Hey all. Wasn't sure if this was the correct place for this question, but I need help so it was the best I could figure - lol. Here is the thing. I need to crop some pictures sized to fit a tv screen. WHat aspect ratio (thats the right term, right?) is appropriate for this. Is there an preset for the crop tool, or can I set one? For some reason what was in my head was 720x480. But I am thinking that is not right for a regular tv screen. Anyone know about this?

THanks!

Dawn

EDIT_ Ok I am thinking maybe this should have been under the software heading? ANyways, So I look through the default sizes for a new document in photoshop, and a few of them are sized for tv screen, that is the NTSC sizes right? But how do I know which one to use? And what would be the best way to go about cropping the images to fit this? Also, when you open a new document there are guides included. WHat exactly are they for? I have noticed they are the same guides in my dvd program for making these slideshows, but I honestly have no idea what they are for. Which guides to I go by to be sure the whole image shows up? And is the regular tv size going to be the 720 x 540? It looks the most correct, but I am not sure. I have been having issues with the whole image not showing up on the tv, so I really want to get this down before I mess another one up - lol. Oh and for viewing on a tv, do I use 72 ppi? Thanks much!

Dawn

640x480 it's what a regular TV displays, always.

There are new High definition TV's that shows 800x600

And there are Wide Screen TV's that shows 853x480

But you need to ask about the software you are using to make your final VCD or DVD. Or are you displaying directly from the PC?

Because there is always some pixels lost in the TV, ALWAYS, and there are different on Sony than Panasonic.

What we do, or my family do, it's render a minute or 2 in the DVD and taste it in a small TV that they have next the computer, so if they need to adjust something, they doit before the final proyect it's done.

Thanks, studioj. I am using Nero. It came with the dvd burner - lol. I am not sure if I will be able to re-burn after the test, but I will look into it Do I use 72dpi for tv, or would it be better to go with 300 like print?

Dawn

Hi Dawn.

TV formats vary across the world
In the UK we use PAL format which is 720 x 576 pixels (4x3 Ratio)
However in the USA the NTSC format is 720 x 480 pixels
A widescreen image is 16.9 (UK)

All this gets very confusing. Some Programs that write DVD’s will try to make any pictures fit the format by enlarging or adding black lines to the pictures.
Also pictures can get stretched because ‘pixels’ on a TV are not square, again depending on the software used.

So it really depends on the software you use.
Make a test DVD of several pictures and make notes. See what happens.

Make a test picture of a circle and see if it’s still a circle on the TV

I tried loads of programs trying to get this sorted. I now use ULead PictureShow which I found the best IMHO.

72dpi is fine. There is no need for a higher resolution for a TV.

Hope this helps.

Ken.

My job is designing graphics for television. So far, everyone has been partially right.

The size that you need to make your images will vary slightly depending on how you're getting them on the TV.

If you're using something like a scan converter to translate the computer screen to video the size you want to make your images is 640x480 or 800x600 if it's supported by your scan converter.

The standard broadcast size of an NTSC image is 720x486. However, if you're editing in DV or making a DVD you want to make your images 720x480.

Another thing you might run into is the square vs. non-square pixel issue. Computer screens only display square pixels while television screens only display non-square pixels. Most of the time you won't notice the difference, but if you look carefully an image designed in square pixels will be a little stretched when viewed on a television screen.

This is because a square pixel image has a pixel aspect ratio of 1, while a non-square image has a pixel aspect of 0.9. This effect is most noticable on circlular areas in the image. The pixel aspect ratio is a separate issue from the image aspect ratio.

When working with images in Photoshop CS or CS2 that you know are destined for a television screen, create an image using the Pixel Aspect Ratio preset of "D1/DV NTSC (0.9)." If your image already exists as square pixels, create a new image using the D1/DV NTSC preset and drag the old image layer into the new image.

Your image might look stretched or squashed on the computer screen, but it will look just fine on TV. You can have Photoshop simulate what the image would look like on a TV screen by going to the VIEW menu and checking and the "Pixel Aspect Ratio Correction" option. This view is not perfect, though, and you should treat it as if you were viewing the image with a proof setup.

If you don't have CS or CS2 then design your images at 720x540 (you'll be using square pixels) and when your image is finished, save a copy of it, and resize it to 720x480 (or 720x486 if you're not using DV or going to DVD). This will squash your image, but the different aspect ratio of the pixels on a TV screen will make your image look correct again.

Hope this hasn't been too confusing. And we didn't even talk about color issues.

--Racc

Great explanation Racc Iria!

Hi Racc.

Thank You. That’s very interesting.
One point I was trying to make was that some software does the stretching for you and some doesn’t.

