FS700 Workflow...What applications allow viewing of clips on Mac

[Scott Ransom]

Just loaded my first card into a Mac with FCPX...Streamclip...etc. FCPX does not seem to want to import the test files to a new event. What apps are you guys using to import footage and screen? There is something simple here I am missing I know.

SR

 

[KMA_WWC]

If I don't use my external recorder and shoot in native avchd, I always use 5DtoRGB to transcode the footage before editing it.

 

[Nirv]

Have you used the import from camera command in FCPX? Why are you bothering with Streamclip? FCPX does everything you need.

FCPX imports FS100 footage perfectly so I'm not sure why FS700 would be different.

Careful with 5dtoRGB it crudely legalises the video and it's much better to do that in FCPX or another NLE for that matter.

 

[Scott Ransom]

Havent tried import from camera command....as I had the SD card in my Mac at the time. Shouldn't X recognize the card the same as if it was still in the camera....and then import files command from the card directly?

 

[Nirv]

Havent tried import from camera command....as I had the SD card in my Mac at the time. Shouldn't X recognize the card the same as if it was still in the camera....and then import files command from the card directly?

You use the "Import from Camera" command to import from the camera or from a card-reader or from a camera archive. That's how FCPX works.

It's all in the FCPX Help file.

 

[Thomas Worth]

Careful with 5dtoRGB it crudely legalises the video and it's much better to do that in FCPX or another NLE for that matter.

Nirv, can you elaborate on this? 5DtoRGB was designed to solve the problem of "crudely" transcoded video, so I'm not sure what you're referring to here. Tests have also shown that 5DtoRGB's chroma upsampling is superior to that of FCPX, so using FCPX for transcoding/native editing instead of 5DtoRGB is a clear compromise on quality.

 

[Nirv]

Nirv, can you elaborate on this? 5DtoRGB was designed to solve the problem of "crudely" transcoded video, so I'm not sure what you're referring to here. Tests have also shown that 5DtoRGB's chroma upsampling is superior to that of FCPX, so using FCPX for transcoding/native editing instead of 5DtoRGB is a clear compromise on quality.

Why would I go to the bother of setting my camera up to capture a full range file only to have that legalised by an automated system?

Given a choice of editing right away or waiting for several hours of rushes to be transcoded for an almost imperceptible improvement in image quality I'd take editing right away. The more time I spend editing, grading and producing graphics the better the project is and a better use of my CPU's time.

I've never had a client say,"Blimey, look at your chroma upsampling! You need to be transcoding better, oh and I don't mind paying the extra cost either." That has never happen strangely.

 

[Thomas Worth]

Why would I go to the bother of setting my camera up to capture a full range file only to have that legalised by an automated system?

I'm concerned about your comment because you made it sound as though 5DtoRGB was doing an inferior job with the transcode when compared to other NLEs. I just wanted you to elaborate on your statement because it was unclear exactly how you came to that conclusion. If 5DtoRGB is not performing up to your expectations, I want to know so I can improve the software.

Regarding your comment about legalization, ProRes itself is not full-range. If it were, its luma levels would be incompatible with all the broadcast video gear out there. The mandatory broadcast levels are perfectly fine in the case of 8-bit H.264 however, because even through ProRes is a broadcast range codec it is capable of 10 bit so none of the original 8 bit data is lost. The actual luma range of 10 bit ProRes is 64-940, not 16-235.

Given a choice of editing right away or waiting for several hours of rushes to be transcoded for an almost imperceptible improvement in image quality I'd take editing right away. The more time I spend editing, grading and producing graphics the better the project is and a better use of my CPU's time.

I completely understand. In many cases, utmost quality isn't required. Certainly videos destined for the web, for example (which end up being highly compressed anyway) are less of a candidate for a 5DtoRGB workflow than say, an independent feature film with the prospect of theatrical distribution.

I've never had a client say,"Blimey, look at your chroma upsampling! You need to be transcoding better, oh and I don't mind paying the extra cost either." That has never happen strangely.

You might like to direct them to some of the 5DtoRGB chroma upsampling tests flooding the web.

