S/PDIF and track markers (echo3g)

S/PDIF and track markers (echo3g)

[Mark Hills]

Hi,

Is there any provision in ALSA (either within the kernel or libasound) for 
extracting track marker information from S/PDIF?

I have a large number of MiniDisc recordings to preserve as FLAC format.

Losing track markers would be a huge inconvenience, though I don't care 
for the rest of the metadata (track names etc..)

There's enough recordings to be worth developing specific software for the 
archival if it can access this information.

I don't really know which path is likely to be successful:

* An ALSA API to access the information; though I can't see anything
  specific to track markers

* Whether a timecode could be used to detect track change; but I can't see 
  any in /proc/asound/Layla3G/status

* Does ALSA provide access to raw AES streams, so I could decode it in 
  userland? I'm un-sure if this relates to hardware AC3 decoding.

* Maybe this sort of status is only available to the firmware of the 
  interface and never makes it to userland (in my case, echo3g); though I 
  can see some constants relating to 'non audio' S/PDIF.

The hardware I have here is Echo Layla 3G (snd-echo3g) and a Pioneer 
MJ-D707 MiniDisc recorder.

Any advice appreciated -- many thanks.

-- 
Mark

Re: S/PDIF and track markers (echo3g)

[Anssi Hannula]

17.06.2016, 16:29, Mark Hills kirjoitti:
> Hi,

Hi,

> Is there any provision in ALSA (either within the kernel or libasound) for 
> extracting track marker information from S/PDIF?

I don't know what track marker data looks like, but see below.

> I have a large number of MiniDisc recordings to preserve as FLAC format.
> 
> Losing track markers would be a huge inconvenience, though I don't care 
> for the rest of the metadata (track names etc..)
> 
> There's enough recordings to be worth developing specific software for the 
> archival if it can access this information.
> 
> I don't really know which path is likely to be successful:
> 
> * An ALSA API to access the information; though I can't see anything
>   specific to track markers
> 
> * Whether a timecode could be used to detect track change; but I can't see 
>   any in /proc/asound/Layla3G/status
> 
> * Does ALSA provide access to raw AES streams, so I could decode it in 
>   userland? I'm un-sure if this relates to hardware AC3 decoding.

ALSA does provide access to the raw stream with drivers that support
that. You'll need to record with IEC958_SUBFRAME_LE or
IEC958_SUBFRAME_BE sample formats.

> * Maybe this sort of status is only available to the firmware of the 
>   interface and never makes it to userland (in my case, echo3g); though I 
>   can see some constants relating to 'non audio' S/PDIF.

Unfortunately the echo3g driver doesn't seem to support
IEC958_SUBFRAME_* sample formats.

The HDA driver, used for most mainboard/laptop integrated audio, does
support that format if the hardware is capable. I'm not sure if the
hardware commonly supports that format or not, though.

> The hardware I have here is Echo Layla 3G (snd-echo3g) and a Pioneer 
> MJ-D707 MiniDisc recorder.
> 
> Any advice appreciated -- many thanks.
> 


-- 
Anssi Hannula

Re: S/PDIF and track markers (echo3g)

[Mark Hills]

On Fri, 17 Jun 2016, Anssi Hannula wrote:

> 17.06.2016, 16:29, Mark Hills kirjoitti:
> > Hi,
> 
> Hi,
> 
> > Is there any provision in ALSA (either within the kernel or libasound) for 
> > extracting track marker information from S/PDIF?
> 
> I don't know what track marker data looks like, but see below.
> 
> > I have a large number of MiniDisc recordings to preserve as FLAC format.
> > 
> > Losing track markers would be a huge inconvenience, though I don't care 
> > for the rest of the metadata (track names etc..)
> > 
> > There's enough recordings to be worth developing specific software for the 
> > archival if it can access this information.
> > 
> > I don't really know which path is likely to be successful:
> > 
> > * An ALSA API to access the information; though I can't see anything
> >   specific to track markers
> > 
> > * Whether a timecode could be used to detect track change; but I can't see 
> >   any in /proc/asound/Layla3G/status
> > 
> > * Does ALSA provide access to raw AES streams, so I could decode it in 
> >   userland? I'm un-sure if this relates to hardware AC3 decoding.
> 
> ALSA does provide access to the raw stream with drivers that support
> that. You'll need to record with IEC958_SUBFRAME_LE or
> IEC958_SUBFRAME_BE sample formats.

Thanks, that's very useful information.
 
> > * Maybe this sort of status is only available to the firmware of the 
> >   interface and never makes it to userland (in my case, echo3g); though I 
> >   can see some constants relating to 'non audio' S/PDIF.
> 
> Unfortunately the echo3g driver doesn't seem to support
> IEC958_SUBFRAME_* sample formats.

Ah, as I'd suspect. And probably there is a lot of logic in the firmware 
that would have to cooperate.

> The HDA driver, used for most mainboard/laptop integrated audio, does
> support that format if the hardware is capable. I'm not sure if the
> hardware commonly supports that format or not, though.

This is the hardware I have on my motherboard, but it doesn't have the 
optical input I would need.

Perhaps I can find a reasonably priced PCI Express card with the 
capability.

Thanks for your help, I have a feeling this is the most promising route. 
But it does assume the MiniDisc deck is outputting this information (it 
definitely acknoledges it on input)

> > The hardware I have here is Echo Layla 3G (snd-echo3g) and a Pioneer 
> > MJ-D707 MiniDisc recorder.
> > 
> > Any advice appreciated -- many thanks.

Thanks

-- 
Mark