Sondra,
Is there a filter or something in or out of Photoshop that would make a frame from a Premiere project print well at 22×35–or even 16×20? Or possibly part of a frame, so it could be portrait rather than landscape?
Well, as you may have gathered from the jolly responses above, this isn’t generally recognized as something you can do. However, that doesn’t mean you shouldn’t give it a try. Whether you will be satisfied with the result depends, in part, on your expectations for "print well".
I print a lot of frame grabs from DV video, but I haven’t tried to print anything larger than an 8×10. I do use a lot of "tricks", both inside Photoshop and outside, to compensate for starting with an image that measures only 654 x 480 square pixels.
If you are willing to regard your 16×20 or 22×35 prints as posters rather than as a fine art prints, you have a chance of getting something you will like, but you will need several tools in addition to Photoshop (I use Neat Image, Focus Magic, and Genuine Fractals). If you don’t mind a sort of "stylized" look, you could convert your image to a vector graphic, which can be scaled as large as you like. If you are really committed to higher yet bitmap quality, you will need to use special expensive software that processes several sequential frames. With that technique, it is possible to retreive some remarkably good stills from video.
That software is called VideoFOCUS from Salient Stills (there is also a VideoFOCUS Lite version with a reduced feature set) and it processes consecutive images from a section of surveillance tape, or from any video, to obtain greatly enhanced stills. Even though single frames of surveillance tapes are virtually useless, VideoFOCUS has been very successful at extracting useful stills from them. Many newspapers and magazines use VideoFOCUS to extract printable stills from their video footage. This website shows how VideoFOCUS works:
<
http://www.upsizing.co.jp/english/videofocus2.htm>
VideoFOCUS does the seemingly impossible by extracting additional information from nearby frames and using a smart mathematical correlation technique to focus and inject that information into the particular frame you have selected. Surveillance tapes are typically very low quality and no single frame is any good, but a sequence of frames, when properly processed, can yield some amazing results. A couple of examples of the quality enhancements achieved are shown at:
<
http://www.upsizing.co.jp/english/videofocus.htm>
A couple of examples of VideoFOCUS used on surveillance tape are shown at:
<
http://www.latimerclarke.com/index.php?option=news&task= viewarticle&sid=39>
<
http://www.videosciences.com/>
VideoFOCUS is not priced for private ownership, because it is quite expensive and is usually purchased by a government agency, a law enforcement agency, or a media corporation. Even VideoFOCUS Lite, bundled with a PC, is priced at $9,995.
<
http://www.salientstills.com/web-brochure/index.html>
The Salient Stills homepage is:
<
http://www.salientstills.com/>
Until a consumer version of VideoFOCUS, or something like it, becomes available, we will have to rely on processing a single video frame with tools like Neat Image, Focus Magic, and Genuine Fractals. If you want to know more about that, let me know.
— Burton —