Capturing DV-AVI via Vegas

mdindestin wrote on 11/23/2014, 2:14 PM
We're digitizing hours of SD 8mm and Hi8 camcorder footage with the goal being top quality archival. File size is not an issue.

For hardware, we're capturing using firewire from the desktop's generic card to a Sony TRV310 camcorder. For software, we're using the capture app in SVP 11.

I have Cineform HD Link and was hoping to use it, but the Cineform capture utility does not recognize SD sources. Perhaps I can convert afterwards, we'll see.

I read where the Grass Valley ADVC-110 captures in 4:1:1 color space (NTSC) while the Black Magic Intensity card will capture at data rates slightly higher than DV at 4:2:2. Moreover, the Cineform HDLink does recognize the stream.

So what color space am I capturing in using the Sony Cam and firewire and should I do anything different?

Comments

videoITguy wrote on 11/23/2014, 2:42 PM
If you are using SD DV source for the capture - you cannot exceed the 4.2.0 native of the file codec. Yes, you could use BlackMagic but it will not offer you anything more than you have got. The most efficient will be the DV capture over standard firewire.

This produces very good files that can be used almost anywhere and is suitable for an archival format on media such as optical disc. You could argue for the merits of transcoding with Cineform to use as great digital intermediate for later use from an archive, but it accomplishes nothing for use with a standard DV file.
mdindestin wrote on 11/23/2014, 5:09 PM
OK great, thanks. We're on the right track.
PeterDuke wrote on 11/23/2014, 5:45 PM
So far as the DV camera is concerned, NTSC uses 4:1:1 subsampling, whereas PAL uses 4:2:0.
mdindestin wrote on 11/23/2014, 8:00 PM
I'm just archiving for now, but one of my concerns was the following statement on dvinfo concerning post processing the DV-AVI captured at less than the Black Magic card can do:

"If you then convert this (4:1:1 color space) to DVD, which is a 4:2:0 format, then you will end up at 4:1:0, which will only contain 12.5% of your original color information."

Not sure if the above is an accurate statement.
PeterDuke wrote on 11/23/2014, 8:15 PM
People have been converting 4:1:1 DV from NTSC cameras to 4:2:0 MPEG2 for DVD for many years without problems. The 12.5% statement is rubbish.

We are talking about chroma subsampling, not colour space. It defines how sharp the colour is in each axis. Going from 4:1:1 to 4:2:0 means that the vertical resolution will be halved, but the horizonal resolution will not be improved, despite the capability.
videoITguy wrote on 11/23/2014, 8:50 PM
The statement is rubbish YES!!, but even more important your source of DV mastered source to a final of SD DVD will have a lot more issues going into the process than the colorspace. In fact colorspace is the absolute least concern in the real world.

Further, if you decide to release an SD from DV source archive later to Blu-ray production (given certain workflows) you will be just fine.
Tim L wrote on 11/23/2014, 9:21 PM
Just as a reminder, "SD 8mm and Hi8" are analog tape formats. The TRV-310 is a Digital8 camcorder that can play back 8mm and Hi8 tapes and convert on the fly to DV output on firewire.

I don't know enough about analog formats to say whether capturing analog output to Black Magic would offer higher quality or not. I would [I]guess[/I] not, as I would also guess that the analog formats were lower quality than DV to start with. But that's just me guessing -- maybe someone else can speak more authoritatively on the subject.
johnmeyer wrote on 11/23/2014, 9:52 PM
" .... even more important your source of DV mastered source to a final of SD DVD will have a lot more issues going into the process than the colorspace. "I'm not sure to what you are referring. What issues??

DVD is a great final format for VHS, SVHS, 8mm, and Hi8 analog tapes that have been digitized. I recommend, without reservations, that you put your results on DVD.

I have digitized and transferred many thousands of hours from these old consumer analog formats to DVD, using essentially the same workflow you describe. In theory, you can get more color by using an analog capture card, but I had so many problems with these, that I eventually gave up and instead settled on using a variation of what you are doing, namely digitizing with a device that takes analog input, and then outputs digital video via Firewire. The process is very friendly, and once you get it set up, it is almost foolproof (OK, I shouldn't say that because it's asking for problems ...).

The biggest issue in any analog transfer is not the colorspace, and not whether you put the final result on hard drive, DVD, Blu-Ray, thumb drive, or floppy disk, but whether you use a decent deck to play the tapes. This whole process has been covered a hundred times in this forum over the past dozen years, and the deck is everything.

So, make darn sure you are using a quality deck, hopefully one which includes some sort of time base correction (TBC). Put the video on DVD, using a bitrate of at least 6,000,000 bps (i.e., don't put much more than ninety minutes on a DVD).

