Can someone please explain what targa sequences of video files are useful for and does Vegas do anything interesting with them?
how does their quality or usefulness compare with other video files?
Targa sequences can be imported as a connected series of stills, Vegas will then see them as one file rather than 30 individual images per second. TGA allows higher resolution to be imported than many cameras can output, and is a great intermediary between applications while suffering no loss. They can contain alpha channels (transparent information) or not.
Vegas doesn't do anything 'different' or 'exciting' with TGA files, it's just a file type that Vegas can import and export.
Some stock footage companies ship their products in sequential TGA file formats.
So if one were to photograph the molding bread on the counter every day, one could convert those photos to TGAs (via Photoshop, etc...) then drag them to Vegas and make a seamless movie of molding bread?
(oooh that was a bit of a run-on)
Or someone could use this approach for stop-motion animation?
thanks Spot - you are always helpful - seems like you are always there..
re: <TGA allows higher resolution to be imported than many cameras can output> does this mean input should immediately be converted input to TGA to have this advantage?
re: <is a great intermediary between applications while suffering no loss. > does this mean that CG files (say from Blufftitler) fare better as TGAs rather than AVIs?
does this mean input should immediately be converted input to TGA to have this advantage?
Generally speaking, no - there's no need to convert to TGA. There is usually no point in converting a compressed format (DV or jpg, for example) to TGA because the image has already been compressed and had its quality affected by the compression - converting to TGA will not undo or fix this. Converting from a compressed format to TGA is probably a waste of time and hard disk space unless:
. you want to add an alpha channel to an image or image sequence,
. you need to import the image or sequence into another program that does not accept DV or that prefers TGA
. You need the image or sequence to be imported and exported from a variety of programs without the loss of quality that comes from re-encoding (TGA, even when compressed, is lossless). Mind you, Vegas' re-compression to and from DV is so good that I'm frequently lazy on this point.
There might be other reasons to convert to TGA, but it is not possible to improve the quality or real resolution of an image or sequence just by switching file formats. Garbage in garbage out. :-)
Of course, if your image sequence is already in an uncompressed format (for example, computer generated images), then it makes sense to keep them in an either uncompressed or compressed-but-lossless format, and TGA has the advantages of being widely supported and allowing alpha channels.
does this mean that CG files (say from Blufftitler) fare better as TGAs rather than AVIs?
Generally speaking, yes - since most people use a lossy codec when they save their AVIs (something to be avoided until final delivery if then) it makes sense to save out as a TGA sequence. This is quite common.
Of course, you could save out of BluffTitler using uncompressed AVIs, which would basically result in an AVI file size about the same as a series of 24-bit BMP images (assuming no audio), But an uncompressed AVI does not support alpha channels (unless there's an AVI codec out there I'm unaware of), and takes up more hard drive space than a series of compressed (but lossless) TGAs (though the size of the difference depends on the nature of the images).
I've not used Blufftitler - does it allow rendering to Quicktime? Quicktime has several lossless codecs available, some of which support an alpha channel. If the idea of generating a folder full of images is bothering you (some people seem to feel it is "messy", even though Vegas treats the sequence as a single event on the timeline), and you'd rather a single animation file, then I suggest the TGA option in Quicktime - it basically wraps up a series of TGAs into a Quicktime file, giving you most of the best of both worlds.
Having said that, there are many advantages to saving out as a TGA image sequence. For example, most image editing programs (Paintshop, PhotoImpact, Gimp, etc) have batch-processing functions that will apply filters and effects to a series of images in a folder - something they generally can't do to a video file. So if you have some favourite effects in your image app that you want to use on a piece of video, a series of TGA files might be the way to go.
I recently completed a simple music video that was entirely computer animated (using Animation: Master (A:M) for the modelling/animation/rendering and Vegas for all compositing and editing). Rendering from A:M to TGAs and importing into Vegas was quite easy, and gave me certain freedoms that rendering to an animation file would not have (for example, stopping and restarting the render at different points, rerendering only certain parts of a sequence and having the new images just show up in Vegas like magic, no need to re-import the new bits, etc). I was essentially able to work in Vegas using uncompressed media all the way through until it came time to render out the DV and MPEG2 versions for the client. And since each render was from the uncompressed Vegas timeline, each codec (DV and MPEG2) were able to work on uncompressed images for the best result.
Okay, that was a long answer - so here's the summary:
. If your images come from an uncompressed source (eg, 3D program), then saving them as a TGA sequences is recommended.
. If your images come from a compressed source (eq, DV), then converting to TGAs is often a waste of time and hard disk space, unless you have a compelling reason to do so.
Uncompressed AVI does support alpha channel, straight out of and into Vegas although Vegas needs a prod to make it read the alpha channel mostly.
