Hi8 capture - best quality

[eedwards86]

Hi All,

I’ve done some searching and I apologize if this was already answered but I didn’t find an exact answer. The goal is to digitize, having freedom of codec choice for best quality, old Hi8 tapes. I have an old Hi8 camcorder but it’s not digital 8 so no FireWire output. I figured there are two ways to do this:

1. Purchase a digital 8 camcorder and use FireWire to capture (hoping through adobe premiere if possible or another program that gives freedom to choose format)
2. Purchase something like Blackmagic Intensity and hook up Hi8 camcorder to that via s-video. Blackamgic im sure has a good video processor so quality may be pretty good.

Would the digital 8 camera capture yield better or more accurate video and audio capture or could black magic perform equally as well?

Thanks for the help.

Eric

[jagabo]

Blackmagic isn't good for consumer tape based formats. It is very sensitive to time base errors and will crap out. A Hauppauge USB 2 or a Diamond VC500 will work much better.

[eedwards86]

Thank you for the reply. Would either of those capture cards be on par quality wise with using a digital 8 camcorder via FireWire?

[LMotlow]

Why would you want to record analog source via Firewire to lossy DV that's PC-only playback, not supported by the internet, and has DV compression artifacts that will involve more loss when you re-encode DV to a desired final delivery format? Isn't the title of this thread ask about Hi8 at "best quality"? If you're serious about that, use the suggested capture devices to capture analog to losslessly compressesd media using huffyuv or Lagarith in YUY2 color and avoid multiple stages of compression and chroma resolution loss. There must be thousands of posts in this forum and others dealing with this issue.

[eedwards86]

I wasn’t aware you were stuck with DV as only encoding option through FireWire,
I figured you could use FireWire connection as mechanism to capture via different compression methods or even uncompressed if needed. If this is indeed a limitation of using camcorder as capture device then I appreciate the clarification. There is already compression running through composite or s-video so I figured there were other options digitally.

[LMotlow]

There is already compression running through composite or s-video so I figured there were other options digitally.

You appear to be confused. What composite or s-video compression are you talking about? Hi8 is uncompressed analog. If you play it via a digital camcorder, it goes through lossy compression. If you play it via Hi8 analog camcorder into a capture device optimized for analog source capture such as a Happuage USB-Live2, VC500 USB, or All In Wonder AGP or PCie capture card into lossless media using Huffyuv or Lagarith , there are no compression artifacts. Lossless media is as close as you'll get to a true copy of your hi8 source.

[EDIT] And BTW, analog video has no pixels, no data bits, no 0's, no 1's, no compression. It's analog waveforms.

[eedwards86]

Thank you that answered my question, I appreciate the explanation and help.

[jagabo]

Some D8 camcorders have a line time base corrector. The losses from DV encoding (mild blocking artifacts, DCT ringing artifacts) will be minimal compared to the improvements form the line TBC.

A raw capture device (uncompressed or lossless compression) without a line TBC will give inferior results. So if you use a raw capture device you'll want to use an S-VHS deck with a built in line TBC or an external line TBC. An old DVD recorder like the Panasonic ES10 or ES15 is good for that (just past the signal through, don't record to DVD). The results can be a little better than the D8 camcorder.

What a line TBC does:

https://forum.videohelp.com/threads/319420-Who-uses-a-DVD-recorder-as-a-line-TBC-and-w...er#post1983288

Aside from looking bad, that type of horizontal wiggle is a killer of compression later on. And it can't be fixed with software filtering.

[Candive]

Jagabo,
eedwards86 indicated that one option was to use a camcorder and capture via firewire. I have a Sony TRV 120E which does just that (and has a internal TBC). But it also can send a signal via composite and S-video. Would this be another/better option? Connect the camcorder to the capture card directly (Hauppauge/Diamond) via composite/S-video or would you still use the Panasonic as the TBC.

[LMotlow]

Candive will also be more than happy to give endlessn tutorials in great detail on how to clean up analog-to-DV compression, color, and noise problems. DV is a lossy codec, and inherent in lossy DV is that you'll be forced to go through a second quality-robbing lossy encode to make your DV capture playable on anything other than a PC. Anyone who is fond of DV technology can tell you with a straight face that multiple lossy encodes are always better than one, and the more times you encode the better it gets.