You can put the result on Blu-Ray, but I don't see any advantages at all to doing so. You certainly do not want to "up-res" to some higher resolution, because that is a waste of time. Blu-Ray is no more likely to last a long time than DVD, and in fact I can show you numerous, scientific, aging tests on DVD that show it to be an excellent medium for long-term storage. In fact, I am not aware of any medium that is better. What's more, because DVD has been the most successful consumer video format in history (based on installed base) and continues to outsell all other formats, it is far more likely that people will have equipment in the distant future which can retrieve the data from these discs.

For information on Blu-Ray vs. DVD sales, see my post in this thread:

How are Blurays going where you are?

For information on archival aging tests of DVD, as well as an attempt to find similar information on Blu-Ray, see my post in this thread:

Recommendations for archival quality Blu-ray media
John_Cline wrote on 11/24/2014, 3:32 AM
I do a LOT of capturing of analog footage (1-inch, BetaCam, 3/4" and VHS/Betamax) for high quality digital archiving and this footage also ends up on DVD. I can often see the difference between video captured at 4:2:2 and 4:1:1, particularly on near-vertical, high-contrast edges. Below is a 72-pixel by 48-pixel crop of a couple of red LEDs sampled at 4:1:1 that clearly shows the limited horizontal sampling of the color: once every four luminance samples. The color sample is co-sited (located at the same place) as the first brightness sample in each 4-pixel group, and the color at that location is simply replicated over the following three pixels.



Of course, 8mm and Hi8 is of pretty limited color resolution anyway, but the original poster asked about the highest quality capture for archival purposes and stated that disc space is not an issue. That would be 4:2:2 capture using something as inexpensive as a BlackMagic Intensity card.

When starting with 4:4:4 full luminance and chrominance, for example, text and generated media in Vegas, and then rendering to 4:1:1 DV and then to 4:2:0 MPEG2, you do actually end up with 12.5% of the original color information.
videoITguy wrote on 11/24/2014, 5:18 AM
Taking issue with last two postings in general. I have access to high-end analog capture that cost me an arm and a leg years ago and is not made anymore, as well as the Black Magic Intensity card. I grant that a good deck transport is an issue in these cases, but given that it works well, the real outcome of any capture is not bested by any value of the colorspace manipulation. In other words, a good Firewire capture will be every bit as usable as the analog pipeline - while indeed their might be theoretical value to merit - the outcome in the real world is not that different.

Then that goes back to the issue of what I was bringing into play - namely other things like deck transport, head-tracking etc. These are very common problems for media and hardware of some age in the tape world.

Last, I never suggest up rezzing SD to some other format. You can include SD productions within a Blu-ray authored disc and not significantly uprez. by custom framing. This keeps color quality and the advantage of using a great media for distribution. Penney for penney spent the Blu-ray is the better archival media over DVD.
mdindestin wrote on 11/24/2014, 10:10 AM
I find myself saying it more and more: Nothing is easy.
johnmeyer wrote on 11/24/2014, 11:01 AM
+1 to everything John Cline posted.

I agree with him that you get better results using an analog capture card instead of a DV/Firewire transfer, and indeed that is what I tried to use for many years. However, for consumer analog video formats, the incremental improvement was (and is) quite small -- barely noticeable for most captures -- but the increased difficulty of getting a good capture far outweighs these benefits.

Granted, my capture back then was with a model year 2002 ATI Radeon video/capture card combo, and the drivers for that had huge problems keeping audio and video in sync, not to mention many other issues. The Blackmagic card is probably a lot more robust.

However, as I have pointed out to dozens of clients, as well as to countless strangers in various forums, the most important thing in archiving old videotape is to get the job done. It is a BIG undertaking, and I can't tell you how many times I've seen people stall out before they manage to get through their pile of tapes. In fact, for most people, I tell them to get a VCR/DVD burner combo, and just use that. The results are pretty decent, and the workflow is a cinch.

All the quality in the world doesn't matter if you don't get the job finished.
mdindestin wrote on 11/24/2014, 11:29 AM
I can appreciate what you say about completing the task.

I completed my personal project of family videos on VHS, S-VHS, 8 & Hi8 a few years ago. It's all neatly organized by year on master and back-up 4TB hard drives.

When VCRs were on their way out, we purchased a new top-notch JVC S-VHS unit with TBC for pennies on the dollar.

This current project is for a friend. The tapes are captured on one computer while working on another. I can do three or four tapes per night.
But on this project, because the job is so large, I didn't want any regrets.

Handheld SD footage from 8mm camcorders operated by zoom happy people under poor lighting is so horrible anyway, we were hoping to do it once and do it right and squeeze all we could out of it.

Incidentally, I've never had audio sync issues and rarely have dropped frames. However, I often have WMP stop with an error message when attempting to play back the AVI files. VLC Player might fix that though.
johnmeyer wrote on 11/24/2014, 2:18 PM
One other thing to consider is "restoration." No matter how good your deck and capture card, you are still dealing with crummy consumer analog video. It will be full of both luma and chroma noise, and will look very fuzzy, especially now that everyone is accustomed to HD.

