If video DVDs are what you're after....
For upsizing you can use whatever [usually ffmpeg-based] video convertor you prefer to go to DVD-spec mpg2, or you can use Virtual Dub &/or AviSynth to resize to a lossless intermediary, which may be higher quality but a bit more learning/work, then go to DVD-spec mpg2. For downsizing can go the same route, but you might rather encode to mpg2 using the higher rez source directly -- in that case you're supplying the encoder with more data so it can make better decisions on what to include/exclude. Be careful though as HD video tends to be in formats without a lot of timing info, so a LOT of convertors &/or encoders &/or video editing apps will drop frames, which is often noticeable & shortens the total length, often also throwing audio/video out of sync. The biggest problem is often getting whatever mpg2 encoding app to import the format you've got.
For encoding to NTSC DVD-spec mpg2, *if* your source is 24 fps, look for an encoder that can use/add pulldown so you're still only encoding 24 fps, which keeps the file size smaller for any given bit rate... Retail NTSC video DVDs most all use 24 fps mpg2 with pulldown. I've read that most PAL DVD players handle NTSC video DVDs, I've read of just changing the fps in the video file from 24 to 25, & I've read of adding a special type of pulldown to artificially bump 24 fps to PAL's 25, but I'm afraid that's something folks in PAL-land will have to check out.
Basic mpg2 encoding needs to be variable bit rate [VBR] -- quiet scenes need & use less than action scenes, pans etc. Mpg2 encoders that can do VBR are harder to write/develop, so low end encoders only use constant bit rate [CBR] -- if someone tells you CBR is all right on a video DVD it is if you're only sticking 1/2 hour or 45 minutes of video on that DVD, or doing a slide show without any picture to picture transitions... otherwise they're either trying to get you to drink the Kool-Aid or have already partaken [IOW it's Bunk]. There are all sorts of calculators but for single sided DVD blanks, for 1.5 hours of video figure on about 6 ave. & 9 peak bit rates, for the 2 hour practical max, 4 & 9. More than 2hours -- go dual layer. You can do a quick check during encoding by watching the new file in Windows' Explorer, refreshing *just* as the progress bar hits whatever %, then using Windows' calculator on the file's size. If you want to really max out that DVD blank, shoot for slightly more than will fit, then use DVD Shrink/Nero Recode to drop the total file sizes down to just what will exactly fit with little or no wasted space.
2 things that are confusing to some folks are Interlaced video & Anamorphic video... Anamorphic video means wide frame [16:9 - or wider] video is stuck in a standard [4:3] frame, & expanded as necessary by the player -- all std. DVD video is 4:3, expanded to 16:9 by the player if wide screen. Adding to the confusion, standard DVD frame sizes will not display the same aspect on a PC as on a TV. It might takes pages to fully explain the how/when/why plus what to do about it, & frankly a Lot of people don't mind if the video on a DVD is slightly wrong so I'll just say this: if you're confused & the info you find with Google/Bing doesn't help enough, take screen shots of the original video, take screen shots of a DVD playing on your PC in Power DVD or similar, open those screen shots in an image editor, & using the re-size tools, basically figure out what resizing you need, if you need to add leterboxing etc. Lots of video tools handle the aspect/resizing for you, & lots of video tools get it wrong at default settings. When I say some apps get it wrong I mean Very & obviously wrong -- there's actually no single wide screen frame size spec for std. DVD-sized video, e.g. Adobe's is different than Sony's. I've scanned film from a trailer shown in theaters, comparing that to official on-line trailers & that film's DVD, & none of the aspect ratios matched -- it's really about what looks good to *you*. As far as interlacing goes, try & ignore it as possible -- it's one of those things that used to be Much more an issue than it is today.
Interlaced video was developed for TVs a bit over 1/2 a century ago. To reduce broadcast bandwidth requirements a std. TV screen refreshes all the odd numbered lines, then all the even numbered ones, 60 times a second NTSC, 50 cycles for PAL. One frame = 1 cycle, so ~30 fps = 60 cycles [it's actually 29.976 fps], while PAL's 25 fps = 50 cycles. Problem -- when you're working with digital video, the picture's changed between 1 field & the next, between the time all the odd lines were captured by the camera, & when the even lines were recorded. Since one frame of interlaced video includes 2 fields, the odd & even numbered lines of pixels don't/won't line up, which usually shows up as ghosting. The graphics hardware on most PCs/laptops along with most video software takes care of this, blending the two fields as needed. When you send non-interlaced [Progressive] video to a std. TV, whatever hardware's doing the sending shows the same progressive frame twice, so no worries. Where you have to double check, be a bit more careful is there are 2 possible frame orders, odd/even or even/odd, & when/if software gets it wrong [or you set it wrong in some apps], high motion scenes can stutter -- nowadays this is a bit rare, but happens. Another thing to watch for is resizing since, truth be told, an interlaced frame cannot be re-sized -- you're increasing/decreasing the number of individual odd/even lines. Usually the software doing the re-sizing handles it OK, de-interlacing [blending] the frame 1st, But, some software tools leave this up to you, & those tools are sometimes incorporated in video apps, so if you're resizing with an unfamiliar app &/or resizing an unfamiliar type of video, run a test & check rather than waiting until everything's rendered. Adding Telecine is somewhat like Interlacing -- basically film's 24 fps has sort of *ghost* frames added to bring it up to NTSC 29.976 -- you can either leave it alone, or try IVT [Inverse Telecine] to [hopefully] dump those extra frames. Caution -- video can be assembled from various sources using different telecine processes, in the same program or film. Video can also take a rather long route to get to your home where you record it -- I've seen samples of video that had been telecined, run through IVT, deinterlaced, had telecine added, IOW all sorts of modifications of the original as it passed from distributors in one country to another.
You generally want your audio in the std. AC3 -- anything else is too big [.wav takes up too much space], or too low quality, or for practical purposes impossible to work with [DTS]. You can fake 5.1 from stereo, which *may* add some ambiance. Usually you'll import audio & video separately into your DVD authoring app -- at a Minimum the app should pass DVD-spec audio & video through without re-encoding. Better authoring apps take more than one track, & better yet if they can deal with subs -- note some take graphical subs, others only take text based sub files. Preferred is creating output folders on hdd vs. burning -- that allows testing, editing/modding, & using burning software that's Much better than that included in most authoring apps. Note that if a DVD authoring app can do menus it includes an encoder -- try not to use it, since it's often a bare-bones encoder with few if any controls/settings... Nero, Roxio, & Sony all have a different encoding setup in their editing vs. DVD authoring portions, even if they use the otherwise same encoder in both.