Threads like this used to be annoying. Now they're just depressing. Carry on, YouTube Central.

[jagabo]

I don't know if the analog out from the camcorder is pre or post TBC. You should try it and see.

[Candive]

Hi Jagabo,
I suspect it is post TBC. The only reason I say this was through testing different captures/transfers. Please note I'm speaking with regards to my own tapes, others may have different results.

In one test I had TBC turned on and in the second test I had TBC turned off. The two video results were basically the same. But when I viewed them side by side using VLC, the video with TBC turned off was sightly more detailed and the colour was a little bit richer. So based on these tests, I decided to process the videos with TBC turned off. Perhaps in my case having TBC turned on didn't improve the quality of the video. My personal impression was that the quality was better with it turned off.

This is not to say that I won't use TBC going forward. After transferring my 8mm video I am planning to transfer old family videos on VHS. My plan is to use the Sony as the TBC between the VCR and computer. Of course this workflow again uses firewire to transfer the signal to the computer creating a DV AVI file.

To help eedwards88 make a decision on the hardware he needs to process his files, I don't have a Hauppauge/Diamond but I do have a Gigabyte GT-PTV-TAF-RH with a video capture breakout box. I can run a file through that using virtualdub to a lossless codec and compare it to the same file in DV AVI to determine if there is any material difference. Who knows we may both learn something new. . I'll then post my results.

[jagabo]

To help eedwards88 make a decision on the hardware he needs to process his files, I don't have a Hauppauge/Diamond but I do have a Gigabyte GT-PTV-TAF-RH with a video capture breakout box. I can run a file through that using virtualdub to a lossless codec and compare it to the same file in DV AVI to determine if there is any material difference. Who knows we may both learn something new. . I'll then post my results.

I'd like to see the results. But... caps from different devices often look very different because they'll have different proc amp settings. So you'll see differences in levels, gamma, saturation, sharpness, etc. Someone once did the same experiment here a few years ago. There was obvious DCT ringing at sharp, high contrast edges, and minor blocking artifacts in the DV cap.

This brings up one other issue with DV capture with a camcorder -- you may not have any proc amp controls to adjust levels, gamma, saturation, sharpness, etc. If brights are clipped, darks are crushed, oversharpening halos are obvious, there's nothing you can do.

[LMotlow]

I'd like to see the results. But... caps from different devices often look very different because they'll have different proc amp settings. So you'll see differences in levels, gamma, saturation, sharpness, etc. Someone once did the same experiment here a few years ago. There was obvious DCT ringing at sharp, high contrast edges, and minor blocking artifacts in the DV cap.

This brings up one other issue with DV capture with a camcorder -- you may not have any proc amp controls to adjust levels, gamma, saturation, sharpness, etc. If brights are clipped, darks are crushed, oversharpening halos are obvious, there's nothing you can do.

That's my point about analog->DV, but stated more clearly and in detail. And, yep, the comparisons have been made ever since Windows 95 and the results are always the same, as described. Unfortunately the comparisons don't include what happens when lossy DV goes thru that second encoding loss for final delivery output.

[jagabo]

I'd like to see the results. But... caps from different devices often look very different because they'll have different proc amp settings. So you'll see differences in levels, gamma, saturation, sharpness, etc. Someone once did the same experiment here a few years ago. There was obvious DCT ringing at sharp, high contrast edges, and minor blocking artifacts in the DV cap.

This brings up one other issue with DV capture with a camcorder -- you may not have any proc amp controls to adjust levels, gamma, saturation, sharpness, etc. If brights are clipped, darks are crushed, oversharpening halos are obvious, there's nothing you can do.

That's my point about analog->DV, but stated more clearly and in detail. And, yep, the comparisons have been made ever since Windows 95 and the results are always the same, as described. Unfortunately the comparisons don't include what happens when lossy DV goes thru that second encoding loss for final delivery output.

But difference in levels, saturation, sharpness, etc. aren't inherent to DV but to the particular devices. If your particular device doesn't have blowout/crushing/sharpness problems it's not an issue. DCT ringing and blocking are mild and typically lessen with filtering (which you will almost certainly be doing with consumer analog tape formats) and encoding. And the 4:1:1 chroma subsampling of NTSC DV (180 lines) is of little consequence with consumer tape formats with only 40 lines of chroma resolutino across the width of the frame. You are way overstating the problems with DV capture.

