Hello everyone, I recently discovered this format and that there is the possibility of drastically reducing the space of a file dsd (dsf or dff) if placed in the "container" wavpack ..... I used this guide [Moderator edit: Link removed ] and I already had several of my albums from dsf to wv, as I have been using Ubuntu for some time I wanted to know if it was possible to do the same thing using Ubuntu and not Windows ....... Another question, I noticed that after I got the wv file the tags are not transferred, is this normal?
Thank you
wavpack is cross platform and free software, so why not
Thanks for the answer but how do I use it with Ubuntu, what I would like to do is turn my dsf into wv, I can only do it on windows, with Ubuntu how do I do it?
Thank you
The batch processor you used on Windows is not available on Linux, so you will probably have to use the command line. I assume that if you like Ubuntu then you must have some familiarity with it.
But assuming you have wavpack 5.1.0 installed on your Linux system, the command to convert a dsf to a wv, in high mode and copying tags, would be something like:
wavpack -h --import-id3 "01 Because.dsf"
Also note that on Linux the music server MPD handles WavPack DSD file natively, and JRiver's Media Center is also available on Linux (although I have not tried that).
Good luck!
Thank you very much for the answer, batchencoder is very simple to use and unfortunately works only with windows, install wavpack 5.1 on Ubuntu as it does? I simply downloaded the zip and extracted the folder, as you say then I have to convert one file at a time, with batchencoder I can upload all the files I want and then I start and do it all by myself ..... Too bad there is not something similar for Ubuntu ..... unfortunately I will have to continue to go on windows ..... With Batchencoder, however, I can not understand how to carry the tags on the wv that I get, they all come without tags ....
Thank you
You actually can do multiple files with the command-line program. Doing a whole folder is as easy as:
wavpack -h --import-id3 *.dsf
As for the batch utility, I don't see any reason that you couldn't add the
--import-id3 option to the list of options. If the DSF file has an ID3v2 tag then it will be imported into the WV file.
I'm sorry but I can not find the option to import tags with batchencoder ....
A question about the file obtained, I saw that mediainfo sees it as a pcm 88.2, by chance the files are reconverted to pcm? If I use the highest compression mode (so get the smallest file) my player will make more effort to read it? Use Foobar 1.4 on a fifth generation Intel Nuc i3 .....
Thank you
In the web page you originally linked to there is a section about adding options to the presets. In that case they talked about adding the option to automatically delete the source file (-d) but what you need to do is add the option to import ID3 tags from the source file (--import-id3). In other words, you have to add that to the list of options that batch encoder passes to wavpack. Note that there are
two dashes before the “import-id3” and there must be a space between the options, but they can be in any order. So, for high mode and importing tags the box would be:
-h --import-id3
It may be that mediainfo has not been updated to properly handle the new DSD files of WavPack, which is why it shows what it does. Don't worry...if you used batch encoder (or the command-line in Ubuntu) to convert from a DSF to WV, then it will be DSD. If you use some other program like Foobar or Jriver Media Center then you
will get a PCM file, so don't do that.
Yes, the highest compression will use more CPU in exchange for a smaller file. However, even in the “high” mode it still decodes fairly easily on any modern CPU. If you are concerned you can use Task Manager on Windows (or System Monitor on Ubuntu) to find out the CPU usage during playback.
I put that option but from error to the end and then the wv file is not created .....
(http://i67.tinypic.com/33z9ro8.jpg)
I finally managed to use my Ubuntu distribution to create the .wv files but also here by mistake on the tag and consequently delete the created file, the error is this "ID3v2 import: not valid ID3v2.3"
So I have to use the "wavpack -h * .dsf" command so I can do the whole folder but without importing the tags ...... it's okay, but of course if you care, I'd like it ....
the command -h indicates that I want to use the higher compression, right? If I put -f the slowest and for the normal one I do not have to put anything? That is, the command would be "wavpack * .dsf"?
At the moment I have the problem on Ubuntu that I can not find a player to play these wv, vlc goes to play but you can not hear anything, DeadBeef does not accept them ..... which player can I use to play them? I installed MPD but I do not know how use it ....
