Comments

Chienworks wrote on 12/26/2013, 3:13 PM
It seems reasonable to me, since most of that information is stored in the file header while smart rendering copies frames, not the file. Reasonable, yes, but i agree unfortunate.

What you'd probably need is an external utility that can 'transplant' header info from one file into another. Keep in mind though it would probably set all the partial clips to have the same start date/time as the original, rather than offsetting them by where that clip started in the original file.

Of course, you could always suggest it to the Sony programmers as a product enhancement.
willqen wrote on 12/26/2013, 4:13 PM
I tested this using HDV files on my system and got somewhat the same results.

Even with smart render straight back to HDV all the meta data Vegas would save were markers, embedded thru the trimmer, or added to the timeline before render, provided I checked the box for it to do so in the render setup screen.

Your post made me wonder about metadata because I have done a ton of DVD's using Vegas to DVD Architect. Always with very good results. It has always been convenient that why because of the way chapters re created from markers, ie., metadata.

It appears though that that is all that is saved. No tape name or take name from media properties, or clip properties.

I haven't had a chance to test any raw camera footage, there is none in my computer presently but I will do so shortly and report back ...
PeterDuke wrote on 12/26/2013, 5:23 PM
OK, I'll show my ignorance. How do you view the HDV metadata?

I notice that mediainfo reports a menu stream in my HDV files. Is the metadata in this stream? If so, then it would not survive passing through Vegas. Smart render is only about preserving the video stream.

Note

I always use HDVSplit to transfer HDV files.
videoITguy wrote on 12/26/2013, 5:40 PM
This has come up time after time in reference to VegasPro working with timecode from HDV sources that have it. It has been my experience that any rendering (smart or not so) always throws away the timecode. And in part this is because VegasPro is an app that can not export timecode by design.
The video stream is apparently stripped of the metadata header.
There are many utilities like DVDate and DVMpro5 that can be used to review the header.
I work with DVMpro5 to get the timecode data (including GPS,etc) out of the stream, keep it in separate container away from the management and manipulation of the A/V streams and therefore have it accessible as a reference at any time.
farss wrote on 12/26/2013, 8:41 PM
PeterDuke asked: [I]"How do you view the HDV metadata?"[/I]

Vegas itself will display this when you RClick the media on the timeline and select Properties.

As I recently discovered even with HDV tapes captured by Vegas this can be very useful information. For example knowing the time each clip started to be recorded allowed me to easily sync clips from different cameras for a multicam edit.

Bob.
johnmeyer wrote on 12/26/2013, 10:43 PM
The easiest way to view the metadata is to simply turn on the display in your camera while playing the tape.

I think the data is present on every frame, or at least every group of frames, because you can see the time, f-stop, and other info change from second to second. Therefore I think it is present in the stream itself, and not just the header. It almost has to be this way because the data is present on a tape, and there is no concept of a header on a tape because it has to work even when there are interruptions (no video recorded) on sections of the tape.

Back when I first used Vegas, over a decade ago, I sent a feature request for the Sonic Foundry engineers to make use of much of this data. That was before I ever realized that they were doing even worse than simply failing to create a few trivial fX that could display the data, and were actually stripping it all away during the render.

I had hoped they would have changed that by now, but I guess not.

BTW, for those wondering how you could make this work in a multi-track environment, my proposal back in 2004 was to designate a "meta-data master track" and any render would use the metadata found in the video on that track, and would copy that to the final render. It would be so simple ...

Rob Franks wrote on 12/27/2013, 6:17 AM
"When I smart render HDV, Vegas loses ALL the embedded metadata in the video stream, including date, time, camera settings, etc."
John, you're speaking like it is the same file and it is not.
Smart rendering avoids recompression, but that doesn't mean it has been COPIED. It is a completely new file which has been produced by Vegas. It no longer has much to do with a camera and it wasn't created when you shot it.
Grazie wrote on 12/27/2013, 7:29 AM
You raise a set of further questions. When I render, does any of the metadata - camera info - remain with the "new" file?

