[AG-TECH] HD on AG
Douglas Kosovic
douglask at itee.uq.edu.au
Sun Dec 14 23:53:09 CST 2008
Hi Gurcharan,
> Really? Wow. You mean I can get vic to do uncompressed HD just like that?
> I've spent months trying to get UltraGrid to work on Macs (I got it
> working on colleagues machines but not mine!). But the goal was always to
> integrate it with the AG later. If this could be done
> now..........hmmmmmm....that would be the answer I've been waiting for.
MacOS X vic can grab uncompressed video frames from the Blackmagic HDMI
capture card and then encode using whatever codec is selected.
For uncompressed HD transsmission, I modified vic to add OptIPortal SAGE
support but there were a number of overheads which resulted in only 21fps at
~700Mbps.
I wrote a simple stand-alone MacOS X SAGE application that was achieving
25fps 1080i at ~930Mbps from a 25fps 1080i HDMI camcorder.
> So, can somebody kind do this quickly and let me try it out? I could
> eventually get someone to do this but it would be from scratch in terms of
> their experience with this software.
>
> If this works it would be so great and save so many hours of work for me.
Did you want a modified MacOS X vic executable with 'large' set to
1920x1080?
Cheers,
Doug
> Douglas Kosovic wrote:
>> Hi,
>>
>>>>>> I, too, am extremely interested in vic supporting HD formats,
>>>>>> (including uncompressed!) One of the issues is the encoding latency
>>>>>> problem. Not sure there's a nice low latency encoding solution out
>>>>>> there yet. Alternatively, we might look at cheap hardware encoders as
>>>>>> the answer or HDMI cards to take the input from the camera.
>>>>>> Can vic be made to take the output of, say, the BlackMagic Design
>>>>>> HDMI card on a Mac?
>>
>> MacOS X UCL vic works out of the box with the BlackMagic Design HDMI
>> capture card, but gets scalled down to 640x480 for the large capture
>> option.
>>
>> A quick hack to add 1980x1080 support is to modify the following width &
>> height lines in video/grabber-macosx.cpp :
>> switch (decimate_) {
>> case 1: // full-sized
>> width_ = SP601_NTSC_WIDTH;
>> height_ = SP601_NTSC_HEIGHT;
>>
>> to:
>> width_ = 1920;
>> height_ = 1080;
>>
>> Windows UCL vic on the other hand crashes with the BlackMagic Design HDMI
>> capture card as it currently doesn't support the BlackMagic YUV
>> colorspace.
>>
>> I wouldn't mind adding support for USB HD camcorders that support UVC
>> (USB Video Class) H.264 streaming over USB, but I currently don't have
>> access to one. Note: Not all USB based camcorders support UVC (e.g. Sony
>> Hanycams).
>>
>>>>>> I am currently using AGDV under windows but would like to move to
>>>>>> linux and use the HDV/DV linux vic for the DV part; i don't think
>>>>>> it's encoding the HDV, just accepting it as input from the camera
>>>>>> already encoded??? [chris?]
>>>>>>
>>>>> It is possible to just packetise the compressed data from a camera,
>>>>> though as Chris points out it tends to have quite high latency. That
>>>>> is why the ultragrid and derivatives used uncompressed HD as their
>>>>> sources. VIC doesn't currently support mpeg2 nor does support
>>>>> packetisation of an existing compressed mpeg2 stream.
>>>>>
>>>>
>>>> wait, i thought chris' vic supported HDV, which i assume is mp4???
>>>>
>>> AFAIK most cameras support HDV - which is closely related to MPEG-2:
>>> http://en.wikipedia.org/wiki/HDV#HDV_720p
>>
>> Mostly just the DV firewire based camcorders, the harddisk USB based ones
>> are now mostly AVCHD (MPEG-4 AVC/H.264).
>>
>>>> also, how hard/easy would it be for vic to support packetizing
>>>> uncompressed HD?
>>>>
>>> Not that hard - one needs to read the [RTP and codec] specs and code it
>>> up - or borrow someone else's....
>>>
>>> Piers.
>>>
>>>> :-)
>>>>
>>>> -gurcharan
>>>>>
>>>>> Piers
>>>>>
>>>>>
>>>>>> My ideal would be to encode the presenter's laptop output and deliver
>>>>>> that as just another video stream along side the presenter's video
>>>>>> and the audience video and have autoplace lay them out nicely in a
>>>>>> framed window, a la Adobe Connect or similar products, so the
>>>>>> presenter, the slides, and additional commentary could be nicely
>>>>>> recorded for later videoon demand playback.
>>>>>>
>>>>>> Just my opinion (and need),
>>>>>>
>>>>>> -gurcharan
>>>>>>
>>>>>> Piers O'Hanlon wrote:
>>>>>>
>>>>>>>
>>>>>>> Hi Derek,
>>>>>>>
>>>>>>> In principle VIC should be able encode and decode HD video - using
>>>>>>> one of the codecs that support such resolutions e.g. H.264 or
>>>>>>> h261as.