Thank you
Ah, so that explains why you were getting the error with the batch encoder...there is a problem with the tag format on the DSF file. The most common reason I've seen for this is that the tag is a ID3v2.4 tag, whereas WavPack currently only accepts ID3v2.3 (which is what the DSF spec calls out, I believe). If you could find some way of providing me the file (short and compressed with WavPack is okay because even without importing the tag it is still stored in the WV file) that would help me identify (and possibly fix) the problem. If you have another way of checking the tags that might tell you if it's v2.4.
MPD requires another program to act as a "front end" to control it and I have never tried to make it work either, although I'm sure sure there are guides online. You could try JRiver Media Center or Foobar under Wine. I see that the Deadbeef developer is aware of a problem with WavPack DSD files with Linux, but I'm not sure where he is with that:
https://github.com/DeaDBeeF-Player/deadbeef/issues/1973 (https://github.com/DeaDBeeF-Player/deadbeef/issues/1973)
Of course, you should be able to encode the tracks on Linux and transfer them to Windows to play them, although until the tags are importable it doesn't make much sense.
As for the compression levels, for DSD there are only two. If you don't specify anything you get the "normal" mode which is very easy on CPU for encode and decode. The only other mode is "high" which is turned on by -h and gives about 10% more compression and uses a lot more CPU, but still not that much. For example, even the high mode easily decodes on a Raspberry Pi or my old Android phone (using Foobar).
Thank you so much for your answer ..... I try to answer everything and I apologize if I use a translator but unfortunately I do not know English ....
The program I use to tag is called Kid3 https://kid3.sourceforge.io/ and is present in the repositories of Ubuntu, it also works with Windows but is not as stable as on Ubuntu, in my opinion it is an excellent program for tags, although undoubtedly it would be better that they were passed automatically on the file wv with this program I put a moment to put the tags on the wv thanks to the copy / paste, I confirm that it is the version ID3v2.4.0, so if you want to do some tests just use this program to tag ......
About MPD I installed it on my Ubuntu and I put two programs that rely on him (Sonata and Cantata) but with neither I can load the library, I do not know if I have to go into the settings of mpd via terminal to do so .. ...
DeaDBeef is just the program that I use for Ubuntu (Foobar I use it with Windows 10 in my Hi-Fi system) and the wv files obtained by the dsf does not load them (I play DeaDBeef, the program opens but does not load any file wv) while instead the dsf / dff reads them without problems .....With VLC instead the files load them but they do not feel ....
Take advantage to ask another question, you can return to the original dsf file from the wv file I got, if you can do it?
Thank you
I update the situation regarding the import of the tags, if the dsf is tagged with version 2.3.0 the tags are imported into the wv file, so the problem is only for version 2.4.0 tags.
The problem I think is only mine so if you can implement the import of version 2.4.0 well, otherwise it does not do anything ..... All my files I tagged them with this program (Kid3) that at the beginning are in version 2.3, when they are then imported to Kid3 and I make changes are saved in version 2.4, that's why I find myself all my dsf files in version 2.4
Thank you
Thanks for letting me know about the tagging issue. I would think that a good tagger program would allow you to use v2.3 tags because those are the most popular and most supported in software and hadware. I have been working a little on v2.2 because I found some old files with them, but I wasn't going to bother with v2.4.
In answer to your last question, yes, you can always restore the original exact DSF file from the WV file. This works even of the tags are not imported...all the original file is stored in the WavPack file. And of course with DSD files WavPack is always lossless.
What is the procedure for returning the wv file to dsd / dff? It's just a curiosity ..... Thanks
What is the procedure for returning the wv file to dsd / dff? It's just a curiosity ..... Thanks
From the command line (Windows or Linux)
wvunpack filename.wv
A batch encoder that uses the command line program should also work. The unpacked file will have the original file's extension (DSF, or WAV, or whatever).
Can I also convert flac files to wv? I tried with the
wavpack * .flac
command, but an error comes out ....
No, flac files are not supported for input. The file types allowed are (from wavpack help):
Formats: .wav (default, bwf/rf64 okay) .wv (transcode, with tags)
.w64 (Sony Wave64) .caf (Core Audio Format)
.dff (Philips DSDIFF) .dsf (Sony DSD stream)
If you want to transcode from flac to wv, Foobar would work perfectly, and would transfer the tags too.
