imported image sharpness

nyee · Post by **nyee** » Sun Sep 17, 2006 5:02 am

I'm importing images (JPG/PNG) into VideoStudio 10 and have tried several things but they turn out mildly blurred in the rendered video. This is most noticeable when there is sharp text in the original image and they come out blurred in the rendered video. I've tried:

- switching between image formats (JPG/GIF/PNG)
- adjusting the resampling quality under preferences to best
- tried right-clicking the image and using the "default/original" sizing options in the preview window

I also have some text in some of these images and it's very distracting to have blurry text. Is there a way around this problem?

I would just use titles to do this, but I have text callouts throughout the video and need a "copyright-confidential" text line at the bottom of whole video. So I figured the easiest way to add that text would be via an image. Let me know if there's a better way to do this.

Trevor Andrew · Post by **Trevor Andrew** » Sun Sep 17, 2006 6:59 pm

Hi

What size are the images you are using, and are you Pal or Ntsc.

Trevor

nyee · Post by **nyee** » Sun Sep 17, 2006 9:18 pm

In the videos I'm working with, the raw footage has a resolution of 1152 x 864, so I made images (in Photoshop) that were 1152 x 100 (for the trademark). I used the "fit to screen" and then "fit aspect ratio" options to scale them to what I thought would be the exact size. When the result came out blurry, I then tried making the images slightly bigger and smaller to see if it would make a difference (1280 x 100 and 1024 x 100), but it didn't seem to.

I'm using NTSC.

When you put an image in the overlay layer, is there a way to specify its width and height (to the size of the original image)? Would that solve this problem?

Jerry Jones · Post by **Jerry Jones** » Mon Sep 18, 2006 12:43 am

nyee wrote:In the videos I'm working with, the raw footage has a resolution of 1152 x 864, so I made images (in Photoshop) that were 1152 x 100 (for the trademark).

Man, I have a hard time understanding this.

If you're editing DV .avi files, then 720 x 480 would be the correct resolution for your video.

If you were editing 720p HDV .mpg files, then 1280 x 720 would be the correct resolution for your video.

In all my years of editing, I've only run into an odd-ball frame size like the one you mention when I encounter people who work with DivX.

Is that what you're trying to edit?

You haven't mentioned your target output...

DVD?

Web?

MiniDV?

There are PREFERENCES settings that can have an impact on how your still images are rendered, but before I even get to that point, I'd like to learn more about what you're attempting.

Jerry Jones
http://www.jonesgroup.net

nyee · Post by **nyee** » Mon Sep 18, 2006 1:01 am

Thanks for the reply.

The raw footage is captured video game-play using FRAPS. 1152 x 864 is the highest resolution that FRAPS can capture at. So that footage is using the FRAPS compression. We're trying to retain as much of the original resolution as possible, so we're using these raw videos to produce videos at 1152 x 864 (big files to be played back locally on computers).

I'm outputting as AVIs using the xVID compression. I know the "image text" is blurred because the text from the titles is sharp, so I know it's not an output rendering issue (i.e., I know it can appear sharper).

---

Also, the raw footage really doesn't have much to do with this right? - because it's the "text image" on the overlay layer that is blurred. Even if I used a solid black background at whatever resolution, the "overlay text image" still comes out blurred.

So I guess the question is really - if you needed to add text via an image in the overlay layer (because the titles layer is already being used or whatever other reason), what are things you can do to make that text appear as sharp as possible in the rendered video?

Or put another way - when you create an image (JPG/PNG) in Photoshop with sharp text and add that to your video in the overlay layer, does the text come out sharp in the rendered output or does it always come out mildly blurred?

Jerry Jones · Post by **Jerry Jones** » Mon Sep 18, 2006 5:56 am

First, export from Photoshop using a non-compressed IMAGE file format.

Based on what I read in your post, I suspect you're losing quality during the export from Photoshop by using compressed image files such as JPEG & PNG.

Use .BMP or .TIF or even .PSD.

Second, you can make your text frame size as big as your video frame size.

Then you could position your text to be where you want it within that large frame.

Have you tried that?

Jerry Jones
http://www.jonesgroup.net

nyee · Post by **nyee** » Mon Sep 18, 2006 6:24 am

I finally found a solution and figured out something weird in the process.

I tried using different image formats again, including PSD (which I didn't realize could be used). I lined up sequentially (in the overlay layer) the BMP version, then the PSD version, then the GIF version, etc. I made each image at 1152 x 864 and used the "fit to screen" option. I then rendered at 1152 x 864 using xVID compression. The only section where the text came out sharp was the GIF section. The PSD section, like the BMP and JPG sections, came out blurry. Weird huh?

The GIF version didn't work out for me at first when I tried this because the image I imported wasn't the *exact* resolution of the output, so it had trouble scaling it somewhere. So now, I'll make a large GIF with the trademark line at the top, use the "fit to screen" and then just move the large almost off the bottom edge till just that trademark line appears. And that does work.

So for me, to make sharp text from image:
- Use GIF format for image
- Use "fit to screen" option to resize
- Make sure the image is the *exact* resolution as the desired output resolution