What’s the real diff. w/ Interlaced and Progressive Scan?

Videomaker – Learn video production and editing, camera reviews Forums General Video and Film Discussion What’s the real diff. w/ Interlaced and Progressive Scan?

Viewing 6 reply threads
  • Author
    Posts
    • #39370
      Avatarnate3po
      Participant

      I sort of know the basic difference between interlaced and progressive scan video? The former alternates between two fields and the later shows the whole frame at once. But is there always only one field on the screen at any give point in time. When the second field is displayed on the screen does the first field disappear? If so, is this how video is video recorded on camcorders that are interlaced. Does it capture images one field at a time or does it capture a whole frame and then interlace it? Or is any particular field captured from a moment in time that occured about 1/60 of a second after when the previous field was captured. The reason I asked is because I’ve had problems with deinterlacing video (esp. animated DVDs). Lines often look jagged when I set programs light VLC player or Nero Showtime to deinterlace DVDs I’m watching. But when I set the players not to deinterlace I’ve noticed that most frames actual look sharp and normal except in cases where there is a lot of movement and I noticed that I can see two fields from two different frames in one frame. And I read somewhere that deinterlacing video actually reduced the quality of the video a little. So what’s the deal.
      Also is there a difference between deinterlacing a video I’m rendering as oppose to outputing it as progressive scan. Does outputing it as progressive scan somehow deinterlace the footage wrong. I have Adobe Premier Pro 1.5 and I’ve noticed that I can output the video as lower field first, upper field first, progressive scan or if I select deinterlace footage those three options are taken away. What’s the deal?

    • #170290
      AvatarAnonymous
      Inactive

      Lets see if I can explain it a little differently and you can make new judgements based on it.

      An image is drawn on any screen (TV or monitor) by a single dot of light, (OK, color uses 3). That dot moves across the screen turning on and off until it reaches the other side of the screen. It then turns off, moves back to the first side, moves down and starts again. The on and off durning the scan creates the image seen. Pretty straight forward. The screen is chopped into 480 lines from top to bottom. Try to image all those lines numbered from top to bottom. Interlaced video will draw all the odd numbered lines first, then go back and draw the even. Progressive will draw them in order. Most televisions use interlaced. It tends to provide smoother looking motion but shakes when paused. Thats because only half of the picture is drawn per field. When the next field is drawn in to complete the image, whatever moving object is in the image will be drawn slightly out of alignment. When these 2 fields are displayed back and forth as in paused video, the shift creates the blinking. Progressive draws the image entirely before starting the next frame. Therefore, paused video won’t blink.

      Typically, deinterlacing will literally delete 1 field and fill in the missing space with an identical duplicate of the remaining field. Yes, some detail is lost therefore reducing the quality of the video. Converting interlaced to progressive will keep all the fields, it will just draw both at the same time. So instead of it drawing 1-3-5-2-4-6, it will mix them into 1-2-3-4-5-6. Progressive doesn’t have fields. Same is true with deinterlacing. It draws the entire image in 1 shot (1 frame). Deinterlaced draws 1-1-3-3-5-5 and so on. Interlace has upper or lower fields first. Upper means it will draw the odd fields first. Lower means even fields first. In some cases where an object in the video is so small that it only appears in 1 field, such a an electrical wire on a telephone pole, it might look like its blinking because its not drawn in the second field. Deinterlacing will either double its size or remove it.

      Watching interlaced video on a computer monitor really reveals the interlacing. Thats why an option for deinterlacing is included with software DVD players.

      I hope this info was helpful.

    • #170291
      Avatarpaulears
      Participant

      you need to go back to the history – interlaced scan exploited the phosphor persistence problem by illuminating one field, then then next – so the phosphors were excited and blurred the swap from one frame to the next – the fact each was only half the maximum resolution didn’t matter too much. As technology got better, they were able to refresh the fields quicker – usually double the rate and this cut down flicker even more. The snag with this interlaced system is that anything that moves between frames tends to twitch between the two fields. Film exposes a complete frame at a time, progressive scan is very similar.

      So your question on the disappearing field is quite right – on a crt, the old image fades away, while the new one is being built – a kind of very fast crossfade, hiding the join. So although we talk of 30fps, we’re talking 60 half resolution frames per second, really. So shooting at full format progressive is similar to slowing the shutter speed – much more care has to be taken with movement – shakes and vibration are a major problem, and you have to modify your shooting technique.

      When TV technology really took off, here in the UK we did a have a few issues with picture quality – our 25fps/50 fields system found fast motion difficult to reproduce. Watching Formula 1 on a modern set with greater resoloution, and less phosphor persistence enabled the snags with hih shutter speed cameras to be seen. The camera technology had got a lot better and high shutter speeds were available that really sharpened up moving images – the problem being the tv’s couldn’t cope!

      I’m not sure that really answered the question – did it?

    • #170292
      Avatarnate3po
      Participant

      Thanks, that does help and the explaination and history was quite enlightening. It does indeed answer one of my questions. I like how you describe it as 60 half resolution frames per second. What I also want to know is if camcorders capture images at 60 half frames per second. Or does it capture images at 30 fps and then record each frame at 2 half frames. Deinterlacing video renders it as progressive scan and theoretically these sets two half frames should be combined to 1 complete frame, right? But if half frame 2 was captured on the CCD 1/60 of a second after half frame 1 then combining these two half frames would be merging two points in time that occured 1/60 of a second apart. Granted 1/60 of a second is virtually nothing but I want to know if half frame 1 and half frame 2 contain information recieved the the CCD pixels at the same 1/60 of a second. Essentially, if I have footage that is interlaced, would it be better to deinterlace the footage, and encoded the video to DVD as progressive scan video or is it better to leave the interlaced video as interlaced on the DVD.

