1
\$\begingroup\$

I am trying to continuously capture and digitise raw composite video signals with sufficient bit depth and sample rate to be able to decode it including the colour channels. I am not looking for a video capture board as i'd like to have the raw signal and decode it in software myself.

So my question: What is the most cost effective way of doing this? i've already looked at digital oscilloscopes (pico etc) which offer capture functionality. but i am not sure if this is the best / cheapest option. it seems to me that capture of analogue signals would be a fairly common requirement but haven't found many easy to use products that do this. Ideally i am looking for a USB connected device with an interface to C++ or similar. any advice greatly appreciated!

\$\endgroup\$

closed as off-topic by brhans, Andy aka, Voltage Spike, ThreePhaseEel, Dmitry Grigoryev Mar 10 '17 at 14:44

This question appears to be off-topic. The users who voted to close gave this specific reason:

If this question can be reworded to fit the rules in the help center, please edit the question.

  • 1
    \$\begingroup\$ The search term you want is probably a "DAQ". The Hantek scopes have all the right bits in (cheaply) but I don't think they can do continuous capture. ( fabiobaltieri.com/2013/07/10/…) The SDR community may have some advice too. \$\endgroup\$ – pjc50 Mar 9 '17 at 12:29
  • \$\begingroup\$ When considering a digital sampling system, noise, volatage and digital resolution need to be considered. \$\endgroup\$ – Voltage Spike Mar 9 '17 at 19:03
1
\$\begingroup\$

Stop and think about the bandwidth requirements. You didn't say what kind of video, so I'll use NTSC as example. The color carrier is 3.6 MHz, so the theoretical minimum sample rate is 7.2 MHz. In practice, you'll probably need 15 MHz samples to do useful things.

There are certainly A/Ds that can convert at 15 MHz rates to 8 bits or more. The problem is then moving and dealing with 15 Mbyte/s, or 120 Mbit/s, without counting any protocol overhead. You'll need at least "high speed" USB for that. "Full speed" uses 12 Mbaud, and the data transfer rate is even less due to protocol overhead.

120 Mbit/s sustained can be done, but you have to think about things carefully. Then how are you going to process the 15 Mbytes/s data stream? This would be doable in the right kind of FPGA, but you're not going to find a programmable processor to handle that, at least not in real time.

If you just want to store a frame or two and decode later not in real time, then things get easier.

\$\endgroup\$
  • \$\begingroup\$ ok yes, very good points. however, processing 15 MB of data a second doesn't seem to me beyond capabilities of a current PC - how did you arrive at this conclusion? \$\endgroup\$ – morishuz Mar 9 '17 at 13:00
  • 1
    \$\begingroup\$ @dr.mo: Maybe the right multi-core PC can do this, but doing phase demodulation at 67 ns per sample sounds like a touch problem. Look into things called software defined radios. What you are asking for is not much different from demodulating a 4 MHz RF carrier in software. Hardware that can handle that should be able to handle your problem too, assuming it's programmable enough. \$\endgroup\$ – Olin Lathrop Mar 9 '17 at 13:06
  • \$\begingroup\$ "15 MB of data a second doesn't seem to me beyond capabilities of a current PC" - The entire frame must be captured with no delays, and PC's don't do real-time over USB. So you need a frame buffer in the capture device. \$\endgroup\$ – Bruce Abbott Mar 9 '17 at 13:13
  • \$\begingroup\$ yes, it is similar to SDR but with a rather manageable carrier frequency i would have thought. and i guess yes, i am looking for a box with fast USB and built in memory buffer - does such a thing exist? \$\endgroup\$ – morishuz Mar 9 '17 at 14:27

Not the answer you're looking for? Browse other questions tagged or ask your own question.