All the testing I did goes back three years so things may have changed, but at the time for some programs I had to do the stretching manually.

I was putting Slides and negs on DVD and many were taken in ‘portrait’ format. When these are viewed on a widescreen (16.9) TV then things can look very squashed.

Ken.

create an image using the Pixel Aspect Ratio preset of "D1/DV NTSC (0.9)." you wanna break that down, please? what is 'D1'? and is 'DV NTSC (0.9)' all one phrase or is it 'DV NTSC' and '(0.9)', where the '(0.9)' would be taken as a multiplication brackets?

craig

From Kraellin...

you wanna break that down, please? what is 'D1'? and is 'DV NTSC (0.9)' all one phrase or is it 'DV NTSC' and '(0.9)', where the '(0.9)' would be taken as a multiplication brackets?

Sure, Craig, no problem.

D1 is a high-end digital tape format used at large TV stations and post-production houses. It was created in the early stages of the digital revolution as a way to store video digitally onto magnetic tape (kind of like DAT tape but much larger). This was necessary as hard drive storage space was prohibitively expensive and networking/internet was in its infancy and out of the question as a transportation medium. As these technologies have matured and become cheaper and newer technologies delevop D1 is falling by the wayside. The image size of D1 is 720x486 (an image aspect ratio of 1.3333) and uses non-square pixels (a pixel aspect ratio of 0.9).

DV, of course, refers to the Digital Video format used by many non-linear video editing systems today and is also part of the DVD specification in the form of Mpeg-2. It's image size is 720x480 (an image aspect ratio of 1.35) and uses non-square pixels (0.9).

NTSC is the standard video format used primarily in the United States and Japan. It gets its name from the National Television Standards Committee that was formed to establish a video standard for use in the USA. Engineers in the TV industry swear that what the acronym NTSC really stands for is "Never The Same Color."

(0.9) refers to the pixel aspect ratio. By far, the two most commonly used in television/computers is square pixels (1) and non-square pixels (0.9).

Photoshop's Pixel Aspect Ratio presets (not to be confused with the image presets at the top of the dialog) found at the bottom of the new image dialog in CS and CS2 lists the preset you want to use exactly as "D1/DV NTSC (0.9)" without the quotes. It should be the second entry in the drop-down list after "Square." The reason is that it just so happens that the pixel aspect ratio for both the D1 and the DV formats is exactly the same... 0.9. And as far as the image size goes... if you're using D1 (which you probably never will) make your image size 720x486 and if you're using DV (which is probably most of the time) make your image size 720x480.

The DVD spec isn't as forgiving, but most video editing/display equipment will simply ignore the extra 6 lines of pixels if you accidentally make your image at 720x486. Which six lines that get ignored only becomes an issue when you have to deal with the field order of the video interlacing, but that's a whole other lengthy discussion.

Did that clarify things or muddy them up even more?

--Racc

Ok. So-- I am thinking I want to pup the top back on this can - lol. No, just kidding. I really appreciate all of the info! Something else that I have been thinking of - as far as DVDs are concerned - I don;t know about other dvd players, but I know ours has many different settings and it will change when you see on the screen depending which is selected. And that is just referring to what gets "chopped" - then you have the color and brightness that changes. I guess my point is that I don't think there is any way to determine what the customer is going to see when they get the dvd home. I guess it is the same as web and all that - never know what kind of crappy monitor Joe and Jane Customer may be viewing your images on. Ugh - this is complicated. lol

The reason I ask all of this is I have begun offering DVD photo slideshows. You know, just as I get all the color issues worked out with printing and web and sending for prints, now I get a whole new set of rules!

Nero is what I have been using for makign these slideshows. I like how easy it was to learn, but I wish there were more to it. Either way though, the software came with the dvd burner, and it has already paid for itself. Anyone else out there making these? I would love to know your experiance with them, and what program(s) you use, and why.

Dawn

thanks, racc :)

craig

In Nero, the DVD is burned according to the settings you've entered. You can tell it to burn either a 4:3 or 16:9 DVD. It will pad your image/video data as needed to maintain the original aspect ratio and fit the target aspect ratio (so portrait photos will have black bands on the sides.) 4:3 is better for photos unless you are doing a bunch of panoramas. Even if your customer has a 16:9 TV, a 4:3 DVD will show all 4:3 and 3:2 photo larger on screen.

The target DVD player will see which format DVD it is and distort things appropriately (my cheap DVD player knows to do this.) Assuming the user has already configured their DVD player to work with their particular TV, your DVD will look correct.

Bart

Craig, you're welcome.