Perfect, thanks ..... in fact I do not need to convert the flac in wv, I tried on windows with batchencoder and the flac files are not reduced as the dsf / dff, I guess because the flac is already a compressed format, but slowly plan my dsd I'm converting to wv because having 40% less occupied space is a great thing, congratulations to you that you have made this program .....
I resume the tread to ask something, with Linux now I create the wv from dsf without problems, but I do not have a player with which to hear them, those based on mpd do not work, deadbeef does not accept them, vlc accepts them but indicates a wrong time and not feel, how can I do to listen to them with Ubuntu?
Thank you
I've never tried MPD myself, but I know that it has full native support for WavPack DSD. You need to choose a client to use it, as described here (https://www.musicpd.org/clients/), and I'm sure you can find help if you run into trouble. I would be interested in hearing about your results.
As I mentioned above, JRiver Media Center has a Linux version, and that handles WavPack DSD natively too. It's not free, but it's not expensive and the free trial is pretty generous, and they have a forum for helping out also.
I think that many people on this forum run Foobar2000 under wine, so that's another possibility. I've never tried that, but Foobar2000 on Windows handles WavPack DSD great.
Requesting user to run Foobar2000 under wine is very nice for open source community.
Right, so ask the Open Source community to update their heccing WavPack dependencies, this has been a solved problem for ages now.
Foobar on windows 10 I use it and that's fine, on Ubuntu unfortunately it does not work anymore, I do not know why but it does not open anymore, I'll try to ask in the Ubuntu community, as regards MPD based programs I can not make them work I tried them all but all have problems, but VLC, DeadBeef and Audacity support the wv with pcm very well but not those with dsd ......
Thank you
I finally managed to use my Ubuntu distribution to create the .wv files but also here by mistake on the tag and consequently delete the created file, the error is this "ID3v2 import: not valid ID3v2.3"
So I have to use the "wavpack -h * .dsf" command so I can do the whole folder but without importing the tags ...... it's okay, but of course if you care, I'd like it ....
the command -h indicates that I want to use the higher compression, right? If I put -f the slowest and for the normal one I do not have to put anything? That is, the command would be "wavpack * .dsf"?
At the moment I have the problem on Ubuntu that I can not find a player to play these wv, vlc goes to play but you can not hear anything, DeadBeef does not accept them ..... which player can I use to play them? I installed MPD but I do not know how use it ....
Thank you
What version of ubuntu do you use? If you use 18.04 you should not have big problems.
MPD must read DSD (in Wavpack) from version 0.20 onwards. Ubuntu 18.04 incorporates version 0.20.18
The issue is that MPD uses a client-server architecture. MPD is the daemon-server (the player itself), and you must install a client (graphical interface for mpd); the client can be on the same machine, on another, on a smartphone, tablet, etc.
You must install mpd and connect the client (one or more) to the server. It's relatively complex, but not so much.
A guide (maybe outdated? I use ubuntu 16.04 yet) http://gmpc.wikia.com/wiki/MPD_INSTALL_USER_SERVICE_UBUNTU
Here they install mpd + gmpc (client). If it succeeds, then you can also install another or several clients to control the player.
Sorry for my english
I use Ubuntu 18.04 LTS, I installed mpd and all the possible clients, I can not load the files to let them read, my music files are all on Nas, however I use Foobar on Windows to listen to my music and wv / dsd read them without problems, in the PC that I use routinely instead I have Ubuntu, fortunately I managed to restart Foobar and even if it is limited to Ubuntu now I can read the wv / dsd files even from Ubuntu ...... Too bad they are not supported by VLC , BeadBeef and Audacious which are my favorite players on Ubuntu ..... I apologize for my english too ....
Thank you
Lack of support from VLC/DeaDBeeF/Audacious would be due to using an outdated version of the WavPack library. And don't expect a newer version to be included with 18.04, either, since it's an LTS release, so it will feature freeze for the next two years. Unless maybe you install an updated WavPack from a PPA, and then you're on your own, and it will still only decode to 24 bit PCM instead of DSD.
The wavpack I have in my Ubuntu 18.04 LTS is version 5.1.0 which is the last version ......
You seem to have a configuration problem in MPD.
There does not seem to be a compatibility problem with wavpack-DSD (yet).
