>
> Basically someone needs to fix the rate plugin and/or add resampling
> into RAT at the audio input side - there is some resampling that RAT
> does elsewhere.
> (there is also a [GPL] resampling library - but GPL brings up
> issues...). To be honest I had hoped that the ALSA stuff would
> improv
The internal audio device handling code in rat is very different for OSS than
it is for ALSA; the OSS one is probably much more mature. I'd hazard a guess
that that is at least part of the explanation.
--bob
On Apr 16, 2008, at 6:01 PM, Christoph Willing wrote:
We've always found that, for so
Ick. I wonder if the full-duplex clocking semantics remains at all in such a
configuration...
--bob
On Apr 22, 2007, at 9:21 PM, Douglas Kosovic wrote:
Under Vista, soundcards are now divided into separate recording and playback
devices, so RAT doesn't list any audio devices as RAT requires t
g/). These
> days, Linux KVM (kernel virtual machine, not keyboard/video/mouse)
> is probably a better way to do the machine virtualisation. I
> believe you need kernel 2.6.20 to do KVM without patches (and
> supported cpu).
>
>
> chris
>
>
>
>> Robert
That's correct, unlike rat which is split explicitly into audio-
engine, controller, and UI components, vic is a monolithic app that
has the GUI stuff wrapped up at a fairly fundamental level.
--bob
On Mar 1, 2007, at 7:44 AM, Derek Piper wrote:
>
> That should be:
>
> vic -t ttl addres
>
> 1) Modify VIC so that it does not join the multicast group - thus it
> should only send traffic into the network and not receive any video
> traffic from external sources. I'm not sure if this will work (though
> it should in theory - We can test it..) but we could add this feature
> fairly eas
It may be historical. When we started out the compressed codecs
seemed to be much less stable.
--bob
On Jul 20, 2006, at 10:45 AM, Derek Piper wrote:
>
> I have wondered before why the default for AG is Linear 16 when a
> compressed codec would require much less bandwidth (that could b
(wow, that code is really old now :-). I wrote the original version
of it way back when).
You're exactly correct on #1 and #2, but wrt #3, there should be a
button that saves the current autoplace settings as the default.
--bob
On May 15, 2006, at 3:20 AM, Breuer, Roman wrote:
> Tom,
> I t
Just wanted to remind folks that the RTCP data stream (where SDES lives) is
pretty low bandwidth. And bridges modifying data going thru to identify it as
having been touched by a bridge is as I recall something that the RTP spec
addresses.
--bob
On Mar 8, 2006, at 10:46 AM, Derek Piper wrote:
One of the ones we built long ago was a standard motherboard that we
put into a 4U case we bought separately ..
I've used the spectra 4-port capture cards with linux, works like a
champ.
--bob
On Jan 31, 2006, at 4:15 PM, Jeremy Mann wrote:
> I have been tasked to build a new portable AG no
Windows, I have had a similar problem. I recall there might have
> been a known problem but...
>
> MYK
>
> Robert Olson wrote:
>
>> is this windows or linux?
>>
>> On Jan 5, 2006, at 3:03 PM, Dr. Marlow Hinton wrote:
>>
>>>
>>> Is anyon
is this windows or linux?
On Jan 5, 2006, at 3:03 PM, Dr. Marlow Hinton wrote:
Is anyone using Realtek HD Rear Output drivers with the AG? My rat is not
picking the device drivers (No Device Found) and thus not sound output, yet the
driver is working with all other applications. The
It's possible for there to be per-multicast-group weirdness in the
routing infrastructure, since the presence of multicast data on the
network creates state in the routers to handle the transmission of
that data.
We actually ran into this quite a bit in the early AG days.
--bob
On Dec 7, 2
You can get by by wiring the unbalanced output from PC directly to
the balanced input by taking the signal from PC and going to + on
gentner. Ground from PC goes to - and ground on gentner.
To go the other way, wire the gentner + to PC signal, and ground to
ground, leaving gentner - floatin
On Sep 26, 2005, at 2:40 PM, Pritesh Shah wrote:
> hi,
>
> i have the gentner (clearone) xap800 and i had two questions
> pertaining to that:
>
> 1) Is anybody out there controlling gentner using a linux/unix machine
> ? Actually i want to find out if there's any way (any software)
> available
Doublecheck your echo cancellation refs. If you have a mic fed to an
EC ref, it can cause the symptoms you are seeing.
--bob
On Jul 18, 2005, at 9:57 AM, Jeremy Mann wrote:
> I'm so close to getting this thing to work. I finally got the serial
> working and setup the machine exactly like the d
The discussions I recall had to do with preemptive scheduling patches
for linux.