[LMotlow]

The final results with DV look different, even with DV source: etched in plastic Kool_Aid color, enough filtering to erase every remnant of inner detail, all you have left with DV is lines and edges. But I'll have to agree, the typical user isn't going to do any work at all except to oversharpern and oversaturate (looks "better". Or, instead, I think the term is "looks great"), even if they do that much. The average user is visually undiscriminating. They'll watch anything, and if it moves and makes sounds it's even better. There are still people around with two good eyes and a sense of good production value -- even my wife can see the difference between good and typically mediocre, even if she can't describe the reasons.

It's one thing to say the difference doesn't matter to some but it does to many, especially to those who value their treasured video memories and movie moments. With most folks it's all expendable, forgotten 5 seconds later for the next predictably in-your-face ugly video. Fortunately the latter isn't my preference and I can just turn away. Suit yourselves, but watching purposely poor video work isn't worth the aggravation IMO. There's still too much good work about.

[JVRaines]

I have done hundreds of hours of DV conversion and I must agree that the colors look different at 4:1:1 subsampling. There are also nasty artifacts around closely spaced diagonal lines, such as a chain-link fence.

One thing we overlook is that an analog image may have only X lines of color or luminance resolution, but the transitions between values may occur at any one of an infinite number of points along the scan line.

[Candive]

This brings up one other issue with DV capture with a camcorder -- you may not have any proc amp controls to adjust levels, gamma, saturation, sharpness, etc. If brights are clipped, darks are crushed, oversharpening halos are obvious, there's nothing you can do.

Jagabo, I just want to confirm that you are correct, with a DV capture using a camcorder, you do not have any processing amplifier controls, either on the camera or the software (eg: WinDv or ScenalyzerLive which I would recommend to eedwards86 if he does decide to capture via firewire). Typically these adjustments would happen in your video editor after the tape has been transferred.

But before further commenting, I have a specific question for Jagabo.

For my own edification, regarding your workflow, am I correct in that, for example, when you capture an analogue signal, do you capture say a 30 second clip, review it, adjust your proc amp controls, review again, tweak the controls and after a number of adjustments eventually capture the entire segment which you think is at the best quality?

[jagabo]

Post capture adjustments won't be able to restore badly crushed darks and blown out brights .

Back when I was doing analog capture I used AviSynth's DirectShow source filter and a Filtergraph to get realtime feedback while adjusting the proc amp in GraphEdit. If you can't get that to work another way is to preview in your capture software while adjusting the controls in GraphEdit. But I wasn't looking to get perfect results during the capture, just to make sure nothing was crushed or blown out. As you pointed out, fine adjustments come later.

[Candive]

Jagabo, so if I understand you correctly you basically make your adjustments in real time as the video is being captured. Unfortunately I don’t have that skill set and I am unfamiliar with the software you used (again ). And I’ll stick my neck out and say Eric (eedwards86) doesn’t either – or he wouldn’t be asking the question.

I’m basically in the same boat as Eric, maybe one small step ahead since I’ve completed my initial captures. I’m going to go against the grain and recommend to Eric to purchase a Sony D8 camcorder and capture via firewire to DV AVI for the following reasons.

The camcorder is the key component to the entire process. You can invest in the best capture card and find a used dvd player on ebay for TBC etc, but if your camcorder fails, you’ll be capturing nothing. So the additional camcorder is a back up.

It also gives you the flexibility to capture via firewire or capture via composite or S-video. Eric can then compare the different results and then he can make your own decision as to which way he wants to proceed.

Last decade, transferring via firewire was the standard process for this consumer format. In another post johnmeyer who said he has been transferring analog for 20 years indicated that the camcorder – firewire – computer - DV AVI is virtually bullet proof. I agree and would also say it’s dummy proof.

What's more, unlike so many other capture methods which are sometimes difficult to get working without dropped frames, audio sync issues, and more, DV "capture" (it is actually just a file transfer, once converted by the camcorder to DV) is virtually bulletproof. In almost twenty years of transferring DV, 8mm, Hi-8, and all manner of analog video, I simply have had no problems at all.