If you can not access your music library, there is probably something wrong with your 'mpd.conf' (mounts the file system, client-server access, permissions, etc.). In the worst case it can be an MPD bug when accessing the remote system.
I'm not a Linux expert... it would not be a good idea to do this query (MPD can not access NAS) in Ubuntuforums or Linuxquestions?
I have DFF files compressed in DST, as the dff does not allow me to use the tags I tried to compress them in wavpack but tells me that the file is already compressed and then stops, I wanted to know if there is a way to put these dff in a wv container so that I can tag them.
Thank you
You should decode DST to DFF first. You need DST decoder - http://cl.ly/0V2O062f3q1m Archive contains coder and decoder for Windows.
Decoder's file is RefDstDecoder.exe in folder C043465e_Electronic_inserts\audio\lossless\mpeg4dst\RefDecoder\export\
How do I use the program? I clicked on RefDstDecoder.exe but it does not open, I see a window but it closes immediately ......
This is command line tool... Like wavpack.exe
RefDstDecoder.exe -i YourCompressedInputFile.dff -o DecompressedFile.dff
And be patient. It is VERY slow.
Excuse me but I can not understand yet how to do it, I put all the dff files in the same folder where there is RefDstDecoder.exe but as I click the window opens for a split second and closes, how should I do? a thing that I've never done and if you please explain step by step, by chance I could do it with Ubuntu using the terminal?
Thank you
Put RefDstDecoder.exe and attached bat file into folder with compressed dff files, run bat file, wait.
Thank you so much, now it seems that it works but it seems very slow, I think I will give up and I will transform these dff in pcm, there is not even a bar or numbers that make you realize where the encoding passes, there is only to wait. .....
Thanks again for your availability
Um. Of course it's slow. And converting them to PCM will not speed it up. You can, however, parallelize it so it's using one conversion per CPU core. The only reason foo_input_sacd is reasonably fast is because it uses multiple cores to execute the DST decompression, so a single file plays back fast enough, but multiple files won't play any faster.
In the end I converted them (it took a long time) and then I recompressed them in wavpack via terminal with Ubuntu and now everything is ok, dsd with tags .......
I have DFF files compressed in DST, as the dff does not allow me to use the tags I tried to compress them in wavpack but tells me that the file is already compressed and then stops, I wanted to know if there is a way to put these dff in a wv container so that I can tag them.
Thank you
Although you have already solved the problem you can also unzip dst to dff natively in Linux with dstcnv (https://github.com/ntaka777/dstcnv). Also on the command line, just as slow, but in this case it reports the progress and does not need Wine on Linux
I use Ubuntu 18.04 LTS, I installed mpd and all the possible clients, I can not load the files to let them read, my music files are all on Nas, however I use Foobar on Windows to listen to my music and wv / dsd read them without problems, in the PC that I use routinely instead I have Ubuntu, fortunately I managed to restart Foobar and even if it is limited to Ubuntu now I can read the wv / dsd files even from Ubuntu ...... Too bad they are not supported by VLC , BeadBeef and Audacious which are my favorite players on Ubuntu ..... I apologize for my english too ....
Thank you
The latest version of DeaDBeef works with wavpack-dsd!
Tested in Debian 10 (it was updated a few weeks ago) and the binaries already exist in the DeaDBeef page
http://deadbeef.sourceforge.net/download.html
Remenor, I had no idea DeadBeeF was out with a new version so thank you for letting me know!
Remenor, thanks for posting about this! I have updated by site with the new information...it's great to see more apps supporting the DSD mode.
One of my pending projects is to port the DSD code into FFmpeg, which should help even more.
Hi Bryant, I take this post again to ask you some things ..... When I convert my dsf to Ubuntu, I can use the -he -hmv option where the checksum is done, is that something you need Because I saw that it takes a lot more time with the -hmv option, among other things I noticed that in the foobar info the checksum of the flac files is displayed while this is not ....
Thank you
That command (-hmv) takes a long time because it's high mode (-h) and verify mode (-v) which means that the output file is decoded from scratch to make sure it's correct, so that takes almost twice as long. I think it's a good idea, and I always do the verify step, but you could skip it. Of course, the "high" mode make it much slower too, but the compression in the regular mode is not as good.
The -m (to store the MD5) adds very little time, and in fact adds no time when combined with the -v option because I use MD5 to do the verify and so have to calculate it anyway (storing it takes no time).
