DVD Authoring Rundown - Digibeta to DLT
So I work for a couple small DVD labels and we are considering bringing some authoring in-house and I'm trying to determine the best way to go about this whole process. I'd like to use DVDSP for the authoring, as a scenarist setup is out of our budget range. The first is mpeg encoding, I'm considering either Bitvice as the mpeg encoder, or using the Optibase card for our needs. The films will come on digibeta and I'll be using SDI for input. As far as I can tell, Bitvice seems to offer more flexiblity than the Optibase card, as well as 2-pass encoding, which seems as if it would yield a higher quality. However the Optibase card is quick and I won't have to deal with massive files. But really the issue is quality of mpeg encoding.
2 Streams of thought on the process:
Digibeta to AJA IoLD via SDI, captured into FCP. Compressed with Bitvice then authored with DVDSP and output to DLT.
Digibeta to Optibase MPEG-2 Encoder for mac via SDI, then authored into DVDSP then to DLT.
Any thoughts on the relevant merits of either of the above setups?
The second issue is in regards to audio. The majority of the titles we will be working on through the end of 06 will be older titles and will be either stereo or mono mixes, however we have in the past put out modern titles that were mixed in Dolby 5.1, which of course presents an issue.
I'm pretty murky here as it seems that once the audio is encoded to Dolby, it is essentially 2-tracks, correct? So after the film is restored (and all our films go through a restoration/touch up house) and put onto Digibeta, I should just be able to pull the Dolby 5.1 audio into my machine as two of the 4 audio tracks on a Digibeta tape, is this right? Or will I need to be able to encode the actual 5.1 signal on my end, which I can do, via compressors Dolbly 5.1 encoder, but that seems a little insane as I would also have to mix the audio, correct? Which seems outside the realm of DVD Authoring. I see that the Optibase card has a '5.1 pass through' capability, which would imply I need a hardware encoder for this, which I have yet to find for the mac, but I haven't looked super hard yet.
As you can see I'm quite vague on the Dolby encoding end of things, any advice you might have on either of the two issues here would be excellent, thanks.
I have not worked with either the Optibase card or BitVice, so I'll leave that question to others. But if the quality is at all comparable, the "real time" (or twice real time for a 2-pass encode)factor of the Optibase sure looks attractive.
As far as 5.1 audio...
No, you don't just put two channels of the Digibeta into an encoder for 5.1 surround. You're right, once encoded, it really is like two channels that get decoded out to 5.1. But to start with, you need six (if you have a Low Frequency Effect channel, 5 if not) discreet files that get sent into an encoder and turned in to the encoded 5.1 track. Typically an audio studio has done the 5.1 mix and provides the 5 or 6 tracks. The normal delivery format is DA-88 tapes. There's lots of great info on dolby's web site at http://www.dolby.com.
Bitvice is is software only encoder and the encode times are really long! Optibase sure - better at lower rates and realtime - there is no real comparison there
At this time, you can't compare software to hardware compression for quality issue. Hardware compression surely gives the best results, especially at low bitrates, as eric mentionned. And the Optibase card (SDI and DE) really worth the money (Compared to Sonic SD-Series). For 5.1 audio, as drumrob mentionned, digital files are often use (That's what we always use at our studio), so they can be easily import in the dolby encoder (Software I mean, I don't think you can afford a Dolby hardware encoder considering your budget).
Excellent - thanks for the advice, you've really cleared some things up for me.
How is the audio normally dealt with in regards to 5.1 encoding? That is if I recieve the digibeta and a DA-88 tape I would need what kind of audio hardware? Obviously I would need a DA-88/98/HR/etc for playback, and then I would like to go out TDIF in order to maintain the digital flow and then into what? Is there an audio interface for the mac that has TDIF straight in? Or would I need to go through a digital mixer (say an 02R) and then out multichannel AES/EBU into an audio interface that is capable of receiving multichannel AES/EBU. Which I can do, but it seems a rather elaborate (and relatively expensive) purchase to buy an expensive mixer just to leave it zeroed all the time and used solely as a TDIF to AES/EBU converter, which I suppose is what I would need. I've looked at a number of sites and have yet to see an interface that supports TDIF directly, outside of the one of the cinewave cards, which I think would put me outside my budget (as if purchasing a multchannel digital audio setup wouldn't, but I digress... :).
What is the likelyhood that I would be able to get the 5/6 audio tracks as aiff/wave/PCM files, rather than on tape? This would be ideal and would greatly simplfy things, is this something most restoration houses are willing to do?
Also - if I am not going to be using an external hardware dolby encoder, then it would make no sense for me to get the DE model of the optibase card (the DE has the 5.1 pass-through) and just stick with the SDI, which is $500 less.
Many thanks again!
DVDSP will include another program called Compressor. One of the things that Compressor does is ac3 encoding, and on a recent PowerMac it will encode quite a bit faster (6X or 10X faster!) than realtime.
Yes, you could receive the 5 or 6 files as WAVs or AIFFs. Then you could put them into a program such as Adobe Audition, and assuming you pay the extra $ for the dolby encoding plug in, you could output a 5.1 stream from there.
As far as hardware, you would need to have some sort of encoder/capture card that would accept 6 inputs of audio. I know the packages from Sonic Solutions do that, but they're VERY expensive. So on a budget, you could get something like the M-Audio 1010 multichannel audio card, and capture the tracks into Audition, then encode from there. I'm sure there are other software solutions besides Audition. That just happens to be what I'm familiar with.
If I were able to recieve the tracks as AIFF files, how would I go about maintaining sync with the video? I don't believe that AIFF would have any sync built-in, correct? I could sync it up by hand, but it seems that it would be more reliable to have some kind of hard wired sync, perhaps a seventh audio track as the sync track, or is there some other way?
Also on the MPEG encoding side of things - I'm pretty certain we are going to go with the Optibase card, my only question here is the encoding (technically transcoded) of assets outside of the film itself. Things like motion menus and so forth, would these need to be transcoded in a different app? Likely I would use Compressor for this task, although I am a bit wary as I've heard the MPEG-2 encoding in Compressor is a bit spotty. Truthfully though, I've used the MPEG encoding on some smaller projects without issue, however I'd like to maintain the same level of quality throughout the disc. One thing that irritates me a bit about some discs is the film itself may be encoded just fine, but then the extras or stills or what not have cleary been encoded poorly. Thanks again for all your help.
Sync between audio and video is normally maintained by encoding using matching timecode on both the video and audio. If you were to be provided just files, the facility that produced them would have to tell you where "0" on your file is as it relates to timecode. Probably 1:00:00:00 since that's the standard way of doing it, but you would need to make sure.
I'm considering selling my DVD Creator System which does everything you need. It's the closest thing to a Scenarist on mac. My email is email@example.com if you are interested.
"everything is broken"