Home of Professional Cinematography since 1996

>
Progressive & Interlaced Scanning

>Published : 9th February 2005

>Hi,

>Can someone explain how exactly is interlaced scanning done? Why was it incorporated into video technology? Also can someone explain progressive scanning too. Is there a difference in sequential scanning and progressive scanning?

>I understand, In interlaced scanning each frame is divided into two fields, even and odd. And in PAL system we see 50 fields per second on out television sets. It is scanned in the same way as well. On the other hand progressive scanning scans one frame at a time. There are no fields in progressive scanning(???) Is progressive scanning done vertically or is the same horizontal way as in interlaced?

>If progressive scanning is better why was interlaced there at all...???

>I know there is a reason and am looking for a detailed explanation. Just want to clear out doubts...

>Thanking you in anticipation.

>Saurabha Goswami


>Saurabha Goswami wrote:

class="style9">>Can someone explain how exactly is interlaced scanning done? Why >was it incorporated into video technology?

>Hi Saurabha,

>Both interlace and progressive scanning are horizontally oriented. Basic technical difference: progressive mode scans every line from top to bottom (1,2,3,4,5,6,etc.) while interlace mode scans every other line (1,3,5... to bottom, then jumps back to top to scan 2,4,6, etc.). The first set of odd lines, then, is Field 1, the second even numbered set is Field 2.

>Interlace scanning was developed for video in order to reduce visible flicker. Though it sacrifices spatial resolution to skip every other line, it improved temporal resolution. Twice as many "full images" are presented per second. This is the electronic equivalent of having a two or three-blade shutter in a film projector. Even though film presents 24 different images per second, the projector's rotating shutter alternately blocks and reveals each of those images 2 or 3 times in order to speed up the number of images presented per second to the eye, thus eliminating visible "flicker".

>Look up "persistence of vision" for a more in-depth explanation of why. NTSC video displays roughly 30 fps (29.97 actual), but when interlaced, results in 60 fields being displayed per second. This method is abbreviated as 60i. Some newer digital systems display at 60p, which would be 60 full progressive scan frames displayed per second (for some interesting reading on the effects such higher frame rates have on the viewing experience, look into the "Showscan" film system developed in the eighties(?) by long-time special effects expert Doug Trumbull).

>Basic aesthetic difference: given the same bandwidth, interlace scan results in smoother motion, progressive scan results in sharper images.

As digital systems improve and bandwidth increases, progressive scan systems are developed with higher frame rates (and thus smoother motion). NTSC interlace was developed to be the best compromise image possible at the time, given the bandwidth limitation of broadcast video.

>Interlace is still with us because old ways die hard.

>Hope this helps.

>Regards,
Michael Petros
Assoc, Professor, Media Arts Dept.
Glendale Community College, CA


>Saurabha Goswami writes :

class="style9">>Can someone explain how exactly is interlaced scanning done? Why >was it incorporated into video technology?

>It was originally incorporated into video technology because it allowed more (subjective) information to be transmitted with less bandwidth. In a given span of time you needed to transmit only half the amount of video information.

>With old-fashioned analog pickup tubes, the odd horizontal lines were scanned first, then the even lines, and so forth. Digital chips do essentially the same thing.

>Each scan constitutes a field. One odd field plus one even field constitutes one frame.

class="style9">>Also can someone explain progressive scanning too.

>With progressive scanning all the lines are scanned in one pass, so the entire frame is transmitted at once rather than as two separate, sequential fields.

class="style9">>Is there a difference in sequential scanning and progressive >.scanning?

>I believe the term "sequential" means the same thing as interlaced -- the odd and even fields are transmitted sequentially.

>With the French SECAM system, "sequential" relates to the transmitting of color information. Each field contains the color information for only one of the three primary colors (red, green or blue). That color information is then held in memory and applied to the following two fields. SECAM stands for "Sequential with Memory" (Sequential avec memoire). It should really be SEQAM, I suppose, but the French love acronyms and would never put up with an awkward one! Besides, the CAM part could refer to a camera, so SECAM it is.

>As you may imagine, the original SECAM system had all kinds of color problems with fast onscreen movements. When digital processing came along this problem was eliminated, but I'm not sure how that was accomplished.

class="style9">>There are no fields in progressive scanning(???)...

>Correct.

class="style9">>And is progressive scanning done vertically? Or is the same horizontal >way as in interlaced?

>Same as interlaced, except all lines are scanned progressively (one after the other), starting with the uppermost horizontal scan lines and lines and ending with the lowermost -- relative to the image.)

class="style9">>If progressive scanning is better why was interlaced there at all...???

>See above. Another advantage of interlaced scanning is that for a given FRAME rate, interlaced will give you smoother lateral motion -- when you pan across a picket fence, for example, there will be less tendency for the vertical slats to "strobe" or "judder."

>I hope that explains it adequately. Others may wish to elaborate...

>Best wishes,

>Dan Drasin
Producer/DP
Marin County, CA


>Interlacing came as a result of two problems: transmission bandwidth and unacceptable phosphor decay solved (sort of) when the entire device (television) was in infancy. In order to achieve proper persistence of motion, when a frame was shown in whole, depending upon the engineering, a very unacceptable result visually occurred. The engineers discovered in the 1940's that they could halve the bandwidth required to transmit a complete frame (logically) if they divided and showed the frame in halves : exciting the phosphors in odd then even "fields" or portions of the frame.

>Progressive scanning is a result of improved technology and the attempt to remove lots of problems with interlacing such as motion artefacts, and blended frames (conceptionally similar to when you have your film footage transferred to video). A progressively scanned frame occurs all the way across the chip (if you're acquiring) and is projected all the way across the screen, top to bottom or (in some cases) bottom to top. As you presumed, in a progressive system, there are no fields. But this gets into murky territory. The problem is that we are in a transitional time in the technology and there are many ways of accomplishing an end result. I'm not entirely sure what you're referring to with sequential scanning, maybe someone can elaborate on this. The last time I heard of "sequential" scanning rather than progressive was with the "EDTV" concept... a progressive version of SDTV usually referred to as 480p. Again, to answer your question as to WHY 525i existed... it was a result of a need for an adequate solution.

All of this mess is why the HD transition is important (and federally mandated) because while our end of the world: the hi resolution film world, looks wonderful the end distribution method for more and more media is the consumer television. This is such a vastly complex topic that there's not any simple explanation...The fact is the current system is fifty years out of date and even the most uneducated consumer is demanding more from their television media: greater size, more quality, and a more immersive experience.

>Hope this helps and welcome and additions or corrections.

>Adoniram Sides
Cinematographer/Photographer
Massachusetts


>HI,

>Thanks a lot to Michael Petros, Adoniram Sides and Dan Drasin. I am really very grateful to all of you for the explanation of the whole thing in detail.

>My doubts are clear and have learnt some new things. Will be sending in some more queries soon!!!

>Thanks again...

>Saurabh Goswami
India