Grazie

Rob Franks wrote on 12/27/2013, 7:46 AM
"When I render, does any of the metadata - camera info - remain with the "new" file?{"
Why should it?
It is a COMPLETELY new file which has been produced by what ever encoder within Vegas you have chosen to use..... and just because that encoder didn't see any reason to open up and completely dissect the file in question, doesn't mean it wasn't scrutinized by another encoder. Maybe that encoder added or altered something we can't see... nulldata perhaps. Change ANYTHING, DO anything other than a direct copy and it can be no longer represented by the original metadata.

Now I do know of some photographic programs that will carry the existing metadata forward after some change or edit has taken place and you have to wonder just exactly how honest that actually is.
Grazie wrote on 12/27/2013, 7:55 AM
Why should it?

Indeed. Is it possible that John was of the opinion that that info would be included in the render of a SMART render?

Thanks Rob.

Grazie

Chienworks wrote on 12/27/2013, 8:07 AM
Well, there's the difference here between forensic integrity and usefulness.

I suspect that probably for the most part (99:1 or more), most of us here probably have a lot more need for usefulness than forensic integrity. Copying all the metadata after an edit, or in the case of smart render probably nothing more than a trim, may not be honest as in acceptable in a court of law, but in many cases it sure would make future reference and edits a lot more easy and happy.
S
ince metadata can be faked easily and often without deliberate intent or malice, it's not really evidence anyway but merely helpful clues. For example, with my first DV camcorder after the CMOS battery died, i never bothered resetting the clock. The last 100 or so tapes i ran through it all claim to be recorded at 12:00:00am, January 1st, 1980. Since it's unlikely that i shot thousands of scenes all in that same second, and in fact i didn't even buy the camera until 2002, 'honesty' of metadata is rather a moot point. Of course, i wrote the date and name of the event on every MiniDV cartridge, and 'honest' or not, misspellings and execrable handwriting aside, that was far more useful to me than the metadata anyway.
JohnnyRoy wrote on 12/27/2013, 8:43 AM
I agree with Rob Franks. Rendering produces a new file that should not contain any of the timecode information, camera settings, etc. from the old file because it's a new file with it's own timecode and modifications. What johnmeyer wants is a program to split HDV files. That's not what Vegas Pro "smart render" is for. Once you "render" camera settings are irrelevant because the file no longer comes from a camera... it comes from an NLE.

~jr
Grazie wrote on 12/27/2013, 9:34 AM
Thanks JR, very clear.

G

johnmeyer wrote on 12/27/2013, 10:12 AM
I agree with Rob Franks. Rendering produces a new file that should not contain any of the timecode information, camera settings, etc. from the old file because it's a new file with it's own timecode and modifications.I'm afraid I don't understand that point of view. It seems like it is motivated by some quasi-legal sensibility that somehow I might be able to fool someone by putting fake metadata into the final render.

That is absolutely true, and what is wrong with that??? After all, every single pixel of every video that any of you have produced is fake.

This includes not only those for whom video is a way to create alternate realities, but also those who are trying to simply record what you see in front of you and transfer it to a viewing medium. In every case, no matter what you do, the video lies because it is not fully three dimensional (even so-called 3D); its colors are not 100% true; it has limited field of view; you cut between multiple cameras, and multiple takes; it is only 24, 25, 30, or 60 frames per second (or whatever ...), and so on.

If the media has the ability to store metadata I should be able to pass along the information that was on the original tape or, for that matter, completely alter it and put new information in its place. We do this all the time when, for example, we copy a computer file. By default, for Windows users, that copy of the original file has the same date & time for the "created" and "modified" metadata that is included with the file.

Arguing that all the metadata in a video file should be thrown away is, to me, exactly like making the argument that a copy of any computer file should have all of its metadata thrown away each time it is copied, and then expect that it be left to the user to remember what date and time the file was created.