On May 24, 2005, at 12:30 PM, sh...@ni.aist.go.jp wrote:
> From: Colin Perkins
>
>
And assuming that the OS was a realtime OS and thus could actually
provide timing assurances to applications.
>>>
>
>
> And assuming that the OS was a realtime OS and thus could actually
> provide
> timing assurances to applications.
I know as of a couple years ago the hardcore linux audio folks were
doing exactly this; I don't know the current state of the realtime
mods for linux.
Anyone know if XP or M
> but it's still quite possible to send the audio quality way
> downhill on a machine that's also doing display of multiple video
> streams. Sharing video capture with audio handling these days is
> quite reasonable.
Of course if the media apps grokked realtime scheduling, as they
really
>
> In particular I think the following would be major leaps forward in
> usability:
>
> * Being able to change the venue in Vic and Rat without
> restarting them
I believe that this can reasonably be done in the context of the
existing vic and rat infrastructure (requiring inter
This just came out:
http://www.ietf.org/internet-drafts/draft-ietf-mboned-routingarch-00.txt
Overview of the Internet Multicast Routing Architecture
draft-ietf-mboned-routingarch-00.txt
Abstract
The lack of up-to-date documentation on IP multicast routing
pro
C Main bld. Room 109
> ======
>
> -Original Message-
> From: Robert Olson [mailto:ol...@mcs.anl.gov]
> Sent: Tuesday, April 26, 2005 13:41
> To: Michael Gates
> Cc: Lyle Carlson; ag-t...@mcs.anl.gov
> Subject: Re: [AG-TECH] Venue client causing network
Is it possible you have a flaky cable or connection? We saw something
like this somewhere, I forget where, where the traffic was triggering
errors like that and a cable was bad.
--bob
On Apr 26, 2005, at 3:23 PM, Michael Gates wrote:
> I'm wondering if any of the network folks have seen a Venu
What you might be seeing is this:
On linux, vic captures via v4l at 352x288. On a BT848-flavored card (I
think the Ospreys fall in this class as well as the Hauppauge cards),
capturing at more than 240 lines causes teh capture to happen from both
fields, as NTSC video is two fields of 240 lines
On Feb 21, 2005, at 10:08 AM, Natalia Costas Lago wrote:
yeah, TFT and screens plugged to the 4 heads of the Nvidia Quadro.
It is PCI... maybe .. is that the problem?
Quite likely ...
--bob
One way to tell what is going on is to mute all of the video streams. when
muted, all vic is doing is parsing the packet streams, and not decoding the
h261 (as I recall anyway). It is definitely not rendering them. If you still
see frame loss with streams muted, it is most likely the network at
On Feb 17, 2005, at 9:41 AM, Fred Dech wrote:
> i've run into the same problem on a unicast network. bridging, RAT
> received
> audio, but VIC did not receive any video. other clients connected to
> the
> same venue could see both the bridged video and audio. i've not yet
> tracked down the
Michael Daw wrote:
> Plans are underway for such a reference site that will be usable by UK
> AG nodes and (but don't tell anyone!) probably everyone else too. I
> don't have a clear idea of how long this will be, but funding has been
> secured and work is starting... No good to you now, Natali
Ivan R. Judson wrote:
>You could enable that rule by disallowing certs from the anonymous CA, if
>that were an explosed functionality. (I'd expect to see it in future
>releases ;-)
>
you can do that now, by just removing the anonymous CA cert from the server.
Make sure you don't have any loopback audio paths configured - if you
tap on a mic, you should not hear anything from the speakers (unless you
have explicitly and carefully configured the setup for reinforcement.
You might get this in the audio computer looping line in to line out.
--bob
Natal
This effect can be explained if you look at the architecture of AC97
sound chip.
Below (if you can see HTML mail, if not, the image is from the
datasheet at
http://www.analog.com/en/prod/0,2934,AD1819B,00.html
you will see the block diagram of a typical AC97 device. Note that all
inputs route t
Wondering if this is sitting in the ag-tech bounced message queue?
Can anyone help him?
--bob
--- Begin Message ---
Hi Robert
Sorry to address this directly to you. I posted several requests to
AG-TECH over the past couple of weeks but nothing is appearing.
I don't know if I have been deleted fro
At 10:25 AM 10/18/2004, Gavin W. Burris aka 86 wrote:
>You do have to wrap your head around the ALSA
>mixer settings, though.
What probably has to happen is the rat code be fixed to make the sliders in
rat do the right thing.
It's fairly ugly what has had to happen (in the rat OSS code anyway) t
At 11:37 AM 10/5/2004, Gavin W. Burris aka 86 wrote:
>Yes it was an insors node. So can this be fixed by dropping in the
>newer RAT executables into the insors directory?
