[goodvibes]

  Low gain has the best SQ imo if absolute neutrality is what your looking for, but it does sound a bit flat and boring (but reference), from my listening low gain also has the best micro detail retrieval of all gain settings. Mid gain is a good compromise between the neutrality of Low gain and the fun factor of High gain. I definitely find high gain to have the lowest SQ, there is a considerable veil on micro detail and certain frequencies are boosted making a shouty, boomy, V shaped sound .
 
I also have a theory about this 

 I believe a big part of this has to do with the BUF634 op amp used in dx90, which in my experience with op amp rolling in the iBasso P4 has this effect of accentuating a V shape sound and with a loss of resolution and veiling the higher the gain is. Basically the BUF634 is degrading the sound with increasing magnitude the higher the gain, even on the lowest gain the BUF634 is limiting absolute transparency of the DX90. So I'm not a fan of it.
 
The main reason the BUF634 is used in DX90 is keeping the power supply stable when large amounts of current are being drawn (its a unity gain buffer), which is only needed when powering bigger or harder to drive headphones. So basically the BUF634 is a compromise to make the DX90 better at driving bigger cans but limiting the transparency of the power supply in the process which can greatly affect iems or any transparent headphone. I would personally prefer the DX90 without the BUF634, for example the iBasso P4 amp sounds considerably cleaner and transparent without it and I'd bet my savings the DX90 would too.
 
If I had excellent soldering skills I would probably try take the BUF634's out and replace them with transistor buffers, if they fit.

That's a bit of bias from a different experience. Those are not absolutes and voicing of components varies by device and surrounding circuit. Low gain is not the most neutral or informative. If I had to describe it, it would be an overstatement but relative to med it sounds a touch overdamped. That is not neutral. High gain is by comparison a little loose. Medium gain is the most natural which is also the most neutral with the best retrieval and contrast. Micro detail is where ambient and spacial clues reside and Med has the easiest to discern/most correct ambiance retrieval and on the highest quality files. Low gain would be preferred Emperor Joseph II that didn't like 'too many notes' getting in the way. No idea but it wouldn't surprise me if the DX90 was first designed at med gain with more options given via the feedback loop.
 
Warning, minor rant not directed at TRANCE: For me, there is no antithesis of natural and neutral even though they are often mistakenly used to describe either colorful or analytic sound, respectively. Both mean normal even if sightly different aspects of it. The second set of adjectives should be used to describe themselves but advocates of certain characters need a good sounding descriptive to promote. It's become all too common usage and accepted by those without an agenda. I'll probably get blasted for it but they are used improperly IMO. Both are good things and can be had in the same device.
 
I do think your earlier experience with a device may have biased your perspective on it's use in the DX90.

 

[lian00]

Just to say the iBasso download page is on line again: http://www.ibasso.com/en/download/ - not the rest of the site.
 
Impossible to contact them for the moment and I wanted to warn them that the DAC beta drivers are interfering with Wacom drivers (!).

 

[Wishmasterx]

Just to say the iBasso download page is on line again: http://www.ibasso.com/en/download/ - not the rest of the site.


 


Impossible to contact them for the moment and I wanted to warn them that the DAC beta drivers are interfering with Wacom drivers (!).

 


When page is down, you can contact iBasso on email, iBasso will respond.

For all: New USB DAC drivers will be in September.

 

[lian00]

 

  Just to say the iBasso download page is on line again: http://www.ibasso.com/en/download/ - not the rest of the site.
   
  Impossible to contact them for the moment and I wanted to warn them that the DAC beta drivers are interfering with Wacom drivers (!).

 


When page is down, you can contact iBasso on email, iBasso will respond.

For all: New USB DAC drivers will be in September.

Do you have their email ?

 

[pauldgroot]

service[at]ibasso[dot]com

 

[lhamp]

Crazy thing is I've owned the unit for about four months now. This was just a classic Brain Fart.

