ESI Gigaport HD+ tweaking

Optimize your system for ultimate performance.

Moderators: MattKingUSA, khz

Post Reply
Sopper
Posts: 2
Joined: Sat Dec 22, 2018 6:40 pm

ESI Gigaport HD+ tweaking

Post by Sopper »

Hello,

I run into some undesired behaviour with this usb dac under Linux.
I am running Debian Stretch AMD64 latest kernel with MPD and Upmpdcli
Everything works fine and this combination sounds wonderful.

But, when we look at the specs:

Features
24-bit / 96kHz D/A converter
8 independent output channels with RCA connectors
different playback modes: 44.1kHz with 16-bit and 8 channels, 44.1kHz with 24-bit and 6 channels, 48kHz with 24-bit and 6 channels, 96kHz with 24-bit and 2 channels
2 independent stereo headphone outputs: 1st output sends out mixed signal of all playback channels, 2nd output sends out different signal from playback channel 1/2 - perfect for monitoring and to pre-listen to signals
compatible with Windows XP and Windows Vista/7/8/8.1/10 supporting ASIO 2.0, WDM, MME (using ESI's EWDM driver technology)
compatible with Mac OS X and supported by the native CoreAudio USB audio support from Apple, no special driver installation required
compact and portable design with around 12 cm x 7 cm x 2 cm


It supposed to support 24/96 files when 2 channels used.

Alsa recognises the dac as 8 channel capability and therefor everything I play gets downsampled to 16/44.
Anyone who knows how I can prevent this?, and play 24/96 files with 2 channels?

Thanks for the input, if you need more information, just ask.
Sopper
Posts: 2
Joined: Sat Dec 22, 2018 6:40 pm

Re: ESI Gigaport HD+ tweaking

Post by Sopper »

This is what I found and reflects perfectly what is happening with Alsa and the ESI.
From: http://www.volkerschatz.com/noise/alsa.html#myst1

The other options allow you to fix the sampling rate (-r), the channel count (-c) and the sample format (-f) before the remaining parameter ranges are displayed. The important thing is that it sets these parameters in the same order in which they are given on the command line. This allows you to find out by trial and error if and how parameters depend on each other. For instance consider the following output for the first device of my second sound card (an Echoaudio GINA3G):

vs@schizo, ~/soft/sound-misc > ./alsacap -d hw:1,0
*** Exploring configuration space of device `hw:1,0' for playback ***
1..6 channels
Sampling rate 32000..100000 Hz
Sample formats: U8, S16_LE, S32_LE, S32_BE, S24_3LE
vs@schizo, ~/soft/sound-misc > ./alsacap -d hw:1,0 -c 1
*** Exploring configuration space of device `hw:1,0' for playback ***
Set number of channels to 1.
Parameter ranges remaining after these settings:
1 channel
Sampling rate 32000..100000 Hz
Sample formats: S32_LE, S32_BE
Significant bits: 32
Let's go through the output line by line. The first execution of alsacap displays the total parameter ranges for the given device. It can handle up to six channels with sampling rates between 32 and 100 kHz and in a number of formats. The second time, alsacap is called with the -c option to set the number of channels to mono. As expected, the only possible number of channels left is one. But surprise, surprise: The available sample formats have also changed, leaving only two 32-bit formats (little- and big-endian).

This is an example of the allowed range of one parameter (the sample format) depending on the choice of another (the number of channels). The last line in the example shows an additional feature of alsacap: If the sample format has been narrowed down to one bit width, the number of significant bits is printed. Some sound cards accept, say, 32 bit sample values, but discard some least significant bits. But all these values only represent the world according to ALSA. My sound card documentation merely mentions 24 bits, as well as a minimum sample rate of 25 kHz (continuously variable mode) and 8 kHz (preset). So the ALSA driver does not include all the card's features. On the other hand, who needs 8 kHz surround sound?

Post Reply