XDCam HD vs. DVCPro HD in Final Cut Pro, help...

DV is 25Mbps, not 35Mbps.

DV is an intraframe codec where each frame is compressed independently, and XDCAM EX is an interframe codec where video is recorded in GOPs (Groups Of Pictures) consisting of a keyframe and a series of partial frames that only encode the differences from the previous frame. That is one of the reasons why it achieves a much higher efficiency.
 
DV uses DCT intraframe compression at a fixed bitrate of 25 megabits per second (25.146 Mbit/s), which, when added to the sound data (1.536 Mbit/s), the subcode data, error detection, and error correction (approx 8.7 Mbit/s) amounts in all to roughly 36 megabits per second (approx 35.382 Mbit/s).

I found this reading wikipedia that's why i said that it has 35mpbs.
thanks for the answer .

What about the computer rig needed to process xdcam
 
If HDV is any point of reference, there were substantially long conforming times when the edit was done in FCP 6.0.2 on a quad core Mac.

With DVCProHD sequences, there is no conforming time. I've edited DVCProHD on a MacBook, not pro. I couldn't do the same with HDV.

HDV is a step done from the EX1, but you still need to conform as I recall for it as well.

Again, HDV is slightly different, but I prefer working with DVCProHD footage.
 
David S. said:
you still need to conform as I recall for it as well.
Click to expand...

There is a move towards different formats for

- Acquisition (DV, HDV, DVCProHD)
- Editing (DNxHD, ProRes422)
- Delivery (DigiBeta, HDCAM, DVD)

I cannot think of a reason to shoot HDV, edit HDV then output to HDV tape, when I could output to VC1 and play back from PC, or output to H.264 and play back from a PS3, or whatever.

XDCAM-HD and DVCPro-HD are, after all, acquisition formats. You could transcode to an intermediate format for editing, but FCP does a clever thing where you edit native footage up to the point at which you 'need more' or 'add more' - and these are done at ProRez.

This is a lovely compromise, and I think it's the equivalent of the old 'off-line, on-line' school of editing. Your rough cut is all done natively in real time. As you add effects, transitions and graphics, so these minor renders are handled in 'eye-blink-time'. When you come to the grading stage, then you start to have serious renders once the heavy lifting is done, but you're at picture lock. And really, if you have to go back and twiddle something, it's not as if you have to re-cut negatives for heavens sake.

It's the best of all worlds - native editing until you need to do something that will regenerate the footage, but it does so at a better colour resolution than XDCAM-HD and a better spatial resolution than DVCPro-HD.

IMHO, originate in a format that suits you. Edit in a format that works with you. Shots wot needs it gets it, those that don't, don't.
 
Thanks people.

Sorry i'll continue bugging you with questions but i'm a novice when it comes to all this editing/camera format thing and i want to get it right. :undecided

The need for a high processing power comes when the material is being copied into the computer or when you want to play it back/edit it in a NLE program ?

Do i need or will it help improve the performance if i have a dedicated capture board ?

@david s. --> what does conforming time mean exactly ?

@Matt Davis --> I don't understand what you are saying. FCP let's you edit in the native resolution (let's say XDCAM) until you output ? Or i switches onto another low-res format when you're in grading stage or... ? i didn't really get it.

And how is online offline editing equivalated in digital terms ? I know that when you work with film you scan it edit it in a program of your choice then you transfer it back to film edited. How about digital ?

Thanks! :Drogar-Happy(DBG):
 
Gogu said:
The need for a high processing power comes when the material is being copied into the computer or when you want to play it back/edit it in a NLE program?
Click to expand...

The need for processing power is to either play back complex formats (XDCam) or highly compressed formats (H.264). The need for fast hard disks and busese is for transferring large amounts of data (DVCProHD).

Gogu said:
Do i need or will it help improve the performance if i have a dedicated capture board?
Click to expand...

Not really applicable for P2 or XDCAM-HD as you're copying across computer data rather than capturing and converting video. SDI or HD-SDI does need a dedicated capture card.

Gogu said:
@Matt Davis --> I don't understand what you are saying. FCP let's you edit in the native resolution (let's say XDCAM) until you output?
Click to expand...

Yes, you're editing the original camera data until you need to do something with it (like filter its colours or superimpose graphics over it). At that point, FCP generates new footage (renders) using a higher quality codec than the one you shot in (ProRes422). When you're done, you render the whole thing in ProRes422 (before converting to BluRay or DVD) or to the format you want to deliver in.

Gogu said:
And how is online offline editing equivalated in digital terms? I know that when you work with film you scan it edit it in a program of your choice then you transfer it back to film edited. How about digital?
Click to expand...

Don't worry. This is not some deep and meaningful metaphor. As I said before: the first pass at editing is about content (what you do in the offline stage with your workprint or rushes) and you're working with the camera originals without recoding them. The second pass is about finessing your content (grading/timing, effects, graphics), so when your rushes are processed, you're creating new copies of the rushes in a higher quality format. But don't get the wrong idea... You can't increase the quality of rushes by moving them from, say, 4:2:0 to 4:2:2 but you can provide more 'room' for grading than the original format would have allowed.