>>>>>>>
>>>>>>> One needs to modify some of the configuration code in VIC to allow
>>>>>>> selection of larger resolutions (ie HD one needs something from 720p
>>>>>>> (1024x720) upwards). And of course an HD source is necessary. We
>>>>>>> don't currently have access to any HD kit so haven't had a chance to
>>>>>>> work on it but anyone is welcome to try!
>>>>>>>
>>>>>>> Cheers,
>>>>>>> Piers
>>>>>>>
>>>>>>> 2008/10/22 Vine, Derek A <Derek.Vine at usd.edu>:
>>>>>>>
>>>>>>>> Hello all,
>>>>>>>>
>>>>>>>> I am doing some research to spec. up a new AG room, and the group
>>>>>>>> with the money is interested in HD video. I see that The
>>>>>>>> University of Queensland has an HDV/DV VIC for Linux based systems
>>>>>>>> already, but we would like to stay with Windows.
>>>>>>>>
>>>>>>>> I was just wondering what kind of time frame I can give them as to
>>>>>>>> when HD will be supported in AG on Windows. In particular, an HD
>>>>>>>> VIC that will take its video from a capture card with an HD-SDI
>>>>>>>> input, like the one attached in the PDF.
>>
>> I didn't see the PDF, what colorspace does it support? If it's YUV 4:2:0,
>> it would probably work with this vic:
>>
>> http://www.vislab.uq.edu.au/ag3/windows/vic-20081121.zip
>>
>> (which contained a fix for 480i DV white line artifacts)
>>
>> Anyway, I started working on adding support for more colorspaces to the
>> Windows code a few weeks ago and I might get back to working on the code
>> soon.
>>
>>
>>>>>>>> Thank you in advance for your help.
>>>>>>>>
>>>>>>>> Derek Vine
>>>>>>>>
>>>>>>>> Communication Network Specialist
>>>>>>>> The University of South Dakota
>>>>>>>> 414 East Clark Street
>>>>>>>> Vermillion, SD 57069
>>>>>>>> (605) 677-8215
>>>>>>>> dvine at usd.edu
>>
>>
>
>
> --
> -------------------------
> Gurcharan S. Khanna, Ph.D.
> Director of Research Computing
> Office of the Vice President for Research
> http://rc.rit.edu
>
> Assistant Research Professor, Ph.D. Program
> Golisano College of Computing and Information Sciences
> http://people.rit.edu/gskpop
>
> Director, Interactive Collaboration Environments Laboratory,
> Center for the Advancing the Study of Cyberinfrastructure
> http://icelab.rit.edu
> ---
> Rochester Institute of Technology
> 1 Lomb Memorial Drive
> Rochester, New York 14623-5603
> Phone: 585-475-7504 ~ Cell: 585-451-8370
> Email: gurcharan.khanna at rit.edu
>
>
>
>
> Douglas Kosovic wrote:
>> Hi,
>>
>>>>>> I, too, am extremely interested in vic supporting HD formats,
>>>>>> (including uncompressed!) One of the issues is the encoding latency
>>>>>> problem. Not sure there's a nice low latency encoding solution out
>>>>>> there yet. Alternatively, we might look at cheap hardware encoders as
>>>>>> the answer or HDMI cards to take the input from the camera.
>>>>>> Can vic be made to take the output of, say, the BlackMagic Design
>>>>>> HDMI card on a Mac?
>>
>> MacOS X UCL vic works out of the box with the BlackMagic Design HDMI
>> capture card, but gets scalled down to 640x480 for the large capture
>> option.
>>
>> A quick hack to add 1980x1080 support is to modify the following width &
>> height lines in video/grabber-macosx.cpp :
>> switch (decimate_) {
>> case 1: // full-sized
>> width_ = SP601_NTSC_WIDTH;
>> height_ = SP601_NTSC_HEIGHT;
>>
>> to:
>> width_ = 1920;
>> height_ = 1080;
>>
>> Windows UCL vic on the other hand crashes with the BlackMagic Design HDMI
>> capture card as it currently doesn't support the BlackMagic YUV
>> colorspace.
>>
>> I wouldn't mind adding support for USB HD camcorders that support UVC
>> (USB Video Class) H.264 streaming over USB, but I currently don't have
>> access to one. Note: Not all USB based camcorders support UVC (e.g. Sony
>> Hanycams).
>>
>>>>>> I am currently using AGDV under windows but would like to move to
>>>>>> linux and use the HDV/DV linux vic for the DV part; i don't think
>>>>>> it's encoding the HDV, just accepting it as input from the camera
>>>>>> already encoded??? [chris?]
