I work with video for a living; editing, encoding, as well as authoring DVD's for production. At my job, I field a lot of calls from people that want to have their old VHS tapes archived to DVD, BD, or as an iPod compatible stream. In most cases, it is less expensive for the consumer to buy the necessary gear and learn to do it themselves, but a lot of people are intimidated by the prospect. No one has yet to make a fool-proof video capture, editing, or DVD authoring software for the consumer, so, yes, there is a learning curve, but it is only as difficult as your project requirements.
If you just want to do simple transfers, a stand alone DVD recorder will usually fit the bill. But if you want to be able to edit or perform some sort of restoration, or have a complex menu system, it can get you into headache mode pretty quickly if you allow it. .
The goal of this guide is to clear the air a bit and give you some clear options for archiving your VHS, 8mm, Hi8, and even BetaMax tapes to a digital format, answer some common questions, and to give a glimpse into the more complex aspects. Most people will find the first three sections sufficient to proceed.
.
Video Capture: Simple Method
The easiest way is to get a stand alone DVD recorder. These can be acquired for right around 100.00, and it will allow you to link a VCR to it via the composite cables (these are the RCA pin plugs that are yellow, red, and white). These machines generally have no extended features, but some will allow basic editing and authoring ( menus, chapters etc) when you use a re-writable DVD (DVD-RW) The only issue, (or benefit, depending on your point of view) is that once you have done a project on a DVD-RW, you need to copy the disc, using a computer, to make a permanent copy on a DVD-R. Re-writable media should never be used for permanent storage.
There were, (and maybe are) stand alone units that have an on-board hard drive. This allows you to record to the hard drive, do basic editing and authoring, and burn the final result to a DVD-R ( or +R). The main benefit: if your recording goes sour in the capture process, you can start over, and no media was lost.
The overall benefit once your DVD is finished, is that you have migrated your analog recording into the digital domain, and it can now be loaded to a computer for further manipulation: trans-coded for iPod, converted to progressive scan for viewing on HDTV, etc...
Most recorders have 5 or 6 bit rate (speed) options, usually listed as modes of 1 hour, 2 hour, 3
hour, etc...or good, better, and best. You should stick to the 1 and 2 hours modes, only. I discovered the hard way on my own DVD recorder, when put in the 3 hour or longer mode it didn't drop the bit rate, but dropped the resolution by 50% and at the lowest speeds, (longest record times) 75%!
NTSC DVD video resolution is 720X480@72 dpi (dots-per-inch). The 3 hour mode dropped the picture dimensions down to 360X240!! While this will display as full screen on your TV set, it's only half the resolution of the better quality capture modes. The 1 hour mode on most machines will actually give from 75 to 80 minutes of record time on a single layer disc. Twice that, if your recorder accepts dual layer media.
When dealing with home movies or other irreplaceable video content, use the highest quality bit rates/speeds available on your recorder. You can always load the DVD's you make onto a computer to make longer running compilations without sacrificing quality at a later date.
DVD-R or DVD+R? DL? DVD5? DVD9?
Most stand-alone recorders require (or prefer) that you use DVD+R. The only difference between -R (that's a hyphen, not a 'minus' sign), and +R, is that the latter has a pre-written finishing file. Back in the days when DVD recorders could only run at 2X max, having the pre-written file would shave about 20 minutes off of the burning process. These days, it's not really needed, except for stand-alone units that typically burn in real-time, or from an internal hard drive at 2X or 4X max.
Most of these recorders will only use single layer media, but there are some that can use dual layer media that will allow you to use higher bit-rates (or speeds as indicated on the recorder) and an increased capacity, (Single layer discs are 4.28 GB and dual layer discs are 8.54 GB). All DVD burners for computers made in the last few years are already compatible.
Writer drives (burners) for computers will accept a wide range of media, and are compatible with -R or +R. For data storage or when creating a DVD master, I prefer to use -R.
Dual layer media doesn't present any special needs in most applications, but there is some misunderstanding on just what DL media is.
