Video Converting and Digital Storage

[FastMotion]

This is my first post to this forum. I have been messing with Video recording/ editing/ converting off and on for years, and now I need to convert, condense and archive this and my photos into clean formats and with codec that can be played on most common devices. We have video files from VHS to DVD to DV(tape) to DV(computer) to HD. Much of the old home video stuff, I just want to get it on our NAS device and make sure our family can access it from smart TVs in the home. I have a fair amount of "stuff " to work with, although some software I have yet to learn. Here is what I am working with:

Several Window 7 computers running i7 or i5 processors
Sony and soon samsung smart TVs
Google TV boxes
Panasonic DMR-ES35V (hope it still works)
Another lesser VHS DVD unit
Progressive SCan Sony DVD
SD card video in HD from Sony Handy Cam and Sony EOS ti3
Very old Pinnalce capture card and unit (I need to up grade here for sure)
Synology DS412+ 8 TB Running Disk station with RAID 5

Corel Video Editor
Adobe Premiere


Much Much Much Video

I am looking for the easiest path to getting this done so that my family can enjoy the memories that are currently lost in a maze of files and boxed up tapes. I also want good quality on the converted video, but do not want to work on every one forever, I need to get this done.

Advice, Help, and recommended services graciously appreciated.

Thanks.

[Krispy Kritter]

Most of the current playback hardware will likely support mp4 or mkv. I'd suggest creating a few test files of each with varying quality to see what works and meets your expectations. I use VidCoder, which is just a gui and uses handbrake. It can accept various input file types or DVD file structure.

As for all of your various sources, capture in the highest quality allowed by your capture card/hardware. Keep a copy of that capture for archiving and future use. Then create a copy as noted above for your NAS.

[FastMotion]

dIf I plan to update my interface hardware to a USB device, what do you recommend

[SameSelf]

I am looking for the easiest path to getting this done

Hire it out to someone.

[FastMotion]

Definitely considered that, but I want to make sure I get quality product in return. Any suggestions of a good person or company for this. I want to be able process new video, and master it, but the large amount of video and the experience necessary to navigate the gauntlet of potential problems, I would love to send it out. I just want it to be done correctly.

[_Al_]

If you find someone, it saves you great amount of time, you'd need to include learning process time as well. Doing it yourself, you can edit your videos, videoediting is very time consuming if done right and that brings some time to do it on the top of things, ... hard to give decisive advises, ..., to do it correctly as you say, it takes lots of time, and sometimes compromise is better than nothing ...,

just to sum it up, perhaps that is what you look for:

VHS - find out how to capture video, properly, using VHS with time base correction , TBC, capture codec would be DVavi or some lossless (very large files, you might want to end up using DVavi), then do one of those below:
a) load captured videos into videoeditor (like Sony Vegas platinum), clean it (with NeatVideo plugin, or do not clean it, it is up to you), export to mp4
b) very advanced - clean the captured tapes in Avisynth> making lossless videos (VirtualDub loads avisynth scripts) and edit those in videoeditor, export to mp4
c) capture DVavi or lossless video (large files, but you might not see a much differences at all as oppose to DVavi capture) and make DVD's encoding those to mpg by you or some DVD authoring software does it for you making DVD's
d) capture card captures mpg, then you author DVD's - the most easy workflow


DVD -
a) you can just extract particular titles out of DVD into Mpg using Vob2mpg for example, this is done without re-compression, which is something you basically want to do for home made videos. With another possible encoding to get mp4 you'd loose some quality. Those are not commercial videos where there is a room for another encoding .

b) can use Vidcoder to encode your titles to mp4 (make video smaler, but you worsen your DVD quality, as I said, home videos are interlaced and not that perfect as commercial DVD's

c) or you just leave it and just copy DVD's VIDEO_TS folder and store it, media players play those, mostly understand DVD structure, if not, it is usually big corp peace of hardware that does not want you to play VIDEO_TS structure

DVavi -
a) use videoeditor ( like Vegas Platinum or Adobe Premiere, or others that can export DV avi losslessly) , it is quite easy, load DV avi clips, join them, edit DV video and export it again as DV avi, this will be without re-compression. That is going to be your archive. This video can be played just by computers or HTPC's , for media players you must encode it into mp4, you can use vidcoder also or directly export mp4 from that videoeditor right away

b)same as a) but you can treat that DV archive export (lossless export), denoise it etc., using NeatVideo or Avisynth (advanced) and encode it into mp4


HD video - similar, load videos into videoeditor , edit clips, export to MP4

Huge task for beginner. Small steps doing one thing at a time if doing yourself.