>>>>>>
>>>>> It is possible to just packetise the compressed data from a camera,
>>>>> though as Chris points out it tends to have quite high latency. That
>>>>> is why the ultragrid and derivatives used uncompressed HD as their
>>>>> sources. VIC doesn't currently support mpeg2 nor does support
>>>>> packetisation of an existing compressed mpeg2 stream.
>>>>>
>>>>
>>>> wait, i thought chris' vic supported HDV, which i assume is mp4???
>>>>
>>> AFAIK most cameras support HDV - which is closely related to MPEG-2:
>>> http://en.wikipedia.org/wiki/HDV#HDV_720p
>>
>> Mostly just the DV firewire based camcorders, the harddisk USB based ones
>> are now mostly AVCHD (MPEG-4 AVC/H.264).
>>
>>>> also, how hard/easy would it be for vic to support packetizing
>>>> uncompressed HD?
>>>>
>>> Not that hard - one needs to read the [RTP and codec] specs and code it
>>> up - or borrow someone else's....
>>>
>>> Piers.
>>>
>>>> :-)
>>>>
>>>> -gurcharan
>>>>>
>>>>> Piers
>>>>>
>>>>>
>>>>>> My ideal would be to encode the presenter's laptop output and deliver
>>>>>> that as just another video stream along side the presenter's video
>>>>>> and the audience video and have autoplace lay them out nicely in a
>>>>>> framed window, a la Adobe Connect or similar products, so the
>>>>>> presenter, the slides, and additional commentary could be nicely
>>>>>> recorded for later videoon demand playback.
>>>>>>
>>>>>> Just my opinion (and need),
>>>>>>
>>>>>> -gurcharan
>>>>>>
>>>>>> Piers O'Hanlon wrote:
>>>>>>
>>>>>>>
>>>>>>> Hi Derek,
>>>>>>>
>>>>>>> In principle VIC should be able encode and decode HD video - using
>>>>>>> one of the codecs that support such resolutions e.g. H.264 or
>>>>>>> h261as.
>>>>>>>
>>>>>>> One needs to modify some of the configuration code in VIC to allow
>>>>>>> selection of larger resolutions (ie HD one needs something from 720p
>>>>>>> (1024x720) upwards). And of course an HD source is necessary. We
>>>>>>> don't currently have access to any HD kit so haven't had a chance to
>>>>>>> work on it but anyone is welcome to try!
>>>>>>>
>>>>>>> Cheers,
>>>>>>> Piers
>>>>>>>
>>>>>>> 2008/10/22 Vine, Derek A <Derek.Vine at usd.edu>:
>>>>>>>
>>>>>>>> Hello all,
>>>>>>>>
>>>>>>>> I am doing some research to spec. up a new AG room, and the group
>>>>>>>> with the money is interested in HD video. I see that The
>>>>>>>> University of Queensland has an HDV/DV VIC for Linux based systems
>>>>>>>> already, but we would like to stay with Windows.
>>>>>>>>
>>>>>>>> I was just wondering what kind of time frame I can give them as to
>>>>>>>> when HD will be supported in AG on Windows. In particular, an HD
>>>>>>>> VIC that will take its video from a capture card with an HD-SDI
>>>>>>>> input, like the one attached in the PDF.
>>
>> I didn't see the PDF, what colorspace does it support? If it's YUV 4:2:0,
>> it would probably work with this vic:
>>
>> http://www.vislab.uq.edu.au/ag3/windows/vic-20081121.zip
>>
>> (which contained a fix for 480i DV white line artifacts)
>>
>> Anyway, I started working on adding support for more colorspaces to the
>> Windows code a few weeks ago and I might get back to working on the code
>> soon.
>>
>>
>>>>>>>> Thank you in advance for your help.
>>>>>>>>
>>>>>>>> Derek Vine
>>>>>>>>
>>>>>>>> Communication Network Specialist
>>>>>>>> The University of South Dakota
>>>>>>>> 414 East Clark Street
>>>>>>>> Vermillion, SD 57069
>>>>>>>> (605) 677-8215
>>>>>>>> dvine at usd.edu
>>
>>
>
>
> --
> -------------------------
> Gurcharan S. Khanna, Ph.D.
> Director of Research Computing
> Office of the Vice President for Research
> http://rc.rit.edu
>
> Assistant Research Professor, Ph.D. Program
> Golisano College of Computing and Information Sciences
> http://people.rit.edu/gskpop
>
> Director, Interactive Collaboration Environments Laboratory,
> Center for the Advancing the Study of Cyberinfrastructure
> http://icelab.rit.edu
> ---
> Rochester Institute of Technology
> 1 Lomb Memorial Drive
> Rochester, New York 14623-5603
> Phone: 585-475-7504 ~ Cell: 585-451-8370
> Email: gurcharan.khanna at rit.edu
>
>
More information about the ag-tech
mailing list