From 1STLITE...
I guess my point is that I don't think there is any way to determine what the customer is going to see when they get the dvd home. I guess it is the same as web and all that - never know what kind of crappy monitor Joe and Jane Customer may be viewing your images on. Ugh - this is complicated. lol
It's true, this can be a very tricky process. But there are some tools that can help you out.

You can't know or control how the customer has the contrast, brightness, color, etc. set on their TV at home. All you can do is make sure that your images adhere as closely as possible to the NTSC standard.

The professionals use two pieces of equipment (among many others) to do this. They are a waveform and a vector scope. The waveform measures the brightness information in a video frame in IRE (a unit of measurement created by the Institute of Radio Engineers). Video black is supposed to be at 7.5 IRE and video white is supposed to be 100 IRE. Computer black (RGB 0,0,0) is 0 IRE. Computer white (RGB 255,255,255) blows past the end of the scale. This is why pure white or bright colors get clipped and darker colors look too dark on a TV screen.

A vector scope is used for color calibration of a video monitor. A standard color bars image is generated with known values and using the scope colors are represented as spikes eminating from the center of the scope (vectors), and the tips of these vectors are supposed to align with marks on the scope. If they don't, adjustments are made to the monitor or the video signal.

Since you probably don't have access to these expensive pieces of equipment, there are a few things you can do in Photoshop to help make sure you're generating color safe images for TV.

1. When making images for video or the web turn color management off. This will give you the closet appoximation of what others will see. This is a complex and controversial discussion that I won't go into here, but trust me on this, it will make life much easier in the long run. If you must use color managment, the next best thing to do would be to use the "NTSC (1953)" color profile. But, it's not perfect.

2. Make sure that any colors you use don't go below RGB 16,16,16 or higher than 235,235,235. For white, 235,235,235 produces 100 IRE. This still causes "blooming" and "bleeding" so white should optimally be closer to 80 IRE, which is somewhere around 217,217,217. These are not hard rules, just guidelines. You can cheat a little here.

3. Once your image is completed, you can also apply the "NTSC Colors" filter from the Fliters>Video section. This kind of does what you should do manually in number 2 above.

--Racc

Racc - Thank you SO much! I am off to test it out! TYTYTYTYTY!!!

Dawn

You're quite welcome! It's my pleasure.

--Racc

Hi Racc.

Thank you that’s really interesting.
I guess the IRE brightness will apply to PAL in the same way.

Do you have any info on PAL?
When I did some of these I stretched the pictures from 702 x 540 to 720 x 540
I can’t remember where I got the figure of 702. Do you know if that’s correct?

Ken.

Sorry, I just saw this part.

From 1STLITE...
Also, when you open a new document there are guides included. WHat exactly are they for? I have noticed they are the same guides in my dvd program for making these slideshows, but I honestly have no idea what they are for. Which guides to I go by to be sure the whole image shows up?

I have been having issues with the whole image not showing up on the tv, so I really want to get this down before I mess another one up - lol. Oh and for viewing on a tv, do I use 72 ppi?

On a TV screen you can't see all the way to the edges of the picture... the bezel of the TV case as well as differences in pictures tubes which change as they age, etc. prevent this. This is refered to as "Overscan." The edges of a TV frame are really quite messy, and can contain ugly little bits of information like timecode blips, or in the case of DV video thin black lines down the left and right sides. So you really don't want to see the edges.

So, if you don't know where the edges are how can you make sure that important stuff will be visible on the screen? That's what those guides are for. The outside guide represents the Action Safe area and the Indise guide represents the Title Safe area.

Inside the Title Safe area is where you want to put any text or important elements that you want to guarantee will be seen when viewed on a TV screen. If the size of your image represents 100% of the screen, the Title Safe area is at 80% of your image size.

The Action Safe area represents 90% of the image size and is generally used to make sure action or movement in the image is visible, but it isn't as critical as text for example so it's get's a little more latitude.

Think of it this way... Anything inside of the Title Safe area is absolutely guaranteed to be seen on the TV screen. The closer you put stuff to the Action Safe line or the edge of the image the greater the chance it could be cropped off by the TV. Some really bad or old televisions sometimes overscan past the action safe line. So, basically the closer to the edge you put something the more you're gambling that it won't be seen.

Regular 4:3 standard definition television resolution is about 72ppi. If you have a 300 ppi image that you want to display on TV you don't have to change the resolution it'll show up fine, just like on a computer monitor. However, if you're creating images from scratch, making them at anything greater than 72 ppi is just wasting file size.

I apologize for missing those questions before.

--Racc

From Cameraken...

I guess the IRE brightness will apply to PAL in the same way.

Do you have any info on PAL?
When I did some of these I stretched the pictures from 702 x 540 to 720 x 540
I can’t remember where I got the figure of 702. Do you know if that’s correct?
Hi, Ken.

