This isn’t going to solve any problem, but perhaps it’ll help others understand and if they're having a problem, hopefully help lead to a cure. Starting with the basics…
There are different color spaces between PCs & std. TVs – more important: std. SD TVs work very differently from PCs , cannot display every color available for PCs, frequently have problems with too much color saturation, & vary a bit in perceived resolution. This effects a lot of things related to mpg2 & DVDs, mainly because both were originally designed & intended for std. SD TV.
Please remember that a std. SD TV is showing you a picture that’s 640 x 480 (if NTSC), & that if you show that same picture full screen on your PC’s monitor, it may look very disappointing – it will often be like comparing 480i (SD) to 720p (HD). Also bear in mind that interlace looks bad, perhaps even nasty on a PC monitor when/if you look at a single frame. Player software & often even your graphics card will turn that same interlaced frame into a progressive one when the video’s playing.
A std. SD TV has a picture tube with an interlaced display that shows alternate line numbers ~60 or 50 times a second. It works fairly well because motion isn’t confined to ~30 or 25 fps, but takes advantage of the higher field rate – every field can show motion.
Still images pose a challenge for std SD TVs, because not only are the possible colors a bit different, but the entire range of color is different. Additionally you have different aspect ratios, and especially with graphics, you can have thin horizontal lines &/or edges that do no span more than one TV line – every other field a line with a sharp edge drawn on it is not going to be refreshed, so you get a bit of flicker as the edge disappears & reappears.
Different codecs store different amounts of data, & store color info in different ways – most behave like jpeg in that they throw out some data to make the file size smaller. Many codecs also have their own shortcomings – mpg2 for example relies on analyzing motion. For info on DV I recommend Adam Wilt’s site: http://www.adamwilt.com/ . Some codecs will also convert better than others into certain other codecs because of the color data conversion that often takes place.
-----------------
Practical aspects:
If you’re using stills please read up on aspect ratios and image resolutions as needed. Scale your image as necessary in either your image editing software or NLE, but avoid extra conversions that lose quality; this includes using images scaled too big that are guaranteed to need further manipulation. Adobe has a number of articles if using one of their products, like this one: http://www.adobe.com/designcenter/premiere/articles/pre7ppimport.html
Many Image Editing applications along with Vegas have color filters that will change or alert you to *out of gamut* colors – colors that a std. SD TV is likely to be incapable of displaying. Use them & for extra safety & no surprises, view your work on a TV before & after rendering &/or encoding. Too strong of a color signal can distort or even cause audio problems. If your project targets only PCs, you might relax this a bit, but remember that software players like Power DVD are expecting TV colors and by default alter the image viewers see so that it looks like your TV – in fact you might look at another [much more PC friendly] delivery format if you only target PCs.
Always avoid as possible lossy intermediate formats, ideally encoding your menus directly to mpg2. If a format is known to cause certain artifacts or edge problems while discarding image data, why would you want to use it when you don’t have to?
Partly because of interlace, partly because of std. SD TV display limitations… As possible use [at least slightly] subdued colors, and for graphics, slightly feather the edges into a more feathered, often semi-transparent shadow. For photos often you’ll need to add a small bit of blur – horizontal edges should ideally span more than one scan line. You might benefit from experimentation with additional filters or effects that will cause edges to blur very slightly. As possible avoid or at least apply effects to *slightly* subdue any busy patterns, whether on a person’s suit, a tree’s leaves, or a field of flowers & grass.
-----------------
More…
Working backwards, a stand-a-lone DVD player will output a spec (or close to spec) signal that a std. SD TV expects and understands. In terms of the signal itself, it will be interlaced, regardless the video source. For DVD menus and stills (as in a picture comp.), the player can read true interlaced video off the DVD, or it can divide a progressive frame into even and odd fields, or [while unusual for NTSC menus] it can follow pulldown directions embedded in a 24p video – these directions make the decision on how fields are created from full frames – the Vegas docs provide a good, illustrated reference. RE: motion menus, at this point [at the DVD player] interlaced vs. progressive issues are more or less confined to smoothness of motion, though some say compatibility of interlaced might be an issue – I have not been able to confirm the compatibility part, & in fact suggest that retail DVDs commonly use interlaced backgrounds. I personally cannot see why decoder electronics would have problems with video in the menu domain that they don’t have with titles.
Moving back a step into DVDA, DVDA & many other authoring programs are able to composite text &/or graphics onto the background still or video. Many like DVDA will encode menu backgrounds to mpg2. If you don’t add or have anything for DVDA to do to a DVD spec mpg2 menu background, it will pass it through untouched to the rendered DVD layout on hdd or DVD blank. I want to stress that point, because if you have a background mpg2 that looks good & DVDA doesn’t touch it, you’ll have the exact same mpg2 included in the finished DVD. DVDA only comes with Vegas, which has many tools for motion menus. Personally I think of their inclusion in DVDA as a courtesy in case I don’t want to bother.
