July 25th, 2017, 10:49 am
DaveR wrote:Cogito wrote:DaveR wrote: If the PC couldn't keep up with the DAC you'd know it.
For example, in redbood, amplitude information has to reach the dac every 0.0226 milli-seconds precisely. About half that time for 24x96kHz PCM. If the info doesnt reach the DAC precisely in time, the resultant analog wave will be slightly different than what is represented in the digital data.
USB audio, as with all USB data, is sent in packets - it is not synchronous. The requirement for precise timing on the source end is limited to ensuring the packets are sent in time for their arrival at the DAC before the DAC needs the data. You can dedicate all resources on the PC to delivering the audio data if you like, but there is no benefit if the PC can do whatever else it is doing and still deliver the audio data to the USB port in a timely manner.
July 25th, 2017, 11:21 am
July 25th, 2017, 11:31 am
July 25th, 2017, 11:39 am
Cogito wrote:If it doesn't work, I can bring over the AK4497 DAC one day. Some CDs have harshness in the upper treble. With this DAC harshness is gone, with more details. Now, I am able enjoy "non-audiophile" sources.
July 25th, 2017, 12:06 pm
July 25th, 2017, 1:58 pm
Grover Gardner wrote:Cogito wrote:If it doesn't work, I can bring over the AK4497 DAC one day. Some CDs have harshness in the upper treble. With this DAC harshness is gone, with more details. Now, I am able enjoy "non-audiophile" sources.
I'd love to hear it but it's a long drive to Oregon. I'll take a look at mpd.conf. In V3 apparently there was a default setting for Cantata. I'm not seeing that in this version. Mpd.conf looks a little sparse right now, with a lot of missing instructions. If you get a minute, could send a screen shot of your mpd.conf? Maybe I've misconfigured mine. I also cannot get my network to see the machine with AP-Linux. Have you solved that?
It's certainly very different from Snakeoil. I don't mind tinkering with it a bit.
For wireless or manual IP setup, open terminal, become root and go to /etc/apl-files directory. There you will find “enable-wless.sh” script. Execute this script with ./enable-wless.sh and reboot. After reboot you will find one small icon on the right in the system tray that you can “right click” and edit your network settings.
July 25th, 2017, 2:46 pm
HAL wrote:Also an async USB DAC interface does the data transfer timing with it's internal clock, not the PC clock.
July 25th, 2017, 2:52 pm
Cogito wrote:Grover Gardner wrote:Cogito wrote:If it doesn't work, I can bring over the AK4497 DAC one day. Some CDs have harshness in the upper treble. With this DAC harshness is gone, with more details. Now, I am able enjoy "non-audiophile" sources.
I'd love to hear it but it's a long drive to Oregon. I'll take a look at mpd.conf. In V3 apparently there was a default setting for Cantata. I'm not seeing that in this version. Mpd.conf looks a little sparse right now, with a lot of missing instructions. If you get a minute, could send a screen shot of your mpd.conf? Maybe I've misconfigured mine. I also cannot get my network to see the machine with AP-Linux. Have you solved that?
It's certainly very different from Snakeoil. I don't mind tinkering with it a bit.
Didnt know you moved.
I am using wired network, it configured itself as DHCP without any issues. From AP-L Instructions:For wireless or manual IP setup, open terminal, become root and go to /etc/apl-files directory. There you will find “enable-wless.sh” script. Execute this script with ./enable-wless.sh and reboot. After reboot you will find one small icon on the right in the system tray that you can “right click” and edit your network settings.
Using the default MPD config as of now. Only change I made are the the Music Directory and USB Destination. I will send it tonight.
That reminds me, last night I was listening Sinatra. Observed his voice sounded more natural, more intimate. The mid-range bloating and huskiness in the voice were reduced. I think I like it, but need a lot more listening to be sure. I will try out format 32 instruction.
July 25th, 2017, 4:40 pm
Cogito wrote:HAL wrote:Also an async USB DAC interface does the data transfer timing with it's internal clock, not the PC clock.
True, but that comes into picture only after the CPU and OS allocated time slot of the music server process, and and the music server tries to send a small chunk of data to the DAC. I do not know the size of streaming chunks used in ASIO. This is the issue I am trying to emphasize. NO matter how fast the USB is, CPU/OS and other processes running in the system will mess up the timing of the data stream.
For example if I am playing a 300MB DSD128 file which lasts 6 minutes, it could be sent to DAC is thousands of small data streams.
Further more, there should not be any other device on the USB for the controller to respond immediately (key word "serial"). If a mouse, keyboard, hard disk are working on the USB bus, the USB controller will accommodate each device by allocating the time slices.
That is why jPlay takes over the entire server disabling all unnecessary processes and keyboard, mouse etc when playing the music to dedicate max possible resources to itself.
July 25th, 2017, 4:54 pm