There are two recordable layers, one on top of the other, and each layer is written with a different wavelength, a that are read back with different focal lengths, on the same side of the disc. When the first layer has been played, the player refocuses to the second layer and continues. You may notice some hesitation or freezing on some players during the layer change, but better players with larger buffers have pretty much made this event rare.
Professional, and some prosumer DVD authoring software will automatically create the layer break for you, by locating a chapter marker or track that is at an optimal position, or will prompt you to create a layer break.
Some consumer software doesn't allow the user to select or create a layer break. Instead, it creates a layer break by splitting the content in half: 50% on each layer.
The commercial designation for a single layer disc is DVD5, and a dual layer disc is a DVD9.
There are some two sided formats too. A disc with a single layer on each side is a DVD10. A two sided dual layer disc is a DVD18. One oddball that you sometimes run across, but only with replicated media, (movies and other store-bought titles) has a single layer on one side and a dual layer on the other. This is a DVD14.
One more note on DL media; when writing the disc from an authoring software, there are two options for the second layer: OTP (opposite tracking path) and PTP (parallel tracking path).
OTP: All disc media tracks start at the hub (center) and spiral to the flange (edge).When the disc is written as OTP, the second layer spirals from the flange back toward the hub. This makes for a very fast layer change, and even a player with a small buffer cache can change pretty seamlessly.
PTP: The second layer follows the same path as the first; from the hub to the flange. This scheme takes more time to effect a layer change as the pick-up (laser/optics assembly) has to physically move back to the hub before it can read the layer.
Time Base Correction
When using consumer equipment to transfer tapes, issues like dropped frames, bad captures with big gaps, or being unable to capture a video at all, even though it seems to play on your TV, are a common occurrence.
Any video capture device requires a solid time code to sync with. Analog tape in general usually requires an extra piece of gear called a Time Base Corrector, or TBC for short. What this does is generate a new time code for the video you are capturing, so that the capture device (stand alone DVD recorder or computer capture device) can sync.
Until a few years ago a TBC would set you back upwards of $1500.00, but now it's possible to find a consumer or pro-sumer version for under $120.00, or a used pro unit for well under $300.00.
In addition to creating a time code, the TBC also allows you to correct the color (saturation and hue) as well as brightness and contrast (offset and gain), in real time.
Some stand-alone DVD recorders are equipped to generate incoming sync and will automatically correct color (not always a good thing on the latter) and so a TBC may not be needed. But all computer capture devices like DV converters, SDI cards and even USB video capture devices will need a TBC added to the signal chain.
Field Dominance and Why it's Important
Standard def TV uses an interlaced picture format, meaning a beam of light scans lines across the screen that are spaced apart by one line width, which is the first 'field', and then the beam scans again in those spaces that were left between the lines of the previous scan, which is the second 'field'. This happens 60 times per second, creating the picture. When each field has scanned the screen, this is equivalent to one frame of video.
In professional and broadcast video, the bottom or 'even' field scans first, followed by the upper or 'odd' field. This is called Lower Field Dominance. Unfortunately, much of the consumer DVD recorders capture devices and software are Upper Field Dominant. When capturing from an LFD source, this can cause colors to slip off target, or in the worse case, create jitter and other motion anomalies, and look fuzzy. When working with LFD files in a UFD environment, these anomalies are exacerbated. Motion becomes very jerky and jittery, picture quality is very soft and fuzzy. This is very obvious to even an untrained eye.
In and of itself, there is nothing wrong with using UFD, if your source material is UFD to begin with.
Stand alone DVD recorders, USB capture devices and software are all UFD.
The best way to get around this issue, is to use a computer with a DV conversion device. This will bring in the files as LFD. For a first time user, try the resident software that came with your computer for capture first, (MovieMaker on PC and iMovie on Mac).
If you ever wondered why something you recorded to DVD from TV looks degraded, it's most likely you used an UFD device.
