THE CONFUSION

For many people, when it comes to getting their home movies converted from the analog domain to the digital can be pretty confusing.  Should they try to do it themselves or hire a professional, such as Trevor Thurlow Productions?

This is pretty understandable.  When you look at a VHS tape or a Digital 8 tape, aside from the obvious size difference, you are just talking about how the video recorder magnetizes the tape.  With VHS it magnetizes the tape in a manner where the signal is recorded in an analog form representing the electrical voltage required to reproduce the image and sound; with digital the signal forms a magnetic signal of the digital information. 

One thing to remember with analog videotapes, especially 3/4" U-Matic, VHS, Betamax (including ED Betamax) and the 8mm tapes is that they were designed to use a composite video signal for storing their video information. These formats (and S-VHS) used what's called the color-under method, where the fullest strength of the Luminance Channel (the black and white video information) was recorded, while the Chroma channel or color information was compressed and recorded to the tape at a very low level in the same channel as the luminance.  For example, VHS recorded it's luminance at a frequency of 3 Mhz, while it's chrominance was recorded at a very low 629 kHz.  S-VHS and S-VHS-ET managed to increase the luminance frequency to 7.0 Mhz, but still recorded the chrominance at the 629 kHz level.  In S-VHS's case, this color under method is used, but then in order to get the full quality of the luminance channel, you need to play the video out of a S-VHS deck through the S-Video channel, due to the two channels being sent separately.  If you played the signal out through the yellow composite or RF composite channels of a S-VHS VCR, you would end up with a signal that contained a high degree of noise due to the use of the composite video signal, which mixes both channels into one.  Even with regular VHS, and for that matter all the other color-under video formats, S-Video will allow the VCR to pass along a cleaner signal than the yellow composite cord could, since the VCR's are able to upconvert the Chrominance channel to NTSC and PAL/SECAM levels and pass it as a separate channel from the luminance channel into the TV or whatever is being used as the digitizer.

4:1:1 vs. 4:2:0 vs. 4:2:2

There is much debate about what the chroma subsampling should be when converting analog video to DVD or other digital formats.  You can check out my arguments for 4:1:1 vs. 4:2:0 at the following link:


 

 Suffice it to say, from all my tests and experiments, with 4:2:0 you get garbage in and then get garbage out.  Even when I shoot in HD, while I use a AVCHD camcorder to just set up in a corner and get shots for cutaways, the camera is shooting in 4:2:0 and the color is horrible when compared to video shot with 4:2:2 DVCPROHD, or even upconverted 4:1:1 DV material.  With 4:2:0 and 4:1:1 you are dealing with color fields that, when you do the math for a standard definition image, are being stretched from 180 by 120, up to 720 by 480; however, DV at 4:1:1 stretches those fields separately, whereas 4:2:0 uses a cosign method and tries to average the levels together in a single field.  This is why 4:2:0 is never used for green or blue screen work, the colors are too combined, and give to much interference.  If you want a good digital representation of your analog video, even on DVD which is natively 4:2:0, importing video through a 4:2:0 method will give you low-quality results.  You start out with a high quality transfer, you will get good quality video on your DVD.

DVD Recorders

For this category, these DVD recorders are the set top DVD recorders that you can buy that either have a NTSC/PAL/ATSC/QAM tuners and were designed to replace the VHS recorder.  Now then not all set top DVD recorders will have all the tuners (most will have a NTSC or PAL tuner, but rarely both; ATSC is the broadcast standard used in the United States and Canada for over the air broadcasts, other parts of the world use other standards for Digital Television Transmission; QAM is used in North America by cable companies to encode their signal for transmission.  There's no guarantee that the QAM tuner will work as, unlike ATSC, there are no laws telling cable companies to use Clear QAM, which allows QAM tuners to decode the signal, so you have to test and see; with satellite you need to use an external decoder.) and there will be some models that do not have any tuners, but will still have input through a yellow RCA composite connection or even Firewire (older models will also have S-Video connections).  Set top DVD recorders record in the 4:2:0 chroma subsampling method, and unless you are going from S-Video or DV via firewire, will only provide low quality images. As of this writing in March, the majority of DVD recorders currently on the market in the US and Canada offer input through only 3 connections: RF coax, RCA (yellow) composite and Firewire.  Also with Firewire, as far as I am aware there are no DVD recorders that will record, via firewire, from MicroMV; thus your current options are to record this via composite video.  Even the manual for my Sony RDR-GDX380 mentions that you can not record MicroMV via Firewire (only DV), so if you have a MicroMV camcorder, and were hoping to get a digital transfer via the Firewire connection, unfortunately that will not work.  But this is for the best, as recompression of an MPEG image will result in a lower quality picture.


