View Full Version : Re: high speed video capture

unknown user
11-13-2007, 11:22 PM
Hi Justin,

the requirements you describe exactly match the systems which are our daily
business. We are delivering video-based systems mainly for biomechanics
research. Our main product is a sophisticated 3D motion analysis software
which integrates synchronous acquisition of high-speed cameras and analog
data. (www.simi.com)

A common video capture system includes 4 to 6 cameras at 100 or 200 Hz at
VGA resolution. Of course we can also offer systems with more than 6 cameras
(at some number you might need a second PC). With mobile PCs/Notebooks you
are much more limited of course.
We store uncompressed video raw data without any quality loss. We also could
easily integrate video compression (e.g. MJPEG) but until now we have not
done this as our customers did not demand for this feature.

Recording 60 minutes with multiple cameras at 200 Hz is not a problem with
our system. We would be happy if you would contact us to discuss your
requirements in detail.

Best regards,

Simi. your movolution.

Charlotte Weber
Dipl.-Ing. (FH) Medizintechnik & sportmedizinische Technik
Support & Project Management

Simi Reality Motion Systems GmbH
Max-Planck-Strasse 11
D 85716 Unterschleissheim/Germany

Phone/Fax +49-89-321459-0/16
Mobile +49-174-3154613
Email: weber@simi.com
Web: http://www.simi.com

On Tue, 6 Nov 2007 14:40:59 -0500, Justin Case wrote:

>I am developing some software for a project that requires continual high
>speed video streaming. We are interested in supporting rates of 100 to 200
>fps, which will come from Basler or similar cameras. I am new to the video
>capture field, so any information on the following topic would be greatly
>I would be interested in hearing from subscribers that have direct
>experience with either commercial software packages or custom software
>developments that successfully stream raw video data to a PC for durations
>of up to 1 hour. I have looked at some companies that specialize in 2D
>video capture for biomechanics, but software specs on data throughputs are
>hard to come by. It seems that common approaches make use of either on
>board camera memory to capture the data (which makes real-time, long
>duration streaming problematic) or highly specialized cameras to identify
>markers and relay only coordinates to the PC (which we cannot use in our
>My impression thus far is that when it comes to long-duration video
>streaming, capture rates higher than 30 fps for long recordings are not
>commonly found in the market place (and hence not easily achieved without
>specialized equipment). Our application is for biomechanics research.
>Firewire, USB and Gigabit-ethernet communication ports are available to us.
>Any comments, suggestions and experiences that you can share will be greatly
>Thank-you in advance,
>Information about BIOMCH-L: http://www.Biomch-L.org
>Archives: http://listserv.surfnet.nl/archives/Biomch-L.html
>================================================== =======================