HDTV, your computer, iPad, etc use a Progressive Scan display. Progressive displays do not use scan lines to produce the picture. Basically, each frame appears on screen as a complete picture. Converting from interlace to progressive requires some processing that I'll touch on in the section on Transcoding/Converting.
Using a Computer to Capture Video
A DV (digital video) converter is an outboard device that usually connects to your computer via FireWire, but there are some newer units that will connect via the USB port, or, better, there are internal capture cards that are easily installed (my favorite is by Black Magic). Be sure the device you are considering is a true DV converter. Interfaces that require proprietary software are generally not.
A capture via a DV converter or Mini DV camcorder that offers feed-thru, will be Lower Field dominant when captured as an AVI DV Stream on a PC or as an uncompressed QuickTime DV stream on a Mac.This will require encoding to MPEG 2 to create the final DVD, and that will add a little time, but it will produce a superior result as opposed to real-time software encoding to MPEG2.
Capture files are going to be quite large. Typically, it's about 13GB per hour of content.
Windows has both a MovieMaker program and DVD authoring software that is geared toward the consumer and is a great way to get started in creating your own DVD content. Mac computers come with iMovie and iDVD that does basically the same thing, but does not offer as many output options, but is easier to use.
ION, Pinnacle and a few others offer USB video capture devices and software bundles, but all of this stuff is UFD and offers very few options to trans-code files to the format you want to end up with. While these apps will make compliant DVD's and MPEG 4 streams for iPod and other devices, they do little to correct problems associated transcoding to or from interlaced video.
Pinnacle software offers the ability to capture and encode in real time, and can produce good results, but not stunning. Not good enough for me!
There are some devices and software apps that will allow you to capture directly to MPEG4, and will even convert your interlaced capture to progressive scan in real-time. But, if the capture device or software offers no real-time processing, (particularly de-comb and de-telecine), your video will be blocky and have a 'venetian blind' effect on any motion.
I use some UFD equipment for one-shot off-air recordings and I will occasionally import video to my computer from DVD's made on my stand-alone recorder for editing. I use Sony Vegas Pro for editing and encoding at home, and it can use any kind of video stream. However, I have found that anything I bring in via my DVD recorder or my Pinnacle Movie Box, which are both UFD, looks horribly degraded.
While I can fix these issues in Vegas, doing so requires more of my time, not to mention adds to the render and encoding process. A 30 minute program can take in excess of 3 hours to export with all of the added processing. A good clean AVI capture, via DV, requires no adjustment, aside from editing, and the same file takes 20 minutes to export.
Surprisingly, using the Windows video capture application produces a cleaner and sharper DV stream as AVI, than does the Sony Vegas software. Vegas has to create proxy files (rendering) to use these, but the result is well worth the extra 10 or 15 minutes of rendering time.
Encoding
Compatible TV format specs for a DVD in the US is in NTSC format, 525 lines, 29.97 fps (frames per second) interlaced, or 23.98 fps progressive.
Full explanations about encoding can be found on-line for those that really want to wade into this subject. What follows is more of a thumbnail sketch.
Assuming you've captured some video, it now has to be encoded to fall within DVD spec. DVD video is in MPEG2 format, and the audio portion can be either MPEG audio, Wave, AIFF, or Dolby AC3 files (the last one is industry standard). Another encoding option is to encode a multiplexed MPEG2 stream that includes both the audio and video in a single stream. Most DVD authoring software can use this type of file, and it simplifies the encoding process for the novice.
Most consumer software allows you to drop in the uncompressed video/audio into the DVD authoring software, and it encodes it for you. Be advised that the encoding defaults for these programs is pretty much down and dirty. You may want to do a little study on getting the best possible result. There are many reference works online that will explain how to set up a particular software for multiple encoding passes, and variable bit rates that can produce stunning results.
Using two passes of encoding produces a better result than a single pass. After the first pass encodes the stream, a second pass cleans up a lot of the pixels that can cause fuzziness or 'feathering'. Professional encoding systems offer multiple passes of encoding, the advantage being that you can create a very clean video stream at lower bit rates, and produce stunning results at the higher rates.