I did check and Foobar2000 does not display the MD5 information for WavPack DSD files, but it does show it for PCM files. That's probably just an oversight because the API works the same for both. Or maybe Foobar2000 will only display the MD5 sums for files that it can verify, and it can't verify DSD files because they get converted to PCM.
Thanks for the reply ..... exactly, Foobar displays MD5 only on wv / pcm not on dsd ...... Could I use only -hm to save time and have the checksum?
I also made many files only with the -h option, so without the checksum, could I have problems? I hope not .....
Yes, you can use -hm to save time by skipping the verify. You can always verify later with wvunpack -v option.
There's no problem not adding the MD5 sum. The files can be verified without it, although in that case we just use the individual block checksums. If the MD5s are present we use that too at the end, so it's sort of a redundant check.
A lot of people use the MD5 as a unique identifier rather than error detection, but you absolutely don't need it.
I apologize for another question ..... I have noticed that using the --import-id3 command does not import all the tags, it would not be a problem because then I reissued everything with Kid3 https://kid3.sourceforge.io / but Foobar also reads the tags of the old dsf file, for example if in the dsf tag there was "editor" I delete it via Kid3 Foobar that tag keeps reading it even though I removed it ....... is it possible to "clean up" the .wv file from all the tags via a terminal command so I can check if Foobar still sees tags?
Thank you
but Foobar also reads the tags of the old dsf file, for example if in the dsf tag there was "editor" I delete it via Kid3 Foobar that tag keeps reading it even though I removed it
Have you tried to reload tags in foobar? Right click->Tagging->Reload tags from files(s)
Also, have you foo_input_sacd installed? There was reported problems with tags of wavpackdsd if foo_input_sacd is installed.
https://sourceforge.net/p/sacddecoder/bugs/108/
https://sourceforge.net/p/sacddecoder/bugs/109/
foo_input_sacd I have to keep it installed otherwise Foobar does not read the dsd files, whether they are wv or dsf / dff
You are aware of that problem but I don't think it was totally solved ...... right clicking on the file to update tagging is a problem for me as I totally control Foobar through an app for Android and I don't use monitors ..... in the app that function is not there. Anyway it is a problem that I have only with wv / dsd files ...
Thank you
foo_input_sacd I have to keep it installed otherwise Foobar does not read the dsd files, whether they are wv or dsf / dff
Actually, foobar2000 does read and play WavPackDSD files without any additional components. And it can play dff/dsf files with foo_input_ffmpeg. So foo_input_sacd is not mandatory unless you need to output DSD directly without decoding to PCM.
Thank you very much for this info, I didn't know of the existence of foo_input_ffmpeg, I will definitely try it ...... I ask you some things if I don't disturb you, my dac (Chord Qutest) accepts input dsd files so I like to play the files dsd native without converting them to pcm, that's why I use foo_input_sacd ..... but I ask you, can the foo_input_ffmpeg plugin install the same right? So I do some tests, in the case I install foo_input_ffmpeg what should I put as an output module? I have [DSD] Asio - Chord ...... with the foo_input_ffmpeg plugin everything will be converted back to PCM? Also the DSF files? On Ubuntu I use DeadBeef and the wv / dsd-64 files reproduce them in PCM 352.8, if instead I have the dsd128 / 256 how are they played? My dac in pcm reaches 384 ......
Thank you
When playing DSD with foo_input_ffmpeg, it will be converted to PCM.
As for other part of you question, i don't completely understand. But if you can successfully play regular PCM files (flac, mp3 and so on) with you current setup, you don't need to change output settings to use foo_input_ffmpeg.
As for samplerate of output PCM when decoding with foo_input_ffmpeg, it can be adjusted
(https://i.imgur.com/nmhna0N.jpg)
Also, notice that you will need ffmpeg itself ( https://ffmpeg.zeranoe.com/builds/ ) for foo_input_ffmpeg to work
Yes, the PCM files reproduce them quietly, so useless I put this component ..... I saw in the list of installed components that Foobar already has ffmpeg by default, what changes with the component foo_input_ffmpeg that you suggested?
Thank you
I saw in the list of installed components that Foobar already has ffmpeg by default
Built-in ffmpeg is used only for decoding of MP3, AAC, Vorbis.