When you're 'cutting', nothing needs rendering, you're editing the original footage as it was shot. When you do something to a clip that requires altering the camera original, Final Cut Pro automatically makes a new version of that bit of the clip using the higher quality format, so the whole sequence contains can ontain either your oringal unproessed clips or modified clips.

When you export the finished sequence, you use ProRes422 for the whole sequence, making one long file in one uniform codec.

FWIW, I am editing HDV and XDCAM-HD on a MacBook Pro 17" 2.33 GHz, 2GB RAM, wth rushes on LaCie D2 FireWire800 drives. I use the 'Bump up everything that needs rendering to ProRes422' technique outlined above to avoid the Conforming nightmare. My main outputs are Flash and H264 for web, Standard Definition DVD, and HD Windows Media 9 files at high bitrate (5000kbps). No tape!
 
Last edited:
I am not convinced that transcoding all your footage to proRes prior to editing produces any real advantage (as you point out, you can't increase the amount of colour information in your footage). You significantly increase your disc space usage (maybe better spent on backing up your footage). Editing, in most situations, on a suitably specced Mac/PC is pretty much real time anyway.
Stu at Prolost describes how you can use your sequence video processing settings to allow headroom to recover out of range values if you apply multiple CC filters (in this case, Colorista). It means you can use XDCAM footage in a larger colour space without first transcoding it. (Note you need to render in FCP to see the advantage of high precision YUV - maybe we will get realtime in the next release)
Obviously there might be an advantage to bypassing XDCAM by capturing 10 bit ProRes via SDI.

I think it is also that case that Apple Color can work in 32 bit, regardless of the bit depth of your footage, again giving you no advantage to converting your footage before grading (anyone able to confirm this?). I have sat in sessions with professional graders using Color on XDCAM footage, and at no point did they notice any format related limitations.
After Effects is a different case - interframe footage seems to confuse it, and frames come out in the wrong order.
 
Last edited:
Short Projects?

Short Projects?

All this talk of long GOP versus I-Frame material .... am I the only one working on long form stuff here? I currently have a timeline with XDCAM and DVCPRO HD footage and they both look great. Nit pick which one you like better until you are blue in the face, problem lies not on editing, but on output. When editing with DVCPro Hd, there is no transcoding on the way out, so my timeline is realtively easy to output and convert to whatever format I wish. My XDCam stuff takes FOREVER on the way out! Unless someone knows some trick I don't, the time saved with long GOP formats like XDCAM and card/disk based transfers is completely nullified when transcoding out. I have done this on an AVID and a FCP system and been equally disappointed. Anyone? Some trick to transcode my XDCAM footage quickly? I am mixing DVCPro and XDCAM right now on a FCP system - best set up? I am begging ... give me a reason not to shoot anyone on sight who gives my Long GOP footage!
 
Just to be clear....

There's a HUGE difference in the amount of time it takes to unwrap XDCAM EX footage into MOV (or MXF if you're on a PC) when compared to transcoding it to ProRes.

Unwrapping is just that, it's taking the raw MP4 footage from the SxS card and putting into a new wrapper. The footage isn't copied, or re-compressed, it's just unwrapped. I like the analogy of calling Quicktime a "bucket". This bucket can hold a bunch of different codecs (XDCAM EX, DVCPRO, DV, etc...). Unwrapping simply takes the SxS card footage and unwraps it from an MP4 bucket into a MOV Quicktime bucket. Same exact footage (I think it's MPEG-2) but it resides inside a new bucket.

Transcoding to ProRes 422 HQ on the other hand is a time intensive process. The computer has to re-compress the footage into another codec. This can take a lot of time. True, the quality of ProRes 422 HQ is better than XDCAM EX, and if you're doing "major" color correction you'll need the 4:2:2 color space. But in my experience, I don't do much "major" color correction, only minor tweaks to the black level, white level and maybe saturation. (also known as primary color correction).

My point is, for "most people" editing in XDCAM EX is the most efficient solution. Quality is still fantastic.

On the other hand, if you're doing secondary color correction, green screen (chroma key) or other graphics intensive work, then it might be good to transcode to ProRes 422 HQ. But honestly, you'd be better off to capture in ProRes 422 using the HD-SDI output and bypass the XDCAM codec all together. You would need a AJA Io HD or the new AJA KI to do this, but for some people it may be worth it.

I've now edited quite a few corporate projects in Final Cut Pro using the XDCAM EX "Easy Setup". The only change I make to the Easy Setup settings is in the Sequence Settings>Render Tab, I change it to render using ProRes 422 HQ. Making this change to the render tab is HUGE!! It's helped immensely with the reduction of render times and having to re-render projects. According to Apple, there's a slight improvement in quality as well.

I'm not flaming anyone, I'm just sharing what has worked well for me for anyone who cares to listen. :)
 
You must log in or register to reply here.
Back
Top