[Matt Campbell]

Thanks Dave. I don’t have a delivery specs as this is for web content. My question is more of workflow and look.

I understand using CT will give me the silky smooth slo-mo, which I want and use all the time for other edits. Unfortunately, this time they did not shoot 23.98 material and I only have 59.94 to work with. My real question is what about the 59.94 footage at regular speed. Since they want the film look with the regular speed footage and didn’t shoot at 23.98 and only at 59.94, I’m assuming either FCP or Compressor route will simply remove the necessary frames for me to edit at 23.98. Then I’ll be able to edit in the few conformed CT clips so that the regular speed footage and slo-mo conformed clips play at the same frame rate together in the 23.98 sequence.

We really want the simulated “film look” but with not having 23.98 regular speed footage and only having 59.94 to work with, will that footage with the frames dropped out still look film-like with the blurred motion rather than the clean, crisp look of 59.94?

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD

[Matt Campbell]

Rafael, I’ve just posted a similar question. So normally you would shoot at 23.98 and edit 23.98 for that so called film look. Any any material, if planned out properly, you wanted for slo-mo, you would shoot 59.94 and conform with Cinema Tools to get silky smooth slo-mo.

What if all you have is Canon 7D footage at 59.94, and they still want that film look. Are you saying to still edit in 23.98 and either take the transcoded material in FCP and either use Compressor to change the frame rate to 23.98 using Fast under Rate Conversions Frame Controls or just drop your 59.94 footage in a 23.98 timeline and under Sequence Setting, Video Processing select Fast (Linear). Will both these methods remove the proper frames from 59.94 to 23.98?

Then whatever scenes we want slo-mo we can simply make a copy of the clip and use CT to conform them, and re-import.

Does this sound right? Or would you edit 29.97 with 59.94 material because its a cleaner conversion and FCP will then drop those frame properly. Then maybe when finished use Phillips Blooms methods to conform your 29.97 sequence to 24 once the edit is complete?

There seems to be many ways to do this, I just want the best workflow that will yield the cleanest looking footage all the way through to output.

Any thoughts?

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD

[Matt Campbell]

Alex, I just had a similar scenario. Our shoot was on iPhones with some film app that shot at a # of frame rates. Our shoot was at 24p, and not 23.976, True 24p. Which gaves us loads of headaches. The footage was actually weird at 24.008 fps or something strange like that. Anyway, this is for broadcast and we had some secondary footage shot with the regular iPhone app at 29.97. Because this is only secondary and the footage used was very quick, like 1-2 sec. worth, we just cut the 29.97 material into the 24p timeline and let FCP pull out the frames. Probably not ideal, as the Compressor method is better, but it has worked for us and the final output at 24p looks just fine.

Our dub house accepted the 24p material then converted to 23.98 and added pulldown for broadcast at 29.97. Nothing was rejected and the spot looks good. But in your case, yes, I would totally convert to ProRes via compressor.

That Filmic Pro app really screwed us with true 24p, but they wanted to use the app to control the iris. For the next one, I’ve urged them to shoot straight 29.97. There is no need for this to be at 24p for the film look since its shot on the iPhone and the spots are for and about cell phones.

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD

[Matt Campbell]

So I made some test and tried an HDV 1440 x 1080 timeline with superimposed subtitles

This is fine for a test, but I wouldn’t ever transcode anything to HDV. HDV is like DVCProHD. They are anamorphic HD formats. DVCProHD has both 960×720 and 1280×1080. HDV is also 1480×1080. I would always transcode this to a full raster codec.

Starting from a DvcPro HD timeline both movies seems to be visually identical though exporting via compressor doubles render times.

Compressor does lengthen the encoding time, but almost always does a better job. However, be careful with those frame controls. Even though there is a best option, doesn’t mean it always yields the BEST results. This is ever so true with rate conversions. Just keep an eye out for wicked frame blending and pixel morphing.

And depending on your deliver, like you said, exporting your DVCProHD timeline straight to ProRes is just fine too.

