[Matt Lyon]

I have watched the sample files and I do see several problems. The main one is the interlacing artifacts. I also see tearing and the occasional (what appears to be) skipped frame (ie, a break in the smoothness of the motion).

But bear in mind that these tests do not represent real world situations. They are designed to show these problems in the most obvious and worst light. I would wager that these problems are more prevalent then you think. They are just much harder to spot in footage shot in real world situations. (Imagine trying to pick out tearing artifacts while being bombarded by the Clash Of The Titans 3D trailer).

I for one DO see problems when playing back DVDs on my laptop — one of the reasons I hate watching DVDs on a computer.

I can’t really speculate on what “big producers do to get around the problem.” My suspicion is nothing. It’s kind of like saying “what do big producers do about the fact the most television sets are horribly calibrated?” There’s nothing they can do 🙂

All this isn’t to say that there aren’t steps one can take to minimize these problems. Maybe there are different software players that do a better job with motion? You could try to get in touch with an engineer at a company like Iridas; they might let you pick their brain on dealing with these sorts of issues.

Matt Lyon
Editor
Toronto

[Matt Lyon]

Ralph,

The problems you are describing are such a moving target. Even if we all watch the same file, there is no guarantee that everyone will see the same motion artifacts. These things are caused by the interaction between graphics cards, specific physical properties of the display, and then all the different software issues and hard drive speed factors.

Not to mention the problems of language. What you call “stuttering” I may call something else. If the issue you are describing is what I think it is, I would call it “tearing.”

The reality of playing back video on a computer display is that you /may/ always get motion artifacts if the refresh rate of the display is not an even multiple of the frame rate of the video. This was easier to combat when everyone used CRTs, since you had more flexibility in changing the refresh rate of the display (for example, using 95.9 hz when playing back 23.976 fps video). Chances are you are watching your 29.97 quicktimes on LCD with fixed refresh rates of 60 hz. And I’m sure there are many other factors, like the response time of the display and the quality of the buffer on the graphics card. I’m not an engineer, so I won’t speculate any more 🙂

But the solution was and will continue to be: when quality matters, monitor video on video monitors, not on computer displays!

Matt Lyon
Editor
Toronto

[Matt Lyon]

[The rushes are a mixture of DigiBeta, DV and HDV. It’s been digitized by firewire at either DV PAL resolution or the HDV. Oh, and I’ve got some XDCAM EX stuff for good measure too! ]

I didn’t realize you had all this mixed footage… now I see the need for the online!

IMHO, I think it is a little unreasonable that the post house wants you to re-create your material with a new sequence setting, just so the keyframes will translate. It’s THEIR job to do the conform, not yours. Unless they are giving you a freebie; in which case you should be bending over backwards to make their lives easier 🙂

In my experience, online editors expect they will have to manually recreate much of the graphics done in the offline cut. And isn’t this what we pay them to do: to bring an extra level of polish to the material?

But I suppose you could export just your split screen stills elements (with the video elements disabled) as a quicktime with an alpha channel, and they could drop it in their timeline. Might be worth a shot if they are at wits end trying to recreate your graphics.

Matt Lyon
Editor
Toronto

[Matt Lyon]

Hi Tim,

Is there any reason you can’t send a hard drive with an uncompressed quicktime of your locked edit? Feed an EDL into the Baselight as well (so it knows where your edits are) and begin grading.

Is re-conforming the whole show on an AVID really going to boost the quality? (it is firewire ingested SD DV, after all). If you change your sequence setting to DV50 before exporting, you’ll get a bump in image quality because everything will be rendered with 4:2:2 chroma sampling.

But if you have to do a conform on the AVID, have you tried using the Automatic Duck plugin to make an AAF?

I wish I could help more, but I don’t have any specific experience with finishing on a Baselight…

Matt Lyon
Editor
Toronto

[Matt Lyon]

Hi Doug,

I would convert your image sequences into quicktime movies using Compressor 3.5 or Quicktime Player Pro. Import the image sequence @ 23.976 fps and export using your desired codec.

I haven’t compared ProRes to ProRes HQ, but the consensus on The Cow seems to be that HQ is total overkill for your particular situation.