 

[nuxos]

  Low gain has the best SQ imo if absolute neutrality is what your looking for, but it does sound a bit flat and boring (but reference), from my listening low gain also has the best micro detail retrieval of all gain settings. Mid gain is a good compromise between the neutrality of Low gain and the fun factor of High gain. I definitely find high gain to have the lowest SQ, there is a considerable veil on micro detail and certain frequencies are boosted making a shouty, boomy, V shaped sound .
 
Thank you
 
I was looking for a good description between all gains. 

 

[T.R.A.N.C.E.]

That's a bit of bias from a different experience. Those are not absolutes and voicing of components varies by device and surrounding circuit. Low gain is not the most neutral or informative. If I had to describe it, it would be an overstatement but relative to med it sounds a touch overdamped. That is not neutral. High gain is by comparison a little loose. Medium gain is the most natural which is also the most neutral with the best retrieval and contrast. Micro detail is where ambient and spacial clues reside and Med has the easiest to discern/most correct ambiance retrieval and on the highest quality files. Low gain would be preferred Emperor Joseph II that didn't like 'too many notes' getting in the way. No idea but it wouldn't surprise me if the DX90 was first designed at med gain with more options given via the feedback loop.

The sound of the ibasso P4 amp is essentially completely dictated by the op amps installed, the op amps are also exchangeable meaning with experimentation you can easily discern what each op amp is adding to the sound. The P4 is a much better amp than the internal DX90 amp, it is quite reasonable to think that if BUF634 affects a stellar amp like the P4 in a negative way that it will also be affecting the DX90. The implementation of the DX90 internal amp is very similar to ibassos other amps, like say the ibasso D12 etc, they are simply op amp chains, the implementation doesn't exactly vary significantly, and the predominant sound signature of all of ibassos amps originates from which op amps are used.

I guess I can agree that mid gain does sound more natural than low gain, but I still believe that it is a coloration from the amp causing this perception, and that low gain is the most flat frequency balance.

 

[goodvibes]

On an amp that you can roll devices or eliminate stages, you cannot optimize a circuit for a given set of deices or topology or at least we don't know what it was optimized for. Perhaps you found it for the P4. Power supplies have a huge affect of an amp as well. In effect, an amplifier is a controlled way to connect said PS to  a transducer. When you have a set circuit topology and guys that listen, you can get a more optimized result, even if more limited in some ways. For instance, the extra added local supply caps used on the dx90, specifically for the buf634s, may be just what the doctor ordered.
 
The DX90s has a normal balance of very neutral except for a small bit of added warmth that tends to fade away with HiDef wav files. It's darn close to right. I think you are trying to compensate that warmth a bit by using low gain and throwing out the baby with the bath water. All IMO. I'm not debating that you are right for you. Just giving my perspective.
 
To those that think we're crazy, well yes, yes we are.

 These are somewhat overstated subtleties that many don't ever notice but important enough for us geeks.

 

[T.R.A.N.C.E.]

End of the day its all good, frankly im just glad i have the option to choose between low and medium gain depending on what my mood calls for, i do enjoy both.

 

[Lurker0]

FYI there are no different gains per se. There is a 3-state resistor divider at the input of the buffer, nothing less, nothing more. Think about it as a good old potentiometer with only 3 states.

 

[jamato8]

  FYI there are no different gains per se. There is a 3-state resistor divider at the input of the buffer, nothing less, nothing more. Think about it as a good old potentiometer with only 3 states.

Nothing much cleaner than a divider, depending on what type of resistors are used. It still may affect the buffer due to the different loads, and subtle effects of frequency shifting. 

 

[goodvibes]

Interesting. I expected three different values in the feedback loop. Definitely some minor difference in sonics. Of course,​ little things tend to be noticed in revealing kit.

 

 

[lian00]

  service[at]ibasso[dot]com

Thank you.

 

[Lurker0]

 

  FYI there are no different gains per se. There is a 3-state resistor divider at the input of the buffer, nothing less, nothing more. Think about it as a good old potentiometer with only 3 states.

Nothing much cleaner than a divider, depending on what type of resistors are used. It still may affect the buffer due to the different loads, and subtle effects of frequency shifting. 

Yes. But, IMHO, the DAC at different volume levels has more impact. And also don't forget an ability to set equal volume levels at different gain switch positions, if ever possible