PowerAmp v3 alpha-703 - How to tune and enjoy your music

Alpha version of upcoming major v.3 Poweramp release has added Hi-Res support for Sony Xperia Z2/Z3 and other devices. Let`s tune it and make it sounds better.

First of all, plug your wired headphones in and set volume level to minimum for safety reasons.
All settings are ideal for Sony Xperia Z2/Z3/Z3 Compact but you can change them according to following notes and use for your device.

Swipe in from the left edge of the screen, select Settings and then...

select Audio.

Replay Gain. In a naive way, we could say that the essence of RG is to to normalize loudness for individual tracks or albums. It concerns major lossy formats such as MP3, AAC and OGG Vorbis.
As you may have noticed Foobar2000 (For PC) have the same RG Options.
Select - Apply Gain/prevent clipping . Click the "Source" and then select Album or Track to normalize each track or a whole album.
RG Preamp  - set gain in decibels for songs which contain RG tags. If you set it to zero - you`ll prevent clipping but volume level of lossy tracks will probably be too low. So, I recommend you set 5 or 10 dB gain.
/Preamp for songs without RG info - leave it "zero" to prevent clipping.

Select Resampler.

Click Resampler Type and choose SoX Resampler.  In that case you`ll have to deal with higher power consumption. Why do we need SoX? Because of its high quality.

Resampler Cutoff Frequency Ratio - leave default values or set 95%.

Dither - renders quantization noise independent of the input signal and generally isn't necessary.
If your smartphone`s chipset can only handle a maximum of 16-bit input (without software conversion 24bit to 16bit), but you want to play 24-bit audio files, then you should enable dither. If your smartphone can handle 24-bit input, then you don`t have to enable dither and SoX resampler is the only one you need. SONY Xperia Z2 has 24-bit 96kHz music chip.

Output -  Commonly available options include 3 output methods.
Hi-Res Output is active at the moment.

AudioTrack Output (Java based) -  transfer the audio data from the Java layer to the native layer and queue for playback. It`s High Latency and low quality output.

OpenSL ES Output (Native optimized) - general output method for all modern Android OS versions. Click Open SL ES Output. You can see what devices is this output used for.
Enable all output devices. This output is limited by 16 bit and 44.1 or 48 kHz (For SONY Xperia Z-series). Remember these figures.
Click Wired Headset/AUX Options button.

All output devices have almost the same options.

No Equ/Tone is usefull because it turns off EQ and other sound enhancements. Generally, you shouldn't`t use EQ because it causes distortion.

Hi-Res Output (Experimental) - is available for some smartphones and supported by SONY Xperia Z2/Z3. Your player and device manufacturer both should support this output to make it available to choose. The most quality output.

Variant shows us "Sony High Res Audio" (In the case of Z2). Some devices (like SONY Xperia Z5) have more than one available options of "variant".
Enable all output devices (it will automatically disable OpenES output triggers for selected devices), then Click Wired Headset options button.
This output is limited by 24 bit 96kHz.

DVC is always enabled means that direct volume control is always enabled (rewrites previous DVC setting).

Select No Equ/Tone for better sound quality if you don`t want to use equalizer.

Sample Rate 

Set 96khz to enable direct play for common hi-res audio files (24bit 96kHz). 16bit 96kHz audio files will be played directly as well. All lossy (MP3, AAC, etc) and CD Quality lossless files (16bit 44,1kHz FLAC/ALAC) will be resampled up to 96kHz. 24bit 192kHz will be resampled down to 24bit 96kHz. The benefit of this method is that we force high quality software resampling (by SoX) and prevent low quality audio driver resampling. If you set 44.1kHz or 48kHz sample rate you`ll have to get on with audio driver resampling to native 96kHz (For Z2/Z3).

Remember that Android always try to choose sample rate (for OpenSL ES Output) divisible by native sample rate. For example, in the case of Z2: 48kHz x 2 = 96 kHz (Native sample rate).

See you later!

p.s. Sorry for my bad English and if you want to correct my mistakes - feel free.