Finally, the EXIF information in a JPEG file is kept intact by my photo editor when I make changes to the photo. So how is OK for every photo editing program under the sun to preserve the entire EXIF package of information, but for Vegas to wantonly throw away exactly the same information??

Answer, it isn't OK for Vegas to throw it away, and it isn't OK to not give us options for using that data even before it is thrown away, especially since it is very, very useful data.

JohnnyRoy wrote on 12/27/2013, 10:34 AM
> "I'm afraid I don't understand that point of view. It seems like it is motivated by some quasi-legal sensibility that somehow I might be able to fool someone by putting fake metadata into the final render. "

Which timecode would you like to use? Which camera setting would you like to use? Let's say you just placed 5 events from 3 different cameras shooting 24p and 60i on the timeline. Now you render. What should Vegas do? What metadata data does it keep? Keeping the timecode doesn't make sense. Keeping the camera data doesn't make sense. The render is a new file completely distinct from it's source.

> "Arguing that all the metadata in a video file should be thrown away is, to me, exactly like making the argument that a copy of any computer file should have all of its metadata thrown away each time it is copied, and then expect that it be left to the user to remember what date and time the file was created."

That analogy doesn't apply here and is the source of your confusion. This is the analogy: I open Microsoft Word. I import some text from another word document. I paste a picture in next to the text. I save the document. Why would that document have any metadata from the original source files? Answer: it should not!

This is why I said that you want a program that splits HDV files. You want to make COPIES! Vegas Pro is not making copies. It it creating totally new content from existing content much like combining information in a word processing document. You are confusing the fact that "smart render" COPIES the video stream with wanting to copy FILES. It is not copying the file. It is placing an exact copy of the video data stream into a completely new file with it's own set of metadata. You are using a "fringe" case of "what if there is only one even on the timeline? isn't that like copying a file?" and the answer is NO it is not like copying a file. 1 event or 100 events from 100 different cameras, a render is a new piece of media with it's own metadata.

> "Finally, the EXIF information in a JPEG file is kept intact by my photo editor when I make changes to the photo. So how is OK for every photo editing program under the sun to preserve the entire EXIF package of information, but for Vegas to wantonly throw away exactly the same information??"

Once again, different analogy. The photo editor in your scenario is editing one photo. I agree 100% that the metadata in that photo should be preserved. This has NOTHING to do with the Vegas Pro scenario. The Vegas Pro scenario is this: (in keeping with the Photoshop analogy) You open Photoshop. You create a new document. You use FIle | Place to add an existing JPG to that document. You save the new document as a JPG. That new JPG has NONE of the metadata from the photo, nor should it. This is EXACTLY how Vegas Pro works.

> "Answer, it isn't OK for Vegas to throw it away, and it isn't OK to not give us options for using that data even before it is thrown away, especially since it is very, very useful data."

That would be true if Vegas Pro edited video files. It does not. It is a content creation tool and all it knows how to do is create NEW video files and the original files remain intact with all of their metadata on your hard drive. I would agree with you if Vegas was actually editing video files and throwing the metadata away. That would be wrong... but that's not what Vegas does. Vegas renders Projects that have their own properties that are separate and distinct from the video content that you place in the project.

So once again... you want a tool that can split HDV files and keep their meta data. Don't use a Non-Linear Editor for that tasks because that's not what NLE's do.

~jr
john_dennis wrote on 12/27/2013, 11:01 AM
@jr

"Let's say you just placed 5 events from 3 different cameras shooting 24p and 60i on the timeline. Now you render. What should Vegas do? What metadata data does it keep?

In the context of a smart render this would not be case.
JohnnyRoy wrote on 12/27/2013, 11:28 AM
> "In the context of a smart render this would not be case."