Sure you don’t have the proc amp controls – but I wouldn’t know how to use them anyway and would probably just leave them on default. I too would like to have the best quality but I am also realistic about my skill set and time. If you have the daunting task of processing 100+ hours of home video tapes you may not have the time or inclination of sitting in front of a monitor, making real time adjustments while the video is being captured. Most neophytes (like myself) who come to this forum initially are looking for the golden ticket;“What’s the best equipment and what are the best settings to capture my 50 – 100 cassettes so I can set and forget and just pump out those files in the best quality!” It’s not like that; there is a lot of work to do!

I don’t know how many hours Eric has to process but my recommendation is just get the job done! Create that initial raw video library. Save it on a hard drive, network drive or personal youtube channel (and make a backup!) so that your audience (your family) can start enjoying reliving their history. It may be raw and not perfect but it’s done and available for viewing. That’s what it is ultimately all about. My daughter opened my eyes when she was watching over my shoulder as I corrected the white balance on a clip. She said, “you know dad, I kinda like it as it was – it looks more authentic!” So while we forum members become anal perfectionists, our audience, our family, don’t necessary have the same standards. They are happy with the grainy, shaky, washed out home movie. It’s a home movie; not a Hollywood block buster. And they are happy and more importantly grateful; not because of the colour of the film, but because you took the time 25 – 30 years ago to film your kids, moments after their birth. You filmed them at every birthday, school concert, Christmas, Easter and every holiday at the beach. They can now go back in time and see themselves before even their first memory. That’s the magic of home movies.

And Eric, after you have captured your initial unedited raw video library in the Godforsaken DV AVI lossy format and encoded it to MP4 so your family can curl up on the couch with a big bucket of popcorn and watch your home movies on TV during a cold winter’s Friday night, you can sit back, relaxed, that you achieved your first critical goal. Now you have the time to go back and hone your video editing skills, practice and learn more from the members of this forum. And slowly you can start replacing the raw video with something you can be proud of. And this again brings the family back to watch the new “Editors Cut” version.

Or you may decide that the original raw capture was good enough for archival purposes and you would rather focus on your current video projects; editing your last scuba dive or ski trip taken with your GoPro. Learning to fly a drone to capture aerial shots or edit that 360 video you just captured. The choice is yours but at least you were able to complete the archival of your analogue footage. And to do it I’m advocating the Kiss principle; Keep It Simple Stupid.

[jagabo]

Jagabo, so if I understand you correctly you basically make your adjustments in real time as the video is being captured.

No, I make the adjustments before the capture to make sure the particular video source will fall within range. Once the coarse adjustments are made I capture. Then fine tune in an editor later.

[Candive]

Ok thanks for the clarification on your work flow. The tapes I have are 90 - 120 min long. Each one can cover a number of scenes; indoor, outdoor, bright beach to shadowy city scape. How granular would your adjustments be? Meaning, would you "ball park" it for the entire tape and then capture the entire tape. Or would you make settings for a specific indoor scene; capture; then make settings for the next outdoor scene; capture etc.

The reason why I ask is because I have already created my initial raw video library for viewing. I am now at the stage where I am trying to determine how can I do it better.

Thanks again for your patience.

[280634157]

Ok thanks for the clarification on your work flow. The tapes I have are 90 - 120 min long. Each one can cover a number of scenes; indoor, outdoor, bright beach to shadowy city scape. How granular would your adjustments be? Meaning, would you "ball park" it for the entire tape and then capture the entire tape. Or would you make settings for a specific indoor scene; capture; then make settings for the next outdoor scene; capture etc.

The reason why I ask is because I have already created my initial raw video library for viewing. I am now at the stage where I am trying to determine how can I do it better.

Thanks again for your patience.

I have a TBC Panasonic AG recorder, which is a version of the end of 1999. I play various variant tapes in it, it doesn't open as well as TBC. I checked the Japanese patent database, including SONY JVC Panasonic explained that the current TBC is actually only tens of lines, not a single frame. And the frame synchronizer just strengthens stability. In fact, I think it is really powerful DV other software algorithms or hardware IC. I asked Chinese electronic engineers, it told me that the AG4700's TBC is just a strobe (four-way election one), in general TBC is The twisted lines are somewhat straightened, but DV's high-precision seems to be not using TBC technology. Videodata's TBC device actually only increases the VHS error. This is what an electronic engineer in China told me. I worked with him to replace the DV TBC with the AG4700 module. It is not meaningful. Because Panasonic's TBC IC is accurate enough.