9 Replies Latest reply: Dec 21, 2013 5:26 PM by Netcommercial RSS

    Pixel Aspect Ratio and ciphering type of footage-CS6

    Netcommercial Community Member

      CS6- updated 12\18\13  V-11.03

       

      Hi All,

       

      A few questions about:

       

      Pixel Aspect Ratio and ciphering type of footage. Bare with me as I touch on two different aspects of what I feel are related:

       

       

       

      1. Setting up a Comp-

       

      Having AE decide on the correct footage being used- Via >Import footage,.(to Folder) then drag it on, or into the 'make comp icon' where AE is supposed to do a 'best guess' for the comp settings. Be it square pix or otherwise. What are the second value in brackets ? See pic

       

      wtf.PNG

       

       

       

       

      2. Interpolation-

      When using two diff footage sources- If using two different source footages in a comp. Is this where I should interplolate my footage to match the existing footage and the comp settings (If I have choosen NOT to pre-comp that secondary footage?)

       

       

      3. Seperating fields-

      How come footage that a camera manufacturer claims to be shooting in Progressive 30 or 24p is showing up as interlaced in my comp's CP? The manual to the cam says 30p or 24p or 60i The footage was shot at 30p (for sure)

       

      pic 7.PNG

       

      Here is a screen grab of AE's 'best guess' of that footage, showing it with an UPPER field render; indicative of HD Interlaced footage and then in the inerpolate settings - 'kinda confirming it.

       

      pic 1.PNG

       

       

      Now for the way out- Render:

       

      4. If that is in fact Interlaced and I then want to reduce the size of that 1920 footage to 66% size... Do I need to do anything in particular at this point?

      pic 6.PNG

       

      pic 2.PNG

       

      For some reason I am getting an error on my end when trying to export as an AVI and reducing size to 1280

       

      pic 3.PNG

       

      5. Is this from the Field render of interlaced challenge?

       

      I ended  up with a couple of errors and then aborted the whole file and re-imported to start over with this error. Which happend again after a redoux

       

      pic 4.PNG

       

      So I sent it over to Prem Pro to export those MTS Files as a movie and noted AGAIN the reference to Interlaced video. "upper and lower" Prem Pro was able to use the same MTS files and export a working movie. (so those files were not corrupt BTW) Yet Note PP's output reference to that footage again as Interlaced would be...

      pic 5.PNG

       

      Thanks for any clarity on those "5" questions... NC

        • 1. Re: Pixel Aspect Ratio and ciphering type of footage-CS6
          Andrew Yoole CommunityMVP

          Netcommercial wrote:

           

          1. Setting up a Comp-

           

          Having AE decide on the correct footage being used- Via >Import footage,.(to Folder) then drag it on, or into the 'make comp icon' where AE is supposed to do a 'best guess' for the comp settings. Be it square pix or otherwise. What are the second value in brackets ? See pic

          It's the pixel aspect, defined as a ratio against 1. So, in this case, (1.00) means the pixels are interpreted as 1:1.00, or square.  A Widescreen pixel might have a ratio of (1.67), meaning each pixel is 1:1.67, or 1.67 times wider than it is high.

           

           

           

          2. Interpolation-

          When using two diff footage sources- If using two different source footages in a comp. Is this where I should interplolate my footage to match the existing footage and the comp settings (If I have choosen NOT to pre-comp that secondary footage?)

           

          You set footage interpolation in the Project window for each footage item.  Whether or not the footage is precomped has no bearing on interpretation of footage.

           

           

           

          3. Seperating fields-

          How come footage that a camera manufacturer claims to be shooting in Progressive 30 or 24p is showing up as interlaced in my comp's CP? The manual to the cam says 30p or 24p or 60i The footage was shot at 30p (for sure)

           

          AE doesn't always guess the interpretation of footage correctly.  It relies on a set of most-common footage rules, but if you know your footage is not interlaced, set the interpolation appropriately.

           

           

          Now for the way out- Render:

           

          4. If that is in fact Interlaced and I then want to reduce the size of that 1920 footage to 66% size... Do I need to do anything in particular at this point?

          pic 6.PNG

           

           

          Don't use the rescale output function except in special circumstances, and NEVER on interlaced output.  Scaling interlaced footage results in field scramble that will destroy your output.

           

          Precomp your output composition into a comp of the correct output dimensions and render that.  Or, render a lossless full-res version and use Media Encoder to resize your delivery files.

           

           

           

          I'm not clear from any of your questions exactly what format you are trying to output, but you will probably have more success rendering using Media Encoder, or rendering a lossless file from AE and compressing your delivery files from Media Encoder.

          • 2. Re: Pixel Aspect Ratio and ciphering type of footage-CS6
            Netcommercial Community Member

            Andrew thanks for hoppin in here and answering some of this.

             

            REF #1 I realize that (1) is a square pixel- What is the second one? See Pic (480x270)

             

            REF#2 I understand how to interpolate footage, so I am deducing that I MUST Interpolate each piece of footage whether or not it is pre-comped or as a 2nd source?