It absolutely could for one of the files. So, for example, if I was rendering to HDV and I placed two events from a Canon 7D and two events from my Sony Z1U HDV camera on the timeline, Vegas Pro would happily smart render the event containing the HDV data, copying the video from the source media into the target file along with rendering the events from the 7D. Smart render will always work for the events containing media formats that it supports regardless of what else is on the timeline as long as there are no FX, cropping, compositing, etc. on those events. My scenario is valid and is the reason that smart render has no bearing on the metadata contained in the new file.

~jr
john_dennis wrote on 12/27/2013, 11:31 AM
Noted.
JohnnyRoy wrote on 12/27/2013, 11:55 AM
Yea, the misconception here is that somehow you can place a single file on the timeline of a project and render it to the same format and smart render will copy the file with metadata intact when what is really happening is that you have created a new project and placed a new event in that new project and rendering the project to a new file with the contents of that event, and just because Vegas Pro has the ability to recognize that the video data in the event can simply be passed thru to the new file unchanged (without rendering), that somehow the newly rendered file should contain the metadata from the media of the single event that was part of your project. That's just not the way it works.

~jr
johnmeyer wrote on 12/27/2013, 12:16 PM
Which timecode would you like to use? Which camera setting would you like to use? Let's say you just placed 5 events from 3 different cameras shooting 24p and 60i on the timeline. Now you render. What should Vegas do? What metadata data does it keep? I already answered that question above: you should be able to designate a "master metadata track" and any video on that track would have its metadata used in the final render. In addition, I should be able to add a metadata fX to that master track that would let me alter and embedd further metadata. Closed captioning is one form of metadata, and most people were very happy when Vegas decided to finally utilize that one specific field of data. Why is that metadata OK to add and carry forward, but other data is not?


This is the analogy: I open Microsoft Word. I import some text from another word document. I paste a picture in next to the text. I save the document. Why would that document have any metadata from the original source files? Answer: it should not!Except that it actually does!! Go to the File --> Properties dialog in Word (at least that where it is in the older version I use). There is a huge amount of data there that is not destroyed and carries forward with the document. So it has both newer data as well as original data. There are many other places within Word where you can access even more document data. So Word actually proves the point and shows how valuable it can be to know when the document was originally created, by whom, , even after other text is imported.


You are confusing the fact that "smart render" COPIES the video stream with wanting to copy FILES.You are taking this to a rather abstract level by worrying about how we define certain terms rather than thinking about what would be a useful feature. Lets instead concentrate on that much more useful approach to product design. All I am asking for is a simple and easy way for me to be able to identify exactly and precisely when the original video that I am watching was taken, without having to jump through hoops and without having to spend time adding text overlays, closed captions, or subtitles. If I play back the original tape, I can push a button on my camera and get that data. I want to be able to edit out the "dead wood" and when I then create the cuts-only version of my original tape I'd like to have that same ability to view the original information on the tape I create from this edited content. This is completely reasonable, and would be tremendously useful. (And it doesn't have to apply just to a tape workflow.)

Once you stop resisting the idea with statements like "why would I want to do that?" and instead explore the possibilities, you'll find that there are a wealth of really useful features waiting to be discovered. You are an excellent programmer and product designer, so you should appreciate that.


You are using a "fringe" case of "what if there is only one even on the timeline? No, I am starting with that, but then with the idea of a master track and a metadata fX, I am expanding the idea to where it can be used for all sorts of additional useful features. In addition, it should be fully integrated into one of the title generators, to closed captions, and to subtitle file export for workflows that don't involve printing back to tape. For instance, I've asked for years to be able to extract the date/time metadata from a master track and then, after editing and rearranging, have that data exported to a subtitle track so I could turn on/off a "date/time" overlay on the DVD or Blu-Ray disc.


You create a new document. You use FIle | Place to add an existing JPG to that document. You save the new document as a JPG. That new JPG has NONE of the metadata from the photo, nor should it. That is not how my photo editor works. If I modify the photo -- whether by simply altering contrast and color, or whether I bring in additional stuff from other photos, it retains all of the original EXIF information from the "base" photo. I just checked to make sure this is so, and it is. So both Word and my photo editor do exactly what I want them to do and not what you say they shouldn't do.


