient the card controller is at handling it tasks ? Like error corrections correct better ? Etc... ?

It only depends on the design. There's space on that chip! I read that the AI in Bitcoin is so intelligent that it creates false weaknesses to safeguard against hackers.

 

[Ojisan]

Just fill it up with files, then play it back. I don’t know why it needs burn-in. But here is from Datasheets

Intelligent controller

errors corrections Algorithms and so on

how intelligent are these stuff on the little chip ? No one know ..... could it be that the more you use it, the more efficient the card controller is at handling it tasks ? Like error corrections correct better ? Etc... ?

Well, after reading all the discussions, I've decided to dive in and order some of the cards mentioned here!

I had my doubts about SD card making any difference. My work involves mixed signal circuits (digital/analog/RF) and I have dealt with error correction codes including ones used for deep space communications. I highly doubt that bit errors are contributing factor to the sound quality or else our file transfer using SD card will be incredibly unreliable. We don't hear people talking about corrupt music/photo/video unless the card is broken.

I was reading the Sony engineer's interview who developed their audiophile SD card. He mentioned one very important factor - noise injected from the SD card. He claimed that certain SD cards inject higher levels of switching noise resulting from memory reading and other "control" circuitry included on the card. Depending on the DAP's architecture, those injected noise can make its way to the audio path, get amplified, resulting in different sound quality. For Sony's card, they apparently found a quiet memory element, or ways to reduce noise injected to the SD reader, or both. I believe this to be the most plausible explanation.

There are also people who claim degaussing SD card is just as effective. Perhaps it's a worthwhile experiment as well

Now the big question with 3Max is: Can the noise from SD slot make it's way into the audio section? The power supplies are isolated and I assume the digital and audio sections are isolated. Then how much noise can make its way to the ear? Perhaps a galvanically isolated SD slot could be the future. I dunno.

In any case, I'm going to give this a try and report back

 

[alsorkin]

September 2, 2021. A new MAX firmware that is OTA.

1. Rectified the bug that there is some background noise when playing MQA via Tidal.
2. Improved Bluetooth compatibility.
3. Improved USB digital output compatibility.

V1.02.184

Local update download: http://www.ibasso.com/uploadfiles/download/DX300MAXFW-v1.02-ota.zip

Download, unzip and then use the FW in the new folder.



The MAX series continues...

​

Let's discuss this beast. @Paul - iBasso feel free to take over the thread

What advantages the Max has over the Dx312 in SQ?

 

[Whitigir]

What advantages over the Dx312 in SQ?

A lot of technical advantages , but about SQ ? It is so subjective that my friend prefer apples earbuds over my HD800S system.

 

[Maxx134]

I highly doubt that bit errors are contributing factor to the sound quality or else our file transfer using SD card will be incredibly unreliable. We don't hear people talking about corrupt music/photo/video unless the card is broken

That is different. That is static data. That is not realtime streaming data, that would be sensitive to jitter/time dependency, like streaming music data & master/slave digital setups in studio gear. So although I am not explaining this very well, it is not simple.

He claimed that certain SD cards inject higher levels of switching noise resulting from memory reading and other "control" circuitry included on the card.

Interesting, switching noises that could be confused or affect data. Maybe affect jitter or data interpretation?

Now the big question with 3Max is: Can the noise from SD slot make it's way into the audio section?

That's not how noise travel. It is contained between digital stages. Look at all the stages in that new card. It would be contained.
If output actually had some noise, It could only get from the card to slot digital interface, and could only try affect "realtime" interpretation of data. Noise from card cannot travel thru system.

 

[Ojisan]

That is different. That is static data. That is not realtime streaming data, that would be sensitive to jitter/time dependency, like streaming music data & master/slave digital setups in studio gear. So although I am not explaining this very well, it is not simple.

Ah ok, so there is file-level error correction and streaming packet error correction mechanism. I can't say what streaming standard they use over SD interface so I may have overlooked something. The stuff I'm familiar with are 3G/4G/5G high-speed cellular data error correction which is basically real-time Gbit speeds, subject to errors from radio interference, clock drifts, doppler, among other things. You can do surprisingly well with the robust algorithms today.

Interesting, switching noises that could be confused as data. Maybe affect jitter or data interpretation.

No, there is no noise interaction with the data. These noise are sufficiently low enough not to impact 0/1 decisions. But when that noise couples into supply and other areas, it's another source of "digital" noise. I would hope that iBasso has figured out a clean digital stage by now such that SD card data is not corrupted before the DAC unless you are using these special SD cards.

That's not how noise travel. It is contained between digital stages. Look at all the stages in that new card. It would be contained.
If output actually had some noise, It could only get from the card to slot digital interface, and could only try affect "realtime" interpretation of data. Noise from card cannot travel thru system.

Unfortunately, it can That's one of the reason why we have separate power source and ground for analog and digital sections in 3Max.

 

[ayang02]

Ah ok, so there is file-level error correction and streaming packet error correction mechanism. I can't say what streaming standard they use over SD interface so I may have overlooked something. The stuff I'm familiar with are 3G/4G/5G high-speed cellular data error correction which is basically real-time Gbit speeds, subject to errors from radio interference, clock drifts, doppler, among other things. You can do surprisingly well with the robust algorithms today.

SD cards use the SPI interface: https://en.m.wikipedia.org/wiki/Serial_Peripheral_Interface

Typically there are send/receive buffers on both the master (SOC) and slave (SD card) sides as the interface is full-duplex.

 

[Ojisan]

SD cards use the SPI interface: https://en.m.wikipedia.org/wiki/Serial_Peripheral_Interface

Typically there are send/receive buffers on both the master (SOC) and slave (SD card) sides as the interface is full-duplex.

Ahh thanks. Pretty basic interface and no standardized error checking. So it's possible that some SD cards are so crappy that when streaming, it has number of errors in the stream If the errors are not always the same, our music should sound different every time, lol. Is it better to carry CD walkman again?

 

[ayang02]

Ahh thanks. Pretty basic interface and no standardized error checking. So it's possible that some SD cards are so crappy that when streaming, it has number of errors in the stream If the errors are not always the same, our music should sound different every time, lol. Is it better to carry CD walkman again?

On the physical layer, yup pretty simple. I think the error checking and all that probably happens at a higher level. I didn’t go through the SD protocol but I thought I saw something involving error checking before.

At least some SD cards show you its block diagram and the better ones out there (Micron, Sandisk, Samsung, Greenliant etc.) pretty much have some form of ECC implemented.