Pseudo-progressive 'frame' modes

[Bnew]

My camera has an option for a pseudo-progressive 'frame' mode, where both fields are captured at once - what are the implications of this?

Does it look any better, and it does it look okay on standard displays, are both fields displayed at once? Or is this just a useless gimmick that should be avoided..

[eddie123456789]

Ive used this with the PD150.
I cant say I noticed any difference though.
This mode captures both fields at once, as you say, so if you take a still frame you dont get
of that horizontal 'striping' (I forget its technical term).
I expect it takes more memory (twice as much?) to save a file, and you may (you may not of course) possibly have dropped frame problems when trying to capture it to your edit suite due to the much higher data rate.
When I used it I was doing some animation, capturing a frame at a time, so it made sense to try and use it, but I didnt notice any difference. Perhaps the resultant image is *slightly* sharper...
Perhaps someone else will know more.

[sn-films]

Originally posted by eddie123456789:
This mode captures both fields at once, as you say, so if you take a still frame you dont get of that horizontal 'striping' (I forget its technical term).

<font size="2" face="Verdana, Helvetica, sans-serif">"Interlacing"

I expect it takes more memory (twice as much?) to save a file,

<font size="2" face="Verdana, Helvetica, sans-serif">Its the same data rate. 30 frames / 60 fields (NTSC) are still written to the tape. The difference is the timing of when the fields are captured by the camera.

and you may (you may not of course) possibly have dropped frame problems when trying to capture it to your edit suite due to the much higher data rate.

<font size="2" face="Verdana, Helvetica, sans-serif">Again, its the same data rate. One frame of standard video is actually two fields (alternating lines). Normally, these two fields are captured a split-second apart from each other, which is why you see the interlacing artifacts when you extract stills from interlaced video. In the psuedo-frame modes, the image is technically still two fields, but instead of the fields being captured a split second apart from each other, they are captured at the same time and written sequentially to the tape. The difference between the two modes (progressive vs. interlace) will be more apparent during shots where there is a lot of movement (either characters or the camera).

There are three reasons I can think of off the top of my head for using the progressive frame mode:

1) You intend to pull a lot of stills from the video

2) Artistic reasons: the motion is supposed to be a closer approximation to film.

3) The final product will be progressive, not interlaced

BNEW - Which camera are you using? Of the cameras I've seen that have this feature, I think the Panasonics do it the best. It's my understanding that the progressive mode is better for making web-based content and progressive DVD's since you won't need to de-interlace the content for delivery.

[curtinparloe]

The horizontal striping seems to be called miceteeth (101fps.com). Interlacing is the technique of alternating the frames, but I believe the terms overlap at times...

I was curious. If progressive means that both frames are captured simultaneously, do the frames capture the same image, or a slightly different one? If it was the same, then surely there wouldn't be any point using it, so I'm guessing the latter, in which case, does a CCD usually capture alternate rows, or is a progressive-capable CCD double-height?

[sn-films]

Originally posted by curtinparloe:
The horizontal striping seems to be called miceteeth (101fps.com). Interlacing is the technique of alternating the frames, but I believe the terms overlap at times...

<font size="2" face="Verdana, Helvetica, sans-serif">Actually, I think you meant 100fps.com.

First, let me preface this by saying that anything written to tape or broadcast over the air is interlaced. Interlaced video signals consists of two fields per frame. When you export 1 frame of video as a static image, the image consists of two fields merged together. With a traditional interlaced video signal, the two fields are from two concurrent moments in time, and any movement between when fields A and field B were created shows up as interlace artifacts, or what your author refers to as "miceteeth".

I was curious. If progressive means that both frames are captured simultaneously, do the frames capture the same image, or a slightly different one? If it was the same, then surely there wouldn't be any point using it, so I'm guessing the latter, in which case, does a CCD usually capture alternate rows, or is a progressive-capable CCD double-height?

<font size="2" face="Verdana, Helvetica, sans-serif">

I think you're confusing Frames and Fields. In the wonderful world of video, there are two Fields in every Frame.

Believe it or not, a CCD on a DV camera ALWAYS captures progressive images. I've been reading about a small group of people who are writing firmware hacks for their video cameras to leverage this, but thats a different story for a different thread. Back to the topic...

The CCD of an NTSC video camera captures 60-progressive frames-per-second, while a PAL video camera captures 50 progressive frames-per-second. However, the buffered signal is written to the tape in a television-friendly interlaced format. This means the every other line of the first progressive image from the CCD becomes Field A of Frame 1 on the tape, and every other line off the second progressive image from the CCD becomes Field B of Frame 1 on the tape. Alternating lines from the third progressive image off the CCD becomes Field A of Frame 2 on the tape, etc. That's it. That's the way it works. Always and forever until tape finally goes away.

On to Frame-Mode...

