sacd_extract DSF output problem

[Synfreak]

... Audiogate will load some of the dff file folders but not others. In fact most of the folders will not load only the occasional one ...

 

Have you set the "Convert DST to DSD" option?

Your problem might be related to DST compression, which is used by a great number of SACDs - mainly those with multichannel content. If you don't perform DST "unpacking", your DSDIFF files (*.DFF or *.DSF) will end up with DST packed content, and Audiogate can't read those.

 

Report back if this solves your problem!

[thm655321]

I just experimented by converting one of the troublesome ISOs using the convert DST option and it worked.

 

Thank you so much, I would not have got that one on my own.

[Synfreak]

A problem which can be solved, isn't a problem ;-)

 

Cheers

[ted_b]

Synfreak, thanks. You beat me to it.

 

My $.02....DST is not my friend; it reduces storage requirements (storage is cheap) for some discs, but at what cost. It causes havoc with some players, Audiogate, some tagging routines, etc. That is why I always say to check for DST 100% of the time. It will even very occasionally show up on stereo layers, and if not converted/decompressed will result in DSF garbage, for example.

[Synfreak]

Ted,

exactly!

 

Thankfully it doesn't harm to set the "Convert DST to DSD" option for all extracts, just to be shure ...

[kmh]

The best and most convenient way would be fixing the bug in sacd_extract ...

@Ted - didn't you email the author about the bug? (I remember you mentioned that in another thread)

Because I don't find the bug on the issue-list of sacd_extract:

https://code.google.com/p/sacd-ripper/issues/list

[ted_b]

The best and most convenient way would be fixing the bug in sacd_extract ...

@Ted - didn't you email the author about the bug? (I remember you mentioned that in another thread)

Because I don't find the bug on the issue-list of sacd_extract:

https://code.google.com/p/sacd-ripper/issues/list

 

I sent Mr Wicked and a couple of his partners an email, as it has been discussed ad nauseum on the trouble logs. Their contention is that the clicks are, again, dc offset which is then audibilized by switching to PCM (i.e the DAC is doing some PCM manipulation/processing internally...probably more like mutlibit SDM, and the clicks are an audible artifact). Example:

https://code.google.com/p/sacd-ripper/issues/detail?id=40&can=1&q=click

 

What I added to the emailed as the fact that DFF files, which should be susceptible to the same issues, are not..at least via the extractor.

[kmh]

Now I'm a little bit confused ...

The statement of this thread was, that just the extraction to .dsf and not to .dff is affected. In issue #40 you linked above, only the playing of .dff files is discussed. Is it really the same bug? Im sorry for my stupid asking - but I'm a beginner in this matter ...

[Miska]

Their contention is that the clicks are, again, dc offset

 

Ted, their DSF writing code doesn't seem to write the last partially filled sample block, only the full ones. How much song tail goes to the last block of course depends on the exact song length.

[One and a half]

I sent Mr Wicked and a couple of his partners an email, as it has been discussed ad nauseum on the trouble logs. Their contention is that the clicks are, again, dc offset which is then audibilized by switching to PCM (i.e the DAC is doing some PCM manipulation/processing internally...probably more like mutlibit SDM, and the clicks are an audible artifact). Example:

https://code.google.com/p/sacd-ripper/issues/detail?id=40&can=1&q=click

 

What I added to the emailed as the fact that DFF files, which should be susceptible to the same issues, are not..at least via the extractor.

 

For the Playback Designs MPD-3 which converts PCM and DSD64 to DSD128 internally, I don't hear any clicks on sacd_extracted iso DSF. An awkward situation where the software does what it does, but depending on the DAC becomes an audible problem or not.

At the other end of the scale, the Sony PHA-2 portable amp about 5% of the MPD-3's price doesn't notice the missing data either and there's no clicks.

 

Perhaps the workaround is to use Jriver's PCM to DSD on the fly conversion, so the DAC receives DSD all the time. Would you suspect the same would occur if the DAC's DSD acceptance was DoP only?

[Synfreak]

The clicks mentioned here are "only" to be heard on some of the DSF rips.

It depends on how the transition between tracks is laid out.

 

BUT, there IS actually some data missing in DSF extractions (as @Miska has pointed out too).

 

Just one example:

 

 

the first entry was extracted directly to DSF files, whilst the second one shown is the same album, but extracted to DSDIFF (*.dff).

 

Foobar shows a difference of 3 seconds total (rounded), and two of the single tracks also show differences (1 second, also rounded).

 

As soon as an album contains data in the transition between tracks, it is likely that a (very) small portion of it will get chopped of.

And this might/will lead to the problem discussed above - at least here on my setup.

 

This has nothing to do with DSD to PCM conversion or the like.

[kmh]

@Miska, @Synfreak

You put so much effort in this subject and are well versed in this. You should open an new issue on the project-website. I think that would be constructive and perhaps we have luck and get a dsf-bug free version of sacd_extract soon.

[ted_b]

Yes, kmh, sorry for my lack of effort on sacd extraction!

 

Miska, Synfreak..go for it. If I hear back from my emails to Mr. Wicked I will let you know, of course.

[airwise]

I'm new to this game but I recently ripped the SHM SACD of What's Going On - which is supposed to be a gapless album.