I think the attraction with these kinds of formats is a bit like 35mm film, it's rather format independant. Say you import a targa sequence, unlike DV which is either PAL or NTSC, the targa sequence can be rendered out to either with no loss.
I don't know though is there any advantage, other than targa being an industry norm, to targa over say tiff or png?
Thanks for the info tarss. Uncompressed AVIs support an Alpha channel? That's (good) news to me. It's been a while since I tried that, I will have a go at it soon.
But I'm not sure what you mean when you say:
it's rather format independant. Say you import a targa sequence, unlike DV which is either PAL or NTSC, the targa sequence can be rendered out to either with no loss.
If we're still talking about computer generated images, if the sequence was rendered for PAL then converting it to NTSC still has all the resolution and temporal issues that converting PAL DV to NTSC DV has, only the recompression issue is avoided (which is great, but hardly format independant).
Unless, of course, the sequence is rendered out at a significantly higher res than PAL or NTSC, then I guess resolution issues are avoided when converting to either format (assuming progressive scan) - but the temporal issues remain. An image sequence designed to play back normally at 25fps is going to either need some conversion to play back at 29.97, or else it's going to run fast (if one PAL frame is copied to one NTSC frame).
But an uncompressed AVI does not support alpha channels (unless there's an AVI codec out there I'm unaware of),
Sure there are. Always have been 32 bit avi files to my knowledge. It's been available in Vegas since the first rev. Even Ulead's toy Cool 3D had this option 7-8 years ago.
To clarify from my perspective; If your images come from a compressed source (eq, DV), then converting to TGAs is often a waste of time and hard disk space, unless you have a compelling reason to do so.
If you are jumping back and forth between a compositing tool and you don't want to wait for uncompressed renders both times through, sTGA (sequential TGA is usually a little faster, depending on which app it is, and it's very clean to import and export. Consider it like a high quality language that both tools can speak. With Maya or Caymen, Ulead or Lightwave, this is really the only great way to get media out while maintaining quality. Of course, if you're working with DV, the sTGA will be eventually compressed down, but the higher starting quality, the higher compressed/finished quality.
TGA files also don't have the shift in color like *some* png's do, I'd like to know what causes this on only some PNG files. So, saving stills from Pshop or other apps as a TGA is overall a better option, IMO.
There was a rather lengthy bit some time ago in the local PC mag which suggested that Adobe's implimentation of PNG was not upto scratch which may explain what you're seeing. I believe there's a plugin from the open source guys for PS that fixes this.
Haven't tried looking for it so this could all be hearsay.
Skevos,
what I'm trying to say is, if you dropped a targa sequence onto the Vegas TL and rendered out to PAL DV25 you get results as good as they get, if from the same TL you render out to NTSC DV25 you get NTSC as good as it gets, all that assumes your TGAs were higher res than than the DV to start with.
I think same goes with film, you can run it through a telecine and get a good PAL copy and run it through again with pulldown and get a good NTSC copy. If all you've got as a master is PAL or NTSC going to the other format is going to mean a quality hit, also if you need a HiDef copy, no sweat, just run the telecine at higher res.
Spot said (referring to AVIs with Alpha channels):
Sure there are. Always have been 32 bit avi files to my knowledge. It's been available in Vegas since the first rev. Even Ulead's toy Cool 3D had this option 7-8 years ago.
Thanks! I can't comment on Cool 3D, but Vegas certainly has this feature (just played with it). Just shows you how long it's been since I tried to make an AVI with an alpha channel (I think it was back in my Ulead MediaStudioPro days). Mind you, since it only seems to work with uncompressed AVIs, the lossless file-size savings of TGA Quicktimes means that I don't see myself ever using this AVI feature. Still, good to know it's an option.
To clarify from my perspective;
I've never tested that, but it sounds like another advantage to a sequence of TGAs in addition to the ones I posted. On your experience, is the speed difference between, say, a Quicktime using TGAs and a sequence of TGA images significant?
Consider it like a high quality language that both tools can speak. With Maya or Caymen, Ulead or Lightwave, this is really the only great way to get media out while maintaining quality. Of course, if you're working with DV, the sTGA will be eventually compressed down, but the higher starting quality, the higher compressed/finished quality.
I strongly agree with this. Staying uncompressed until rendering the final delivery item (DV or MPEG2 etc) is definetely the way to go. Recompression, even with the nice job that Vegas does, is to be avoided wherever possible.
TGA files also don't have the shift in color like *some* png's do, I'd like to know what causes this on only some PNG files. So, saving stills from Pshop or other apps as a TGA is overall a better option, IMO.
