1 person found this helpful
Two streams cannot be captured simultaneously using Premiere.
If you're capturing DV and HDV tapes, then why bother at all with the SDI capture? You're gaining nothing, quality-wise, and it's just going to take up more space (presuming you're capturing to uncompressed or MJPEG or similar). The transfer from DV and HDV tapes is bit-for-bit across a FireWire/IEEE1394 connection, so you're losing nothing in the capture.
It sounds like you know what you're doing with this, so I'm not making any assumptions, but thought I'd throw it out there. I'm actually kind of curious about the workflow as I'm entering into an archival project myself, though with analog Betacam SP tapes.
There is a lot of discussion in the world of archiving digital video about the pros and cons of SDI vs. Firewire. Many echo your opinion - why not do only firewire. This opinion is held by those that have what might be called a digital purist point of view. In other words with firewire you get the pure digital information. Whereas those who favor SDI are of the opinion that with an SDI transfer (using hardware conversion and error correction) the result is closer to what was actually seen at the time the video was shot. In other words 20-years from now no one will remember how a scene actually looked and so the digital video file may be the only artifact left. The closer the video is to the actuality of the moment the more historically accurate the details will be. There is also the issue of sustainiblity of the digital files as technology advances. For archival purposes many "experts" recommend capturing both the SDI signal and the firewire signal. We have elected to do that. You can find a lot of lively discussions with "experts" taking every side of the issue on the WEB.
those who favor SDI are of the opinion that with an SDI transfer (using hardware conversion and error correction) the result is closer to what was actually seen at the time the video was shot.
But to answer the original question, use two computers, both running Premiere.
Not a helpful answer. You must know more than most about all of this to brush off the consideration with such glibness. When someone does that I feel that they cannot be taken seriously.
Is there any discernible (eye-ometer) or measurable (waveform /vectorscope monitor) difference between same video captured via both methods?
If one was to have 2 computers both running Premiere.
Could you capture both formats at the same time with only one pass of the video ?
And if so how would it connect to the camera ?
Just curious: GLenn
Since the deck in use outputs both DV and SDI, each would be connected to a different computer for simultaneous format capture. You couldn't have deck control using both, obviously. But theoretically at least, capturing both should work fine.
You must know more than most about all of this
It doesn't take much more than simple logic. If you were recording live, before compression, than I'd concede the superiority of SDI. But from tape, after compression, the best you can hope for is that the missing data has been correctly restored to the SDI signal. And there's no guarantee that the hardware you currently use will do a better job of that than the software used on the DV file down the road (or even currently).
Here is a message posted on the AMIA (Association of Moving Image Archivists) in reply to a similar question. The writer is recognized as one of the leading authories on the subject. A link to his website is included in this quote:
"Date: Wed, 24 Feb 2010 12:42:46 -0500
Reply-To: Association of Moving Image Archivists <AMIA-L@LSV.UKY.EDU>
Sender: Association of Moving Image Archivists <AMIA-L@LSV.UKY.EDU>
From: Jim Lindner <jim@MEDIA-MATTERS.NET>
Subject: Re: Best practices for Mini DV preservation and digital transfer.
I think that the points that David makes need to emphasized a bit. In particular his reply to Lee's comment regarding that the file output will be an exact replica because no re-encoding is done. For years some people thought that just because video was "digital" that all "digital" copies would be identical. That is of course not the case. It is also not necessarily the case that a firewire output from a digital video recorder will be the same as the digital video output from the same recorder or even a "replica". Digital Video recorders are optimized to provide a good looking VIDEO output for the human eye, and that means that a great deal of the technology in these machines are devoted to doing things to the picture that "massage" the data in such a way that it is pleasing to the eye. Many things are done to the data after playback that makes it look better. It is a very long list.... dropouts are corrected for, motion artifacts are smoothed in some situations, edges may be enhanced or blurred in others, color spaces may be changed, entire sections of images are interpolated, there are a whole list of things that happen - all in real time, and it actually is very fancy stuff. Depending on the specific machine and the output type a great deal of image processing occurs and that directly impacts what the image looks like when you view the video. When you take the direct data output you are not getting the benefit of all of the image processing that has been done. Truly Video â‰ File for many reasons - in fact they frequently are very different and will look very different. There is another area of difference which David mentions which are the issues that relate to error correction and concealment. Correcting for these errors are fundamental to the design of digital video recorders in general and mini-dv in particular because mini-dv is a very error prone format due to the density of the recording and the media size. Correction and concealment are fundamental to the performance of that format and you easily can get different correction and concealment on every section and on every playback run - that is how error prone the format really is. I am saying that even though it is "digital" when you play back the same exact section of the tape - say only 10 or 20 seconds worth, it is extremely unlikely that you could ever get two passes that were bit for bit identical data wise. These are not clones or anything even close to clones. When you are moving just the data stream you are getting most of the errors (which can and do change) and not getting the benefit of most of the correction circuitry. You may get error flags that are telling you that the machine knows that there is an error, but correcting for that error outside of the machine on the file itself is a non-starter for many reasons including the fact that the exact algorithms that the manufacturer uses in hardware in the deck for correction are not available in software or on any editing system that I have ever heard of (even those made by the same manufacturer). So while the flag telling you that there is an error is nice - you really cant do anything with it and there is little if any benefit to it because it cannot be used to know what the circuitry did to make the picture look correct. It is sort of like lots of metadata - nice but what can you do with it? What does it really tell you? I suppose in aggregate it is useful to know how lossy or bad the tape is, but this really has little to do with the original point which was trying to have a best practice for MiniDv preservation. What you have "preserved" is in fact not what the machine would have reproduced as video at all. It is something different. Frankly I feel the approach offered is too simplistic and is not best practice at all. IF one of the goals of preservation is to preserve the material in such a way that it is at least a fair representation of what the viewer would have seen at the time of playback on the deck, then I am afraid that I would have to argue that in this case a straight data capture that does not include all of the image processing technology used to create the picture is in fact a very poor preservation master precisely because with that information you can not recreate the viewing experience. In fact the best preservation master would be the SDI or HDSI output from the deck because what is coming out is digital video information that has the benefit of all of the image processing as well as data correction that the direct data stream does not have. The SDI stream has "video" and the firewire output has data - and they are far from identical and this can be very easily shown to be the case.
Email: firstname.lastname@example.org Media Matters LLC. 450 West 31st Street 4th Floor New York, N.Y. 10001
eFax (646) 349-4475
Mobile: (917) 945-2662
Media Matters LLC. is a technical consultancy specializing in archival audio and video material. We provide advice and analysis, to media archives that apply the beneficial advances in technology to collection management."
Have you tested these claims?
Capture from one single tape using scene detection, so you have short clips. Do it once with a DV deck, repeat with a DV camera and repeat with a HDV camera set to DV output. Then do a hex or binary file comparison. What are the results?
I just tried it and on three clips with exactly 8,648,736 bytes in each, one captured with a Panasonic deck, one captured from a Sony PD-150 and one from a Sony Z1, the total difference between these files were at most 13 bytes and all at the end. This accounts for maybe a difference of 0.000001 frame and it is not unthinkable that is caused by mechanical differences with the three devices.
Given that short test, I think that your authority is making things up for the data transfer side of matters.
The writer is recognized as one of the leading authories on the subject.
This short test has just proven that this "Leading authority" does not know what he is talking about.
"This short test has just proven that this "Leading authority" does not know what he is talking about.."
Thanks for clearing that up.