Capturing tapes that contain both 4:3 and 16:9 content?
 

Hi Guys,

I have to transfer a bunch of miniDV tapes which contain both 16:9 and 4:3 content. Just wondering if there is a capture card which can read the aspect ratio and automatically capture the correct aspect. At the moment we have to do the tedious process of capturing each section and editing it together.

Thanks,
Sean

Capture cards are noit used to transfer DV and miniDV. DV is copied 1:1 via Firewire using software such as WinDV, a Firewire cable, and a Fiewire input.

Would WinDV/Firewire capture handle different aspect ratios without requiring multiple captures?

Firewire is not a "capture". It's a copy operation.

Consumer miniDV frames are 720x480 NTSC (720x576 PAL), whether the DAR is 4:3 or 16:9. The image in both sections of the tape will appear differently in an editor. To combine them into a DVD or other final format, you have to process them separately as two different videos. DV is no longer supported by most operating systems, so you will need your original camera's DV codec or a newer codec such as Cedocida. DV cannot be posted on the web or played on smart tv's or desktop players. Both parts of the original video will have to be re-encoded into newer formats.

Users of the miniDV format can give you more details if necessary. My reply is a quick general caution to prevent spending money on capture gear that will damage the final transfer of your original.

There is no such 16:9 DV, The original camcorder was just adding black bars on top and bottom of the 4:3 to give it a cinematic look, If you are exporting to a 16:9 format you will have to separate those scenes and crop the black bars but the resulting picture would look blurry as hell. But if you are exporting to 4:3 format then just leave it alone.

Pro camcorders back in the 4:3 era that recorded anamorphic widescreen picture on a 4:3 imaging sensor used a special lens that squeezed the field of view horizontally and in the editing stage the pixels were stretched horizontally to fill a 16:9 frame. Those lenses sold for thousands of dollars.

HDV camcorders used a similar technique but on the imaging sensor itself not the lens, The resolution is 1080x1440 which is an exact 4:3 ratio, but the pixels are not square which gives a perfect 16:9 frame with only fewer horizontal pixel count. This and using the new then MPEG-2 compression they managed to use regular DV tapes for this new pseudo HD format.

1080x1440 is not a horizontal rectangle. It is a 3:4 vertical rectangle. I think you mean 1440x1080? The width is usually stated first, not the height.

There is 16:9 DV, it doesn't have black borders in the DV stream itself, it's the same resolution as 4:3 but more squeezed horizontally. In most cases it was achieved by the camcorder cropping the image rather than an anamorphic lens I think. A camcorder may add black borders on analog playback though to make have the right aspect ratio for standard SD video (which isn't ideal for capture). Applications transfering dv via firewire can usually be set to divide the stream into clips. I don't know if there are any that can be set to only split the transfered DV stream on an aspect ratio change though.

The frame is 16/9 horizontal made from 4/3 CCD so the pixels are stretched horizontally during playback, therefore horizontal rectangles not vertical. When they started using 1080x1920 the imaging censors became 16/9 wide to keep the pixels square or 1920x1080. I just like to start with the smaller number.

Yes its just a digital zoom, if you account for the missing lines there is actually black bars that the camcorder just don't show them with widescreen TV's that are equipped with iLink input (I had an LG TV back in the day with iLink input) but they do show up in the view finder while recording or playing back tapes as well as thru analog outputs. So it remains to be seen how the transfer software will deal with those missing lines from top and bottom of the frame, The poster should upload a sample from the scene to examine.

The (NTSC) DV stream is nominally 720x480. How it looks on the screen is based on how the display system interprets the signal; i.e, the pixel shape or aspect ratio. A 640x480 computer screen image is 4:3 and calls for a square pixel (aspect ration of 1:1), as does a 16x9 FHD image 1920x1080, (However and FWIW the old Hercules, AT&T DEB, and the IBM EGA displays among others were not based on square pixels.)

To display NTSC DV on a 4:3 screen one has to define the image in terms of non-square pixels. This was the typical mode of all analog and pre-HD/HDV digital camcorders including HDV.

Some camcorders offered a so called CINEMA mode to achieve a pseudo wide screen that retained the traditional SD pixel shape and in the process ignored about 120 scan lines of video (made it black). This liked like a letter-boxed wide image on a 4x3 SD set with a black band at the top and bottom of the image.

Some offered a 16:9 wide/Full screen option which implied sufficient pixels in the image sensor to retain a full 480 line image, and then fit the wider screen into 720 pixel by using a different pixel aspect ratio (wider pixel). HDV did this as well. On a 4:3 display this look squashed in at the sides.

Some camcorders offered all three modes as a user selected option, which can account for mixed modes on a tape. The display system (and editing system) needs to know how to interpret the image, and that information may or many not be encoded in the data stream or file header. Many TV sets allow the user to specify an aspect ration (and zoom level) for display as well. However, all this is transparent to a firewire copying process itself.

The attached extract from a Sony manual may help visualize this.