ALSA for audio
 

In article , Andy Furniss
wrote:
Jim Lesurf wrote:
I've added the digital-tv group to this as someone there may be able
to clear up points I'm not sure about!...



How did you tune the EZCap and tell it which 'station' you wanted?

I think the ezcap is just analogue input - there's probably a way to
select between s-video and composite unless it only takes one at a time.

There is no tuning - it doesn't have a tuner.

I guess it just outputs yuv + pcm and lossless compresses it, sends it
over over usb and the specific v4l driver decompresses it and presents
yuv and pcm.

To summarise what has 'gone before' on the Linux and audio groups...

Davey has been trying to get audio output from the EZCap via ALSA. The
'arecord -l' command does seem to list what he says is the EZCap. But
attempts to read in the audio and capture it give either errors or an
all-zero-sample-values recording.

I've been over the basic ways to determine the device ID under ALSA and how
to connect. But it seems not to co-operate. So I've been wondering about
the cause of the problem. In particular if it isn't really providing a
fully demodulated audio output.

Thought this because the FUNcube I'm using does output over ALSA, but the
output is IQ samples *not* audio, so need post-processing. In an earlier
posting Davey said his Windows software for the EZCap gave mpg output...

Hence I'm currently trying to decide if his working on having EZCap output
audio (without some further processing on the computer) is chasing a
boojum. So not an 'alsa' problem at all. But I dunno as I know zero about
the EZCap. My experience is only with the 290e and FUNcube.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

ALSA for audio
 

Jim Lesurf wrote:
Hence I'm currently trying to decide if his working on having EZCap output
audio (without some further processing on the computer) is chasing a
boojum. So not an 'alsa' problem at all. But I dunno as I know zero about
the EZCap. My experience is only with the 290e and FUNcube.

I have the Easycap dongle and a couple of USB tuner sticks and they are
very different devices. The Easycap (or EZCap) is basically a digitizer
which samples an analogue composite video or S-video input, with audio,
and sends PCM data to the USB. I don't know to what extent that data is
compressed, but I do know that it requires a fast USB connection and a
fast processor otherwise it simply won't work.

EZcap can be used with VLC if a driver is installed, but this is as a
video capture device and not as a tuner device such as the Hauppauge,
which supplies an MPEG2 or H.264 transport stream.

ALSA for audio
 

Jim Lesurf wrote:
In article , Andy Furniss
wrote:
Jim Lesurf wrote:
I've added the digital-tv group to this as someone there may be able
to clear up points I'm not sure about!...



How did you tune the EZCap and tell it which 'station' you wanted?

I think the ezcap is just analogue input - there's probably a way to
select between s-video and composite unless it only takes one at a time.

There is no tuning - it doesn't have a tuner.

I guess it just outputs yuv + pcm and lossless compresses it, sends it
over over usb and the specific v4l driver decompresses it and presents
yuv and pcm.

To summarise what has 'gone before' on the Linux and audio groups...

Davey has been trying to get audio output from the EZCap via ALSA. The
'arecord -l' command does seem to list what he says is the EZCap. But
attempts to read in the audio and capture it give either errors or an
all-zero-sample-values recording.

I wouldn't expect that to work without having the v4l drivers setup and
running the device.


I've been over the basic ways to determine the device ID under ALSA and how
to connect. But it seems not to co-operate. So I've been wondering about
the cause of the problem. In particular if it isn't really providing a
fully demodulated audio output.

Unlikely - but even if it were you would hear/record something.


Thought this because the FUNcube I'm using does output over ALSA, but the
output is IQ samples *not* audio, so need post-processing. In an earlier
posting Davey said his Windows software for the EZCap gave mpg output...

The windows software will probably (badly) convert to mpeg on the fly
because it's not practical to save as raw video. Of course it's possible
the device has a hardware encoder - but more likely not.

Some probing with v4l-ctl should give more info.

Hence I'm currently trying to decide if his working on having EZCap output
audio (without some further processing on the computer) is chasing a
boojum. So not an 'alsa' problem at all. But I dunno as I know zero about
the EZCap. My experience is only with the 290e and FUNcube.

Looking at Johns' post it seems that ezcap will work with VLC - I think
getting this to work should be the first thing, and also not testing
initially with a rare format like NTSC 443.

ALSA for audio
 

Andy Furniss wrote:

Looking at Johns' post it seems that ezcap will work with VLC

Forgot about the crosspost - maybe that's with windows.

ALSA for audio
 

In article , John
Legon
wrote:
Jim Lesurf wrote:
Hence I'm currently trying to decide if his working on having EZCap
output audio (without some further processing on the computer) is
chasing a boojum. So not an 'alsa' problem at all. But I dunno as I
know zero about the EZCap. My experience is only with the 290e and
FUNcube.

