Degraded Renders from Import of .veg files.

drdancm wrote on 3/16/2015, 3:09 PM
For a lengthy project it makes sense to work on segments saved with appropriate file names (xxxxx.veg) and then put these segments into the proper sequence for the finished full video, and then render it all.

My understanding is that these *.veg files contain simply information and instructions about the original video clips (name and location) as well as all of the editing information performed.

When you import such files there are only 2 options:
Files / Import >> Media
Files / Import >> Media from Project Files

When I am ready to build a finished sequence of these "chapters" the program stops to render them, I was thinking that perhaps the render is simply for the playback on the computer and would have no effect on the quality of the final render to create the finished video file which I would either upload or burn to optical media.

However, for video that consisted of credits with text created using the Rolling Display etc. the smaller text (12 to 14 points) resulting final video was horribly degraded, fuzzy, and difficult to read with either 720p or 1080p HD internet rendering.

I assume all video is degraded but one can best see the degradation with smaller text.

To get around this I had to fully re-create the credits, rather than import prior segments created and saved as xxxx.veg files. Hence my speculation that the poor final rendered quality was because the final render was made from a prior render rather than from the saved *.veg files.

Am I doing something wrong? Or is this how it was designed to work? I find it almost inconceivable that it was designed this way. It's like having a word processor that you use to write separate chapters, saving each, but when you paste these saved chapters into a long sequence that makes up an entire book, the text is degraded when you print (render) them. The only way to avoid this is to write all of the chapters in single file.

Thanks for clarifying this matter.

Dan

Comments

videoITguy wrote on 3/16/2015, 3:17 PM
too many details about your workflow, tools, and objectives are missing for any imformative comment.

Are you nesting? Are you wanting to use DI's? not clear at all where you are headed.

Text has to be entered in a project at its final rendering setting - in other words a text created in HD project full frame should only goto DI and output at that same size. Not reduced and certainly not scaled up.
drdancm wrote on 3/16/2015, 3:47 PM
I'm guessing that nesting is what I'm doing, but you need to tell me what nesting is.
What are DI's ?


Let me clarify. I edit by combining and modifying several video clips and save it with filename A. I render it and it looks good.

I then do this again and save it as file B. I render and it looks good.

Then I create the credits using Rolling Titles and save it as file C. I render and it looks good.

I'm now ready to assemble files A B and C.

The only way I know to do this is to choose File, Import (Media or Media from Project).
Is the official term for this "nesting" ?

When I bring these files in they are quickly rendered during the import (or possibly when I place them on the time line. I don't remember).

I save the whole thing and call it file D. I render it.

The rendered video looks pretty good except for the credits which look degraded, the text is fuzzy and hard to read, very poor relative to the render of C by itself.

All the render are with the identical settings, for example internet 720HD. I don't see why scaling is an issue at this point. There is no scaling as far as I can tell.

Now is that clear enough?


Since you find my detailed descriptions of what I am doing to hard to understand you might do me better by making your comments easier to understand. Please do us a favor and do not use abbreviations and terms that beginners may not understand.

Thanks for your quick reply,

Dan
videoITguy wrote on 3/16/2015, 4:18 PM
Di = digital intermediate - search the forum. DI as you are using should always be rendered in a lossless or near lossless codec. MagicYUV - search term in this forum.
All renders in intermediates before final output should be in same codec - especially the lossless type. Your final output render may be a highly compressed type but if at an artificially low-bitrate - all previous prep for naught.
johnmeyer wrote on 3/16/2015, 7:43 PM
I did a huge amount of testing on nested VEG files (that is what you describe in your first post). This was seven years ago, back in 2008. I posted in several threads, including this one: Converting to progressive when nesting, but my tests -- and the resulting posts -- were confusing. I did uncover a bug, which Nick Hope later verified was finally fixed in V12 (I think). However, while the bug may have been fixed, the other thing I did uncover is that the project settings in the project you are dropping on the timeline matters. Think about it, since you haven't rendered the project you are now putting on the timeline, what frame rate, resolution, and field order should be used? The only thing Vegas can use is the project setting.

Normally, for fast timeline performance, the recommendation is to match the project properties to the source footage, and Vegas even provides a button for this inside of the Project Properties dialog. However, in this case, before you drop that VEG file on the timeline of the project that you are using to assemble all the pieces of your project, you need to open each source VEG file, and make sure that project properties represent the quality of the video you want to appear on the timeline. In most cases I would still recommend matching these properties to the source footage. However, if you are just using whatever project properties you happen to have set as the default, and this doesn't match either the source footage or the project properties of your master project into which you are dropping all the VEG files, Vegas will have to make a lot of changes to your footage, and this could cause problems.

One thing I have definitely confirmed over the years is that Vegas does a terrible job of changing field order. This can (and should) be a trivial thing to do, but the way Vegas does it often results in a very soft-looking final render. I confirmed this to be the case, many years ago, if you used TFF (typical for HD sources) as the source, but then rendered this to BFF for MPEG-2 prior to making a DVD. After extensive testing, I found that much of the soft DVD look that many were complaining about when rendering from HD to standard def DVD could be overcome merely by rendering the MPEG-2 using TFF (the DVD spec accepts either). The difference in quality was quite noticeable.

To be clear, the problem was not BFF or TFF, because Vegas can read and render to either. Instead it was Vegas' improper handling of the conversion from one to the other, something that can easily be handled by just dropping the first field in the project (although there are other ways).

To my knowledge, I don't think Sony ever fixed the underlying bug that cause this problem, but they did change the default field order in the MPEG-2 template from BFF to TFF, probably because most people by now, when making DVDs, are creating them from source material that is TFF HD. So, since most people have long since moved to HD for acquisition, this "fixed" the problem for most of them, as long as they didn't use older custom templates that still used the BFF default.

So, look at the project properties for each VEG, and change them if necessary so they are all consistent with the project properties in your final, master VEG.