             

            REF#3 Separate fields, So AE by the pic, is guessing wrong, and IS in fact guessing it is Interlaced by the field render UPPER in the project window pic--Then Prem Pro pic is not what I thought it was, as far as Interlaced via Upper to Lower (see pic last one in first post) One would think it is my camera manufacturer not being truthful if BOTH AE and PP guessed it as Interlaced. OR am I mis-interpretating the panels?

             

            Thanks for the suggestion on a resize of nesting in a smaller comp size or using ME on an Uncompressed AVI

            • 3. Re: Pixel Aspect Ratio and ciphering type of footage-CS6
              Rick Gerard CommunityMVP

              Ref #1, When you have some formats open in a media viewer  (qt player or windows media player) and  you view it at less than full screen and close the media player the meta data of the last state of the player is saved. That's what you are seeing there. It can foul up AE. Open the file in the media player and view it at full size, close the media player and import the file again in AE and the problem should go away. (this is an old bug that adobe has nothing to do with.

               

              Ref #2, You should never change the interpretation of Pixel aspect ratio on footage  unless you know for a certainty that it is wrong. For example, open Photoshop or Illustrator and create a new document that is 720  X 480 pixels and AE will always interpret the image as rectangular pixels because that is a standard rectangular pixel frame size. Unless you specifically created the image in Photoshop with rectangular pixels it is square pixels so the interpretation must be changed to avoid distortion. The same holds true for the other non square pixel frame sizes. If your image is a standard non square frame size AE will always assume that it is non square pixels. If you know for sure that your image is square pixels then you must change it. If the non-square image came from a camera then it is non square pixels. The only time you may need to change it is if the footage is NTSC or DV widescreen and it is being interpreted as 4:3 instead of wide screen.

               

              The frame rate is usually properly interpreted. Only if you want to match frame for frame, or slow down HFR footage would you change the frame rate. Changing the frame rate WILL CHANGE the audio playback so sound will be out of sync or have other problems. Again, you must know what you are doing here and you must make intelligent decisions based on the requirements of your project.

               

              Ref#3, Separating fields must be done right. Some 1080P footage is interpreted as having fields. This may be incorrect and if there is a question you MUST test the footage. Test the footage by making sure that the fields are being separated then selecting the footage in the project panel and choosing Create Composition from Selection, then opening the Composition settings, doubling the frame rate and stepping through the footage a frame at a time. If the footage is really interlaced then each of the frames will be different. If the field order is reversed then motion will be in the right direction then reverse, then go back to the right direction.

               

              Field order problems generally come from SD footage from capture cards. Most of the time interlaced footage from HD cameras is correct. You can get field order problems from trans-coded footage so here again, you have to know what the source is and if in doubt, check the footage.

               

              Film Transferred to tape or early 24fps DV footage may have 3:2 pulldown applied. This is a whole other kettle of fish that must be completely understood and if the footage is to be moved in the frame, scaled, or manipulated in just about any way except simple color correction, the 3:2 pulldown must be removed and the footage converted to progressive footage with individual distinct frames. You must learn how to do this correctly and verify every take in the project before proceeding with your work in AE.

              • 4. Re: Pixel Aspect Ratio and ciphering type of footage-CS6
                Dave LaRonde Community Member

                At 1080, a lot of footage is interlaced, and evidently your camera made interlaced footage.  Even if you know for a fact that it was shot 30p, the camera captures the entire frame... but it records it as two fields. They give it a fancy name: Progressive Segmented Frame.   AE will treat it as an entire frame if you interpret it as having a field order of None.

                 

                Old DV cameras capable of shooting 30p recorded the footage as interlaced, too.

                 

                How about your camera shooting 24P?  You'll have to remove the pulldown before you begin any kind of work on it.

                • 5. Re: Pixel Aspect Ratio and ciphering type of footage-CS6
                  Netcommercial Community Member

                  Thanks for the details Rick. 

                   

                  Thanks for the insight on  #1

                   

                  Ref #2, You should never change the interpretation of Pixel aspect ratio on footage  unless you know for a certainty that it is wrong. For example, open Photoshop or Illustrator and create a new document that is 720  X 480 pixels and AE will always interpret the image as rectangular pixels because that is a standard rectangular pixel frame size.

                  Unless you specifically created the image in Photoshop with rectangular pixels it is square pixels so the interpretation must be changed to avoid distortion.

                   

                  --Did you mean to write:

                  Unless you specifically created the image in Photoshop with SQUARE pixels using  a rectangle layout?  I am not trying to bust your blz here, I am trying to understand the unknown. (for me)

                   

                   

                  So it may very well be AE mis-guessing the footage and the Upper first is a mistake on AE's part and I need to test it to find out? On that note...

                   

                  Ref#3, Separating fields must be done right. Some 1080P footage is interpreted as having fields. This may be incorrect and if there is a question you MUST test the footage. Test the footage by making sure that the fields are being separated

                   

                  How do I make sure the fields are being seperated? Does the window as in pic two from the top of my post confirm that or is there another test?

