Head-Fi.org › Forums › Equipment Forums › Portable Source Gear › iBasso DX90. Dual Sabre, 1st page: Downloads, info & instructions- New dac software! & 2.1.5 FW. . . . . . . . . .
New Posts  All Forums:Forum Nav:

iBasso DX90. Dual Sabre, 1st page: Downloads, info & instructions- New dac software! & 2.1.5 FW.... - Page 478

post #7156 of 10107
Quote:
Originally Posted by IH8RDNX View Post
 


OK I'll bite.

 

Gain typically refers to the audio signal level. It is often expressed in dB, where 0 dB is the maximum voltage level, -3 dB is half level, etc. It is related to signal amplitude and is sometimes incorrectly expressed in mV. Measure with multimeter when expressed as mV but dB is more useful.

 

Volume typically refers to the sound level. It is also expressed as dB, where 70 dB is very quiet and 140 dB will deafen you. Speaker volume varies according to input level and speaker impedance. Measure with sound level meter.

 

Best noise rejection is highest gain + low volume. Did I get it right?

This is not true form many reasons:

- DX90 have SW volume control. It mean, that on lower SW volume with higher amp gain You lost more music information.

- Circuits before amplifier have any noise. This noise is independent on SW volume. Higher gain of amp = You got higher this noise.

- "Gain swith" set gain of opamp (I think swith between different resistors in loop). Higher gain of opamp can produce more distortion, noise, etc. But this can not be say generaly, it is more difficult (any opamp are more stable at higher gain).

post #7157 of 10107
Quote:
Originally Posted by borrego View Post

The "SD Card" in recovery actually means internal memory. You can do the wipe cache before or after the update. But since you are reverting to older version firmware, I would say do a wipe cache before update.
Has anyone done this procedure before? I don't want to risk bricking my DX90.
post #7158 of 10107
Quote:
Originally Posted by athmospheric View Post

I would suggest batch tool -> http://www.ibasso.com/download/DX90-tools.zip

Just follow instructions in PDF plus read paragraph If you are having problems loading the batchtool from this post

Here you have 2.0 FW: http://www.ibasso.com/download/DX90V2.0.0Apr7th.zip
Do I need to use batchtool to load the original 2.00 FW?
post #7159 of 10107
Quote:
Originally Posted by athmospheric View Post
 

 

I would suggest batch tool -> http://www.ibasso.com/download/DX90-tools.zip

 

Just follow instructions in PDF plus read paragraph If you are having problems loading the batchtool from this post

 

Here you have 2.0 FW: http://www.ibasso.com/download/DX90V2.0.0Apr7th.zip

OK thank you. I had to use the restore option this time. That worked so I am back to 2.0. Lets see how it sounds.

post #7160 of 10107
Quote:
Originally Posted by wizard327 View Post


Do I need to use batchtool to load the original 2.00 FW?

Yes Thats how I did mine. Just use the restore option instead of upgrade when your using 2.05 or 2.1 fw

post #7161 of 10107
Quote:
Originally Posted by Lurker0 View Post
 

I don't want to do it. Below is a long explanation.

 

I don't like any kind of software processing. In hardware, I don't like equalizers and other "beautifizers". I prefer bit perfect playback of digital content, when the quality of the sound mostly depend on the hardware. This way, an ultimate recording sounds ultimate, a good recording sounds good, a bad recording sounds bad.

 

I never tried the original 2.0.0 firmware. I've flashed 2.0.5 as the frist step after unboxing. Then I ran a test to check if DX90 is bit perfect, which succeeded. Same with 2.1.0.

 

The test is quite simple, I use it over past 10 years with sound cards, and I wonder why other people don't use it. I connected DX90 S/PDIF to my pre-pro (Onkyo PR SC-5509), and played back a flac track from a DTS Audio CD rip. The processor has detected DTS stream and has decoded it into 5.1 sound. But only if EQ is turned off. Because DX90 software treats DTS-encoded stream as a usual PCM sound data, it would alter it the same way as any other music records, if it alters sound bit stream either way. But the fact that the processor decodes DTS stream flawlessly, exactly means that DX90 player outputs sound stream unaltered, if EQ is turned off. This is called bit perfect playback :)

 

Now you might wonder what if the player has two different, parallel, paths for output to DAC and to S/PDIF... I'm sure it hasn't. To confirm that, I even looked into MangoPlayer in a disassembler (IDA).