    • #170294
      AvatarAnonymous
      Inactive

      A half frame is called a field. 2 fields make up the frame. "Half frame 1" is the odd field. "Half frame 2" is the even field. I guess you missed my explaination about losing resolution when deinterlacing. If your camera is set to record interlaced, the image sensor scans that way. 30 FRAMES per second (60 FIELDS). So in answer to your question, NO. The 2 fields are not scanned during the same 1/60th of a second. The first 1/60th scans the odd, and the next 1/60th scans the even. Thats why a paused image "twitches". Converting interlaced to progressive is NOT the same as deinterlacing. Deinterlacing loses resolution, converting to progressive does NOT. Converting to progressive will not remove the "twitches" between fields. It simply changes the way the fields are stored in the video file.

      In all my years of editing video I’ve never seen a real need to convert interlaced video. I’m not saying a real need doesn’t exist, but for most home/consumer needs, its really not necessary. If your watching DVDs on your computer monitor, you’re going to have either the interlaced lines that you’ve been seeing, or the jaggedness of deinterlacing. Solution: watch them on a tv.

    • #170295
      Avatarnate3po
      Participant

      kkmac Wrote:

      A half frame is called a field. 2 fields make up the frame. "Half frame 1" is the odd field. "Half frame 2" is the even field. I guess you missed my explaination about losing resolution when deinterlacing. If your camera is set to record interlaced, the image sensor scans that way. 30 FRAMES per second (60 FIELDS). So in answer to your question, NO. The 2 fields are not scanned during the same 1/60th of a second. The first 1/60th scans the odd, and the next 1/60th scans the even. Thats why a paused image "twitches". Converting interlaced to progressive is NOT the same as deinterlacing. Deinterlacing loses resolution, converting to progressive does NOT. Converting to progressive will not remove the "twitches" between fields. It simply changes the way the fields are stored in the video file.

      In all my years of editing video I’ve never seen a real need to convert interlaced video. I’m not saying a real need doesn’t exist, but for most home/consumer needs, its really not necessary. If your watching DVDs on your computer monitor, you’re going to have either the interlaced lines that you’ve been seeing, or the jaggedness of deinterlacing. Solution: watch them on a tv.

      Thanks kkmac. Sorry, somehow I completely missed your first message and only saw paulears. You first and second explaination both help a great deal.
      I guess I was also thinking of the future and how televisions will eventually all be progressive scan. But if I understand you right, if I were to convert my videos to progressive scan I’d be combining two half images that were actually shot at two different points in time, right? So outputing to progressive scan alters the image of each frame as well, right? This means that interlaced video isn’t even 30 fps (or rather isn’t 29.97fps) because of frame is actually actually the combination of two fields that were shot at different times.
      But then what happens when you play an interlaced DVD on a DVD player and output the video to the TV as progressive scan. Does it convert the video to progressive scan or does it deinterlace the video show each frame as 113355 as you put it? Actually I suppose it would 224466 for me since I work with NTSC and I believe the lower fields are drawn first.
      Thanks again for your help and I hope you don’t mind me asking more questions for further clarifications.

    • #170293
      AvatarAnonymous
      Inactive

      Video is 30 frames per second, period. This means 60 fields per second. (NTSC of course). The 29.97 fps is (more or less) the way timecode labels each frame. Actuality, there is a frame that gets dropped every so often, but they are far apart and no one every actually sees when it happens.

      When video gets deinterlaced, an option is sometimes presented as to which field to remove, odd or even? If odd is removed, then 224466 will be the correct drawing method used. The first even line will replace the first odd line, then the first even line is drawn again where it should be. 22 instead of 12. 44 instead of 34, and so on. Thus the loss in resolution.

      Converting to progressive won’t change the image shift between odd and even fields. Instead of drawing all the odd and then all the even, it will draw 1 odd then 1 even, then the next odd then the next even. If you could freeze 1 frame of interlaced video and the same frame of converted to progressive, they would look identical. The only difference is the order in which they are drawn. The only real way to get progressive video is to film it that way. This means using a camera that uses progressive image scanning. As like a tv, a video camera utilizes the same method when seeing the image. Behind the camera lens is an image sensor which is basically a grid of light sensitive pixels. In order to save the info onto tape, there must be some type of scanning. If you fish through your camcorders settings, you’ll find the framerate control. This can be set fron 1/30 to 1/4000. Some cameras have a bigger range. This is the actual scan rate. Now don’t think that 1/4000 means 4000 frames per second. It scans at 4000 frames per second but only stores onto tape 1/30. Why? For fast moving action, 4000 will create a very sharp and non-blurred image versus 1/30. Sure, 3999 scans get lost and playback does look a bit choppy, but even with interlaced video, you’ll never see the interlaced lines. Some may even argue that it tends to represent film since film displays 24 fps and there are no line to draw. Just the whole image. Experiment and you’ll see. By the way, you need alot of light at that framerate.

      Technology is advancing so fast its tough to keep up. TVs of the future will probably all be progressive scan. Planning for it now is good, but understanding the difference is more important. Once your content is on tape, not much can be done to change it. Conversions work mostly for display purposes only. They really don’t change the content itself. Thats why DVD players usually have options. Until the format is standardized, users have the options to select what’s best for them. It really comes down to trial and error. Now that you have a more than basic understanding of this, experiment with it. You will ALWAYS remember what you’ve done versus reading what others have done!

Viewing 6 reply threads
  • You must be logged in to reply to this topic.

Best Products

Stock media: How to find and use the best stock photos, videos, music and more

Acquiring stock media and editing it into your project is very often faster, cheaper and much more convenient than creating that media yourself.
homicide-bootstrap