                   

                  Then I do the following:

                  then selecting the footage in the project panel and choosing Create Composition from Selection, then opening the Composition settings, doubling the frame rate and stepping through the footage a frame at a time. If the footage is really interlaced then each of the frames will be different.  (IE:  29.94 to 59.94fps) OK

                  If the field order is reversed then motion will be in the right direction then reverse, then go back to the right direction. ---I did not understand?  Are you referring to Upper and Lower reverse?   Direction, as in fwd and back in Time line? Sorry unlclear on my end...

                   

                   


                   

                  Oh and HFR footage? HFR? High frame rate??

                   

                  One last thing, Do all cameras shoot anamorphic, or could they as AE is showing my MTS files; are Square? (1.00) Pic #1

                   

                   

                   

                  @Dave-  You -

                  At 1080, a lot of footage is interlaced, and evidently your camera made interlaced footage

                  Me- How do you know? because of the 2nd pic and upper field first indication?

                   

                  You-Even if you know for a fact that it was shot 30p, the camera captures the entire frame... but it records it as two fields. They give it a fancy name: Progressive Segmented Frame. 

                  Me- Is this typical false advertisement of the sales divison of brands like not telling you what CMOS or CCD is in a cam just slapping HD 1080p on the side of the box?

                   

                  You-AE will treat it as an entire frame if you interpret it as having a field order of None.

                  Me- Won't that be a problem mis leading AE on an interpret?

                   

                   

                  Thanks Gents for the precise details.

                   

                  Best

                  NC

                   


                  • 6. Re: Pixel Aspect Ratio and ciphering type of footage-CS6
                    Dave LaRonde Community Member

                    Me- How do you know? because of the 2nd pic and upper field first indication?

                     

                    Yes.

                     

                     

                    Is this typical false advertisement of the sales divison of brands like not telling you what CMOS or CCD is in a cam just slapping HD 1080p on the side of the box?

                    It's not any worse than camera makers saying their cameras shoot 23.98 frames/sec.  Well, they don't: they shoot at 23.976.  Besides, your footage apparently IS progressive video; it just happens to be frames of progressive video recorded as two fields, and when you look at such video on a TV set, it looks like progressive-scan.

                     

                     

                    Won't that be a problem mis leading AE on an interpret?

                     

                    No. 

                     

                    You just have to keep it straight in your head what footage was shot in progressive scan and what was shot in interlaced scan.  If you can do that, you'll be fine.  People have been dealing with it for years, since the first DV camera came out with progressive scan.  There's no reason why you can't.

                    • 7. Re: Pixel Aspect Ratio and ciphering type of footage-CS6
                      Rick Gerard CommunityMVP

                      On the 720 x 480 issue, unless you specifically tell Photoshop that you are using rectangular pixels, they are all square pixels. No one should ever intentionally create rectangular pixel artwork for motion graphics.

                      • 8. Re: Pixel Aspect Ratio and ciphering type of footage-CS6
                        Rick Gerard CommunityMVP

                        Sorry for the double post. I'm using my phone, so I can't go back and edit.

                         

                        About the interlacing issue, just read my instructions. If the footage is interlaced you will see a difference on every single frame that has motion in it when you double the composition frame rate. The info at the top of the project panel will tell you whether or not the footage has been separated into fields. You will also see if the footage has been separated and it feels when you open up the interpret footage panel.

                         

                        Let's just simplify everything here. You have to know where your footage came from and how it was produced. If you don't, then you need to figure it out by running some tests. If you properly interpret your footage and always work in square pixel standard size compositions your artwork and your project will never be distorted and always look right on any display. If you work with non-square pixel compositions your rendered project will be distorted and be wrong on every device except the specific device for which the non-square pixel format was originally designed.

                        • 9. Re: Pixel Aspect Ratio and ciphering type of footage-CS6
                          Netcommercial Community Member

                          Thanks Rick

                          I did not know you could have AI or PS create in anything but Square Pixels. Now I know how to create even more problems. LOL! Got it.

                           

                          To be clear-The test to see if it is Interlaced is the doubling of Frame Rate and will be reflect in the window as in pic 2 of seperating fields.

                           

                          However with Dave's reply...My Progressive footage is really interlaced- but shot progressive. Which should be treated as Progressive, albeit quazi interlaced ( in each frame)

                           

                          Then Why is Ae seperating fields (pic2) if the footage is Progressive and only Interlaced inside of each frame as opposed to frame1-upper frame2-lower Interlaced.

                          Or am I off base?  Thanks for staying tuned BTW.

                           

                          Interesting tid bit there Dave about yet another thing to look for when buying a camera(s). One now has to try and figure out... Is it really P or is it IP as you stated. Is it only FIlm or a 40k camera that will shoot real P?  This explains my challenge when tracking in MOCHA if they are Interlaced progressive frames and not progressive as I thought I was using...Or does that even matter? As you stated tell AE or MOCHA it is P even though it is interlaced Progressive due to CCD or CMOS digital I am assuming...

                          Thanks again Gents

                          NC