Preserving EDL with Offline/Online files

[kiril]

Hi all,

I am new to Edius and I have a basic question.

I have an AVCHD camera. I tried to play the MTS files directly in Edius, but they do not play well (jumpy, clipped, sound is bad too).
I have a powerful machine - an HP 8710w Mobile Workstation, dual core 2.5 Ghz, 3 GB Ram and I expected that I'll be able to play them directly.

So, I use now the AVCHD2HQ codec, I get great results, but there is one drawback - the resulting online files are huge...

My question is if it is possible to decode first the MTS files to offline quality, edit and then re-"capture" the relevant files (or the used parts of them only) with online quality.

Or may be there is another work flow that I'm unaware of?

Thanks in advance,

Kiril.

[Bluetongue]

Kiril
They havn't invented the computer with enough grunt to edit native AVCHD yet!
The best bet is what you are using with the CanopusHQ files, BIG but just great as they loose no quality and take the load off the procesor.
So far as proxy files are concerned take this bit of advice, in HD FOCUS is so important I just cannot stress enough and with an SD proxy file (ala VAGAS) FOCUS problems are difficult to detect so you can end up with a final product that has glaring change in focus through out the video.
When you edit in full HD this and other small problems that can ruin the final stand out glaringly so keep with the HQ files.
I take it you are using a laptop for editing so I can relate to your problem as I use a Toshiba Satelite A100 with a 100G HD for use in the field for rough edits before I return home and find the HD is just not enough but I take a 250G USB2 drive with me which is adequate for rough editing but is just on the limit for serious editing and am looking at an express card to connect an external SATA drive to the laptop
Hope this helps
Regards Barry

[kiril]

Thanks a lot.

Ok, so it seems that having enough disk space is the only solution.
BTW - I read few posts saying that today's CPU's cannot do the job of realtime decompression of AVCHD.
Well, if the processor in my AVCHD camera can do it to play them back, I guess that we are close to the moment in the future when there will be something (hardware or software) capable doing so.
I have an HG10 and the software coming with it plays the files quite Ok, but it is useless in terms of editing.

[Bluetongue]

BTW - I read few posts saying that today's CPU's cannot do the job of realtime decompression of AVCHD.
Well, if the processor in my AVCHD camera can do it to play them back, I guess that we are close to the moment in the future when there will be something (hardware or software) capable doing so.
I have an HG10 and the software coming with it plays the files quite Ok, but it is useless in terms of editing.

Playing and Editing are two different processes.
The MPEG format is the problem as a FULL picture frame appears every 15 frames in DVD format, every 30 in HDV and I believe every 60 frames in AVCHD.
All the frames in between are the encoded difference between the current frame, the frame before and in some cases the frame to come, so even placing two clips together on the timeline will violate the MPEG structure at the junction with either 2 picture frames together or insufficient encoded frames from the last picture frame, so the processor has to be continually updating the video reference points instead of just keeping up with the flow.
The hardware chip in the camera is programmed to do just replay in sequence and does not have to update memory etc so has a pretty easy time of it.
The moment you cut a MPEG clip during editing the structure is further violated and adds additional load to the processor.
Editing in MPEG does not gain anything as the timeline has to be completely re encoded on output to get rid of the mess and place the "I" picture frames at the correct spacing
Hope that answers you question, tried to put it as simple as possible.
Regards Barry

[redgum]

The MPEG format is the problem as a FULL picture frame appears every 15 frames in DVD format, every 30 in HDV and I believe every 60 frames in AVCHD.
All the frames in between are the encoded difference between the current frame, the frame before and in some cases the frame to come, so even placing two clips together on the timeline will violate the MPEG structure at the junction with either 2 picture frames together or insufficient encoded frames from the last picture frame, so the processor has to be continually updating the video reference points instead of just keeping up with the flow.
The moment you cut a MPEG clip during editing the structure is further violated and adds additional load to the processor.
Editing in MPEG does not gain anything as the timeline has to be completely re encoded on output to get rid of the mess and place the "I" picture frames at the correct spacing
Regards Barry

Just a few minor corrections to Barry's statement. Sony HDV is 15 frame long GOP (Group of Pictures) and another common one is 12 frame long GOP and as you progress up the format (professional) level the GOP's get shorter.
The load on the CPU is little different between native HDV (M2T) and an intermediate AVI file. There is certainly a great deal of difference between native HDV and DV. An HDV file is generally 4 times smaller than an equivalent intermediate file like Canopus HQ or Cineform.
Editing in Mpeg requires less encoding than intermediate AVI files. The only Mpeg frames encoded are the ones that are changed whereas with AVI intermediate every frame is encoded.
There are two reasons why people favour working with intermediate AVI files. In the early days computers did not have enough power to process Mpegs but more importantly software developers had not developed efficient Mpeg algorithms. The first has certainly changed and I believe the second will change in the next twelve months and Mpeg editing will be the norm.
You only need to look at all the professional formats to see they are all Mpeg based already and Avid is just about to release consumer level Mpeg editing without options.
AVCHD is ahead of its time (as was HDV) and software is slow to develop but that usually changes pretty quickly. Think of it like this, when every other camera format is dying your will be just coming into fashion.

[kiril]

Thanks guys.. :)

Yes, AVCHD will be in fashion, some day, just to be on the top for 12 months until the next thing pops up...

Kiril.