Encoding is measured on MBPS (Mega Bits Per Second). The average bitrate you will find on most commercially released titles is right around 7MBPS, variable bit rate with a peak nearly at 10.00 MBPS. Commercial titles are usually dual layer discs, ( DVD9), which affords the luxury of higher bit rates.
EXAMPLE: Not all software will have the options related, but many do: A two hour video that is to go onto a single layer disc with either .MPA or AC3 audio can be encoded at 4.2 MBPS, variable, two passes of encoding, with a peak setting of 9.2 to 9.8 MBPS (depends on who's software you are using). This keeps the video within the disc capacity and produces a best quality result.
In the same scenario, but the audio is in an uncompressed format like wave, or AIFF, it will require using a lower bitrate for the video, as the audio portion will take up 3 to 4 times more space than an encoded audio like AC3 or MPA. A lowered video bit-rate will produce a lower quality picture. If you have no option to encode the audio, then you might consider using a dual-layer DVD-R (DVD-RW-DL). This will allow the use of even higher video bit-rates as well (an average of 6.5MBPS would be a safe bet). Be advised that a high video bit-rate combined with an uncompressed audio track my present playability issues on many DVD players. It may exceed the player's ability to transfer, buffer and stream.
One of the options for encoding is CBR, or Constant Bit Rate. This is not really suitable for motion, and is best used for limited motion, still presentations, or when capacity is a concern over quality.
Transcoding/Converting
Conversion simply means converting one video format to one that differs in picture format, scan rate, frame rate, or format platform (NTSC to PAL, or interlaced video into progressive, for instance), while transcoding means changing the file type or codec to a different one (such as transcoding from MPEG2 to MPEG4, or Windows Media Video into H.264, etc...).
Converting video from one format to another takes considerable computer resources and can take as long as 10 minutes for each minute processed. So far, the most effective software I've found for format conversion is Apple's Compressor 4. The results look great, but may take days to process, unless you've got a network of MacPro's you can tap into. Hardware conversion is best, but also very expensive.
Transcoding from MPEG2, AVI or MOV files into streaming video or for iPod will be in MPEG4 format with the H.264 codec. There are other compressed formats such as DIVX, WMV, etc, but I recommend sticking with the iPod or MOV files, as these seem to be more universally compatible.
The simplest way to create this type of stream is to download a copy of HandBrake (totally free software, just put in a search). HandBrake's default settings leave a bit to be desired, but the accompanying help file is very clear on how to get the best result, without getting needlessly technical. After a little homework, and some experimentation, you can produce some really great looking MP4 or M4V streams that will look great on your iPad, computer , iPod, and even on a large HDTV set. Also, the processing is relatively fast when compared with other transcoding/encoding solutions.
Handbrake offers the necessary processing (filters) to convert an interlaced video into progressive scan, without any artifacts or motion anomalies.
Handbrake also offers creation of larger files types that can be used to up-convert interlaced video to a compatible file for HD viewing.
Professional and consumer editing software apps do offer up-converting processes, in order to put old SD content into the HD realm. But these require a computer with a lot of processing muscle, or even multiple computers via a network, to transcode files in a reasonable time frame. Results will vary according to the limits of the software and the user. Also, an up-converted SD video is not going to look any better in HD than it did in SD. It's usually enough just to convert the interlaced video to a progressive scan format, as this will produce the same end result, without a lot of time consuming processes.
Authoring
This is what the process of creating the actual DVD is called, as you are 'authoring the programing and elements' to make the disc perform and behave in a particular way. The earliest authoring tools were created by and for programmers. The abstract interface versions we use now (DVD Studio Pro, iDVD, Pinnacle, Vegas DVD Architect) made it possible for the rest of us to author DVD.
Still, this part of the process can be the most confusing step for a lot of people, which is why we have programs like Mac's iDVD, and Windows DVD Maker. These will usually meet the needs of most people that just want to burn off some home movies, and with a little experimentation, you can get a very good result.