I've heard of PNG issues on deveral forums, but never done any tests. In fact I rarely use PNG at all - if I want lossless compression I use TGA sequences or TGA/Animation Quicktimes (because they work in just about everything), and if I can stand lossy I use DV for video and JPG for images. I should investigate PNG, as it's meant to be a file format that is more open and unencumbered by any potential patent issues, yes?
Okay, I think we're in agreement then - if when you say "you get NTSC as good as it gets" you essentially mean "you get a PAL-NTSC conversion that's as good as it gets". That I would agree with - but the NTSC version would still have the temporal issues that come with frame-rate conversion from 25 to 29.97. Obviously, if you rerendered a second image sequence from your 3D app so that it was made to run at 29.97fps (ie, same running time but more frames), then this would result in a much better "true" NTSC version. But it's rare (for me and my clents anyway!) to have the time and budget to render two versions of every animation sequence, generally (being in Australia) I render to PAL and convert to NTSC.
I think same goes with film, you can run it through a telecine and get a good PAL copy and run it through again with pulldown and get a good NTSC copy.
Sure - but if your final product is primarily aimed at NTSC markets, then you'd often shoot film at 29.97fps so that there would be no frame-rate conversion necessary (pull-down looks okay, but nothing looks better than one frame of film for every video frame IMHO). But I think you know this and we were talking slightly across each other - I was referring to wether or not using and/or converting to TGAs was best for quality reasons. I think we all agree that using uncompressed or lossless formats is definetely the way to go if your material is not already compressed.
If all you've got as a master is PAL or NTSC going to the other format is going to mean a quality hit, also if you need a HiDef copy, no sweat, just run the telecine at higher res.
Yup - This is what Kodak are talking about when they describe film as "future-proof". With every upgrade to the resolution of home screens, all you need to do is rescan your neg and you're right there with stuff shot with the latest tech. Though it doesn't address the framerate issue of course - what if a future video format runs at 100fps? But from a resolution point of view, it will be some time before home screens exceed film's resolution.
Mind you, developments like the Kinetta camera (assuming it ever goes into production) it means that acquisition in greater-then-HD resolution is just around the corner (the prototype is HD, supposedly the shipping version will be more like 2K - and you can pull the CCD out and upgrade the resolution). The Kinetta even has a crank handle for Pete's sake! We live in interesting times. :-)
Spot, when loading a TGA seq into Vegas 4 or 5, is there a setting other in "preferences" for tga time spacing. ie., I have 15 frame/sec TGA and 25 frame/sec TGA, that I want to put on a 30 frame/sec timelime,( like just drag and down). Thanks in advance.
No, you'll go to file>open and select the first tga in the sequence, click the Open in Sequence checkbox, and then click the last TGA in the sequence. Vegas automatically opens the sequence as a single file so you don't have say...900 separate images that you would be challenged to apply filters to.
thanks Spot, I have always treated the "file>open" seq just for veg files. I quest I'm been traveling too much in the country by the side roads vs using the freeway!
I use targas all the time and swear by them - but there is one little bugish thing in vegas using any file seq .. i should warn you about ..
that is - if you use the 'replace' feature , as i often do, to swap out different sources .. file seq. do not 'replace' //
the work around for this is to frame serve the targa 's into vegas using virtualdubmod and create a fake avi ... these can be swapped at will ..
an added bonus is that you can change the frame rate of the file seq in v-dub before importing as a frame served clip into vegas - this has some great advatages (to me anyway) .. v-dub also has better and faster resizing and cropping than vegas -- so this function also makes sense to do outboard ..
i just again tonght have a 3D right and left eye render going right now (overnight) using these methods ...
frame serve each targa seq (2k files) into 2 x vegas's using 2x v-dub , in each vdub is a resize , smoother (acting as a dither) and crop and resize (again) and a frame output fps adjustment .. the resulting 1080p HD files are sent to the 2x vegas where a track motion adjustment and color matching operation is performed to sync and match up the 3D (using the frame served output of these two to a third temp vegas to check 3d) , also a cut down edit and new ending credits are added .. i use the swap function of vegas also in this proccess ..
the primary 2 vegas 's are then rendered out as satish frame served avi to 2 more v-dubs , which then render out new targa seq 's ... one job after the other for each eye ..
there is another app running which checks when the final v-dubs are finished and starts up a pre-configured encoder to encode these new targas to a wavelet compression file ...
all of this runs on one dual xeon system ...
anyway -- targa's and other file seq.'s are used a lot in the film industry...
B JM, You're description of the TGA process, for VEGAS, is exactly what I'm interested in. I'm been working on a large number of TGA files, and your approach for variable rate is right on the number! The "replace" note will be helpfull also. The 1080p project your working on sounds similiar to my project, but better structured! I like the way you managed the TGA processes. I've considered using Vegas as a frameserver, but have not yet tested it. it would be interesting to see the "flow model" of your setup.