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD

[Matt Campbell]

Do you think ProRes LT will be good enough to export out my final project or are you suggesting I edit with it and take my fine cut out in ProRes or ProRes HQ?.

This wouldn’t make sense because if you edit ProRes LT, you won’t get any better than that. Just a larger file. Unless your thinking about doing an offline/online edit. But with this type of footage, there is really no need. Stay online and edit PR LT. Export your master as PR LT.

I still don’t understand why the AVCHD to ProRes transcode gets so inflated with no discernible increase in quality over the smaller XDCAM files.

AVCHD is a camera codec not designed for editorial work. Its the same with DSLR’s that shoot H.264. The footage looks great, but the bitrates are not as high as the higher end cameras. No matter what codec you transcode to for editing, in any system, the footage won’t look any better than the originals. So I would transcode to a codec of a similar, but higher bitrate for editing. This won’t improve the quality, but it will prevent it from getting any worse.

ProRes files are larger, but that’s just the nature of the beast. Again, choose a ProRes codec that’s is close to the bitrate of the camera codec. I often use ProRes LT for XDCamEX footage, but use regular ProRes 422 for just about everything else. Just be glad your not working with full raster 1080 uncompressed or even larger like 2 or 4K. Remember the better the codec, the larger the file and the more CPU power and more importantly disc speed you’ll need to edit it.

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD

[Matt Campbell]

I’ve used this workflow loads of times. Working with and editing DVCProHD 720p formats then transcoding on output. However, I didn’t use Rafaels method, I simply changed the output settings under Export Quicktime movie. I made a self-contained file and left re-compress frame unchecked. I simply matched the ProRes flavor of choice to the frame rate of my sequence and let FCP render out the 960×720 DVCProHD timeline to ProRes 422 1280×720. This worked for everytime and it only had rendered gen-loss (generation loss) once. The other method seems like 2 render passes for 2 gen-losses.

Now, I don’t know if I’d do the latter for broadcast but for corporate videos and web deliver it was just fine. Not saying my method was write and theirs was wrong, cause either method will work. Its just my 2 cents.

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD

[Matt Campbell]

Chris, it could just be a glitch, but it sounds to me like frame doubling. This will often happen when working with 23.98 footage in a 29.97 sequence. FCP doesn’t add in the proper pulldown with 1080 material (only with 720p material), so it will add in 4:2 pulldown, versus 3:2. You’ll see 4 clean, different frames and 1 duplicate frame, repeated for the length of that clip.For web stuff, I usually don’t care much, but that prob won’t fly for broadcast.

But that’s only a guess. Without seeing the footage I can’t really say.

Anyway, sounds like you’ve found a workflow that works for you. Compressor is great for a lot of things. For me, I rely on Cineman Tools for a lot things like this. Both work.

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD

[Matt Campbell]

Excellent summary Andrew. Couldn’t have said it better myself.

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD

[Matt Campbell]

Chris, here is the link to the thread i posted about 24p to 23.98 conversion you wanted to see mentioned. this is the opposite of what we’re talking about but the prodecure are about the same. And where we netted out was to have our dub facility here in NYC do the standards conversion on their hardware converter, like an Alchemist.

https://forums.creativecow.net/thread/8/1160456

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD

[Matt Campbell]

Sean, you might want to try converting your MTS file ProRes LT. Those camera’s don’t shoot at a very high bit rate and ProRes LT is probably the equivalent to that. Plus it will save you some storage space on file size without a noticeable difference in image quality. ProRes 422, IMHO, is just overkill for YouTube web delivery.

We often use ProRes LT for even pro-sumer level camera’s like the Sony EX3, which shoots XDCamEX footage at 35 mbps. ProRes LT is right around that bitrate I think, at about 45 Mbps. ProRes 422 is like 110 and 422 HQ is about 220. Again, you’ll save storage space.

OS 10.6.7, Mac Pro 2 x 3 ghz quad-core intel xenon, 4 gb ram and AJA IoHD