While everyone can certainly benefit from 5DtoRGB's high quality output, it's the user's job to decide whether the 5DtoRGB workflow is appropriate for any given project. There have of course been times I've used typical workflows (like log and transfer) instead of 5DtoRGB, but for any critical work I will use 5DtoRGB, no question.

 

[Nirv]

Thomas

I wasn't trying to knock your software and the term I used could have been better. Your software does not crudely top and bottom slice the illegal levels away like most NLEs do given a chance.

I shoot with the FS100 to either cram as much info in between 0 and 109 or to work around the CineTone curves Sony has provided so the extra 10% headroom can come in handy. When I legalise the super brights I can do so in a number of ways, just drop the gain or pull the mids down and roll the whites off. An automated system can't make that sort of decision. And that was the point I was making with poorly chosen wording.

Regards

 

[Thomas Worth]

I shoot with the FS100 to either cram as much info in between 0 and 109 or to work around the CineTone curves Sony has provided so the extra 10% headroom can come in handy. When I legalise the super brights I can do so in a number of ways, just drop the gain or pull the mids down and roll the whites off. An automated system can't make that sort of decision. And that was the point I was making with poorly chosen wording.

Assuming you could have a feature in 5DtoRGB to deal with this, how would you like it to work? Something like a "clamp" and "soft rolloff" option, perhaps?

 

[Dermot]

Regarding your comment about legalization, ProRes itself is not full-range. If it were, its luma levels would be incompatible with all the broadcast video gear out there. The mandatory broadcast levels are perfectly fine in the case of 8-bit H.264 however, because even through ProRes is a broadcast range codec it is capable of 10 bit so none of the original 8 bit data is lost. The actual luma range of 10 bit ProRes is 64-940, not 16-235.

Thomas, i have piles of ProRez on my system that are 0-1023.. most have come from Alexa or F3+gemni, i have a couple off feature films worth of them.. i also have a show that used the 5D-RGB and we had to throw it all out, and re-conform from the camera orignals to avoid clipping that the software seems to introduce

d

 

[PDR]

Thomas, i have piles of ProRez on my system that are 0-1023.. most have come from Alexa or F3+gemni, i have a couple off feature films worth of them.. i also have a show that used the 5D-RGB and we had to throw it all out, and re-conform from the camera orignals to avoid clipping that the software seems to introduce

You guys are actually saying the same thing - Alexa doesn't allow recording extended range when using internal prores , and the actual recorded luma range is 64-940 as Thomas says . It's the way prores works

The data recorded is clamped to Y' 64-940. Superbrights/darks aren't clipped , everything is just "squished to legal range" . This is true for all 10bit Prores flavors (64-940) and 12bit prores flavors (256-3670), not just the Alexa internal recordings. It's a general quicktime/mov behavior. When using software such as FCP or color correcting suites, the prores values are scaled back to 0-1023 (if using 10-bit) , or 0-3760 (if using 12-bit) , thus all the data is retained, no data is clipped . Had the actual prores recorded data 0-63 or 941-1023, those values would be clipped - so you can think of it as a "safety" feature

IIRC, 5D2RGB has a full and limited range setting, perhaps there was user error in footage you recieved

 

[timmytimetravel]

In answer to the original question, as well as using FCPx log and import thing (which I dislike in being forced to work that way) I've been using clip wrap. Hadn't used it for years but it just rewraps the source into a mov.

I could also open bits of footage using vlc

Clipwrap Takes a min per card, but then I can add selective files directly into FCPx
i can also dump the card and clear it in finder without either importing the whole card or pulling up the interface to selectively import

I also tried copying the structure of the card to a finder folder, deleting the files in stream, then using that as a dumping ground for select clips, which can still be opened with "open camera archive"

Anyway, not sold yet. Probably comes back to changing my practice for FCpx where I have big attached hdds of dumped files/cards
It's an attractive idea to stay in avchd but still open and view original files in mov.
Waiting on more info about external recorders and 4k before I hit go on something more I have to wrangle on shoots just to get prores