Image file format recommendations

[ted_ludford]

I plan to create over 300 image files from my photographic file storage in currently in .TIF file format. The files are at a physical size of 6H X 4W inches and at a resolution of 2400pixels per inch. The files are 130M to 395M in size depending on they are black and white or color. I plan to convert the .TIF files for use in making four different BluRay movies using VideoStudio X2. My planned files would be as follows:
Physical size of photographs: 6 H x 4 W in inches
Pixel size of photographs: 1080 W X 720 H in pixels
Photograph resolution: 180 pixels per inch
Image file type: .TIF

I understand there will be a large area of "space" on the screen. Most of that "space" will be consummed with typed narrative about the picture subject. I will add sufficent "canvas" to make the image a total of 1920 pixels wide.

Do you have any advice about my proposed image file type or other image file chacateristics?

[Clevo]

If you have a decent photo editor this is what I have done with success but with images you will find there will be several suggestions.

Assuming you have calibrated the monitor, OS and

I have set my photo editor so that the crop tool maintains a 16:9 ratio. In some cases even to portraits I will crop. In some cases I will leave as is, as I will play with it later in the VS pan & Zoom feature

I make adjustments in the levels and saturation if needed. I know 300 is a lot of work but I find I get much better results.

I always save as BMP's even if the originals are jpegs. Safest choice I think.

I tend to go for 300 dpi minimum

[Trevor Andrew]

Hi Ted

Are you sure your images are a Resolution of 2400pixels per inch¡¦. Cos wow that¡¦s big. That would allow a 4ft print at full quality.
OK
Still images normally use square pixels, whereas video doesn¡¦t, it uses ¡¥non square pixel rendering¡¦ to stretch the pixels to a 16:9 shape.
With the exception of hi quality HD which uses 1920 X 1080 frame size.
If you want to fill the video frame then you have to use an image at 16:9 ratio, --- 1920 x 1080 would do just that.
A little large but would allow for ¡¥Pan & Zoom¡¦ editing.
Otherwise use 1280 x 720 again a frame size using 16:9 aspect ratio.

If you use 1080 x 720 (this is not 16:9 but 3 x 2) you will get a border to the right and left of the image, the vertical will fit to screen.
Try a few image options to see how they look in the preview screen.

[Ilene]

I'm working w a similar issue.... I think. So I thought I'd add to this thread before starting a new one.

I general create photo montages in 4:3 ( NTSC) .. and scan all my still images as jpgs, 720 x 480. I need to make a project using wide screen format....(NTSC) 16:9..... Should I still scan may images at 720 x 480 or do I need to scan them at a different resolution. ( 1024 x 576, 1280 x720 ( but I think this is larger than an original resolution scan of a 4x6 picture and that is what I commonly deal with)

Should I expect to see the images in letterbox after it is rendered or is there a way to fill the screen?

Also - do I keep setting non-square pixel rendering checked or unchecked. I always keep it checked when I'm in 4:3.

lastlly when I render the file....to mpg. do I still choose 720 x 480 ( I assume yes if I scan 720 x 480) or do I choose another setting if I scan images at a different resolution?

thanks.



Ilene

[Trevor Andrew]

Hi Ilene

Leave Non square pixel checked , as default.

Scan your images to 16:9 ratio.

Use a calculator you will find that 1024 x 576 is 16:9 ratio.
This will fill the video frame.

If you use smaller 4:3 ratio images they will sit in the middle of the widescreen frame, producing a black border left and right.

Use the overlay tracks to place more than one image to the frame.

Render your project to your normal template NTSC-DVD 16:9
Assuming you are trying to make a dvd.

[sjj1805]

......
I always save as BMP's even if the originals are jpegs. Safest choice I think.
....

BMP is a "Flat Format" - whilst it is Non Lossy it has the disadvantage of merging any layers. If at a later date you decide to re-edit a previously edited image then it is better to have kept a copy with all of the layers intact.

Here is what I do to EDITED Images.
Pictures from my digital cameras are recorded in JPG format (Yes I know I can use RAW and do on the odd occasion)
Scanned Images.

1. Save the image in the editors native format.
For PhotoImpact this is UFO for Photoshop this is PSD.
PhotoImpact will open and save files in the Photoshop PSD format.
I tend to do most of my work in PhotoImpact.

2. Edit the images as required and save in the same UFO or PSD format.

3. Preserve those EDITED images somewhere such as a DVD or External Hard drive.

4. Make a copy in JPG format at highest quality and use that much smaller file for storage in my filing system

NON EDITED photographs are simply kept in the original JPG format - nothing to be gained by saving it as something else - the save as something else step is only necessary if you intend to edit the picture.

Take this Image as an example:



Each view of the house is a cut out from its original separate full sized image 3008 x 2000 pixels - using the photoimpact "Object Extract" tool. The master copy of the image is also 3008 x 2000 pixels and each view has been resized using the Photoimpact Transform Tool. This enables me to go back into the Master copy and move the houses around and resize them. Something that you would not be able to do if I had saved it as a BMP.

The Master copy will be quite huge - it is a collection of full sized images. Not much good it you intend to share the images with someone - especially on the Internet. So you make a "Distribution copy" in JPG format.

If I was intending to place these images into a Video File then I would create a Canvas of the correct proportions (PAL/NTSC - 16.9/4.3) and then place the images within those boundaries.

Please view:
What dimensions etc should I use for still images in Videos?

[Ilene]

hi...
thanks... all good information!

I actually had done a test on my own. I scanned pictures at 720 x 480, full resolution ( ie whatever the scanner took the image to be usually 1176 x 772 standard 4 x 6 images....) and I also scanned at 1024 x 576 ( which inicidentally is the same as 864 x 486 which I've seen mentioned ( by Steve and others?))

I put all the images in the time line and put a title on them as to what they were.... I also put each image twice... once left full, and a second time w a pan/zoom on it.

I burned a dvd in 16 x 9.... when I looked at the images on the TV they all looked well... however, as expected the 720 x 480 ( non zoomed) had a right/left black border as did the original or full resolution images. The 1024 x 576 images had a top/bottom letter box... but all the images looked well and "non-stretched"

Then a colleague of mine gave me an important tip.... he said not only did my TV need to be set to 16x9 but the DVD player also needed to be set to 16 x9 to avoid the letter boxing....

Lo and behold, he was absolutely correct. with both the DVD player and TV set to 16:9 mode the images that I had scanned at 1024 x 576 filled the entire screen and looked great. 720 x 480 still had right/left border issues but the borders were narrower.

Conclusion... 1024 x 576 or 864 x 486 which are both 16:9 ratio is the best method for 16:9 projects.

Steve... I agree with you to keep the files sizes smaller... the rendering time is quicker and the project looks just as well.. clearly you don't need high resolution for video as you do with prints.

thanks...

Ilene

[Trevor Andrew]

Hi

I think you summed that up.

If you are using Pan and Zoom then some reduction in quality may be seen........
If you use images larger than the 16:9 frame size (say double) 1024 x 576 to 2048 x 1152 should allow for zooming in, the quality should be better.