Using AVISynth, I can make dramatic improvements in the quality of the captured video. While there is no such thing as a "one size fits all" restoration script, I get pretty good results with the one posted below. I simply run this on the captured video, before any editing, and use the "restored" version as the input to my editing. For big projects, my editing only consists of removing blank spots and other bad video, as well as trimming head & tail.

I can provide before/after comparisons, if you are interested.

#Denoiser script for interlaced video using MDegrain2

SetMemoryMax(768)

Loadplugin("C:\Program Files\AviSynth 2.5\plugins\MVTools\mvtools2.dll")
LoadPlugin("c:\Program Files\AviSynth 2.5\plugins\CNR\Cnr2.dll")
Import("C:\Program Files\AviSynth 2.5\plugins\Film Restoration\Script_and_Plugins\LimitedSharpenFaster.avs")

SetMTMode(5,6)
#Modify this line to point to your video file
source=AVISource("E:\fs.avi").killaudio().AssumeBFF()
SetMTMode(2)

#Only use chroma restoration for analog source material
chroma=source.Cnr2("oxx",8,16,191,100,255,32,255,false) #VHS
#chroma=source.Cnr2("oxx",8,14,191,75,255,20,255,false) #Laserdisc

#Set overlap in line below to 0, 2, 4, 8. Higher number=better, but slower
#For VHS, 4,0 seems to work better than 8,2. Most of difference is in shadows
output=MDegrain2i2(chroma,8,0,0)
#output=MDegrain2i2(chroma,8,4,0) #Retains a little more detail, but slower
#output=MDegrain2i2(chroma,4,2,0) #Better for retaining detail on flat surfaces, but slower still

stackhorizontal(source,output)

#return output

#-------------------------------

function MDegrain2i2(clip source, int "blksize", int "overlap", int "dct")
{
Vshift=0 # 2 lines per bobbed-field per tape generation (PAL); original=2; copy=4 etc
Hshift=0 # determine experimentally
overlap=default(overlap,0) # overlap value (0 to 4 for blksize=8)
dct=default(dct,0) # use dct=1 for clip with light flicker

fields=source.SeparateFields() # separate by fields

#This optional line gets rid of chroma halo, a problem with 2nd gen tapes
#fields=MergeChroma(fields,crop(fields,Hshift,Vshift,0,0).addborders(0,0,Hshift,Vshift))

#This line will shift chroma down and to the right instead of up and to the left (an uncommon problem)
#fields=MergeChroma(fields,Crop(AddBorders(fields,Hshift,Vshift,0,0),0,0,-Hshift,-Vshift))

super = fields.MSuper(pel=2, sharp=1)
backward_vec2 = super.MAnalyse(isb = true, delta = 2, blksize=blksize, overlap=overlap, dct=dct)
forward_vec2 = super.MAnalyse(isb = false, delta = 2, blksize=blksize, overlap=overlap, dct=dct)
backward_vec4 = super.MAnalyse(isb = true, delta = 4, blksize=blksize, overlap=overlap, dct=dct)
forward_vec4 = super.MAnalyse(isb = false, delta = 4, blksize=blksize, overlap=overlap, dct=dct)

#Increase thSAD for more degraining
MDegrain2(fields,super, backward_vec2,forward_vec2,backward_vec4,forward_vec4,thSAD=400)

#Optional sharpening steps
#unsharpmask(50,3,0)
#sharpen(0.1)

Weave()
}


mdindestin wrote on 11/24/2014, 4:11 PM
I'm very interested.
johnmeyer wrote on 11/24/2014, 4:39 PM
Here's a "before/after" clip, using some really, really bad VHS LP (6-hour mode) video that was captured without a TBC. It shows the kind of improvement you can expect from the script shown above.



I up-res'd this to HD in order to keep YouTube from "fuzzing up" the noise in the before clip. Therefore, to really see the difference, you should play it on YouTube, and then make sure the resolution is set to 1080p.


johnmeyer wrote on 11/24/2014, 4:52 PM
P.S. If you want to upload 10-15 seconds of DV video from one of your captures, I'd be happy to process that and upload the result to Megaupload or Dropbox.
mdindestin wrote on 11/24/2014, 7:05 PM
Looks a lot better. I had to laugh, ours has that same video look.

I'll upload a sample when I get past this deadline that's got me racing against the clock tonight.
johnmeyer wrote on 11/24/2014, 10:16 PM
BTW, here's an even better example, although it was done on HD video. The script is somewhat similar. The person had an underexposed clip. He was having trouble making it brighter because he didn't use color curves. Even with the method he used, the noise in the video after brightening was awful.

This clip shows the original video side-by-side with my brightened version. Then, the second part of the clip shows the brightened version side-by-side with the same bright version, but with the noise removed. As before, the effect is best if viewed full screen in full HD.