[PuzZLeR]

You have quite the adventure in front of you.

I've been there too (still am), so here's some advice before anything.

VHS:
Formats change all the time, MPEG-2/DvD, H.264 MP4/MKV, DivX, Xvid, etc. My advice is to capture, and archive, your VHS tapes to a lossless format first (ex:HuffYUV). Yes, it's much larger in file size, but so is HDD space today, and after all, we may be talking about precious stuff here. But the advantages of lossless are worth it - easy to work with, highest quality, flexible if trying several filtering options, etc. As well, it's always easy to convert to the playback of choice. And keep the lossless as your Master Copy always - especially if you're getting rid of the tapes.

Capture Devices for VHS:
All you need is a good USB stick. Some good ones: Hauppauge USB-Live2, ezcap.tv (the REAL one from the site just mentioned), StarTech SVID2USB2. Another good one is the ATI 600 USB, but it's not made anymore and may be tricky on modern systems.

DV:
You won't need lossless here, or any of the above capture devices. Just get a firewire, use an app like WinDV, and copy it to your HDD in DV format. This is less than half the size of lossless, and all the full quality bit-for-bit. Again, for similar reasons mentioned before, keep the DV as your Master Copy.

Photos:
You mentioned photos? Use the TIFF file format for a similar reason to using lossless for VHS. You can always convert this to JPG for any device when needed. If looking for a great scanner, I recommend the Epson V600. IMO, this is the best scanner that is below the ultra-expensive professional models. It can do source film too. If it's alot of paper documents, just get a feed scanner. (But they are not cheap.) I personally love my Fujitsu ScanSnap S1500. It can do photos very well too if ~95% of the quality of a good flatbed is good enough for you, but it will do them super fast.

Capture Houses:
As for outsourcing this project, be careful. For video transfers, some of these so-called professionals use basic consumer equipment, like to a DVR, and do it cheaply - you can do it yourself at a savings and with better quality. I'm not saying they're scammers since they do provide a service to a certain group that really wouldn't be able to do it on their own, but may not be for you. As for the higher end professionals - expect a bigger price tag.

If it's DV, then you'd be throwing money away with a capture house when it's dirt cheap, and even more simple, to do on your own.

Some will do photo transfers, but all they do is use a scanner. I still say it's a waste of money here too, even if they have ultra professional scanners that will give you that < 0.5% extra quality, big deal.

[SameSelf]

Let's not forget: VHS video is ugly, interlaced, and all that and needs to be de-interlaced with a quality algorithm like QTGMC for proper display on modern, native progressive viewing doohickeys. https://www.youtube.com/watch?v=EFEmTsfFL5A

[PuzZLeR]

Maybe this is just me, but honestly, I find de-interlacing ugly. Not only does it take (much, much) longer to process, but over the years I don't think I've ever enjoyed a result fully, even with the best of tools and slowest of settings.

Yes I get good compression, and yes I get rid of the ugly raking for a given display, but the video still looks beaten up regardless for these advantages. After all, we get rid of half the chroma/luma to accomplish this, so there are disadvantages.

Today I just don't bother, especially when using interlace-aware playback (ex: DvD, BD, etc). And if I really need to de-interlace anything now, I just de-interlace at the field level (double the fps). Yes, bigger file size, and the motion interpolation gives it a "soap opera effect", but it looks so much slicker and life-like.

At any rate, whatever the O/P decides to do is fine if he/she enjoys the result - but I still remain with keeping the source, AND I will also add to keeping it interlaced (for VHS).

[SameSelf]

That's all well and good. But my experience in side-by-side comparisons is: when given the choice of letting a native progressive display de-interlace on the fly in real-time (using whatever black box algo it has) and me doing it myself using QTGMC, doing it myself always gives better results. And I find it is not terribly slow on my machine when using the MT version. But then again, between denoising, color and gamma correction, and everything else I do on VHS quality footage, the de-interlacing is certainly not the slowest step in my workflow. And yes, I also double the frame rate to maintain the temporal resolution. Plus, de-interlacing can't be avoided when de-noising.

[PuzZLeR]

That's all well and good. But my experience in side-by-side comparisons is: when given the choice of letting a native progressive display de-interlace on the fly in real-time (using whatever black box algo it has) and me doing it myself using QTGMC, doing it myself always gives better results.