As for P vs PSF: (and someone please correct me if I’m wrong here…) My understanding is that PSF is a concept that only really applies to a Digital Video transport stream as it is recorded to a Tape deck. For all intents and purposes, it is still a Progressive image, but the electronics of the transport treat it as having fields, which just happen to occur at the same instance in time.

So I’m not aware of a specific “PSF” setting in FCP. In the past, I’ve always just edited and treated 23.976 HD material as fully progressive. It is only when laying off to tape that the deck gets set to PSF.

And I would also stick with your TIFF images, for maximum quality. As you’ve already observed, MJPEG tosses away a lot of visual information.

Hope this helps,

Matt Lyon
Editor
Toronto

[Matt Lyon]

This could also be a serious I/O issue. I had this happen to me recently with an external FireWire drive that was on it\\\’s last legs. The drive would still mount and appear in the finder, but would flood the bus with read/write errors. I don\\\’t even think I was playing media off it, but it was still chocking FCP.
Also, make sure you aren\’t connected to any servers or other machines on your network, in the off chance FCP is trying to find media on a remote drive (unless you are specifically using a SAN Or similar).

Matt Lyon
Editor
Toronto

[Matt Lyon]

Hi George, I don’t really have any specific experience with the camera you are using, so I can only speculate about the field issues…

That being said, it does clearly sound like the computer is having interpretation issues, since the footage is definitely interlaced, yet always comes up as “progressive.”

If you open a clip in Compressor, you can pull up the Inspector and look at the “native field dominance” setting. (Make sure to click on the clip thumbnail, so you are modifying the clip attributes, NOT the compression settings). You should be able to override the automatically selected setting (progressive, in this case). I’m not sure which is the correct setting to go with (“upper,” I suspect — but you should test both ways).

Hopefully after changing the field dominance setting, Compressor will properly remove the pulldown! Please let us know if you have any luck.

Matt Lyon
Editor
Toronto

[Matt Lyon]

The fact that you had to uncheck the “standard upper/lower” box suggests that something is funky with the field order/field dominance of your clip. Could you have flipped the dominance inadvertently during the transcoding process (the first one, not the reverse telecine step)?

It also sounds like Compressor can’t automatically detect “non standard upper/lower” so maybe you are stuck using cinema tools.

IIRC, when you select the “remove pulldown” option in FCP, it will make use of whatever your last settings were in Cinema Tools. So you should find a setting in CT that works, then you can run a batch in FCP. HOWEVER, i believe FCP uses the FIRST frame of the clip as the “start point” for your A/B/C/D frame setting (as opposed to the “parked frame” system that CT uses). So, for the batch processing to work properly in fcp, EVERY clip has to start at the same point in the pulldown pattern. If your clips are all over the map in terms of cadence, you’ll need to do them one at a time in CT.

I guess the best way to speed up the process would be to try to fix the “non standard upper/lower” issue so you could use Compressor and have it auto-detect the correct pulldown removal setting.

Hope this helps,

Matt Lyon
Editor
Toronto

[Matt Lyon]

Another workflow that I\’ve found fast is using markers. Load a source clip in the viewer, the insert a marker at the start of each take. When you\’re finished marking the whole clip, navigate to the browser and twirl down the arrow next to the clip\’s name, exposing all the markers you\’ve just created. Select them all, then select the \” make subclip\” command. You will now have subclips for the range of frames between each marker. Now throw them on a timeline, sync the audio and link the audio to picture for each take, using command-L.
Now drag all the clips in a scene bin and rename as needed.
You only need to set the timecode and reel name on the original QuickTime file, since that info lives in the file itself, not in the fcp project.
This is just one way I\’ve done it when capturing tapes in one go. My preference is to have separate files on disk for each take (easier to do if you\’re getting a telecine log file, for example).
Hope this helps,

Matt Lyon
Editor
Toronto

[Matt Lyon]

Ron, have you considered a more general policy that everyone sign their posts with a full name, job title and location? I’ve wanted to suggest this even before this flame war erupted. Maybe this is too hard to enforce? I’m on another mailing list that does this and it seems to keep standards of decorum a little higher. (Not to suggest that things have gotten out of hand on this forum — far from it). Maybe this would also encourage some folks to do a little more googling before posting about well tread topics.

Just my two cents,

Matt Lyon
Editor
Toronto