sequence setting confusion

[Alan Okey]

SD video uses nonsquare pixels – pixels that are taller than they are wide. 720×480 will need to be displayed as 4:3 on a computer screen (i.e. 640×480) in order to appear correct, as computers use square pixels. 3:2 only describes the ratio of horizontal to vertical pixels, not the visual proportions of the image. The visual proportions are always 4:3, regardless of whether square or nonsquare pixels are used.

[Anthony Faulkner]

Right, but why am i having to use two different pixel ratios to get images that display correctly?

are there settings somewhere that i am neglecting?

[Beau Tardy]

Alan is right about the pixel sizes. This will impact how your image displays. Not sure why you are using 2 settings? What do you mean? Are you rendering and then re-importing?

You should find a sequence setting that displays your movies correctly, but be aware that when you export at ‘Current Settings’ it might look different depending on the monitor you are watching it on. What is your final output and display device? DVD? Bluray? SDTV? Web…

[Alan Okey]

Anthony,

If you select the first sequence in the browser and right-click (command-click) it and choose “duplicate,” then put your second clip in the duplicated sequence, does it display correctly?

I’m not sure why you created two SD sequences with different settings. It’s probably better to duplicate an existing sequence and replace the media in the timeline than to create a new sequence. It can be easy to miss something or mistakenly choose the wrong setting when creating a new sequence.

In any case, you should be aware that you will get better quality by not using FCP to downconvert your HD clips to SD. FCP’s resizing quality is poor. It would be better if you used Compressor to transcode your HD media to SD first, then import the SD media into an SD FCP project. It’s also not necessary to use ProRes HQ for anything but very high end RGB HD, such as from a Red camera or a Sony f900. In fact, several users on this forum report that using ProRes HQ when it’s not appropriate may cause problems.

Use Compressor to transcode your HD footage to standard definition ProRes, then create a new SD ProRes project in FCP and import your transcoded footage.

How are you handling the conversion from 16:9 to 4:3? Are you letterboxing or cropping?

[Anthony Faulkner]

Neither video will copy into the other sequences without distorting. That is what is confusing me. It is like the source material was actually different, yet I know the cameras were both set to 1080P30.

I am cropping my images. And i shot with that in mind.

All videos are for web release.

A lot of the work i do is composite work, so i wouldnt leaving it in HD until i had keyed would yield a better quality key?

I just tested the compressor setting you recommended and they look very nice.

now here is where it gets funny….

If i take the clip i have downsized in compressor (50%,960×540 pro res) and dump it on top of the same clip that displays correctly in the sequence (3:2, NTSC CCIR601, square) it stretches the new clip vertically. yet the same clip underneath is perfect.

If i highlight both clips they are noticably different dimensions. yet they are copies of each other???

I really appreciate you taking the time to respond. Thanks.

Anthony Faulkner

[Alan Okey]

What is your intended delivery size (pixel dimensions)?

After hearing more detail about what your intended output is (4:3 web delivery), I think the best workflow would be to simply create a new project in FCP using the easy setup for Sony XDCAM EX 30p, then import, edit and key your clips in a native HD project. FCP shouldn’t distort your clips at all when importing them into a sequence with settings that match the recording format. If you’re adding any text or graphics, keep them within the 4:3 safe area of the 16:9 frame.

Once you’ve finished the edit, export your project as a self-contained Quicktime file using the export to Quicktime setting. Open the exported file in Compressor and choose the h.264 Quicktime preset. In the inspector window of Compressor, you have full control over the output size (use square pixels), and you can crop the image to the 4:3 frame.

In summary, edit and finish in full 1920×1080 HD, then as a last step, downconvert and crop to your desired delivery size.

I can’t answer your question as to why the two different source clips are behaving differently in FCP, but I think you’re adding a needless layer of complexity by trying to edit in an SD video format. Since SD video formats use nonsquare pixels and your ultimate destination is web delivery (square pixels), there’s no reason for you to edit in an SD video format. Simply complete your editing and compositing in the native HD format of your footage (XDCAM EX 30p), then scale and crop the final output to 4:3 in Compressor.

Suggested resolutions for 4:3 web delivery: 480×360, 640×480, 960×720

[Anthony Faulkner]

That seems to make perfect sense to me.

That is the way i will go from now. Thanks again for responding, i seem to have a habit of over complicating things!

[Anthony Faulkner]

Hmmm, I can’t seem to find any 4×3 overlays in FCP. I shoot with them in the viewfinder. But i use a lot of images and texting. Is there any way besides manually marking the safe areas?

cheers

Anthony

[Alan Okey]

[anthony Faulkner] “Hmmm, I can’t seem to find any 4×3 overlays in FCP. I shoot with them in the viewfinder. But i use a lot of images and texting. Is there any way besides manually marking the safe areas?”

Unfortunately there’s not a 4:3 safe canvas overlay for 16:9 projects, but you can create a custom matte and overlay it on V2(V3, V4, whatever). It’s a bit cumbersome, but it would give you peace of mind for maintaining safe areas for your overlays.

Method:

Click the Generator button on the Viewer window and create Matte > Color Solid. Place the color solid on the highest level video track in your timeline. Add a matte filter to the color solid (Video Filters > Matte > Mask Shape), choose “rectangle” as the shape and set the horizontal and vertical scales at 75% and 100%, respectively. Check the “invert” box. This should create black bars on the sides of your 16:9 image, leaving you with a 4:3 center cut. You can switch the video track on and off for checking the 4:3 safe area.

[Anthony Faulkner]

you are a champion Alan,

Thank you very much!