I have the Easycap dongle and a couple of USB tuner sticks and they are
very different devices. The Easycap (or EZCap) is basically a digitizer
which samples an analogue composite video or S-video input, with audio,
and sends PCM data to the USB. I don't know to what extent that data is
compressed, but I do know that it requires a fast USB connection and a
fast processor otherwise it simply won't work.

That may mean the problem is that Davey either can't identify the format
for the audio, or that it has to be processed on the host machine.

ALSA can routinely do things like shuffle between 16bit and 24bit and LE
and BE byte sets. Even resample/convert the sample rate. But not do things
like decode mp3, say. That needs the addition of some other software
process in the chain.

I don't have an EZCap. So can't do any tests. Davey reports the EZCap as
giving alsa device ID numbers and strings. But then says he gets 'empty'
streams (0 values) when using these with arecord. Which indicates something
somewhere isn't set up correctly by him. But I can't tell what the problem
may be from here. :-)

I've had such empty results from another device. But that was IIUC because
the device was being controlled incorrectly. That may also be true here.
The problem there seemed to be a bug in some versions of the HID USB
libraries. I changed library and the code worked. But if that's the problem
his software may need re-compiling and linking against an alternative
library.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

ALSA for audio
 

Jim Lesurf wrote:

I don't have an EZCap. So can't do any tests. Davey reports the EZCap as
giving alsa device ID numbers and strings. But then says he gets 'empty'
streams (0 values) when using these with arecord. Which indicates something
somewhere isn't set up correctly by him. But I can't tell what the problem
may be from here. :-)

If things are not correctly set then it's certainly possible for the
device to initialize OK but to give a completely blank recording...

ALSA for audio
 

Andy Furniss wrote:
Andy Furniss wrote:

Looking at Johns' post it seems that ezcap will work with VLC

Forgot about the crosspost - maybe that's with windows.

Yes, Windows XP.

However, I've just tested it on a 1.3 GHz laptop and find that it will
not work when a wireless dongle is also plugged into a USB port, and it
will not work correctly when connected through the (supplied) short USB
extension cable! It needs to be plugged directly into the port, and
other USB devices should be disconnected (except mouse/keyboard).

Needless to say, the correct input (S-video or composite) and PAL or
NTSC also need to be selected.

ALSA for audio
 

On Sun, 17 Feb 2013 15:37:01 +0000
John Legon wrote:

Jim Lesurf wrote:

I don't have an EZCap. So can't do any tests. Davey reports the
EZCap as giving alsa device ID numbers and strings. But then says
he gets 'empty' streams (0 values) when using these with arecord.
Which indicates something somewhere isn't set up correctly by him.
But I can't tell what the problem may be from here. :-)

If things are not correctly set then it's certainly possible for the
device to initialize OK but to give a completely blank recording...

Jim and John and Andy,

Ok, back after yesterday's alcoholic birthday party (not mine).
I've also had some time to do some thinking:

Thought 1. Since this laptop, for whatever reason, doesn't want to run
Audacity, which is the one editing programme I am familiar with, and it
is having all these problems getting audio using either the EZCap or
the Hauppauge, I think I will abandon this approach. I have other
things to fix, such as no audio when playing a video clip in Firefox, so
this one I will lay aside.

Thought 2. The old Desktop, which is now used for Zoneminder in a
different room, used to run Audacity, AND it has a real soundcard.
Using it would mean moving the (quite large) cassette player down
there, and finding somewhere to put it, but it is likely to work. I
will be loading Audacity back onto it later this morning.

Thought 3. I have an old HP/Compaq laptop, which just sits keeping cardboard boxes anchored to the floor, so I might just try that in the same
way that I was trying to get this laptop to work, with the dongles.
Since it has no other duties, I can do whatever I like with it, without
worrying about messing anything up.

I appreciate all the help so far, and have learned a lot, but I am
making little useful progress towards my original goal, and it might
be an impossible task anyway.

I will report progress.
--
Davey.

ALSA for audio
 

On Thu, 14 Feb 2013 15:10:54 +0000 (UTC), Tony Houghton
wrote:

In ,
Jim Lesurf wrote:

You may also find that the mic input clips 'line' level voltages since a
mic input may have a preamp stage which can't cope with such 'high' levels.
Any hardware gain adjustment may be after this stage. This, again, all
depends on the specific hardware in the box.

In general I'd be wary of the analogue stages of hardware in a normal
computer. Might be OK. But a dedicated special purpose external 'sound
card' may well give much better results.