DVDA allows you to integrate into & exit video with menus. If you use these, progressive mpg2 allows better frame matching – your intro will appear to be pretty seamless with the menu. DVDA also allows loop-points, where the intro is part of the menu video – at the end of play the menu video goes directly to the loop-point, skipping the intro. This allows more seamless audio from intro to menu, but the end of menu pause is in my experience more noticeable [not a fault of DVDA but the way it works].
Moving back a step into Vegas or whatever editor you choose to work in, you can create a menu background using things like text overlay tools,&/or import graphics for overlay from your choice of image editing software – these are composited or combined with the still or video background, which can be rendered to DVD spec mpg2.
Using video for menus can be less trouble than stills, as chances are the video is already TV friendly; graphics, PowerPoint slides, &/or photos often aren’t. Depending on the quality level you’re after, before use in SD video they’ll probably benefit from being processed, maybe run through something like Photoshop in batches. With experience you’ll be able to spot many problems, but the surest way is to display images and graphics on a SD monitor. If you’re working with SD video, having one hooked up to your PC has always been strongly & universally recommended.
Vegas and some other NLEs can process your stills – reducing interlace flicker is a method of blurring slightly in Vegas – but if you’ve got a lot of them, running a batch process on a folder(s) using Photoshop can be more efficient. And personally I believe that Vegas does best with Video, Photoshop with images, Word with text, Nero with burning, and so on…
-------------
Example…
One workflow for pro-level, commercial or retail type results is to grab a still from the title video, cropping, resizing, & altering it artistically as desired in Photoshop or similar. Then design any text, if appropriate using a vector art program like CorelDraw! or Illustrator, putting the results on a new layer with transparent background in Photoshop. On a 3rd layer (also with transparent background) either copy your text and make everything black, or otherwise design highlight graphics in black. Export all 3 layers separately as png files.
In a new Vegas project add a generated white clip & then a black one -> drag the black partially over the white, and then add Vegas' built-in Iris transition -> adjusting lengths and settings, render a clip of the transition up to a point before the transition is complete, where the black oval takes up much or most of the screen; you now have the basic mask for a menu plus an intro. In a new project, or a new version of your title project, make a 30+ second trailer out of clips from your title, perhaps including some outtakes. Open another new Vegas project.
On one video track lay down your menu still, having it loop for ~3 5 – 45 seconds. Add a fade from black at the start, then as you like set a marker for the menu to start [5 – 10 second?]. Import your mask video to a higher track at the marker, and with the mask track solo, take a snapshot of the last frame, then loop this new still to the end of your menu – add an inverted mask generator to the track. Add a bottom track and import your trailer clip at the marker, applying any FX you like (maybe film etc?) to the still & trailer tracks. Import your text layer from Photoshop into a new top track at the marker, for an overlay that loops to the end of project.
While you can change it as you like, the idea is black screen fading to your background still, then menu text fading in, & at the same time your trailer clip starts growing through the center. Using your own artistic preferences, set the text overlay track fade in and other timing…
Very common practice is to have 30 seconds of menu video that repeats for a total of 1 minute worth of menu, so you want to isolate 30 seconds worth of your stabilized menu (text faded in, center window as big as it’s going to get etc), & it’s probably easiest rendering it to uncompressed or Sony YUV or lossless avi . Also render a separate clip from the beginning up to the start of the 30 second clip you just rendered. If you want a separate intro, optionally add audio & render this clip to DVDA mpg2 & [if including audio] ac3. If using loop-point, import this clip onto a track in a new project, then the loop clip twice, laying down an audio track for the complete duration – if separate intro, just use the loop clip twice. Render your mpg2 & ac3 using DVDA templates.
Import this clip or clips and audio into DVDA for your menu & optionally intro, set the loop point as needed, & trim the end of your clips by setting DVDA’s video endpoints. Edit the name of your menu page, and then delete the text object. Set the button areas by using image buttons with no image, set the button highlight to Image Mask Overlay, under media set the Highlight mask to the 3rd Photoshop layer png, check your button navigation if more than one button is used. Check Optimization under DVDA’s File Menu, and you should see all green checks for your menu & if used, your intro clip. With your cursor at the beginning of the separate or integrated intro, try the preview, making sure to let it go past the end to check the loop behavior.
There are some shortcuts I omitted in trying to keep this as easy, short, and generic as possible, and there are lots and lots of variations that can be used with the same basic process. It’s not terribly creative – text fades in instead of morphing or swooping – but I tried to make it similar to what I think most people have seen fairly often. Hopefully this will give some folks a start or at least an idea or 3.
