[Simon Wood]

Hi Jon

I sent your question to the Grader his response was –

Well, it is is true that the camera is shooting less quality from LT but in the process of importing to prores HQ, we found that there are much less banding and sharper edges in the picture and that makes grading much easier , if you look on the original footage from the 5d magnified by 100% especially on the edges, there is a difference . Also for even better result there is a standalone software that gives a better result like the 5dToRGB, so, to sum it up, it is much better to “open up” the footage as much as possible and not use LT or H264…

[Simon Wood]

We’ve spent an intensive four days manually transcoding and reconnecting all the media for our 52 minute documentary. Going forward, here are some of the workflow flaws we’ll know to avoid:

Error number 1 was moving the original disk images off to an external drive for storage. The one disk image that had remained on the internal drive (despite that drive being renamed, and all the folders around that DMG moving and changing names) worked like a charm to do a batch capture. The clips were upgraded to HQ and reconnected in minutes.

However, for the other 1.25 TB of DMGs we had, life was not so fun. It was impossible to put them back to the directory they came from because by then, each of our three internal drives had been replaced and renamed – and we hadn’t kept record of how the various project folders within folders had originally been labelled in any case.

Unfortunately, FCP Reconnect only had about a 50% effectivity in matching our LT clips to the re-converted HQ ones, and on top of that, a significant number of clips had their in and out points badly adjusted when we reconnected the media in FCP. This is because we were reconnecting the media to a media managed version of the project with 1 second handles, so where the original media had been one long clip in certain instances, it was broken up into several fragments that needed to relate back to one media source. Moreover, the audio had come from a different source and had been synced using PluralEyes. Where, for example, a shot would start at 12:00:00:00, but the audio extended before that, and the editor had used a section at the beginning of the clip, the media managed clip (with 1 second handles) would have a timecode of 11:59:59:00. Therefore, not only would FCP Reconnect not recognise it as the same clip, but FCP would also slip the clip out of sync when reconnecting to it.

The process we ended up using was to check the timecode of each clip, open the appropriate disk image, select the right shot based on that timecode, copy the name across, and then transfer that clip as pro res HQ. Clip by clip, we then reconnected the media in FCP. There were many instances where FCP referenced the wrong timecode when reconnecting (because of different clip lengths?) so every single shot needs to be checked and slipped into sync.

The final hurdle was Motion and After Effects projects reconnecting, and timelapse that had to be painstakingly redone.

We don’t think that to convert to HQ in the first place would be a viable solution for all future projects: the disk space is just unrealistic, and Compressor seems to take much longer than the canon plug in. Also, having timecode on the clips made it MUCH easier to sync the sound files. We did have some footage that was done in the manual way of dragging H264 files off the card, using Compressor to convert to LT, and then later on to HQ. FCP did not reconnect this easily, particularly when one clip was used multiple times.

The best possible workflow for our company’s circumstances would be to use and store the DMGs on a single RAID; keep all effects in the highest quality possible; and budget time to recapture the media.