From what I've read you're quite right to be wary of standard sound
cards or onboard audio. A few years ago one of the regulars on
uk.comp.homebuilt used to point out that a reference design for sound
cards had an over-severe clipping circuit on the line-in (not just mic)
which would clip at well under the usual maximum level. And this design
was used by the vast majority of sound cards.

I think that must have one of my postings in another news group from
nearly a decade back when I first discovered this problem with
on-board and a few different makes of cheap PCI soundcards that I had
to hand at the time.

A google search unearthed the fact that MoBo and PCI soundcard makers
had 'blindly' used the "reference" cct design the audio chip makers
had issued with the data sheet, blithely ignoring the significance of
the reduction from 12v to 5v supply rail for the buffer input amp
feeding the ADC chip which had been hard jumpered to desensitize it by
6 db in order to reduce the noise floor by that same factor.

I did marvel at the 6db improvement in the "Dynamic Range" of the
line input on the PCI and MoBo mounted chips compared to the old
SoundBlaster16 ISA cards but quickly realised that it was an articial
improvement that introduced line input clipping a good 3.5db below the
FSD limit of the ADC which created a bigger problem of monitoring for
clipping than the benefit of an overall increase of 2db dynamic range
offered.

At the time, I was in utter despair of resolving this issue with the
newer MoBos that were then appearing sans ISA slot of any sort so
selected the most recent Socket A MoBo with single ISA slot to
accomodate my AWE64Gold soundcard to build a 'special' PC to allow me
to continue digitising my tape (reel to reel) and vinyl collection.

The machine still resides in the front room with the decks and is
still in regular use by the XYL to this day. SWMNBO complains
occaisionally about it's 'sluggishness' but only because my 30 yo
'teen age' son likes to stirr things up and thinks ( in that teen
agers' sensitised to fashion way) that it needs to be updated to
something like a quad core 8GB system with win7 inflicted upon it.

I take a very dim view of this 'waste of resources upgrade' just to
'stand still' - the win2k 512MB single core 1.2GHz Athlon setup works
just as fast as it ever did (and probably faster than the suggested
win7 upgrade if my experience is anything to go by) so I'm not about
to rush into such an upgrade any time soon (but see below - I might
change my mind on that, but not win7).

About a year ago, I bought some usb audio dongles cheaply at a
fleamarket stall which, on testing, seemed to be lacking this defect.
Out of interest, I ran tests on, what was then, a 3 year old MoBo
design and discovered likewise. It seems the penny finally dropped
with the sound card and MoBo makers about half a decade or so after it
had been introduced by their lack of attention to detail attitude.

I think the problem was finally resolved around 5 years back on
pretty well anything designed since then so you should be safe with
recent kit but it's just as well to test line inputs for this issue to
be sure.

--
Regards, J B Good

ALSA for audio
 

On Mon, 18 Feb 2013 11:07:19 +0000
Davey wrote:

On Sun, 17 Feb 2013 15:37:01 +0000
John Legon wrote:

Jim Lesurf wrote:

I don't have an EZCap. So can't do any tests. Davey reports the
EZCap as giving alsa device ID numbers and strings. But then says
he gets 'empty' streams (0 values) when using these with arecord.
Which indicates something somewhere isn't set up correctly by him.
But I can't tell what the problem may be from here. :-)

If things are not correctly set then it's certainly possible for
the device to initialize OK but to give a completely blank
recording...

Jim and John and Andy,

Ok, back after yesterday's alcoholic birthday party (not mine).
I've also had some time to do some thinking:

Thought 1. Since this laptop, for whatever reason, doesn't want to run
Audacity, which is the one editing programme I am familiar with, and
it is having all these problems getting audio using either the EZCap
or the Hauppauge, I think I will abandon this approach. I have other
things to fix, such as no audio when playing a video clip in Firefox,
so this one I will lay aside.


I fixed the no-audio with video clips by disabling the .asoundrc file!


Thought 2. The old Desktop, which is now used for Zoneminder in a
different room, used to run Audacity, AND it has a real soundcard.
Using it would mean moving the (quite large) cassette player down
there, and finding somewhere to put it, but it is likely to work. I
will be loading Audacity back onto it later this morning.

It certainly loads Audacity, and appears to be ready to do something if
I had a connection line-in.

Thought 3. I have an old HP/Compaq laptop, which just sits keeping
cardboard boxes anchored to the floor, so I might just try that in
the same way that I was trying to get this laptop to work, with the
dongles. Since it has no other duties, I can do whatever I like with
it, without worrying about messing anything up.


It loads Audacity, and appears to be ready to do something if
I had a connection line-in. Later on.

Now to work out how to have the benefit of the .asoundrc file, without
having the disadvantage of having the .asoundrc file.

--
Davey.