HDV: Drop Frame or non-drop frame & why?
 

Please, I need a quick education on the topic of using/choosing drop frame versus non-drop frame on a Premiere Pro HDV project especially as it’s related to or may directly impact on rendering/compression quality or may even cause compatibility issues with the source footage.

Bottom line: Is there any visible or discernable difference seen on the resulting video after being rendered/compressed when played back?

What should I know and keep in mind in regards to this aspect of video processing going forward?

Thank you.

Relax, the drop/non-drop just affects the appearance of the SMPTE time code numbers. It doesn't affect your video and no frames are dropped. It all comes from the fact that 30 frames in broadcast is really 29.97 frames and to keep the timecode from going wonkey in broadcast timing, there is an adjustment in the INDICATED time. There is a good explanation here: SMPTE time code - Wikipedia, the free encyclopedia

The term “Drop Frames” refers only to how frames are counted in NTSC, there are no actual frames being dropped from the video.

Here is my explanation of 29.97fps in Drop-Frame and Non-Drop:

If you use “Non-Drop” method, you would count 30 frames in a second. So in one hour of video you have 30fps * 60sec * 60min = 108000 frames.

In the “Drop Frame” method, you still count 30 frames per second, but you skip counting 0 and 1 for frames in each minute except of those minutes that divide by ten.
Here is an example of frame count when passing through a minute in the Non-Drop and Drop Frame method:

ND: …27, 28, 29, 0, 1, 2, 3…
DF: …27, 28, 29, 2, 3, 4, 5…

In Drop-Frame method by skipping to count frames as described, the running time is more precise (true), where in Non-Drop method longer your video is more error you have in your time vs. real time.
Example, one hour video in non-drop format and drop-frame format:

Non-Drop format:
30fps * 60sec * 60min = 108000 frames in an hour.
The actual frame rate as we all know is 29.97 frames per second, or the duration of each frame is 1 / 29.97 = 0.033366666 of a second. So if we multiply number of frames in an hour by duration of each frame, we should get number of seconds in an hour: 108000 frames in an hour * 0.033366666 = 3603.6. There are 3600 seconds in an hour, but we got 3603.6, an error of 3.6 seconds or 108 frames. That translates into error of 86.4 seconds or error of 2592 frames in 24 hour period.

Drop-Frame format:
As described above, each hour needs 108 frames correction. So DF hour has 10800 – 108 = 107892 frames. Therefore 107892 frames * 0.033366666 (duration of a frame) = 3599.996328 seconds, an error of -3.672mS per hour or -88.128mS per 24 hours (2.6 frames per day).

As mentioned in above post, the drop-frame method is used in broadcasting to indicate true running time.

Good explanations.

Here's the deal in a nutshell.

Time codes are just counting schemes that help you keep track of frames.

If you are doing something where you have video from multiple cameras and are working with other folks so and you all need to have a common frame of reference, drop frame is usually the lowest common denominator. (That is what most consumer cams use to report time code). For example, maybe you need to tell another editor on the other side of the contry to work with the footage from camera number one at timecodes 29:11;15 to 36:00;02. You both need to use the same time code format to find the exact clips you want to work with.

Apart from that kind of book-keeping,, it will not matter. The choice of time code has no effect on compression or rendering quality and will make no discernable difference to the edited video.