Exporting stills at print resolution

[Jeff Carpenter]

Follow this:

FILE> EXPORT USING QUICKTIME CONVERSION> FORMAT: STILL IMAGE > OPTIONS> TIFF> OPTIONS> COMPRESSION: NONE

Then open that in Photoshop and run the filter VIDEO: DEINTERLACE

That’s the best you’re gonna get, although it won’t be much. Printing at 300 dpi you’ll get an image that’s probably in the range of 3 inches wide. You can scale up and see how far you can push it, but there’s only so far you can go. At any rate, following my plan will give you the best possible image to work with.

[Tony]

FYI: Betacam SX is a compressed tape format. It is 4:2:2 sampling in NTSC.

Tony Salgado

[Scott Thomas]

There are a few things to keep in mind. SD video has a different pixel aspect than what a printer is going to need. When I have to take video captures and print them, I have a fairly long process that has to be done in a set order.

1. Take your video capture into Photoshop.
2. Determine if the image NEEDS to be de interlaced, (i.e. Is there motion in the frame?)
If the image has no interlace fringing from motion, there is no need to deinterlace
3. If deinterlacing must be done, it must be done before anything else is done.
Make a copy of the image layer so that you have two identical layers. On the top layer, choose Filter-Deinterlace-Upper Field. On the bottom layer, use the same filter but choose Lower Field. Take the top layers opacity to 50%. As long as there wasn’t too much motion you should have a much cleaner looking frame.
4. If random noise from the camera must be overcome, and there wasn’t too much motion. You can pick extra frame captures adjacent to the first, run the same deinterlace process as above and mix them into the original document. (I’ve done five or six frames together to reduce microwave feed noise with reasonable results.)
5. Now we need to correct pixel aspect. A 720×486 frame is simply scaled to 720×540. With DV footage, I add 6 pixels to the hight of the image with the canvas tool and then do the scale to 540.
6. I usually scale up the image to it’s final print size x 300 or 400 dpi and attempt some more processing.
7. The Median filter is helpful for reducing noise sometimes. Not too much.
8. Unshape Mask can sometimes get back some lost detail.
9. Working in LAB color space sometimes is useful for softening the chroma without affecting the detail.

The main thing is, don’t do it unless you have to. If it’s a video shoot that is staged, you should consider having a still camera on hand for such future needs. If you must use a video capture, scale up the image as little as you have to.

I had to get a ton of video captures preped for our annual Hurricane Guide. I had somewhere over 50 video captures that went into the magazine. I built a Photoshop action that did most of the steps above and also cropped enough of the image to get rid of the blanking. Thankfully, none of the images were printed too large. Some looked pretty good, some looked like video.

Scott Thomas

[Chris Poisson]

Khooper,

There are several apps out there that allow you to blow up stills much better than Photoshop. I use PhotozoomPro, and I have blown up video stills to print resolution with much better results than just uprezing with Photoshop. Genuine Fractals and Smartscale are also good I’ve heard.

Still, most stills from video kinda suck at best, so you should not get your hopes up too much.

Have a wonderful day.

[Michael Alberts]

In your case the exported frames from FCP (or any edit system) will be 720×486 at 72 dpi. That’s the resolution of standard definition NTSC video. There is no way around this. If however, you imported some high resolution still frames into FCP, those will retain their original resolution though the export process.

Michael Alberts
Ambidextrous Productions, Inc.

[David Bogie]

Nice work, Scott, thorough and concise.
Strongly urge you to paste it into a sticky on your machine so you ca paste it into the forum every once and a while. The topic comes up very often around here and on every other FCP forum.

bogiesan

This is my standard sigfile so do not take it personally: “For crying out loud, read the freakin’ manual.”

[John Fishback]

Check out Genuine Fractals at https://www.ononesoftware.com It’s a plug-in for Photoshop 7.0 or higher. Genuine Fractals converts bitmap still frames to fractal files. These files can be resized dramatically without the jaggies created by simple up-resing. This plug-in can’t create detail where none exists, but the results in the 400-500 percent blow-up range are often significantly smoother and less blocky than enlarging a pixel-based file.

John

Dual 2.5 G5 4 gigs RAM OS 10.4.3 QT7.0.3
Dual Cinema 23 Radeon 9800
FCP Studio 5
Huge U-320R 1TB Raid 3 firmware ENG15.BIN
ATTO UL4D driver 3.50
AJA IO driver 2.1 firmware v23-28
SonicStudio HD DAW, Yamaha DM1000, Genelec Monitors

[Scott Thomas]

Thanks for the link to the OnOne website. I downloaded the demo and did some tests on upresing a SD video image. It has some interesting and nice image qualities. Some detail is enhanced well, like the white coller of a shirt and the pupil of an eye. It does seem to over enhance some DV 4:1:1 chroma sampling. Now that I know that, it’s an easy workaround to use LAB color mode and blur the chroma a tad.

I wasn’t aware that the software was still around. It was 8 or 9 years ago that I first heard about it.

[Felix Bach]

Hello there.

Even if you guys have been discussing this some 6,5 years ago, I will continue this discussion and would like to have some things I dont get clarified.

We are currnetly working with ALEXA 4444 Files, one the highest resolutions available. We have shot old pictures from a museum (from a tripod, no movement in picture) and want to reprint them in order to use them as props for in some scenes. As it was forgotten to take photographs from the picture, we are now working on getting 300dpi printable stills from our video footage. So we dont need to go back to the museum. 🙂

The stills are to be printed on canvas and then get an old proper wood frame. The largest of those paintings has an original size of 64x81cm
and the smallest one 33x60cm.

My export from the original camera files has given me 72dpi Tiffs with a size of around 5,5mb. I have forwarded these stills to two graphic ladies who will try to get the max out of it and im still waiting for their feedback. Somebody in this forum mentioned that there are some nice fotoshop plugins to pimp the resolution. I have also applied the test version of Photo Zoom Pro 4 (file size after transforming to 300dpi from the 5mb-tiff-cp-export=105mb, which sounds good 🙂 and forwarded the result to the graphic ladies and also wait for a feedback concerning the quality comparism between the fotoshop-pimp and the photozoom-pimp. To be shared later.

I still would like to get some clarifications on a few points:

-Exporting from 4444 Timeline with original camera footage as TIFF is pretty much the best I can get out of FCP, isnt it?

-Why does this high resolution footage give me the same results in dpi (72) as DV material? Or is that wrong? It seems so natural to me that an ALEXA should produce something far more outstandind in quality than my old DV-bone. And that this should also be reflected in still exports. However I am not strong grahicwise and my fotoshop skills are even below basic.

-As we are talking square inches and non-interlaced footage no resizing/converting is needed in fotoshop right?

-Is this whole idea of getting printable stills from prores4444 a stupid idea anyway? Or my attempt wrong in the beginning?

Thanks for any help people. I will post all my final results and gained knowledge in this matter whenever I have all the information together.

Best regards, Felix