Thanks again, to use this component, do I need to have ffmpeg installed on windows 10? How do I install it or do I just need to put the component on Foobar?
Thank you
AFAIK, external ffmpeg will only decode stereo WAVPACK/DSD at the moment. I think
@bryant has a patch in the queue for multichannel files to be decoded by ffmpeg but so far it hasn't been voted in.
What should I select for wv files?
(https://i.postimg.cc/k6fZRsWg/Schermata-del-2019-08-25-20-45-44.png) (https://postimg.cc/k6fZRsWg)
AFAIK, external ffmpeg will only decode stereo WAVPACK/DSD at the moment
ffmpeg doesn't decode WavPackDSD at all (yet). Tested on build from 2019.08.22. You can test it too. As for DSD in dsf/dff, ffmpeg can decode all variants.
Thanks again, to use this component, do I need to have ffmpeg installed on windows 10? How do I install it or do I just need to put the component on Foobar?
Yes, you need to download ffmpeg - https://ffmpeg.zeranoe.com/builds/ No installation needed. Just unpack archive and point FFmpeg Decpder Wrapper (in its settings) to folder containing ffmpeg.exe and ffpropbe.exe.
What should I select for wv files?
One more time: to play WavPackDSD in foobar2000 you do not need additional components. ffmpeg is needed to play dsf/dff without foo_input_sacd
You need to point FFmpeg Decpder Wrapper to folder containing ffmpeg.exe and ffpropbe.exe. And create new profile for dsd decoding.
(https://i.imgur.com/CpUPA3O.jpg)
(https://i.imgur.com/WbAVGDX.jpg)
One more time: to play WavPackDSD in foobar2000 you do not need additional components.
Then surely I'm wrong something because I can't reproduce wv / dsd in any way except with the foo_input_sacd component and in the decoding window sacd decoder must be on top of wavpack, in this way I reproduce them, in any other way Foobar makes me wrong .....
One more time: to play WavPackDSD in foobar2000 you do not need additional components.
Then surely I'm wrong something because I can't reproduce wv / dsd in any way except with the foo_input_sacd component and in the decoding window sacd decoder must be on top of wavpack, in this way I reproduce them, in any other way Foobar makes me wrong .....
With foo_input_sacd you can have some problems. When you select a wavpack-dsd file from your media library, what does the codec field say: DSD64 or WavPack?
DSD tells me, also because I have no other way to reproduce them, in any other way Foobar gives me an error, I have to use an output module [DSD] Asio or Wasapi, then forcefully use foo_input_sacd, if I put any other output module it gives me error ...... so I don't reproduce them natively, I have to have foo_input_sacd ..... I would like to try to see if there is a difference between the foo_input_sacd component and the reproduction in Wavpack but I can't. ...
Thank you
Try moving the wavpack decoder above the sacd decoder. Then drag and drop a wavpack-dsd file to a playlist and try playing it back... the output module should be just a normal one (for pcm).
If this works you will need to delete foobar2000 media library and rescan your files. If this doesn't work I don't know what the problem is.
AFAIK, external ffmpeg will only decode stereo WAVPACK/DSD at the moment
ffmpeg doesn't decode WavPackDSD at all (yet). Tested on build from 2019.08.22. You can test it too. As for DSD in dsf/dff, ffmpeg can decode all variants.
FWIW, the patch to allow FFmpeg to decode WavPack DSD is here (https://github.com/dbry/FFmpeg/tree/wavpack_dsd) and still merges and probably still works. It handles both stereo and multichannel, but whether it would work with FFmpeg is not clear.
The reason it was rejected is that it creates a new codec ID for WavPack DSD, which seems to be against the rules in some circumstances (no explanation given). I hope to create a new patch in the next month or two.
Try moving the wavpack decoder above the sacd decoder. Then drag and drop a wavpack-dsd file to a playlist and try playing it back... the output module should be just a normal one (for pcm).
If this works you will need to delete foobar2000 media library and rescan your files. If this doesn't work I don't know what the problem is.
