Iím an oceanographer that uses analog standard-definition cameras underwater to capture real-time video of the seafloor using a towed sled. Historically, this video has been recorded using a Sony mini-DV recorder, but we are interested in recording the video direct to hard drive, hopefully at the same resolution we get on the mini DV tapes. I believe the video captured on the tape is .dv format.
Specs on the video generated on the camera: 1 volt p-p; NTSC; 460 TV lines; 768 horizontal pixels, 494 vertical pixels. The video is transmitted through a 1500í length of cable (so some unavoidable loss in quality there), and the conductors are UTP, so we run the video through a passive transmitter balun (brand: Nitek) inline after the camera and then an active balun receiver at the surface inline just before we run it into the mini DV recorder. We use composite video cables for the connection.
We have tried the El Gato USB capture device, but the resulting video quality was surprisingly terrible. The video quality we get on the video tapes is actually decent considering being run through 1500 feet of cable. We donít have any preferences about what format, compressions, codecs, etc. are used Ė we just want to record the video in real time on a hard drive at the quality weíre used to getting on the tapes. And just to be clear, my question is about recording the real-time video we generate in the future, not converting the mini-DV tape library we currently have. Any suggestions on hardware/software to accomplish this? Thanks!
+ Reply to Thread
Results 1 to 13 of 13
Thank you. Yes, it does have a firewire port - we have used it in the past to export video from previously recorded tapes to DVD. I was not aware that it could also pass through the real-time video as well. I imagine in order to do this, there would still need to be a tape in the deck and to use the physical buttons on the deck to initiate the recording and activate the video pass-through?
you may have to go tapeless and leave the tape door open for the cam to stay on allowing you to record over the firewire to the computer's hard drive in dv format as long as you have space available there, it's 13GB/hr.--
"a lot of people are better dead" - prisoner KSC2-303
iirc you can initiate recording using the control panel of windv without touching the cam. if the cam is on and detected by windv you should be able to just click it's capture button. make sure to set the options in windv something like this to maximize the time it will run.
[Attachment 46848 - Click to enlarge]--
"a lot of people are better dead" - prisoner KSC2-303
I may be missing something but why are you transmitting the video through 1500' of cable? What I would do is use a camera that has built in storage and record to that, start the recording while the camera is on the boat and let it run as it's submerged and towed and stop the recording once you pull the camera up. Or is it that you want to see the footage in real time?
Sorry for the delayed follow up - was out at sea for the past couple weeks. Just wanted to thank you guys for your thoughts. To answer some of the questions raised:
* aedipuss: I will try to do as you suggest, but you mentioned leaving the tape door open so the cam will stay on - just to be clear, the unit we are using is a Sony GV-D1000 which is a "clamshell" type mini DV player/recorder but isn't a camera itself like a Sony Handycam. The video comes from a third party analog camera. At any rate, I'll give the firewire pass-through approach combined with WinDV a shot as this seems like the most straightforward way to record the video directly.
* sophisticles: We run the video through the cable for the reason you suggest - so we have a live video feed which we use to "navigate" the towed sled - lifting it up and over obstructions as they come into view. We do also have a second camera (a GoPro) that we use to capture higher resolution video locally to an SD card. Another reason we want to record the live video feed is that we're able to pass the video through an overlay board that burns the time/date/GPS/depth/etc. onto the video - something that is not easily accomplished in post processing (the claims of some software programs to the contrary).
* KarMa: I will check to see what sort of settings are available with the ElGato device's software, but if I recall, it's pretty bare bones without the ability to tweak the brightness/saturation/etc. You mention there are plenty of good alternatives out there - if you have one that you prefer, would you mind passing that along?
Thank you again!
Do you calibrate (color chart) the camera before shooting ?
No, and to my knowledge, there doesn't appear to be any way to calibrate color on this camera. This is the camera we are using:
I would also add that such calibration, even if possible, would be challenging as we are collecting underwater video. Any calibration done at the surface in air wouldn't necessarily be applicable to the video taken at depth given the different scattering and absorption of light underwater. We have done a couple experiments with a plywood colorboard bolted to pipes a couple meters in front of the camera to see how some of the different lights we use (Halogen, LED, HID, etc.) on the sled affect how the camera captures those colors.
That said, the poor video quality from the ElGato device isn't about poor rendering of colors, it's about grainy, washed out video quality in general. I'll see if I can get a couple screen grabs from the mini DV tapes and the ElGato device for comparison.
Yeah, calibration should be done underwater, ofcourse, otherwise it would be of no use, if it where possible at all.
My guess also is that a lot of the resolution of the camera is lost through the DV format.
another recorder should have been used, a camera with SDI output is also a better choice.(With SDI you can have long a cable and no video delay)
My guess the Gopro camera records better quality footage, so the low quality of the dv recordings are not important? only for documentation, or to sync with the
Gopro footage ?
camera needs underwater case though,
Last edited by Eric-jan; 21st Oct 2018 at 17:29.
White balance will be of limited use underwater. The farther away on object is the more reds will be attenuated. So only objects at the same distance at which the white balance was set will have the right colors.
One trick that might work around the red attenuation problem in underwater video is to use heavily red shifted lighting. So if you know reds are attenuated by 90 percent at 10 feet (and that's the distance that's most critical to you), use lighting where reds are 10x brighter. Whether that's practical in your situation I don't know.
Last edited by jagabo; 21st Oct 2018 at 20:24.
The live video (stream) needs to be converted/captured, at the location in this case, not in post.