Jump to content

Snarii

  • Posts

    2
  • Joined

  • Last visited

  • Country

    France

Retained

  • Member Title
    Newbie
  1. Brilliant, thanks a lot ! After exploring the different options, I'll be using these ones to generate the DSF files : sacd_extract -2 -s -c -C -P -i <file> -2, --2ch-tracks : Export two channel tracks (default) -s, --output-dsf : output as Sony DSF file -c, --convert-dst : convert DST to DSD -C, --export-cue : Export a CUE Sheet -P, --print : display disc and track information -i, --input[=FILE] : set source and determine if "iso" image, Nethertheless I think some things are lost during the conversion as the the files size doesn't sum up to the size of the ISO. That's a bit scary, embedded things are lost from the ISO, but I don't know what. According to the .dsf file size, the bitrate of the .dsf is OK at 2.8mbit/s on each channel and they play well on A+, so I think that's OK.
  2. Hi guys, Is any of you using a network share to store SACD ISOs for use with Audirvana+ ? Does the playback start immediately for any track or do you have to wait, like me, for many seconds during track bufferization ? On my network, with Audirvana+, whatever the version 1.5 or newer, it's something like 30 seconds to open a track on a directly attached gigabit ethernet local network, and more than 2 minutes to open the same track on a slower wireless network. That's a problem, unless you don't change track too often... Instead of that, a .dsf single track of similar size (say 130MB for example) is opened at full networks speed and starts to play after only a pair of seconds... That's the awaited comportment. So whatever the network speed, gigabit ethernet or slower wifi, the bandwidth seems only used at 1/10 of its nominal network capacity during SACD ISO track bufferization. Not having this problem when listening non-ISO files, I looked a little deeper at the network transport and I've seen 2 different kind of file bufferization, resulting in 2 totally different levels of performance... A bit of network technic now... - While reading a normal file, Audirvana+ manages to read a huge quantity of data at once, and in this way benefits of TCP window scaling, allowing large quantities of data to flow from the server before having to be handshaked (TCP ACK) by the client. - While reading an ISO file, Audirvana+ accesses the file at a 4k byte block rate. Each block received has to be acknowledged by the client to the file server before the next one is requested, and so on... This introduces latencies that drop the network bandwidth to 1/10 of its capabilities. As a consequence, these huge SACD ISOs can almost not be used on network shares, but instead rely only on local -noisy- large disks... Damien and forumers, could you take a look at that and confirm this comportment ?
×
×
  • Create New...