CONCEPT: Pre-render in camera. Is it possible?

[net]

When the information first writes to the card while recording in the camera; how much of that information is later being written twice during render with a video editing program?

Do you think it's technically possible, in a future camera option, to:
...Select either your NLE (i.e. Sony Vegas) within the camera menu or even a universal standard which the software makers could choose to adopt...
...during recording in the camera, pre-render and/or be able to use this written information later in the rendering process as a chunk of information being attached to the final render file, but not actually re-written? Or however it could be conceived to include this information without actually re-writing.

It might already be doing something similar now in the actual program itself to what I am proposing.
I was just wondering if there was a way to utilize the write process within the camera to be more advantageous than it is now.

This is concept only, I am not claiming to actually know the mechanics involved.

[Janke]

Editing software is designed to handle the format standard(s) that cameras use, not the other way around.

[net]

Editing software is designed to handle the format standard(s) that cameras use, not the other way around.

Yes.

Just thinking, for example, if I know I am going to be rendering to Sony AVC later on my computer, there must be away to bring that write information over from the camera without compressing in-camera or having to re-write in the rendering process.

[net]

To illustrate more what I'm saying (in lay man's terms):
If the common information that I am talking about, ("common" meaning both contained in the mts file and file after it is rendered) were contained in a folder - it would be the equivalent of pasting the info from that folder into the rendered file instead of re-writing that information.

So, now that I think of it, it would actually be about having the camera make extra information and the NLE would do the job of being able to attach to the file without the long extra render time.

If everything just ends up ones and zeros in the end (the rendered file) - couldn't there be a way of (the equivalent of the short process of ) pasting information into a folder (a rendered file in this case) instead of re-writing it?

Anyone get what I am saying?

[Janke]

without compressing in-camera

Then you could only get about 5 minutes (or so) on a 16 GB card... and it would have to be a blazingly fast card.

[Daniel Rutter]

have to be a blazingly fast card.

It won't be long until we have cards that can write at higher speeds than our current 30-40mb/s...
Think of the Red cameras and their hard drives, and P2 cameras and their cards...

[Janke]

Sure, but they aren't consumer models... and they aren't optimized for Vegas, which was mentioned in the OP...

[Daniel Rutter]

and they aren't optimized for Vegas, which was mentioned in the OP..

Yeah I know that, I wasn't pointing out whether or not either of you were wrong... simply applying my knowledge to a part of the argument that I understood. I'm within the understanding that a company gives us a camera, and our editing suite works around this. Not the other way round. If this were the case, you'd have several different versions of the software which were tailored to each type of camera.

Give it some thought though. Sony has delivered Vegas Pro to us, right? Vegas is optimized (pretty much) to work with Sony Cameras, it has features there for specific Sony cameras and products (Blu-ray, PSP, Bravia TV's etc). But Sony (I believe) is the only camera manufacturer that also delves into designing editing suites. Unless RED comes with it's own software??

[cgbier]

Editing software is designed to handle the format standard(s) that cameras use, not the other way around.

Not necessarily true. The JVC HM100 delivers an MPEG2 .mov that was designed to work natively in FCP.
... and that cam still falls under consumer to me.

[Playing]

be able to use this written information later in the rendering process as a chunk of information being attached to the final render file, but not actually re-written?

You are talking smart rendering, where a part of a video clip that has not been edited in any way is just copied without re-encoding.
A version of Vegas prior to the current one used to be able to do this for AVCHD.

What you need is not a camera that creates clips for different programs but a standard where every AVCHD camera produces the exact same output irrespective of manufacturer.

[net]

You are talking smart rendering, where a part of a video clip that has not been edited in any way is just copied without re-encoding.
A version of Vegas prior to the current one used to be able to do this for AVCHD.

What you need is not a camera that creates clips for different programs but a standard where every AVCHD camera produces the exact same output irrespective of manufacturer.

ok, good, so it's already possible. My thinking was that there could be some way to take advantage of the writing in the camera - but I guess, if it's all just data in the end, it's the same thing. I have heard that term before, but it never occurred to me to see what it was about.

What made me think of this recently was during the the long render times of a file that there must be some redundancies going on. For example, if there is an image of a house in the video, then the data of those lines are in the mts files and being re-written again during the render. And so I thought - why not have the camera mark key points and tell the editor not to do it again...

Platinum 10 doesn't mention this in the help file, but I see now Platinum 11 and Pro 10 have smart rendering. It might be worth the upgrade, depending on if it's only reserved for certain output files or not.

[MoDu]

That only makes sense if you're planing to output in the same format as the camera stores the video. And it disables you from doing anything more than cutting en joining, do a small color correct and BAM, you need to reencode everything.

Besides, real-time codecs, even the best, are no match for a double pass h.264 encode with x.264.

What we really need is a good recording codec (like Cineform, for example) with little to no losses and adequate for editing.

[Playing]

why not have the camera mark key points

That is exactly what the camera does. It's called a key frame. Look it up.
There is nothing wrong with your thinking just google a bit.

You mention mts files. Unfortunately Vegas Pro 10 will not smart render mts files.

[ejolson]

Not necessarily true. The JVC HM100 delivers an MPEG2 .mov that was designed to work natively in FCP.
... and that cam still falls under consumer to me.

The Apple Iframe mode in Panasonic and Sanyo camcorders is another example of a consumer camcorder designed to work with editing software.

[cgbier]

Thanks, ejolson. Never looked into them.

[racer-x]

The problem with trying to "smart render" AVCHD format, is that it (AVCHD) is not a standardized format. Way too many variations. It's not even finalized yet, there are constant new variations added all the time........

Smart Rendering is easy with Locked standardized formats like HDV and DV-AVI just to name a few.

[_Al_]

Use HV20, HV30, HV40, and smart render from Vegas for example.
I know stabilization and narrow angle sucks (wide angle needed) but editing is relaxing matter even on older computers.

Getting AVCHD camera for home user is a classy example of a chase for a coolest gadget, the same one that a friend has, so I must have the same one too.
Apple and IPhone uses this drive also. No need to be in a hurry to get latest stuff.