[Bob Linsdell]

Please let us know how the double sampling works out.

The most significant improvement I made to my VHS captures was to use the AVT-8710 Time Base Corrector. It cost me $230US new. It reduced the video noise across the frame, reduced jitter, reduced dropped frames to zero, and improved the sharpness a little.

Example

Direct from VHS player:

Using Time Base Corrector between VHS player and PC:

After filtering Time Base corrected capture (the zoom/cropping is mostly as a result of the deshaker filter):

[Bob Linsdell]

The ‘Chroma Noise Reduction’ filter (see my above post) may well remove much of the noise from video in the original post of this thread.

[Bob Linsdell]

Deshaker was used, so two passes were needed.

PASS 1
The Filters

Shift Filter: Set to match the shift needed in Pass 2. (For interlaced video ‘X’ must be an even number else the field order gets messed up.)

Deshaker Filter – Pass 1

PASS 2
The filters:

VHS Filter: Used to bring colors into alignment.

Chroma Noise Reduction: Removed typical VHS color noise.

Shift Filter: Centers image within screen area.

Hue/Saturation/Intensity Filter: Some manual color balance adjustments.

Brightness/Contrast Filter: Manual contrast adjustment.

Neat Video Filter. The version I used couldn’t handle interlaced video so the ViewFields/UnViewFields filters were used to get around the problem. Later versions can handle interlaced video. Extensive testing found that Neat Video produced better results than any other noise filter or combination of filters. Note that I didn’t remove all the noise to retain as much detail as possible, and I also did some sharpening in Neat Video.

Deshaker Filter – Pass 2: The Extra Zoom Factor setting removes the black boarder from the video capture as well as hides borders resulting from deshaking.

ACOBW Filter: Auto white balance adjustment.

That’s it.

The video source was very poor having been recorded in February 1993, and been copied from the source VHSC tape to VHS. Overall, I’m happy with the results, which do look better when the video is playing vs. a still image.

I didn’t used any compression for the audio. Compression for the video was DV for standard definition video, and x264 for deshaked HD video.

[Bob Linsdell]

I used VirtualDub and various filters to achieve the below:

SAMPLE 1

Unfiltered from VHS tape:

Filtered in VirtualDub:

SAMPLE 2

Unfiltered from VHS tape:

Filtered in VirtualDub:

—

[Bob Linsdell]

I think your decision is good, a fair compromise, and should meet your requirements. I’d just check that it streams smoothly with a large file.

Regarding choice of resolution (720 vs. 1080). The decision was made for me by the Blu-ray standards, see below.

Resolution—Aspect/fps—-Type
1280 × 720—16×9 23.976—Progressive
1280 × 720—16×9 50.00—-Progressive
1280 × 720—16×9 59.94—-Progressive
1920 × 1080–16×9 29.97—-Interlaced
1920 × 1080–16×9 25.00—-Interlaced
1920 × 1080–16×9 23.976—Progressive

From the table you can see that I had to use 1920×1080 for my interlaced video. The standards just don’t allow 1280×720 interlaced – very odd I thought. Had Blu-ray allowed it, I would have gone with 1280×720 interlaced for my converted SD video.

I use 1280×720 at 59.94fps progressive for digital camera video (1280×720 30fps progressive). I had to create a 1280×720 Vegas template from one of the existing 1920×1080 templates.

Side note: To de-interlace or not?
I had done a number of tests converting my interlaced video to progressive, including converting the interlaced fields to progressive frames (doubling the frame rate), but overall I felt I wasn’t gaining much, if anything. I found that my equipment does a good job of presenting the interlaced video on our HD screens.

[Bob Linsdell]

“I need to determine which is better, the Sony or the MainConcept codec”

If the MainConcept is Mpeg-2, then for the same size file as an AVC+audio, the AVC gives better quality. I read that Mpeg-2 has to be 20+ Mbps to equal the quality of a 16Mbps AVC. Having said that, this was probably realized when the source was true HD.