With some monitors, or playback software that de-interlaces, yeah, it's better with QTGMC. But that doesn't mean I like the result. Just choosing among two evils.

On DvD? I prefer it without any de-interlacing. But that's just me.

And I find it is not terribly slow on my machine when using the MT version. But then again, between denoising, color and gamma correction, and everything else I do on VHS quality footage, the de-interlacing is certainly not the slowest step in my workflow.

Yes, a recent i7 machine makes a world of difference here on my end, and maybe even the QTGMC algorithm has improved in quality lately too. Honestly, it's been a couple of years since I've given up on it, but I should re-visit it again. However, I expect speed only will improve, and still have low expectations as per the quality.

I still say anything less than an i5 would warrant many, many overnights using the better settings of QTGMC.

And yes, I also double the frame rate to maintain the temporal resolution.

Yes indeed, this makes a huge difference. I was actually referring to a bigger quality drop when de-interlacing interlaced 29.97fps down to progressive 29.97fps. I'm sure you'd agree somewhat here.

Plus, de-interlacing can't be avoided when de-noising.

I agree partially. When denoising, I wouldn't say "de-interlacing can't be avoided". I would say "working at the field level shouldn't be avoided". You could still end up with denoised interlaced content at high quality if done properly.

For example, NeatVideo gives you the option of working via field. Also, RemoveSpotsMC3x(), as a personal favorite of mine from here on VideoHelp, needs special care when working with interlaced content. In other words, when applied, you should never do it at the frame level otherwise you'd mess with the interlacing and have problematic motion. Yes, this would be disasterous, but applying the denoiser with an interlace-aware method avoids this tragedy, and still keeps the interlacing:

SeparateFields()
Ev=SelectEven().RemoveSpotsMC3x()
Od=SelectOdd().RemoveSpotsMC3x()
Interleave(Ev,Od)
Weave()

I'm sure this would apply similarly to other denoisers if wishing to keep the result interlaced.

[hello_hello]

PuzZLeR,
Out of curiosity.... there's some small sample encodes here along with the original video. If you play the original and compare it to QTGMC at 50fps, how does it look? To me, most hardware/player de-interlacing is comparable to Yadif at 50fps (there's a sample of that in the zip file too) and QTGMC looks much better than both. The tray on the desk does a bit of shimmering, as do the blinds over the corner window as the camera pans. QTGMC takes care of most of it, but I'd be interested to know how that compares to what you see when the original's being de-interlaced on playback.

[PhilNorman]

xxxxxxxxxxx

[PuzZLeR]

PuzZLeR,
Out of curiosity.... there's some small sample encodes here along with the original video. If you play the original and compare it to QTGMC at 50fps, how does it look? To me, most hardware/player de-interlacing is comparable to Yadif at 50fps (there's a sample of that in the zip file too) and QTGMC looks much better than both. The tray on the desk does a bit of shimmering, as do the blinds over the corner window as the camera pans. QTGMC takes care of most of it, but I'd be interested to know how that compares to what you see when the original's being de-interlaced on playback.

Yes, nice demonstraton. Thanks.

I like the 50fps QTGMC result best, even better than playing the interlaced original on a de-interlacing display. It still looked a tiny bit beaten up to the discriminating eye, but it by no means wasn't unwatchable. Great result nevertheless.

However the de-interlacing display pretty much is indeed comparible, or maybe slightly beats, the 25fps QTGMC, and the 50fps Yadif result, in my eyes. But this is arguable. But I do see what you mean when you compare 50fps Yadif similarly to a de-interlacing display.

The 25fps Yadif was just brutal to watch, and it was no contest against my de-interlacing display. This is prettty much why I've given up on de-interlacing at the original fps, and with older de-interlacers. I remember when Yadif was new some 8-10 years ago, and was supposed to be the best ever, and it was then, which tells you how much more ghastly previous methods were.

Please keep in mind folks - I never said that using a good deinterlacer, and doubling the frame rate, was bad - I actually recommended it if you wish to de-interlace. But, honestly, I've never gotten a satisfactory result deinterlacing to the original fps.

I've been using more DvD, and blu-ray playback lately, so I've laid off de-interlacing in recent years. However, now with an i7 for faster processing of QTGMC, and a bigger interest in newer playback methods, able to support higher fps, higher level AVC encodes, etc, I am re-visiting this again. Your 50fps QTGMC example should be encouraging to many.