 

I get very a worrying thump between some tracks (not a tick), and Pure Music skips the next track on occasion. These are DFF's.

 

On another note, Pure Music is somehow importing the tracks into iTunes in the wrong order - this only with DFF's that I am ripping. Looking at the Extracted folder on the HDD the order is correct. Is anyone else experiencing this?

[Esprit]

Summarizing:

1) which is the "correct" format we have to use: DSF or DFF?

2) what program/software/procedure TODAY is better on Mac OS (ISO --> DSF/DFF)?

Thanks

[Esprit]

Is "version 3.0" the newest Ted's guide?

[Synfreak]

1) which is the "correct" format we have to use: DSF or DFF?

 

According to the "lost bits" when using direct extraction to *dsf files, I would recommend going to *.dff.

 

And - if tagging doesn't work with these - using some software to "repack" these into *dsf files.

 

As I am on Windows, using Foobar, I can use DSDIFF files with tagging.

 

 

 

2) what program/software/procedure TODAY is better on Mac OS (ISO --> DSF/DFF)?

 

I can not really help here, but ...

You surely can use the Sonore GUI, can't you?

[Skeptic]

Summarizing:

1) which is the "correct" format we have to use: DSF or DFF?

2) what program/software/procedure TODAY is better on Mac OS (ISO --> DSF/DFF)?

Thanks

It seems like it would be easier to just keep things as ISO files rather than extracting them to individual tracks, and use a player which supports that. (e.g. JRiver)

[ted_b]

It seems like it would be easier to just keep things as ISO files rather than extracting them to individual tracks, and use a player which supports that. (e.g. JRiver)

 

Yes, I've been playing around with that idea myself. There are three issues that I thought might be obstacles:

* those of you who love to put playlists together. From what I can tell, ISO selections fit in JRiver playlists very nicely.

* ISOs can often contain multichannel content, which you would not want to play on your 2 channel DAC. Create a custom view called ISO (or whatever) and make a rule that says "channels less than or equal to 2", so the 5.1 content doesn't even show up on the album track list

* ISOs are slightly better than DFF at tags, but nowhere near as good as DSF. And JRiver, up until Windows version 115, would take any tag edits in the ISO and reset them back to stock once JRiver restarted! This has been fixed, so if you need to add or fix tags in the ISO you can, and JRIver will remember them. It ain't ID3V2 compliant but its a taggable solution.

[khmusi]

1.) Synfreak wrote:

And - if tagging doesn't work with these - using some software to "repack" these into *dsf files.

2.) For me - the ongoing thems are confused. Is the reason of the "click" problem failure SACD_extractor or the Barchfile "extract DTS Stereo DSF files".

 

I use "YATE" for tagging dsf-files on MAC. This app is very good.

 

3.) There are some hope to solve/fix the problem in next future?

 

Thanks

khmusi

[Miska]

According to the "lost bits" when using direct extraction to *dsf files, I would recommend going to *.dff.

 

And - if tagging doesn't work with these - using some software to "repack" these into *dsf files.

 

Technically there's no reason to not have full ID3v2 tags on DFF files too. I added support for this, but it just won't be "standard" since it is not in the official specification. Same as ID3v2 tags for WAV or AIFF...

[Synfreak]

2.) For me - the ongoing thems are confused. Is the reason of the "click" problem failure SACD_extractor or the Barchfile "extract DTS Stereo DSF files".

 

The reason for the "click" sound on some (!) track transitions is missing data in the *.dsf files.

There is some fault in the SACD_extractor software, which does not affect extracted *.dff files.

The batch files - or Jesus' (Sonore) GUI for the extractor - are fine.

 

3.) There are some hope to solve/fix the problem in next future?

 

Maybe - maybe not ...

We will have to wait and see if Mr.Wicked comes up with a solution/fix.

[Synfreak]

Technically there's no reason to not have full ID3v2 tags on DFF files too. I added support for this, but it just won't be "standard" since it is not in the official specification. Same as ID3v2 tags for WAV or AIFF...

 

Yupp.

Seems to work fine over here.

[Synfreak]

It seems like it would be easier to just keep things as ISO files rather than extracting them to individual tracks, and use a player which supports that. (e.g. JRiver)

 

Ted has already posted some of his findings, to which I would only add one:

 

- SACDs are often encoded in DST (a packing scheme like ZIP or RAR or the like), which needs quite some computional power on "extraction".

If playing those files on a low power machine, it might be not fast enough to get the decoding (from DST to DSD) done in realtime.

 

So there are new culprits in there. Take your poison ;-)

[ljames]

My painful processes following:

 

My finding of DSF individual tracks output from sacd_extract will not do gapless perfect, I use mac mini+myteck, or Sony Z1ES, both are supposed to be gapless, but I still hear gaps.

I have quite a lot rips need gapless playback, so I'd rather have one big track. Unfortunately, sacd_extract only output DFF as whole disk track, I don't think DSF can. So the only way I found, is outputting a single DFF, using DFF2DSF coverts to DSF, tag by JRiver…when you have hundreds of rips, it's really painful.

 

I don't know if there is a better method. I have tried couple of rips, although one DSF per disc is the best solution for me, I just don't have that much time to finish the process.