Does re-rendering lose quality in image?

R

Ruby Tuesday

Active member
Hi. I've heard differing views on this, so I'd like to hear your take.

If I re-render a file that's already been rendered, does it lose ANY quality in definition? What if I re-render it several times?

Thank you,

Ruben
 
Are you talking about altering the footage with, for example colour correction and then rendering, or just re-encoding?

If you alter an image in any way and render it, you will loose quality, because the image is being processed. If you are using uncompressed footage, the loss will be extremely small, however, if you are using DV or HDV, or any other high compression codec, there is a visible loss.

If you are outputting to DVD and shooting on anything but a DVD camcorder, you HAVE to recompress you footage to MPEG2. In order to keep the best quality, you want to encode to MPEG2 straight from you orginal footage, with no extra compression. You can add a stage of chroma smoothing, but do a search to find one of the threads which discusses that.

From you NLE you should be able to colour correct and add effects and then encode to MPEG2 without an additional render step. You may still need to render to pre-view your work, but the render files ARE NOT what is used to encode the DVD, the original footage is. The NLE does the render straight to MPEG2. In FCP you do this by using "Export using Compressor."
 
Sorry, just re-read your post.

Do you mean rendering a clip, not liking what you see, tweaking the filters etc, re-rendering, teaking, re-render...etc?

In that case you will not loose more and more quality each time, it is just a single stage compression. Each time you render, your NLE renders from the original footage. You will loose some quality from the recompression (see above) but you can re-render the same clip a million times if you want and it is the same loss as rendering once.

If however you render the file and then export/save it as a new clip and then add effects to that NEW clip, you will loose further quality.
 
Once a clip is rendered in DV format, it isn't rendered again, but the frames copied when being re-rendered. You'll see this if you pre-render half the video, then render everything -- the half that was pre-rendered will render almost instantly.

You can use this to speed up rendering of large projects by pre-rendering parts that are done, so by the time of the final render most is already rendered.
 
Hi, guys! Thank you for your response. Actually, what I'm referring to is, editing together a .veg file, rendering it, and then editing the rendered file (not adding FX, just editing clips), and then re-rendering the new edited-rendered file. Would that lose any definition at all?
 
But isn't a vegas file simply an organized compelation of 'mpg, avi, etc..' clips (that would already have some form of compression or encoding to them)? I know when I use vegas to capture HD, it saves them as a m2t file or something (that's an mpg file right?) Or when capturing as a DV, does it save them as a some sort of raw file?
 
James,

You're right, the .veg file is a compilation of existing rendered files, but I was just trying to explain that if I rendered that edited .veg file and THEN re-rendered the new rendered file, would I be losing any resolution. I just wasn't sure if "mmm" was kidding with me or what about not knowing what a .veg file was.
 
Ruby Tuesday said:
You're either pulling my leg or I posted on the wrong forum. .Veg files, you know, the ones that Vegas uses?
Click to expand...

Haha! Sorry, i've never used Vegas, I just picked up the thread from new posts because the same applies with all NLE's pretty much.:cheesy:

I'll stick with my .movs in FCP :beer:

I can't help on the specific .veg question then I'm afraid - my blab was just about rendering in general.
 
RubyTuesday:

If any of the data changes, then yes you will lose quality as the video gets recompressed. If the data doesn't change (which means you're doing cuts-only editing, no transitions no dissolves no color correction no cropping no *anything*) then the data will not be re-rendered, it'll be copied over -- and there will be no loss in quality. But if you do *anything* to the video at all other than straight cuts-only editing, then yes those sections that have been "touched" will need to be re-rendered. If you render to a compressed codec (like DV) then that means you will be losing some quality when you do so.

If all you're doing is cutting the timeline and re-ordering clips and snipping out sections, things like that will involve no re-rendering and you will lose no quality whatsoever; what you render out of that will be bit-for-bit identical to what you imported.
 
My experience with a .veg file is that when you bring it in and park it on a timeline, it's not being Re-Rendered ... so it's no different than having everything in it's original state (all the individual audio/video tracks) in that project ... instead, you're just streamlinging one project onto one video/audio track.

The reason I believe you'd not lose any quality with bringing in the .veg file is that when I have text that's in a .veg file over a video clip/track in my current project, the background is transparent, meaning it's not been rendered yet ... otherwise, if I WERE rendered already, the background wouldn't be transparent.

cc'ing & FX to video is a different issue, and yes there will be SOME loss depending on how you render, what you render to, etc. But bringing in a .veg file to another will not cause that imported .veg file to have any loss. ZERO. At least that's my experience.
 
I'm just glad there are 437 different opinions on it personally. Otherwise I'd feel we actually helped Ruby Tuesday
 
Actually... this pretty much makes things crystal clear to me:

Barry_Green said:
If all you're doing is cutting the timeline and re-ordering clips and snipping out sections, things like that will involve no re-rendering and you will lose no quality whatsoever; what you render out of that will be bit-for-bit identical to what you imported.
Click to expand...
 
I was wondering...what if you render it 10 times (applying different FX and CC) using uncompressed format, will there be a lesser loss compared to rendering it using DV codec?
 
bLueButterfLy said:
I was wondering...what if you render it 10 times (applying different FX and CC) using uncompressed format, will there be a lesser loss compared to rendering it using DV codec?
Click to expand...

Yes, you will see a smaller degradation of the file by rendering 10 times to uncompressed vs rendering a DV code 10 times back to a DV codec.
 
You must log in or register to reply here.
Back
Top