iBasso DX300 Qualcomm Snapdragon 660 Octa-core 6GB RAM ******NEW Firmware 2.00 Android 11******
Feb 13, 2021 at 9:40 AM Post #3,977 of 14,596
So pure mango cant play mqa. Seems strange to me

MangoOS can recognise, unfold and play MQA files

Go here:

http://www.2l.no/hires/

There are a heap of files for you to compare with. I download them straight from dx300. Suspect dx300 will run all the file types.

20210214_004346.jpg
 
Last edited:
Feb 13, 2021 at 9:44 AM Post #3,978 of 14,596
I prefer MangoOS and the High gain with D3 filter

I'm finding low gain and D3 as my soft spot. In low gain vocals seem to drop down to eye level for me. Seems to set the stage for me when I get a nice layered recording.
 
Feb 13, 2021 at 9:48 AM Post #3,979 of 14,596
I'm finding low gain and D3 as my soft spot. In low gain vocals seem to drop down to eye level for me. Seems to set the stage for me when I get a nice layered recording.
Yes, the staging has a better cohesiveness and in some cases it has better presentations, just at the cost of some dynamic. I guess it is all based on the synergies of the gears, but D3 is a definite sweet one for me as well
 
Feb 13, 2021 at 9:54 AM Post #3,980 of 14,596
You've not answered anything I'm afraid.
Digital gain would imply changing the gain in the digital domain before DAC stage which is not what's happening as the background noise gets louder with increasing volume setting with no audio playing.

An electronically controlled analog gain would make more sense to explain that behavior, but where is it in the signal chain?

Same for the gain settings, is it just modifying the range of the volume setting? Or switching in different circuits? The former would make sense and matches with the impression that noise levels on different gains are equivalent for similar listening levels.

And finally, there is no info on the Max output for different gain settings in the manual.

I think Po adds current amplification vs Lo. It's not a separate buffered circuit which means it is not as pure as a dedicated line out. But its not like the AK players which just force volume to max. If you look at iBasso's website it even lists line out specs separately.

But the gain aspect is confusing. The circuits create 8V rails which is higher than the battery output. I imagine low gain is best to be closest to a true line out.

I hope to see iBasso respond on this topic
 
Feb 13, 2021 at 9:54 AM Post #3,981 of 14,596
Meh, seemed a bit dull, flat at around 250 hours....

280 hours it seems to be approaching maturity. And is quite listenable and enjoyable. More coherent overall.

Even Coldplay sounded good. :)

i will have to compare with my Max notes and see if theres any parallels.
It seems on the cusp of of dap self actualization at the moment.lol.
Thinking beyond 300 hours there will not be any significant changes.

@jamato8 probably knows as he is ahead of me.

Low gain, D3 filter 95% of the time (except periodic switching between em. To just check what's happening). Mostly using Andromeda MW10.

usually popping in to listen every 24 or 48 hours.


- just got interrupted before by a big as earthquake...

You do realize that once you stop burn-in playback and resume later, you will have to wait another 280hrs before it will sound good again? :p
 
Feb 13, 2021 at 10:30 AM Post #3,984 of 14,596
Anyone DSD512'ing on DX220 can try the 8x (352 / 384) oversampling on DX300 in NOS mode. I just did it for the Nguyen Le / Purple album and was able to correctly identify every single time which version (44.1 kHz or 352 kHz) is playing on a blind test (randomly selected by my wife). The difference is more audible than filters.
 
Feb 13, 2021 at 10:35 AM Post #3,986 of 14,596
Anyone DSD512'ing on DX220 can try the 8x (352 / 384) oversampling on DX300 in NOS mode. I just did it for the Nguyen Le / Purple album and was able to correctly identify every single time which version (44.1 kHz or 352 kHz) is playing on a blind test (randomly selected by my wife). The difference is more audible than filters.
what program are you using for the OS ? NOS is Non Oversampling, so if you can hear the differences between NOS vs OS, that is normal. It is the same as if you try D3 VS D5, there are a clear differences

The idea behind the DSD512 offline processing is that there is less “errors corrections” while using the better “algorithms” to compute with Sigma Delta techniques VS Live processing of a mere small players.

So by using DSD512 offline processing, it isn’t that people are seeking the differences, but rather is seeking the most accurate way to process PCM from digital into Analog, and the last thing that effect it is only LPF circuitry and the amplifier tuning rather than the DAC Chips, and the whole little systems
 
Last edited:
Feb 13, 2021 at 10:45 AM Post #3,987 of 14,596
what program are you using for the OS ? NOS is Non Oversampling, so if you can hear the differences between NOS vs OS, that is normal. It is the same as if you try D3 VS D5, there are a clear differences

The idea behind the DSD512 offline processing is that there is less “errors corrections” while using the better “algorithms” to compute with Sigma Delta techniques VS Live processing of a mere small players

I didn't check the spec of the DAC but normally DS DACs, as I understand, apply 2X interpolation in 3 steps to reach the 8X resolution. As this is such a low power DAC, maybe this is one of the cut corners. Anyway, it works well. I am using the AuI ConverteR 48x44.
 
Feb 13, 2021 at 10:47 AM Post #3,988 of 14,596
Most of the desktop DAC Chips can only do 256x oversampling or simply put, DSD256 at it end points, and as you said 8X oversampling. These number is coming from 32 bits x 8(OS). I have not seen anything to be able to do 16X or DSD512 at it ends points. But even so, offline processing will always be at a technical advantages.

The only downfall is that different genres will need different filters and modulations to sound (better to personal preferences), and only HQPlayerPro can offer that but not cheap

The best Line out Portable source that can also utilize this features will have to be MAX. But the heat dissipation is a little “cooking” lol
 
Last edited:
Feb 13, 2021 at 10:48 AM Post #3,989 of 14,596
I didn't check the spec of the DAC but normally DS DACs, as I understand, apply 2X interpolation in 3 steps to reach the 8X resolution. As this is such a low power DAC, maybe this is one of the cut corners. Anyway, it works well. I am using the AuI ConverteR 48x44.
More likely a secret recipes of their own DSP algorithms that is also super efficient? Chord MScaler for an example is also a very efficient machine Vs HQplayer
AUI is great!!!
 
Feb 13, 2021 at 10:50 AM Post #3,990 of 14,596
More likely a secret recipes of their own DSP algorithms that is also super efficient? Chord MScaler for an example is also a very efficient machine Vs HQplayer
AUI is great!!!

AUI is great also because it is much cheaper. :beyersmile:
 

Users who are viewing this thread

Back
Top