Best bang for my buck

[doolhoofd]

@Thelucifer1978:

First of all, I was unfamiliar with MusicBee, so thanks for introducing me to it.

I’m guessing that any bottlenecking from 32 bit to 24 bit will be done automatically either way.
I would recommend you read the last post in the thread I linked, the technical data shared by FxSound’s lead engineer.
I’ve also copy-pasted that full post below - it’s not the first blockquote, but the second, long one.
Because things are actually a bit (pun not intended) more complicated than I thought.
Also keep in mind that these results depend on your audio output devices, your speakers or headphones, and their frequency settings and capabilities.

@bvijay

24 bit 48 kHz is what the FxSound device supports, and the application renders audio in this format. After processing this audio stream from the application, the output will be in the format that the FxSound DSP and the output device supports. That is, for example, a 48 kHz audio can be processed and rendered as 192 kHz which the output device supports.

Let’s say if the content itself is a Hi-Res audio content of 24 bits 192 kHz sampling rate, and the hardware supports this format, then it can be directly played without FxSound.

Here are some example scenarios where FxSound comes into the picture:

1. Content is of lower quality, let’s say 16 bit 44.1 kHz, and OS does the dithering to stream it to FxSound at 24 bit 48 kHz. FxSound takes care of any loss in audio quality in the dithering process done by OS, boosts volume, etc.

2. Content is 16 bit 44.1 kHz or 24 bit 48 kHz, and the output device requires 32 bit 96 kHz. FxSound takes care of improving the audio to meet the output device requirement.

My apologies if all of this sounds confusing.
Let me be perfectly clear: this is wizardry, and I do not fully understand it myself. I am only resharing what I was told.
In the end, I would advise you to try different settings, and stick to the ones that work best, for you.
MusicBee also has its own forum, with a subsection for questions.

Questions

@bvijay

I went through the driver code and DSP code and compiled the audio formats supported by FxSound. Here are the details,

FxSound Audio Enhancer is a virtual audio playback device and applications stream audio to this device in the formats supported by this device. FxSound DSP which is part of the application processes the audio and outputs it to the audio output device in the format supported by the device.

FxSound supported audio format is, PCM audio, 2 to 8 channels, 24 bit audio with sampling rates 44.1 kHz or 48 kHz. This is the format that we see in the FxSound Audio Enhancer properties.

The processed audio output can only be in the formats that FxSound DSP supports. So, if the audio output device is configured in a format that FxSound cannot support, the device cannot be listed or the audio cannot be output to the device. The output format supported is PCM, 2 to 8 channels, 16, 24, or 32 bits, 16 kHz to 192 kHz sampling rate.

This means that 24 bit audio can be sampled to 16 bit, 24 bit or 32 bit depending on the format supported by the output device and in the range or 16 kHz to 192 kHz sampling rate.

Also note that only PCM format is supported so, if the audio output device expects Dolby Digital or DTS encoded audio, then we cannot support that.

There are two types of multi-channel audio like 6 (5.1) or 8 (7.1).

1 . Multi-channel audio encoded in Dolby Digital or DTS formats and transmitted over S/PDIF or optical digital audio ports. This method is definitely not supported. Because I see that the DSP code does not process audio and create audio output in this format.

2 . Discrete multi-channel audio where the audio card on the PC has 6 or 8 individual analog audio ports and the ports are connected to individual speakers. In this case theoretically FxSound processes and generates multi-channel PCM audio. But I don’t have this type of setup and I have never tested this.

I have created a block diagram to explain the audio flow.