Workflow advice for VHS home movies?
 

Got a loads of home movies from mine and my partners family. Captured a lot of videos already using crap equipment then I found this forum and now trying to do things the right way or atleast way better! You really do learn something new everyday on this forum.

Goal - Capture home movies -> restore best I can -> encode in an accessible format -> deliver videos to family members through USB stick/external HDD/something physical that anyone can plug in and watch (open for suggestions)

Sub Goal - archive videos for potential safety backup if I ever want to redo anything

Capture setup: JVC HR-S7900 -> Cypress AVT-8120 -> ATI 600 USB clone -> VirtualDub w/ Huffyuv

1. Capture - JVC settings are Video Cal - Off, Picture Control - Normal, R3 - off, Video Stabilizer -off
2. Deinterlace with qgtmc - I have a simple script and I'm slowly experimenting and learning the ins and outs.
3. VirtualDub filters and masking - CCDenoise and Chroma Noise Reduction are the main filters I go to. I mask using the stickied masking/cropping post. As well as trimming video where needed. Save as AVI
4. Open in Adobe Premiere - Export audio as a WAV and color correct. Already pretty comfortable with Premiere's color correcting tools.
5. Cleanup audio with Izotope RX7 - Already have the program and am fairly versed with it. Export audio back out as WAV
6. Export from Premiere. Line up new audio. Export as H.264

Still unsure on the best delivery method for all these videos for family members when all is said and done. Or the best way to keep these files around for safe keeping. Anything better or more efficient than just keeping all the raw captured avi's?

Any critique or suggestions welcome. Really enjoying jumping into this hobby.
Sorry if this post is messy!

- Hardware is fine.
- Goal is attainable.
- Family probably needs deinterlaced 59.94fps SD 640x480 MP4 on HDDs.
- Keep lossless for yourself on large HDD, then back it up to another HDD.

JVC settings fine.

QTGMC best.
Are you using Hybrid? AvsPmod+VirtualDub? Writing your own scripts?

VirtualDub masking.
Remember to set VirtualDub in/out as YUY2, not RGB. (Video > Color Depth or Decode Format)

VirtualDub ColorMill is pretty good for color correction.
Premiere is more Photoshop-like.
I use both. I try ColorMill first.

My experience with Izotope has never been great, it's always been too weak for heavy restoration. But it may be fie. I have RX7 demo now, not used it much, but I can already tell it's no Sound Forge. But for minor things, like removing hiss, it seems fine.

Premiere's export for H.264 has always been weak. It's the SDK version of MainConcept. Hybrid would be my choice for the x264 encoding, as it has all the quality options. Premiere has always been best as only the editor (and same for FCP, Vegas, etc). Lossless in, lossless out.

You seem to be on the right path. Your method will work, but a few tweaks would make it better. :2cents:

- Why 50 fps? Don't ever see that number much for fps mentioned.

- Using avspmod for script. Found some examples on the forum and avisynth wiki to get me started while I figure it out

- I'll checkout ColorMill , Soundforge, and Hybrid. No reason not to if I wanna get the best results.

- I assume I should use Hybrid for making the final Mp4's?

Thank you!

Edited my post, meant 59.94fps (29.97 deinterlaced) NTSC, 50 is PAL (25).

I use an older Sound Forge, pre-Magix. I think it's changed a lot in recent versions, potentially neutered some, at least for the cheaper $$ versions.

Should I be capturing in 640x480? Or resizing in Hybrid?

Edit: After doing some searching on the forums and reading about square and rectangle pixles the consensus is capture at 720x480 then encode to 640.

Capture 720, resize later.

Ideally you would crop off the sides of the 720 pixels to 704 before you do any resizing (by D1 standard active video area is around 704 pixels only), it is the only way to get a perfect aspect ratio of the original contents, Your capture card (any capture card for that matter) captures an extra 16 pixels to the horizontal resolution during capturing (8 on each side) to make sure the active video area is safe and not being cut off, Those 16 pixels will change the aspect ratio slightly, so to get an accurate 4:3 you can either set the pixel aspect ratio flag of the 704x480 to 10:11 for NTSC and the 704x576 to 12:11 for PAL, Or resize 704x480 to 640x480 and 704x576 to 768x576 for a square pixel mode.

I've already proven the aspect ratio accuracy of the 704 with an experiment over at VH.