 

Now, bearing in mind that both 2.0.5 and 2.1.0 are bit perfect, what is the difference between them? There are two:

  1. Changes in MangoPlayer to fix playback order of files and tracks in CUE sheet (that's what I asked them for ;)).
  2. Removed 5 libraries from system/lib/soundfx.

 

The #2 is the most questionable change. Well, probably I have to tell you that I'm professional programmer for past almost 30 years, involved into mobile and embedded platforms since 1995, and for past 4 years Android is one of the platforms I work with. That's to avoid speculations about my experience.

 

The soundfx filters are not used by MangoPlayer. They can be used only by Android sound path, but MangoPlayer does not use Android sound path. I've read Android programmer's documentation on those filters, and I'm pretty sure they are not used. Moreover, with 2.0.5, I tried to delete the core Android sound files, and it changed nothing: MangoPlayer continued working.

 

So why iBasso decided to delete them? In my opinion, to stop speculations floating around.

 

Why somebody hear different "sound signatures" of different firmware versions? It is up to them. I do understand that there are different "sound signatures" depending on temperature, humidity, time of the day, day of the week, mood, tiredness, how good/bad IEMs inserted etc. etc.etc. That's why I said in comments to "my" firmware that it may change the "sound signature".

 

But hey, all in all, even it is truly confirmed that soundfx files being present, even not used, affect the sound signature, for me it's exactly mean they must be deleted and never put back! Because I like bit perfect sound!

 

If you like to listen not to music but to distortions, wait for RockBox to appear, and you'll have them all :D Frankly, when I was waiting for my DX90, I've read a lot of forums, built RockBox for Android myself to try, had read its sources and decided that this is not the way to go with players like DX90. At least - not for me.

 

By the way: apart for specific test tracks, I didn't listen lossy files of even CD rips on DX90 so far, only hi-res records. That's what I purchased it for.

 

My university degree diploma was also in computer science. So I believe 100% in you on what you have said on FW2.0.5 and FW2.1 that there are no different coding that would affect the sound signature.  However,  I also believe in my own hearing that there were indeed sound signature difference during the time I tried 2.0.5 and 2.1 firmware. Like you said, it might as well be temperature, humidity, time of the day, mood, etc. But if I have to find a more technical/logical reason after your explanation, it could be the different processor loading % of the Rockchip processor when playing different tracks which will result to different sound signature.

 

When I started using my Windows PC + E-mu 0404 USB as the source of my stereo system 4-5 years ago, I could actually experienced slightly different sound signature every other time I boot up my PC. Even updating to a newer version of anit-virus software with a new background process would affect the sound this and the other way. My explanation was in a multi-tasking OS running multiple processes, you would never know what priority the processor is assigning to a particular process at a given time, or at what time the OS process decide to do certain garbage cleaning to purge memory/release resources.

 

But my view on DSP is different from yours. While many people believe any form of DSP is "Bad" for high end audio, and demand "Bit Perfect" playback, I care the least for that. Being a supporting vocal singer and occasional sound system controller handling live recording in my church, I have learned how inevitable and how much DSP effects were already added during the recording processes. There is little point of insisting "bit perfect" when the original bit was already the result of countless altering. As long as it sounds good, I don't mind altering those "bits" again during playback.

 

And with a multi-threading capable processor running muti-tasking OS, sometimes it is actually good to keep a DSP as active high priority process , just to keep the processor "focus" on that task alone and not giving priority to other irrelevant tasks.

 

I know nothing about Android programming. However, I understand with any high level programming languages, different compiler options used during compiling can also cause different processor resource optimization/power saving policies, etc. I am not sure if there are such options with the Rockchip kitchen resources though.

 

As always, Thanks for your great work!


Edited by borrego - 7/20/14 at 7:26am
post #7162 of 10107
Thread Starter 

Extremely interesting reading and thanks to all that are contributing to this. Most all strive for the best sound possible and to this end, it is great to read what is going on here. 

 

I tend to always like the least processed as possible and have wondered, if bit perfect, how is the sound changing. To me it would have to be the processing and how the resources are dealt with as borrego mentions. And they the fascinating work by Lurker0 adds a dimension. Now to somehow figure this out. Hopefully it won't take as long as trying to figure out the true source of gravity. :^) But maybe? 

post #7163 of 10107
Quote:
Originally Posted by Wishmasterx View Post
 

This is not true form many reasons:

- DX90 have SW volume control. It mean, that on lower SW volume with higher amp gain You lost more music information.