Yes, most of what I talked about will work for PAL, too. The only differences are the image size and that PAL is much more forgiving with the colors. By following the guildlines for NTSC, but using the PAL image sizes you should be okay.

The PAL DV (non-square pixel) image size is 720x576. To make images for PAL using square pixels, make the image 768x576 and when you're finished, resize it to 720x576. PAL requires horizontal compensation, while NTSC requires vertical compensation.

I'm in the U.S. and so only work in NTSC, so I'm afraid I don't know as much about PAL, especially concerning the differences in color. I do know that one of the reasons PAL was created was to correct the color issues of NTSC. I'd have to do some research to learn more about it begfore recommending PAL safe colors. NTSC colors, though, are PAL safe since NTSC has a smaller color space.

Hope that helps.

--Scott

Yes. It does help.
Thanks Racc.

Ken.

racc,

interesting.

ok, here's a newbie question for you. what about the newer higher resolution tvs? are these using the same aspect ratios in general or something else? and are they using the same raster techniques as before only with more rasters or are they doing something else?

also, on a related question, what does s-video do, like if you hook up your computer to your tv via s-video? if the aspect ratios are different and the pixels are different, then how is s-video handling all this?

craig

You're welcome, Ken.

From Kraellin...
what about the newer higher resolution tvs? are these using the same aspect ratios in general or something else? and are they using the same raster techniques as before only with more rasters or are they doing something else?

also, on a related question, what does s-video do, like if you hook up your computer to your tv via s-video? if the aspect ratios are different and the pixels are different, then how is s-video handling all this?

The newer TV's are a whole different ballgame. At this point, the color issues are still pretty much the same, but everything else is tossed on its head. The easiest way to tackle it is to decide beforehand what you need to design to. Standard Def, Letterbox, HD, 4:3, 16:9, etc. The rule of thumb is to design to the largest size/resolution you need and scale down from there.

The image dimensions and aspect ratio are largely determined by the television standard you choose. If you look in the stores, many of the HD 16:9 TVs have different maximum image sizes. Some are 720 vertical, some are 1080, and others have odd sizes. Most of this confusion is due to the HD standards war. Standard def is 480i. But now comes along 480p, 720p, 1080i, and now 1080p.

But, generally, for 16:9 image aspect ratios the width in Pixels for HD should be 1280 and Standard Definition 940 pixels. The vertical is determined by the standard. If using 480p or 480i, then the vertcal size should be 480 pixels. If you're using 720p, then 720 pixels, and 1080i or 1080p should be 1080 pixels. The most common flavor of HD in the U.S. at the moment is 1080i.

It can all be quite confusing. And this is also why DVDs are generally made two ways... 16:9 or letterboxed and 4:3 full screen.


Now, on to your easier question about S-video.

The NTSC standards are the same regardless of which type of connector you use. The difference between Composite, S-Video, and Component are basically a function of bandwidth.

A Composite signal (like that found using a single RCA connector) crams all the color information (chrominance), the brightness information (luminance), and timing information onto a single pair of copper wires (a positive and a ground). With all this info crammed in there, there isn't a lot of room, so the colors bleed and the signals interfere with each other making for a pretty noisy picture.

S-Video (also sometimes called S-VHS as it evolved from those days on VCRs) improves the picture by separating the chominance and luminance and sending each on a different set of wires, thus allowing more bandwidth for each resulting in a cleaner, better picture.

Component goes a step further and sends the signals for Red, Green, and Blue on separate wires vastly improving the bandwidth for each color resulting in a much clearer picture.

When you hook up your computer to your TV using an S-Video connector (or any other analog connector for that matter) the graphics card is basically performing the function of a scan converter, though usually not as well as a dedicated device.

If you're using a graphics card designed for video editing that has S-Video out to a monitor, it's probably not scan-converting your computer screen, but sending out the video signal from the editing application in which everything we've covered in this thread will have been taken into account, either by you or by the editing application.

If you're sending out the signal over a firewire card, then you're using DV and need a device like a DV tape deck, or a DV camera to convert the fireware DV signal back into video.

--Racc

racc,

thank you! a most excellent explanation of a somewhat confusing subject.

craig

THank you so much Racc. This is exactly what I needed to know. Thank you for being so very helpful!

Dawn

It was my pleasure, guys.

In this world of converging technologies this kind of information is going to become more and more important not to mention confusing. Perhaps it will get better once standards have been decided.

Probably not, though.

Like I said at the beginning of this thread, I do this for a living. Making graphics for television/video, that is. So don't hesitate to ask if you have any video or TV related questions.

If I don't know the answer, I know a bunch of other people who probably do.


--Racc