Check with the insors folks; I don't know if that is a good thing to do or
not. A workaround I believe would be to not mute t
Was this with an insors node by any chance? The insors audio tool sends
zero-length packets when it is muted I blieve...
--bob
At 11:02 AM 10/5/2004, Gavin W. Burris aka 86 wrote:
>Today in a meeting between five sites, one site connected through a
>quickbridge over unicast. When they joined, r
I had problems with the alsa rat and an ensoniq 1370.. assertion failure in
rat.
--bob
At 10:27 AM 9/30/2004, Ti Leggett wrote:
>I've seen this behavior with some cards and the later kernel OSS
>drivers. It might be worth moving to ALSA and using the rat/ALSA patch
>to detect the card.
>
>On Wed
At 10:12 AM 9/27/2004, John Hodrien wrote:
>PCI Express. Even more bandwidth than AGP (similar initially), and
>symmetrical bandwidth. Alienware have a dual 16x PCIe motherboard, I'm not
>sure who else do. That'd enable you to have four high performance outputs
>with any old pair of PCIe graphic
At 10:06 AM 9/27/2004, John Hodrien wrote:
>Is the Rushmore significantly nicer than the G200 MMS? Anyone looking at
>using twin PCIe cards to get four nicely 3D accelerated outputs?
Unless system designs have changed since I last benchmarked this stuff, you
really want your video to be AGP-conn
>- This is done by certificates?
Yes, the AG software uses X.509 identity certificates for the purposes of
authenticating communications endpoints.
>* After I notice someone has broken in to my venue, I can "blacklist"
>him/her, or rather that particular certificate.
Yes, by manipulating the a
[Background for those who are coming in in the middle: we use the Cisco VPN
for offsite access at ANL; it has always caused problems with using rat].
In looking at debug output from another libcommon app, I realized what the
problem has been with rat and the VPN.
The libcommon code I added to m
>so, does this mean my certs are supposed to work, assuming i install them
>correctly?
They are properly formed; have you been able to use them to connect your
client to your server?
--bob
This was with the prototype code, so it won't port directly over; it would
be a trivial shared app - the shared data would be the multicast group to
be used, and all the client app would do is start up wbd on that group.
--bob
At 10:59 AM 6/22/2004, Adam Carter wrote:
>Robert Ols
At 10:17 AM 6/22/2004, Randy Groves wrote:
>A cursory search in Google doesn't reveal any sources for things like wb
>and wbd - are there others that might be available?
wbd sources are available at
http://www-mice.cs.ucl.ac.uk/multimedia/software/wbd/
we did a demo at sc a couple ye
At 09:00 AM 6/22/2004, Jennifer Teig von Hoffman wrote:
>One question I have about all this: I'd been assuming that if your clients
>and server are all behind a firewall, that you'd need a CA of your own
>since the clients and servers couldn't communicate with a CA. But I'm
>starting to wonder i
Hi Josh -
I'm sorry about the late reply on this - I tried to find it in my ag-tech
mail but didn't see it there, wonder if it bounced somewhere. Anyway:
> We are trying to set up a local bridging service adjacent to our wide
> area router so that AG traffic will be unicast on our lan. I woul
If it is set, it overrides any local determination of hostname. The
toolkit never sets it (in the registry; it will set it in the process if
necessary and have it inherited by the process's children). I've given
trying to set it for future use, as a check at startup time is sufficient.
--bob
On F
No, they are not. There shouldn't be any client-IP-address-specific
behavior in ag2.
--bob
On Fri, 18 Jun 2004, Richard Coffey wrote:
>
> Hi all,
>
> Are the certs IP address specific? I'd like to run our AG 2.0 machine on a
> test network with DHCP before I move it into production.
>
> Tha
At 03:41 PM 6/17/2004, Matthew Martinez wrote:
>I ran AG under linux, and i'm pretty sure my timestamps are correct now.
>My Ca cert is installed, and the timestamp on that is also within the
>correct boundaries. could you explain this error to me? Could this error
>be generated by the fact that i
sorry, I should explain the question more: if the clock was a day ahead on
biochem when the certs were issued, the start-time on them would be incorrect.
--bob
At 09:41 AM 6/17/2004, Robert Olson wrote:
>is the clock right on biochem?
is the clock right on biochem?
At 09:35 AM 6/17/2004, Matthew Martinez wrote:
> > They typically should be; I think the default is to make the starting time
> > the current time. How are you creating them?
> >
> > At 09:29 AM 6/17/2004, you wrote:
> >>i understand now, i apologize. my certificate
Check the clocks on your computers, and the good-from date on your certs:
error:80066420:lib(128):verify_callback:remote certificate not yet
valid
--bob
At 08:31 AM 6/17/2004, Matthew Martinez wrote:
> >>what if i'm in linux and still encountering the same error???