The original frame-modes on video cameras still buffered 60 progressive frames-per-second off of the CCD. The first progressive frame off the CCD gets written to Field A of Frame 1, the second image off the CCD gets written to Field B of Frame 1 on the video tape, etc. There is nothing different about the way the information is written to the tape. The trick occurs when the images get buffered off of the CCD and before the buffer is written to the tape. A copy of the first image off the CCD overwrites the second image in the buffer, the third image copies itself over the fourth, etc. resulting in duplicate pairs of images in the image buffer. Essentially, the even numbered images off of the CCD are discarded and replaced by the preceding image. When the buffer is written to the tape, the first image off the CCD is written as Field A of Frame 1, the second image (which is a copy of the first) is written to Field B of Frame 1, etc. The result is that you do not have interlace artifacts in a frame since both fields represent the same moment in time, instead of being from two different moments in time. This results in a crisp picture, but the fast motion will be less fluid than a traditional interlaced signal since every other progressive frame off the CCD is discarded. This mode is ideal for "talking-head" content that you want to stream over the internet.

Then comes slow-shutter...

This is an attempt to overcome the stutter that results from discarding half of the progressive frames off the CCD in Frame-Mode. In slow shutter mode of 1/30 on an NTSC camera, the "shutter" is opened for a longer period of time (1/30 of a second instead of 1/60). Technically speaking, the CCD's sample data over a longer period of time (one of the benefits of slow-shutter is that the light accumulates over a longer period of time on the CCD longer and allows you to shoot in darker conditions). Using 1/30th shutter speed, there are still 60 progressive frames buffered off the CCD, with the first and second progressive images being identical. This is similar to the frame-mode described above, however each image represents 1/30th of a second instead of 1/60th. Since the CCD collects data over a slightly longer period of time, any motion will have a more noticable blur than a video signal based on the standard 1/60 shutter speed.

The difference between the original frame-mode and slow-shutter modes is a trade-off between sharpness and smooth motion. A frame-mode that drops every other progressive frame coming off the CCD will be sharper than something shot with a slow-shutter mode, but the slow-shutter mode used in conjunction with the frame-mode on the camera will create a closer approximation to the properties of film in regards to motion.

As far as I know, all new cameras that feature a progressive or frame-mode use a slow-shutter rather than dropping frames off of the CCD.

Hope this helps.

[Bnew]

Thanks sn-films, a lot of help.

Mine's a panasonic MX300B [img]graemlins/thumbs_up.gif[/img]

Just to clarify, if kept interlaced (say if i had to output to VHS) it will still display fine on a TV, but also a progressive DVD will be fine? Never understood how progressive DVD's display okay on TV's.

[curtinparloe]

Originally posted by sn-films:
Actually, I think you meant 100fps.com.
The CCD of an NTSC video camera captures 60-progressive frames-per-second, while a PAL video camera captures 50 progressive frames-per-second. However, the buffered signal is written to the tape in a television-friendly interlaced format. This means the every other line of the first progressive image from the CCD becomes Field A of Frame 1 on the tape, and every other line off the second progressive image from the CCD becomes Field B of Frame 1 on the tape. Alternating lines from the third progressive image off the CCD becomes Field A of Frame 2 on the tape, etc. That's it. That's the way it works. Always and forever until tape finally goes away.

On to Frame-Mode...

The original frame-modes on video cameras still buffered 60 progressive frames-per-second off of the CCD. The first progressive frame off the CCD gets written to Field A of Frame 1, the second image off the CCD gets written to Field B of Frame 1 on the video tape, etc. There is nothing different about the way the information is written to the tape. The trick occurs when the images get buffered off of the CCD and before the buffer is written to the tape. A copy of the first image off the CCD overwrites the second image in the buffer, the third image copies itself over the fourth, etc. resulting in duplicate pairs of images in the image buffer. Essentially, the even numbered images off of the CCD are discarded and replaced by the preceding image. When the buffer is written to the tape, the first image off the CCD is written as Field A of Frame 1, the second image (which is a copy of the first) is written to Field B of Frame 1, etc. The result is that you do not have interlace artifacts in a frame since both fields represent the same moment in time, instead of being from two different moments in time. This results in a crisp picture, but the fast motion will be less fluid than a traditional interlaced signal since every other progressive frame off the CCD is discarded. This mode is ideal for "talking-head" content that you want to stream over the internet.

As far as I know, all new cameras that feature a progressive or frame-mode use a slow-shutter rather than dropping frames off of the CCD.

Hope this helps.

<font size="2" face="Verdana, Helvetica, sans-serif">Thanks for the info, exactly what I was getting at. Interesting that half of the information entering a DVCam is just discarded. What a waste. If I were a cut-throat cam producer, I'd half the size of the CCD and use all the information...
pseudo-progressive sounds like the slow-shutter method to me, I'm after true progressive (PD170?) when I shoot my next short (second week in June touch wood).