That would be true if Vegas Pro edited video files. It does not. It is a content creation tool and all it knows how to do is create NEW video files and the original files remain intact with all of their metadata on your hard drive. I would agree with you if Vegas was actually editing video files and throwing the metadata away. That would be wrong... but that's not what Vegas does. Vegas renders Projects that have their own properties that are separate and distinct from the video content that you place in the project.Once again, what you are saying really isn't any different from how these analogous tools work. Word also incorporates content from other programs -- photos and line art and "objects" -- and then prints out the results. All the metadata from the first file opened is preserved (and/or updated) as these edits are made. It is not thrown out and abandoned. My photo editing program lets me combine together material from many different sources, and it leaves it to me whether I still want to retain the original EXIF information, alter it, or dump it. If the underlying photo is still the photo I took of the Eiffel Tower at 9:00 p.m. on the night I proposed to my wife, I'd rather like to have that exact date and time remembered.


riredale wrote on 12/27/2013, 12:27 PM
Hmmm...

Interesting discussion with different points of view.

In any event, Merry Christmas to you all. I got a winter jacket and a sweater (which I'm taking back) and two kinds of socks (which I'm keeping). These days I get a lot of satisfaction seeing little people getting stuff they really wanted. Like the bicycle I got when I was seven.
johnmeyer wrote on 12/27/2013, 12:31 PM
I got a winter jacket and a sweater (which I'm taking back) and two kinds of socks (which I'm keeping).I got a few nifty things as well, but the best thing was a gift I got for my daughter which made her dance. What a treat:



JohnnyRoy wrote on 12/27/2013, 1:19 PM
> "...you should be able to designate a "master metadata track" and any video on that track would have its metadata used in the final render."

That's an interesting concept. So the final render inherits metadata from events on this track. Hmmm... I'm just not sure how useful it would be. You said yourself in your first post that you using Vegas to archive tapes. That's not what Vegas' main purpose is which is why I suggest that you should use a program, for splitting HDV files which does maintain metadata because it really does edit the actual file. I use Vegas to create new content and I'm not that concerned about the metadata from the source files once I've created my new content because it probably bares little resemblance to the original files. In your case, you are archiving files so metadata is more important.

> "Once you stop resisting the idea with statements like "why would I want to do that?" and instead explore the possibilities, you'll find that there are a wealth of really useful features waiting to be discovered. You are an excellent programmer and product designer, so you should appreciate that."

How do I argue with you once you compliment me like that? ;-) Yes, I can appreciate that being able to save metadata is important. I understand how in your scenario of archiving tapes this could be extremely important. I don't see how having EXIF data from images that I dropped into my timeline that are now video, or camera settings from multiple video clips that are now composited and indistinguishable from each other are valuable in the final video but you did say that what metadata gets saved should be configurable. I guess if you just wanted metadata in the final output that said this video file came from footage shot on these 3 cameras and 20 images with these EXIF settings, it could be interesting but I'm not sure how useful it would be in general since you wouldn't know what parts of the video the metadata was referring to. But you did just give me an idea for a new add-on to Vegas Pro so I will keep an open mind and keep listening.

BTW, did I mention that you'd probably be happier with a program that could split captured HDV tapes and preserve the metadata? lol :-D

~jr
NormanPCN wrote on 12/27/2013, 1:27 PM
A photo editor edits "A" photo so preserving EXIF if you want can make some sense. Not all EXIF would still be applicable but much will still apply.

If you brought two photos into an editor and saved a single photo, then what EXIF to save?

This latter situation is what the Vegas position seems to be. It is an editor assumed to be multiple input, single out. When you will have multiple sources, what metadata do you try to keep.

I suppose Vegas could detect the special case when only one media clip exists and preserve metadata in that situation.