There are different color spaces between PCs & std. TVs – more important: std. SD TVs work very differently from PCs , cannot display every color available for PCs, frequently have problems with too much color saturation, & vary a bit in perceived resolution. This effects a lot of things related to mpg2 & DVDs, mainly because both were originally designed & intended for std. SD TV.
Please remember that a std. SD TV is showing you a picture that’s 640 x 480 (if NTSC), & that if you show that same picture full screen on your PC’s monitor, it may look very disappointing – it will often be like comparing 480i (SD) to 720p (HD). Also bear in mind that interlace looks bad, perhaps even nasty on a PC monitor when/if you look at a single frame. Player software & often even your graphics card will turn that same interlaced frame into a progressive one when the video’s playing.
A std. SD TV has a picture tube with an interlaced display that shows alternate line numbers ~60 or 50 times a second. It works fairly well because motion isn’t confined to ~30 or 25 fps, but takes advantage of the higher field rate – every field can show motion.
Still images pose a challenge for std SD TVs, because not only are the possible colors a bit different, but the entire range of color is different. Additionally you have different aspect ratios, and especially with graphics, you can have thin horizontal lines &/or edges that do no span more than one TV line – every other field a line with a sharp edge drawn on it is not going to be refreshed, so you get a bit of flicker as the edge disappears & reappears.
Different codecs store different amounts of data, & store color info in different ways – most behave like jpeg in that they throw out some data to make the file size smaller. Many codecs also have their own shortcomings – mpg2 for example relies on analyzing motion. For info on DV I recommend Adam Wilt’s site: http://www.adamwilt.com/ . Some codecs will also convert better than others into certain other codecs because of the color data conversion that often takes place.
-----------------
Practical aspects:
If you’re using stills please read up on aspect ratios and image resolutions as needed. Scale your image as necessary in either your image editing software or NLE, but avoid extra conversions that lose quality; this includes using images scaled too big that are guaranteed to need further manipulation. Adobe has a number of articles if using one of their products, like this one: http://www.adobe.com/designcenter/premiere/articles/pre7ppimport.html
Many Image Editing applications along with Vegas have color filters that will change or alert you to *out of gamut* colors – colors that a std. SD TV is likely to be incapable of displaying. Use them & for extra safety & no surprises, view your work on a TV before & after rendering &/or encoding. Too strong of a color signal can distort or even cause audio problems. If your project targets only PCs, you might relax this a bit, but remember that software players like Power DVD are expecting TV colors and by default alter the image viewers see so that it looks like your TV – in fact you might look at another [much more PC friendly] delivery format if you only target PCs.
Always avoid as possible lossy intermediate formats, ideally encoding your menus directly to mpg2. If a format is known to cause certain artifacts or edge problems while discarding image data, why would you want to use it when you don’t have to?
Partly because of interlace, partly because of std. SD TV display limitations… As possible use [at least slightly] subdued colors, and for graphics, slightly feather the edges into a more feathered, often semi-transparent shadow. For photos often you’ll need to add a small bit of blur – horizontal edges should ideally span more than one scan line. You might benefit from experimentation with additional filters or effects that will cause edges to blur very slightly. As possible avoid or at least apply effects to *slightly* subdue any busy patterns, whether on a person’s suit, a tree’s leaves, or a field of flowers & grass.
-----------------
More…
Working backwards, a stand-a-lone DVD player will output a spec (or close to spec) signal that a std. SD TV expects and understands. In terms of the signal itself, it will be interlaced, regardless the video source. For DVD menus and stills (as in a picture comp.), the player can read true interlaced video off the DVD, or it can divide a progressive frame into even and odd fields, or [while unusual for NTSC menus] it can follow pulldown directions embedded in a 24p video – these directions make the decision on how fields are created from full frames – the Vegas docs provide a good, illustrated reference. RE: motion menus, at this point [at the DVD player] interlaced vs. progressive issues are more or less confined to smoothness of motion, though some say compatibility of interlaced might be an issue – I have not been able to confirm the compatibility part, & in fact suggest that retail DVDs commonly use interlaced backgrounds. I personally cannot see why decoder electronics would have problems with video in the menu domain that they don’t have with titles.
Moving back a step into DVDA, DVDA & many other authoring programs are able to composite text &/or graphics onto the background still or video. Many like DVDA will encode menu backgrounds to mpg2. If you don’t add or have anything for DVDA to do to a DVD spec mpg2 menu background, it will pass it through untouched to the rendered DVD layout on hdd or DVD blank. I want to stress that point, because if you have a background mpg2 that looks good & DVDA doesn’t touch it, you’ll have the exact same mpg2 included in the finished DVD. DVDA only comes with Vegas, which has many tools for motion menus. Personally I think of their inclusion in DVDA as a courtesy in case I don’t want to bother.