The basic DVD structure is comprised of tracks and a menu (or menus), with buttons linked to tracks or markers within tracks. Every video file brought into most DVD layouts will become a track by default. A DVD can contain up to 99 tracks. Within the track, markers can be added to the time line for chapters that can be linked to menu buttons, button prompts or a layer break. A track can have up to 99 markers. Some advanced authoring apps, like DVD Studio Pro will allow you to put multiple files (elements) on a single timeline, so many files can become one track.
Professional DVD authoring software like DVD Studio Pro for Mac, or Sony's Vegas DVD Architect for PC (the latter can also author BD) will give you full authoring capabilities, but with a major learning curve.
A good intermediate application is something like the Pinnacle Studio series (made by Avid), which will give you many options that the professional software has, but in a more simplified interface. Roxio has similar software for Mac users.
Any of these applications will produce a playable disc for you, but if you want to author a DVD for commercial distribution, you really need to use the professional software or hire a pro. The prosumer and consumer DVD authoring apps don't always make a DVD master suitable for the replication process, and will often fail the mastering process (a glass master is made in order to create 'stampers' used to replicate discs).
The final step in authoring is to do a 'build' or 'create'. This will mutiplex the audio/video streams and build the final DVD folder that is to be burned to disc.
You create a target folder, and the software will create the following folders within it: VIDEO_TS and AUDIO_TS. Only the VIDEO_TS folder will have any contents (the AUDIO_TS folder is for DVD Audio discs). I would recommend just creating the folders on your hard drive first and opening the VIDEO_TS folder via your computer's DVD player app. This will allow you to preview your disc before you burn it, and allow you to go back, correct any errors, re-build and preview again. Once your sure everything is right, go ahead and burn your DVD.
For more detailed instruction, there are many freely available tutorials that will teach the finer points of DVD and BD authoring, as well as books.
Burn
While all authoring software has the ability to burn the disc for you, I usually recommend that a third party burn software to be used like Nero, Roxio Toast, NTI, etc...as they offer a verification option. Be sure to set the burn preferences as DAO (disc-at-once) as many of these default to TAO (track-at-once).
In the case of Roxio Toast, use the Data layout and set the properties under the advance tab as UDF (Universal Disc Format). In Nero, just use the Create Video DVD selection. Nero will only import the VIDEO_TS contents and will not permit any additional ROM data. The NTI software is similar to the Roxio software made for PC, and has more options than Nero.
When making a DVD9 (dual layer) disc: If your layer break selection is critical, let the authoring software burn the disc. When authoring a DVD9, it's common practice to create a layer break in a section that has faded to black or at the beginning of a track, so it appears seamless. Third party apps will ignore your selected layer break, and split the content evenly over both layers, creating a layer break at the 50% point. This will make for a very fast layer change in a good player, but still cause a visible stutter on a cheap player.
Regardless of which app does the burn, be sure to set your write speed for 4X for all disc types, or as low as your app will allow. Burning below 4X is not advised, as this can actually introduce errors, or not burn successfully. Equally, burning too fast can create readability issues in addition to physical burn errors.
For many years, 4X has been the sweet spot in the professional realm. I'm sure there is some mathematical explanation for this, but as I'm not a mathematician, I've never ventured to find out.
NOTE: If you plan to create a DVD9 for commercial distribution, make sure your software can create a file image in DDP2.00 format. This creates two folders; LAYER0, and LAYER1 that contain the .dat image files. These can be transfered to a thumb, hard drive, two DLT tapes, or to two single layer DVD's. This is important if you have a critical layer break, or require any encryption, as a playable dual layer disc can not have protections added, and the desired layer break will not be maintained.
For example, if CSS encryption is desired, for DVD 5 or DVD9, the flags for CSS have to be turned on in the authoring process, before writing the DDP 2.00 image file. The actual encryption is applied in the glass mastering phase. CSS can not be added after the fact. Rovi (formerly Macrovision) encryptions for copy prevention can be added to an authored DDP2.00 file image, but not a playable DL disc.