- Circuits before amplifier have any noise. This noise is independent on SW volume. Higher gain of amp = You got higher this noise.

- "Gain swith" set gain of opamp (I think swith between different resistors in loop). Higher gain of opamp can produce more distortion, noise, etc. But this can not be say generaly, it is more difficult (any opamp are more stable at higher gain).


I think the original question was open to interpretation and we are interpreting it differently. Let's assume for a moment the original question was not a joke. Then, anyone using an IEM from the HO cannot control gain, only volume. Therefore, the only situation that both gain control and volume control can occur would be using an external amp.

 

The external amp will generate noise with or without an input signal applied. The noise is generated in the higher voltage back end and is proportional to the volume control position on the external amp. Now assuming the upstream source (the DX90) is clean, the best S/N ratio will occur with the DX90 LO gain at 0 dB (volume = 255). This will mask the noise generated by the external amp. Then by turning down the volume on the external amp, the amp noise is further reduced.

 

So you are right and I am right but I noticed none of the famously smart posters on this thread tried to answer the original question. Perhaps this is the wrong thread for this type of discussion. Let's agree to disagree.

post #7164 of 10107
Quote:

Originally Posted by borrego View Post

 

But if I have to find a more technical/logical reason after your explanation, it could be the different processor loading % of the Rockchip processor when playing different tracks which will result to different sound signature.

That exactly is. More CPU load means higher temperature, and more power consumption by CPU, which, in turn, affects output amplifiers. Also, I'm curious, may it lead to bigger jitter, but I'm lazy (or busy?) enough to start investigating this question;)

 

You might notice that all the discussions around different firmware versions are mostly concentrated on bass. Temperature and power supply have more impact exactly on bass.

 

And this is exactly the reason that I like removing unneeded libraries and services: it makes the system more stable and predictable.

post #7165 of 10107
Quote:
Originally Posted by Lurker0 View Post
 

I don't want to do it. Below is a long explanation.

 

I don't like any kind of software processing. In hardware, I don't like equalizers and other "beautifizers". I prefer bit perfect playback of digital content, when the quality of the sound mostly depend on the hardware. This way, an ultimate recording sounds ultimate, a good recording sounds good, a bad recording sounds bad.

 

I never tried the original 2.0.0 firmware. I've flashed 2.0.5 as the frist step after unboxing. Then I ran a test to check if DX90 is bit perfect, which succeeded. Same with 2.1.0.

 

The test is quite simple, I use it over past 10 years with sound cards, and I wonder why other people don't use it. I connected DX90 S/PDIF to my pre-pro (Onkyo PR SC-5509), and played back a flac track from a DTS Audio CD rip. The processor has detected DTS stream and has decoded it into 5.1 sound. But only if EQ is turned off. Because DX90 software treats DTS-encoded stream as a usual PCM sound data, it would alter it the same way as any other music records, if it alters sound bit stream either way. But the fact that the processor decodes DTS stream flawlessly, exactly means that DX90 player outputs sound stream unaltered, if EQ is turned off. This is called bit perfect playback :)

 

Now you might wonder what if the player has two different, parallel, paths for output to DAC and to S/PDIF... I'm sure it hasn't. To confirm that, I even looked into MangoPlayer in a disassembler (IDA).

 

Now, bearing in mind that both 2.0.5 and 2.1.0 are bit perfect, what is the difference between them? There are two:

  1. Changes in MangoPlayer to fix playback order of files and tracks in CUE sheet (that's what I asked them for ;)).
  2. Removed 5 libraries from system/lib/soundfx.

 

The #2 is the most questionable change. Well, probably I have to tell you that I'm professional programmer for past almost 30 years, involved into mobile and embedded platforms since 1995, and for past 4 years Android is one of the platforms I work with. That's to avoid speculations about my experience.

 

The soundfx filters are not used by MangoPlayer. They can be used only by Android sound path, but MangoPlayer does not use Android sound path. I've read Android programmer's documentation on those filters, and I'm pretty sure they are not used. Moreover, with 2.0.5, I tried to delete the core Android sound files, and it changed nothing: MangoPlayer continued working.

 

So why iBasso decided to delete them? In my opinion, to stop speculations floating around.

