data recording with analog NTSC

I've used a number of NTSC color video tape machines that had to have the color burst to record properly. They had a switch to allow you to record in B&W by using either its internal burst, or synching to an external burst signal.

Our needs were pretty primitive - the signal we were recording was an ultrasound image, so there was no colour information at all, and the ultrasound frequency was either 2MHz or 4MHz which didn't demand much in the way of time resolution.

One of the video recorders we used was one of the first consumer VHS machines from JVC. In the lab we used Sony U-matics, with 19mm tape casettes, but they were a bit bulky when the ultrasound scanner was being used at a hospital.

We also used a rather more compact open reel machine (AWA - IIRR) with little bits of reflective tape stuck onto the rim of the take up reel for an application where we played backed a sequence of scanned images some twenty-odd times to translate the images into a format that the cardiologist was used to interpreting. A bit Heath-Robinson/Rube Goldberg but it worked very reliably.

--------------- Bill Sloman, Nijmegen

I've used both industrial and professional video recorders. Both types needed external color burst to function properly. This came from an external sync generator, or could be synched to another stable video source, depending on the model.

Low end consumer crap with no time base correction capabilities used it own color burst signal to generate sync signals. Their performance was just too poor for our needs.

The point remains that the original request was totally bogus.

Complex differential equations are often solved by digital means such as Newton's Method. These digital solutions have existed for many centuries.

Any analog "solution" that significantly disagrees with the digital answer is wrong.

And the analog recording is nothing but a totally pointless means of further corrupting the bad data.

See

I must agree. What the OP wants to do is foolhardy but possible. I accomplished this in 1989 with the help of my co-workers and a high-end VHS recorder (Super-VHS). We used it to record a 1.4 MHz portion of the FM broadcast spectrum so that we could bring the "road" into the RF lab. We were testing our car radio design for multipath, dropouts, etc ... worked great. If I had to do this today, I would use a dedicated wideband high dynamic range ADC (LT, Ana, Nat, for example) and record the data to DDR2 memory. Playback's a bitch though.

Tips:

2) Don't use the lower frequency color subcarrier (it's lower than the video because it's down-converted before processing.)

The switching noise will probably trash your data, unless you can work around it. In our application, the FM noise blanker took care of the switching noise just fine.

OP: I hope this helps, but I still think you are wasting your (and this group's time). If you are like me, you will need to prove that to yourself before trusting what someone says. Best of Luck.

Frank Raffaeli

If you wouldn't mind (on, or off list) please explain to me how Newton's method is _digital_ ? It obviously proceeds by discrete steps as classically formulated. But it does not depend on computation with _digits_ and therefore is not digital. Also, it is possible to construct analog feedback loops which have all the character of discrete iteration. Such iteration is a special case of continuous feedback.

Mike

That's easy to demonstrate... just add some mild hysteresis to a linear feedback system, along with a restrictive slew rate.

...Jim Thompson

On Wed, 28 Sep 2005 11:14:11 -0700, Mike wrote: [snipped attribution]

You're right, there's nothing inherently "digital" about Newton's method, other than that it's amenable to digital computation, since number-crunching is involved.

But then again, doesn't an ordinary differential amp or gain stage "obey" Newton's method (in a manner of speaking) when it "solves" its own output? ;-)

Cheers! Rich