VideoHelp Forum
+ Reply to Thread
Results 1 to 8 of 8
Thread
  1. Member
    Join Date
    Mar 2005
    Location
    United States
    Search Comp PM
    Long time poster/lurker here but it's been a while since I was last year.

    I have a ton of VHS, SVHS, Hi8 and MiniDV content that's been captured. The captures are all top quality for the analog stuff... AG1980p/W5U/9911U source decks on the VHS/SVHS side, EV-S7000/TRV-D820 for the Hi8 stuff, and everything in Huffyuv except for the MiniDV stuff and the TRV-D820 Hi8 stuff, which is obviously DV-AVI.

    I'm keeping the originals, with backups of it, which amounts to a good number of terabytes.

    I've started to use cloud based backups for some off-site safety, and want to start getting this video into that regimen. Unfortunately the data volume is obviously too high. Since I am keeping my original tapes + original digital "lossless" captures anyway and am looking for a safety net in case something happens to them, I would like to convert them to a format that's much easier on disk space (for the sole purpose of sync to cloud), but gives me something that doesn't give up much detail/information should I lose the originals. Encoding time is not an issue, only disk space and quality.

    All of the content except for the MiniDV is interlaced.

    I'm a bit behind on the current state of the art; can anyone recommend a strategy for someone who is keeping the originals, but needs a smaller copy for offsite archival storage? For these purposes,
    - should I de-interlace the interlaced stuff?
    - if I go with something like h264 over handbrake in a MKV container, is there a constant quality setting that I should use for VHS vs. SVHS vs. Hi8 vs. MiniDV, so that I can keep everything the same except vary the quality setting based upon the amount of source content actually captured?

    Any feedback would be appreciated.
    Quote Quote  
  2. Member
    Join Date
    Mar 2005
    Location
    United States
    Search Comp PM
    Related question... a thought exercise: Let's say I took some source material captured to a lossless format. Say, VHS.

    If I were to encode it to h264 using CQ=30, then use a script to compare frame by frame the "difference" between the original video and the decoded re-encode, sum up the cumulative measure of differences across all the frames, I might get a number that tells me how different my encode is. Then if I were to repeat this exercise for CQ=50 to CQ=10 in increments of 5, I might get some kind of curve that would help me identify the "sweet spot" for the particular kind of video and source material I was encoding.

    Has anyone done something like this for modern codecs like h264 to determine their sweet spot for storing various older analog video formats?
    Quote Quote  
  3. Member yoda313's Avatar
    Join Date
    Jun 2004
    Location
    The Animus
    Search Comp PM
    Just a thought here on the extra backups that you are trying to ensure survive:

    how about burning the original files to bluray?

    Since you are already keeping the originals themselves the odds of having a bad bluray burn won't be that problematic. Also you could more easily store blurays.

    You might also consider investing in a bank safe deposit box and storing said blurays in that box. Also if its big enough you could stash an extra harddrive in there for safety.

    That seems a bit extreme to me but this isn't my material so I can't place a value on it obviously.

    h264 will be a great choice if a standard dvd output is not your final destination. If it is then h264 will be a waste of time.

    Basically as you probably know you want to convert as few times as necessary to maintain quality. So if you are planning on playing back just the h264 files in either a media player or just a computer than converting from your original caps to h264 is just fine. Just don't go the route of converting to h264 then to dvd. Go from the original caps to dvd if you so desire.

    As far as a sweet spot that is really kind of up to you to decide. Just like anything with video use a bitrate calculator to determine file size output. Obviously the higher the bitrate the larger the file size will be. The trade-off between filesize and quality will be your determining factor. Do test samples with high motion content and low motion content and use a different range of bitrates to see what is acceptable to you. Than you can do a full conversion of a test video to see how you like it. Once you are set on a rate you can do the rest of them.

    Other choices for software include format factory and ripbot (I'm nearly certain ripbot can use multiple input sources - it is frequently used for bluray to h264 conversion (in mp4/mkv containers) but I think it can do other inputs also).

    There are plenty of guides for whichever software you choose. Trial and error should be the rule you live with. Best to use small samples to make for quick conversions. Make note of any settings you change and continue that way.

    Good luck and please post questions when you have them.
    Donatello - The Shredder? Michelangelo - Maybe all that hardware is for making coleslaw?
    Quote Quote  
  4. Member
    Join Date
    Mar 2005
    Location
    United States
    Search Comp PM
    Thank you for the reply. I am not interested in alternative backup methods. I already have quite a bit of in-home redundancy and a proven workflow for cloud-based backup/recovery for all my other data. I simply want to incorporate this video into the existing workflow.

    I also have a pretty good idea what tools to use, and how bit rate results in different file sizes. I'm long since done with DVD and have no interest in ever going to DVD. If I did, I'd go from the lossless source anyway.

    Anyway, I think what I'm trying to get at is,
    - has anyone tried to figure out for themselves what the sweet spot is for h264 with current encoders with various analog video formats?
    - when archiving to h264, are people sticking with interlaced h264 (and if so, any encoding/playback issues?) or simply biting the bullet on de-interlacing?

    Certainly, there's a part of me that wants to take advantage of this opportunity to also get everything onto dual-layer blu-ray for distribution to family... but I'm not sure any modern players handle interlaced h264 particularly well? Or do they?
    Quote Quote  
  5. I do not think there is sweet spot for backing up your encoded video. It is either encoded good CQ 16-18, (sure it always could be better) , or whatever you prefer, or just already compromised video if you lower quality.

    I'd just give some hardisk to a friends house, he puts it into his closet, much better, faster.

    If you are looking for sweet spot, there is none if you set CQ to 30, for example, you'll ruin your video. We are talking about SD video. Not worthy to compromise quality, there is not enough pixels already and it is interlaced on the top of it.
    Quote Quote  
  6. missed that interlace - deinterlace question,
    You deinterlace only to double frame rate if you decide to, to 60p not to 30p, but it has to be done well. QTGMC is a good deinterlacer, but it is not straight forward. You need avisynth, QTGMC plugin and everything takes much longer.

    One weird thing though, even absurd to realize it. If you encode it with x264, constant quality with x264, that 60i encoded video will be bigger in size that same 60p QTGMC video. Interesting in your point of view because you try to shave off some data.
    As an example with that 60p you can do CRF16 and you can get pretty much same volume if you encode 60i CRF 18. Normaly there is about 1000kbps difference between CRF16 and CRF 18. I made "executive decision" to encode old SD videos to 60p, I needed it to be available for playback. I would not even bother but all players will not play good old DVavi (Which is a crime if you ask me )
    Quote Quote  
  7. Member
    Join Date
    Mar 2005
    Location
    United States
    Search Comp PM
    FWIW, I decided to deinterlace to 60p, x264 encode to MP4 with CQ20. I'd like smaller sizes, but I'll live with what I got which is still some 30:1 better compression than the lossless source.

    Handbrake gets the job done, but I really wish there was an easy tool that did not only what Handbrake does (the simple, fast mindless re-encode) but included a basic trimmer/splitter with chaptering ability. Many of these encodes are straight captures off a 1 or 2 hour tape, with various gaps and abrupt transitions from one scene to the next. I'd have loved to splice out the junk/noise, and insert chapter marks at all the scene transitions.
    Quote Quote  
  8. use virtual dub , trim off, cut off not desired parts and direct stream copy audio and video
    Quote Quote  



Similar Threads

Visit our sponsor! Try DVDFab and backup Blu-rays!