Today I started to do some tests, I had a PC with windows 10 just installed and I put fb2k 1.4.6, actually reads the wv / dsd files but I discovered that a dsd64 sends it as pcm 352.8 while instead I made a try with a dsd256 and fb2k goes wrong, then I went into the windows audio settings and noticed that if I set as output over 192 on the dsd128 / 256 foobar it goes in error, so paradoxically a dsd64 can reproduce it in pcm 352.8 while a dsd128 / 256 no more than pcm 192 ...... with dsd64 I can use wasapi as an output module while with 128/256 no, I can only use ds.
The test was done with windows just installed and with the only wasapi component on fb2k, I didn't put anything else, not even the drivers of the dac (chord qutest) that windows recognizes ......
In the other PC where I have 1.5 b11 and various components installed including the sacd decoder I have no other way if I don't play wv / dsd as native dsd, in any other way fb2k gives me error ......
Today I started to do some tests, I had a PC with windows 10 just installed and I put fb2k 1.4.6, actually reads the wv / dsd files but I discovered that a dsd64 sends it as pcm 352.8 while instead I made a try with a dsd256 and fb2k goes wrong, then I went into the windows audio settings and noticed that if I set as output over 192 on the dsd128 / 256 foobar it goes in error, so paradoxically a dsd64 can reproduce it in pcm 352.8 while a dsd128 / 256 no more than pcm 192 ...... with dsd64 I can use wasapi as an output module while with 128/256 no, I can only use ds.
WavPackDSD decoder in foobar2000 outputs PCM with samplerate equal to samplerate of DSD divided by 8. So, for DSD64 you will get PCM with samplerate 352800, for DSD128 - 705600, for DSD256 - 1411200. If you use WASAPI exclusive output and your sound card doesn't support such high sampleraters, you need to use resampler in foobar2000's DSP. As for DS output - resampling is done automatically, when needed.
Thanks for the answer ... In fact it is difficult to find a dac on the market that goes beyond 384 in pcm ... On fb2k where should I act to activate the component to reduce the frequencies? Then will it resample me all over? I would like to use wasapi as an output module and resample only the wv / dsd 128/256 because they go beyond the frequency of my dac .... As far as DSF files are concerned, should I install the ffmpeg component right?I would like to do some tests without installing the foo_input_sacd component and see how it goes .... I would practically send everything in pcm ...
and resample only the wv / dsd 128/256
You can do this with foo_dsp_resampler_mod2 - https://hydrogenaud.io/index.php/topic,67373.0.html
@Bryant, for the first time it happened to me on two files that the md5 verification codes do not match, in fact the program signaled it to me, I had to use the -hm option without the v otherwise the .wv file was not created for me , practically what have these files wrong? Can I use them or are they not reproducible?
Thank you
The only time I've seen this is with .dsf files that are invalid because they don't have zeros at the end of the final, incomplete block. The DSF spec says they are supposed to, but a few don't. Since WavPack does not store that final bit of data, the MD5 ends up not matching even though all the actual audio is correct.
Assuming that your source file was a .dsf file, everything should be fine. I would recommend not using the -v option (obviously, because it won't work) and also not using the -m option because then a "verify" operation in the future will still fail.
Yes, two .dsf files are correct but this is the first time something like this has happened to me ...... I code my files with this command "wavpack -hmv --import-id3 * .dsf (not always amount the tags because I often have problems with the sacd encoder so I prefer to put them after), so you advise me not to use the mv option anymore and then to use only -h?
Thank you
I still believe that the -hmv is the way to go.
But for those few invalid files that fail I would recommend skipping the -m and -v.
I have two dsf files that give me this error, what could it be? is not a valid .DSF file (by total size)!
Now I'm a little confused. In your post above you said that you had two .dsf files that would not match MD5 and you have to skip -v but otherwise they worked. I explained that this was because bad padding made the files invalid, but they were otherwise okay.
Now you mention the error is not a valid .DSF file (by total size)!. Are you actually getting that error? If so, then that is something different and I have never seen a file like that, and I don't understand how you were able to encode it.
These have nothing to do with the files that gave the md5 error, there are two other dsf files that wavpack does not encode by returning this error, it is the first time I see such an error .....
Thank you
Ah, that makes sense then! I have never seen a file with that error.
Did it come like that or did you edit it or create it somehow? Would it be possible for me to get a copy of one of the files?
Thanks!
Unfortunately now I threw them away, if I recapture that kind of error I see you get them .....