>
> >Hm. Can you send along yo
>ok, this is 2.1.2, and yes, i installed the CA cert as a trusted cert
>before i installed the identity certs.
>
>i'm heading home, cya guys tomorrow
>
>(thanks for the heads up bob, hows that?)
If you're on windows, you're probably hitting a known bug with added CA
certs. It's fixed in 2.2.
He
[resending, forgot the ag-tech cc]
The bulk of the tools is written in Python, using the C-based Globus
toolkit via a python binding called pyGlobus. It also uses the C-based
OpenSSL library via the pyOpenSSL binding.
The media tools (vic and rat) are written in a mixture of C, C++, and Tcl.
-
At 09:54 AM 6/15/2004, Matthew Martinez wrote:
> Just one more thing, and i'm sorry to keep pestering everyone, but
>does anyone know if Access Grid uses OpenSSL certificates, or does it
>make its own kind?
We use standard X509 certificates, as generated by either OpenSSL or the
Globus toolk
not specifically; the linux AG code never set GLOBUS_HOSTNAME (It really
can't in the way the windows code did, as it could do it via the registry).
--bob
At 11:44 AM 6/2/2004, Michael Miller wrote:
>Would this apply to linux as well?
At 10:45 AM 6/2/2004, Frank Sweetser wrote:
>I just removed AG 2.1.2 from my laptop, and then installed AG 2.2. I've had
>two issues so far
>
> - The GLOBUS_HOSTNAME environmental variable was still set to an
> old hostname. Previously, when the venue client started up, it
>
yes; that is how it works. Are you sure the mail address that was
submitted was valid?
--bob
On Fri, 21 May 2004, Jason Hill wrote:
> Has anyone else had problems with the certificate request process? I had
> two users request certs last week and never recieved an email that they
> were appro
At 08:30 AM 5/13/2004, George Estes wrote:
>If you're stuck with florescent lighting, as most of us are, try color
>balanced tubes for your fixtures.
I'll second that; we had daylight-spectrum bulbs installed in our node
rooms a couple years ago and it made a big difference, both in the feel of
Doublecheck via the cert mgr that you have CA certs installed; if they are
indeed there, please send me your venueclient log file and I'll have a look
at what's going on.
--bob
At 09:09 AM 5/4/2004, Darran Edmundson wrote:
>I've been struggling to resurrect AG2 on our node
>since the combinati
That's often how it's done, yes.
--bob
At 03:30 AM 5/3/2004, Natalia Costas Lago wrote:
>let say a laptop with AG software and Power
>Point installed so he can click Next&Previous buttons, am I wrong?
On Wed, 28 Apr 2004, Darran Edmundson wrote:
> Bob Riddle wrote:
>
> > ahhh ... it is generating IGMP v3 requests?
>
> I don't think so. We traced the problem
> down last night to XP's propensity for
> configuring extra network interfaces.
> In our case, interfaces for the two
> firewire ports.
At 04:55 PM 4/26/2004, Deb Agarwal wrote:
>Jabber is instant messaging similar to MSN and IRC but it is based
>on open standards. We have found it to be very reliable and the
>developer community is very active adding new features and capabilities.
>Using Jabber for AG chat would be more like the
At 02:43 PM 4/16/2004, Randy Groves wrote:
>Until such time as I (or someone) can modify the certificate system in
>such a way what it pays attention, for instance, to certificates already
>installed in IE (for example), or in Exchange/Outlook,
I've contemplated support for this, and didn't know
At 12:54 PM 4/8/2004, Ta Quoc Phong wrote:
>I installed AG ( on Windows ). When I request a certificate, I got an
>error: "Certificate Request failed: Connection failed. Did you specify the
>proxy server correctly".
>I think the program need a passphrase to authenticate, but I don't know
>where
That is likely to be difficult; rat has some fairly deep assumptions within
that the device it's getting samples from is a full-duplex audio device.
However, I think there might be some code in there that handles the binding
of two half-duplex devices to make a single full-duplex one, though I'v
At 12:10 PM 4/7/2004, Dwight D.McKay wrote:
>We're experiencing a strange problem with a new room node. The audio
>machine is using a new, soundblaster live! card that seems to route the
>analog line input to the analog line output no matter what we do. We've
>tried both the RedHat 9 sounds driv
At 04:37 PM 4/5/2004, Randy Groves wrote:
>(Man, I'm getting tired of the 'time warp' in the AG list. I can't seem
>to get any of my messages from Boeing to show up in any reasonable time
>frame. For example, this message was originally sent from my work at 10AM
>today)
This particular timewarp
67 matches
Mail list logo