 

Why somebody hear different "sound signatures" of different firmware versions? It is up to them. I do understand that there are different "sound signatures" depending on temperature, humidity, time of the day, day of the week, mood, tiredness, how good/bad IEMs inserted etc. etc.etc. That's why I said in comments to "my" firmware that it may change the "sound signature".

 

But hey, all in all, even it is truly confirmed that soundfx files being present, even not used, affect the sound signature, for me it's exactly mean they must be deleted and never put back! Because I like bit perfect sound!

 

If you like to listen not to music but to distortions, wait for RockBox to appear, and you'll have them all :D Frankly, when I was waiting for my DX90, I've read a lot of forums, built RockBox for Android myself to try, had read its sources and decided that this is not the way to go with players like DX90. At least - not for me.

 

By the way: apart for specific test tracks, I didn't listen lossy files of even CD rips on DX90 so far, only hi-res records. That's what I purchased it for.


Thanks for hopefully putting an end to the DSP silliness.

 

Artists manipulating a recording to better express themselves is completely different than altering their message after the fact and those manipulations done by the artist are done to tracks and not an entire recoding. DSP as used after the fact robs the music of time and amplitude relationships that are key to understanding the intent of the artist. If someone else wants to observe a Picasso (which the master probably altered before presenting) with rose colored glasses, that's their prerogative but it's certainly not for me.


Edited by goodvibes - 7/20/14 at 7:49am
post #7166 of 10107
Thank you Lurker0 for the international version firmware. So does this version display Chinese or Japanese characters properly? Thanks..
post #7167 of 10107
Is the gain on the DX90 digital or analog?  The Calyx M has digital gain I believe.  The volume on the DX90 has a certain range, what happens to that range when you put it on high gain?  Does it stay the same?  I believe the Calyx changes the volume curve, and I'm guessing the curve changes from logarithmic to more linear depending on the gain.
 
I think the software gain settings are all the same.  It changes the volume curve.
post #7168 of 10107
Thread Starter 
Quote:
Originally Posted by Lurker0 View Post
 

That exactly is. More CPU load means higher temperature, and more power consumption by CPU, which, in turn, affects output amplifiers. Also, I'm curious, may it lead to bigger jitter, but I'm lazy (or busy?) enough to start investigating this question;)

 

You might notice that all the discussions around different firmware versions are mostly concentrated on bass. Temperature and power supply have more impact exactly on bass.

 

And this is exactly the reason that I like removing unneeded libraries and services: it makes the system more stable and predictable.

Hi frequencies can have a curious drain also. And to my ear it isn't just the bass but also the openness of the soundstage that is affected, which to me, often equates to transparency, that elusive quality of 3D effect where sound shimmers through sound but does not get tangled in it. And to me, it still boils down to how the power supply is handled and home many power supplies are used, therefore more effectively isolating the effects of power drain resource from one component to another. The Dac I built some years ago, based on Audio Note Uk, had 4 distinct power supplies for the board and then there was a massive few for the analogue stage. 

post #7169 of 10107
Quote:
Originally Posted by Lurker0 View Post
 

That exactly is. More CPU load means higher temperature, and more power consumption by CPU, which, in turn, affects output amplifiers. Also, I'm curious, may it lead to bigger jitter, but I'm lazy (or busy?) enough to start investigating this question;)

 

You might notice that all the discussions around different firmware versions are mostly concentrated on bass. Temperature and power supply have more impact exactly on bass.

 

And this is exactly the reason that I like removing unneeded libraries and services: it makes the system more stable and predictable.


I'm sure less CPU cycles etc does lead to less noise but the question is always how much extra dig noise (jitter) is actually noticeable or at what level to be considered noise when it may actually be within the floor. That would take up a dozen pages in a forum. Whenever it comes down to #s vs hearing, any discussion seems to become circular.


Edited by goodvibes - 7/20/14 at 7:56am
post #7170 of 10107
Quote:

Originally Posted by Lurker0 View Post

If you like to listen not to music but to distortions, wait for RockBox to appear, and you'll have them all :D Frankly, when I was waiting for my DX90, I've read a lot of forums, built RockBox for Android myself to try, had read its sources and decided that this is not the way to go with players like DX90. At least - not for me.

 

Lurker0, can you explain what you mean here regarding "distortions"? 

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Portable Source Gear
Head-Fi.org › Forums › Equipment Forums › Portable Source Gear › iBasso DX90. Dual Sabre, 1st page: Downloads, info & instructions- New dac software! & 2.1.5 FW. . . . . . . . . .