If you do use separate video and audio files, and then use them in a Blu-ray project, they will be combined into one file when written to disc. The combined file can be copied from the disc and renamed as required. Such files are saved on the disc in the BDMVSTREAM folder, and have the extension ‘m2ts’. My media box plays the m2ts files; they don’t however contain the subtitles. FYI: I use subtitles to provide information e.g. who, what, where, when.

Please let us know what format you decide to use.

[Bob Linsdell]

Here’s the solution that worked for me:

Background: I have 16 hours of home video; comprised of 213 edited movies, most of which is standard definition captured as DV from tape (PAL & NTSC). The video was cropped, filtered, color corrected and deshaked using VirtualDub. Editing and final color correction was done in Sony Vegas. The project setting used was the final HD output that I desired.

I’d previously created DVD’s. They were okay, but I could see some loss in quality compared to the DV version. When Blu-ray became affordable and I’d done some tests, I decided that this was the way to go. I’ve now rendered all my video from Vegas to AVC video and PCM audio. I wanted the best quality possibly and damn the space; I used the following:

Video – Sony AVC, Blu-ray 1920×1080-60i, 16Mbps (for PAL use 50i)
Audio – Wave, 48,000 Hz Bit, Stereo, PCM

Some may say this is overkill, but I have piece of mind that it’s as good as I can get it.

My 16 hours of video uses 111GB of space.

Around our home we have Western Digital TV Live Plus media players. Because AVC files don’t contain audio, I use mkvmerge with the ‘mkvmerge GUI’, to combine the Video, Audio, and subtitles converted from DVDA, into one file that streams perfectly to our HD TVs.

Note: Although I had captured PAL video, it was rendered from Vegas to 60i as I live in NTSC land now.

[Bob Linsdell]

“Why are you putting this many video files into DVDA? What is your workflow and what are you trying to do?”

The finished Blu-ray disc will be a video family tree. It comprises 80 video clips from several branches of the family. I have one playlist for ‘All’ video, this plays all the clips in date order. I have playlists for each branch of the tree. I guess the best way to imagine what I’m doing is buttons arranged in a typical family tree structure. When a button is clicked, it just plays the video relevant to that part of the tree. I also have playlists for special events e.g. weddings, holidays, etc.

Occasionally I’m asked to add another clip, and it’s a lot simpler to add a new clip to playlists, than it is to insert it into a large compilation. I also use subtitles to identify the event, the place and key people. These would have to all be moved if I was working with one big file; not so with discrete clips.

Since you’re running XP, my first offhand guess would be that you are running out of RAM.

I’m monitoring physical and virtual RAM and I’m not seeing it max out. But, I’m not ruling out that it is a memory issue.

Of course, “crash” is used to describe some thousands or so conditions. How is the “crash” presented?

The first sign of a problem is the background image goes black, then some of the button images disappear, and then it starts giving me an error message telling me that the media I’m attempting to load is of the wrong type; but that’s not true. At that point I have to exit the program.

[Bob Linsdell]

I’ve just read what I wrote and need to clarify what I said in the following paragraph:

“As for fixing the scene markers, you need to ensure you have the same number in the re-edited movie as you had in the original, and in the same order, else you have to manually select them in DVDA. FYI: You can have markers for which you don’t have a scene button. I use them all the time to allow skipping through long chapters.”

When I said you need to ensure you have the same number, I mean that you have to have the same number of markers (not makers that are numbered the same as they were in the original movie). Also, when I said they have to be in the same order, what I mean is that the scenes need to be in the same order (not the numbers on the markers have to be in the same order as they were in the original movie).

[Bob Linsdell]

If you had used the original video for the scene button images, they will be messed up from the point you edited the movie forward.

If you have the original Vegas project file, create a very low resolution movie file, drop this in your DVDA project folder and then for each scene button change the media to the low resolution video file – you will now how the original image back.

For the low res movie I use one of the Memory Stick templates of Sony AVC. To save space you can disable the sound.

If you have lost the original Vegas project file, but you have the original finished movie file, you can re-render this to create the lower res movie.

As for fixing the scene markers, you need to ensure you have the same number in the re-edited movie as you had in the original, and in the same order, else you have to manually select them in DVDA. FYI: You can have markers for which you don’t have a scene button. I use them all the time to allow skipping through long chapters.