Also most DVD recorders only have 2-D color separation when you bring a video in via RCA (yellow) composite.  This results in oversaturated and bleeding colors, and makes everyone look like they have high blood pressure or a very bad case of sunburn.  Also, in many department stores and photo studio's that offer video to DVD transfers they will be using DVD recorders to hook up to their own low-quality VCR's and camcorders.  If you've ever wondered if it was possible to fit an entire T-200 tape on 1 (one) DVD, that was recorded in the SLP/EP mode where you have 10 hours on one tape, you can be sure that for that $5.99 that those stores charge you'll get a DVD that was made on a set top DVD recorder at a very low picture quality (this quality is much lower than VHS SLP/EP quality!).  My previously mentioned Sony DVD recorder is able to pack 14 hours of video on 1(one) 4.7 GB DVD disc; I've tested the level and really, it is so soft that on the curling game I tested it on I could not even make out what the graphics were saying, except for the large station logo that came up a few times before and after a playback of a shot that the commentators wanted to take a second look at.

Video Capture Cards

This is an interesting subject.  The average person will see in the electronics section of their favorite department store, devices that look like USB keys with connections coming out of them for connecting a VCR by RCA (yellow composite), or some will even have RF connections and a few have S-Video connectors. With analog video, these capture cards capture MPEG-2 video (or in a few cases MP4/H.264) at a maximum variable bitrate of 1.5 Mbps.  This is extremely low---most DVD burning programs will not burn a DVD if your bitrate is going to dip below 1.5 Mbps, as the resulting image would be very pixelated and soft.  And if you are planning to edit your videos, this is not the way to go.  And very amusingly, this is how many so-called "Professionals", who claim to have been in the transfer business for decades, transfer videos.   


Now then when you are using a Professinal level Capture Card, such as the Canopus ADVC-300 or the Matrox MX02, you can, in the case of  the ADVC-300, capture at a Constant Bit Rate (CBR) of 25Mbps, or even higher.  A Constant Bit Rate can give you a better transfer due to the bit rate only moving in a very narrow range of +/- 1.0 Mbps for the desired target.  So with DV you have a CBR of 25 Mbps, so the bit rate only fluctuates between 24.0 and 26.0 Mbps.  WIth a Variable Bit Rate (VBR) you can have a range ranging from 0.0 Mbps up to 9.4 Mbps for DVD.  So a VBR gets rid of more information due to not keeping area's that have very little information (such as titles with just a black background).  And with DV the computer compresses each individual frame by itself, and does not rely on the frame on either side of it to determine what should be kept.  DV is also not as CPU intensive as MPEG, and even the average consumer editing program can edit DV material.


But with other capture cards, such as the Matrox MX02, when the video is imported the codec can also be changed to help give a much higher quality transfer.  Other codecs that can be used are DVCPRO50, which imports your video at a 50 Mbps rate, and at a 4:2:2 color subsampling.  Uncompressed AVI or MOV can also be done, where your video is brought in at a rate of 96 GB per hour, or about 90 Mbps, and a 4:4:4 color subsampling.


HuffYUV and Lagarith

I've seen many people say to use HuffYUV or Lagarith for Uncompressed video.  The problem is that HuffYUV has not had an official update since 2002, and Lagarith is also getting pretty old as well (Lagarith was a build off of HuffYUV) and both require good CPU's to process them.  And Lagarith seems to work only with Windows.  Whereas DVCPRO50, DV and Uncompressed will be supported for a long time to come and work with both Windows, Linux and MAC computers.  As well, while many professional editing programs can handle the codecs, many of the consumer editing programs are unable to handle the codecs, or have extreme difficulty working with them.

H.264

While H.264 can support all the chroma subsamples, this codec really, really compresses the video.  Some satellite and cable companies are recompressing their signals with H.264 (thus the reason why you'll get a letter from your company telling you that you need to upgrade your satellite/cable box if you have had your box for many years) from the MPEG-2 that the ATSC standard requires television stations to broadcast at.  Unfortunately broadcast TV uses a chroma subsampling at the 4:2:0 level---this is one of the reasons that HD content looks so much better on Blu-Ray. 

Blu-Ray uses H.264 as one of it's compression codecs (MPEG-2 was also used on early Blu-ray titles, as it could support 1080i, and then the VC-1 codec, based on the Windows WMV codec, is also used), at a 4:2:2 subsampling level.  So when you are watching a TV show on Blu-Ray in HD, you'll notice that the Blu-Ray will have a better picture quality due to the lower compression and improved chroma subsampling.  Unfortunately this is only for HD material, SD material (such as VHS video) gets encoded in MPEG-2 4:2:0.

Suffice it to say, but MPEG-2 and MPEG-2/H.264 are fine for delivery, but if you want to be able to see your videos at their best, then neither are the way to capture the video.  And if you are looking to edit your video, the best way is to get it transferred to a format that offers the least amount of compression. 


So for all your transfers, you should go with a place that captures your video at a higher quality than the end device.  And going with a place that captures at 4:2:0 and then burns to DVD at 4:2;0 is not going to be able to give you the picture quality that you are looking for.