DVDA allows you to integrate into & exit video with menus. If you use these, progressive mpg2 allows better frame matching – your intro will appear to be pretty seamless with the menu. DVDA also allows loop-points, where the intro is part of the menu video – at the end of play the menu video goes directly to the loop-point, skipping the intro. This allows more seamless audio from intro to menu, but the end of menu pause is in my experience more noticeable [not a fault of DVDA but the way it works].
Moving back a step into Vegas or whatever editor you choose to work in, you can create a menu background using things like text overlay tools,&/or import graphics for overlay from your choice of image editing software – these are composited or combined with the still or video background, which can be rendered to DVD spec mpg2.
Using video for menus can be less trouble than stills, as chances are the video is already TV friendly; graphics, PowerPoint slides, &/or photos often aren’t. Depending on the quality level you’re after, before use in SD video they’ll probably benefit from being processed, maybe run through something like Photoshop in batches. With experience you’ll be able to spot many problems, but the surest way is to display images and graphics on a SD monitor. If you’re working with SD video, having one hooked up to your PC has always been strongly & universally recommended.
Vegas and some other NLEs can process your stills – reducing interlace flicker is a method of blurring slightly in Vegas – but if you’ve got a lot of them, running a batch process on a folder(s) using Photoshop can be more efficient. And personally I believe that Vegas does best with Video, Photoshop with images, Word with text, Nero with burning, and so on…
-------------
Example…
One workflow for pro-level, commercial or retail type results is to grab a still from the title video, cropping, resizing, & altering it artistically as desired in Photoshop or similar. Then design any text, if appropriate using a vector art program like CorelDraw! or Illustrator, putting the results on a new layer with transparent background in Photoshop. On a 3rd layer (also with transparent background) either copy your text and make everything black, or otherwise design highlight graphics in black. Export all 3 layers separately as png files.
In a new Vegas project add a generated white clip & then a black one -> drag the black partially over the white, and then add Vegas' built-in Iris transition -> adjusting lengths and settings, render a clip of the transition up to a point before the transition is complete, where the black oval takes up much or most of the screen; you now have the basic mask for a menu plus an intro. In a new project, or a new version of your title project, make a 30+ second trailer out of clips from your title, perhaps including some outtakes. Open another new Vegas project.
On one video track lay down your menu still, having it loop for ~3 5 – 45 seconds. Add a fade from black at the start, then as you like set a marker for the menu to start [5 – 10 second?]. Import your mask video to a higher track at the marker, and with the mask track solo, take a snapshot of the last frame, then loop this new still to the end of your menu – add an inverted mask generator to the track. Add a bottom track and import your trailer clip at the marker, applying any FX you like (maybe film etc?) to the still & trailer tracks. Import your text layer from Photoshop into a new top track at the marker, for an overlay that loops to the end of project.
While you can change it as you like, the idea is black screen fading to your background still, then menu text fading in, & at the same time your trailer clip starts growing through the center. Using your own artistic preferences, set the text overlay track fade in and other timing…
Very common practice is to have 30 seconds of menu video that repeats for a total of 1 minute worth of menu, so you want to isolate 30 seconds worth of your stabilized menu (text faded in, center window as big as it’s going to get etc), & it’s probably easiest rendering it to uncompressed or Sony YUV or lossless avi . Also render a separate clip from the beginning up to the start of the 30 second clip you just rendered. If you want a separate intro, optionally add audio & render this clip to DVDA mpg2 & [if including audio] ac3. If using loop-point, import this clip onto a track in a new project, then the loop clip twice, laying down an audio track for the complete duration – if separate intro, just use the loop clip twice. Render your mpg2 & ac3 using DVDA templates.
Import this clip or clips and audio into DVDA for your menu & optionally intro, set the loop point as needed, & trim the end of your clips by setting DVDA’s video endpoints. Edit the name of your menu page, and then delete the text object. Set the button areas by using image buttons with no image, set the button highlight to Image Mask Overlay, under media set the Highlight mask to the 3rd Photoshop layer png, check your button navigation if more than one button is used. Check Optimization under DVDA’s File Menu, and you should see all green checks for your menu & if used, your intro clip. With your cursor at the beginning of the separate or integrated intro, try the preview, making sure to let it go past the end to check the loop behavior.
There are some shortcuts I omitted in trying to keep this as easy, short, and generic as possible, and there are lots and lots of variations that can be used with the same basic process. It’s not terribly creative – text fades in instead of morphing or swooping – but I tried to make it similar to what I think most people have seen fairly often. Hopefully this will give some folks a start or at least an idea or 3.