I can take a longer look at your new samples tomorrow, it's been very long and hectic today. If you're wondering why your huffyuv files are so large, compared to how small they were earlier, it's because your new samples aren't huffyuv. They've been converted to uncompressed RGB. They are 4 times the file size they would be if they had remained losslessly compressed Huffyuv YUY2:

Your TBC "On" position is still contracting video from the bright side inward toward the middle left of histograms. Which version between on and off is "correct" is up for debate. But even with levels correction the "off" position still makes highlights and midtones look washed out. TBC "On" looks more natural and is less noisy.

Avisynth is downloaded from its home site at SourgeForge, but now that all of its sites have been "corrected" and "updated", it just looks confusing. Programmers refer to this confusion as "improvements" (LOL!). The latest version is a Beta dated May 2016, but the more stable, faster-running, and popular version is Avisynth Rev-2 v2.6o dated a year earlier in May/2015. The download link for the Avisynth_260.exe installer for May/2015 is https://sourceforge.net/projects/avi...0.exe/download.

People are surprised that the entire install is less than 7MB, but who cares? When you run the installer the program asks where you want Avisynth installed. The default program folder will be "Program Files" in 32-bit systems, or in "Program Files(x86)" in 64-bit systems. Use 32-bit Avisynth and 32-bit filters if you ever add any (the program already has several dozen built-in functions anyway). 64-bit has no speed advantage and offers very few 64-bit filters or enhancements.

There is a small install window of extra features. The only extra box you need to check "on" is associating .avs scripts with Windows Notepad, which will come in handy. Don't associate .avs files with media players or with Virtualdub. If you maintain a file associations with .avs and Notepad, then double-clicking an .avs file will automatically open it in Notepad for editing. For running scripts, you can always open .avs files directly in VirtualDub.

You can also change the default install folder. My Avisynth is installed in "D:\Avisynth" so that I won't lose all my plugins and other additions if I lose my operating system drive. Or you can always just take the default folder.

Avisynth installs its own online help. It can be accessed by going into the Windows "All Programs" or "Programs" listing in the Start menu, and look for the Avisynth program group. Expand the group, and inside you'll see "Avisynth Documentation", which runs in your web browser. Most users just make a new desktop shortcut pointing to that program item (by copying the program item itself to your desktop). A very similar but more elaborate update is on the 'net at http://avisynth.nl/index.php/Main_Page. Its "getting Started" page will have an image that shows you the latest version of Avisynth being run, which means they might show you the v2.6 2016 version. Never fear. If you have May/2016 installed instead of May/2015, you'll just have slightly slower running (let us know if you could possibly tell the difference without actually measuring it with a specialty utility called AVS Meter). If you've followed this post you downloaded and installed the May/2015 version. If you want a later one you just download and install it over the old one. But the later 2016 version is still a bit buggy in some respects and in some machines.

The Getting Started help also says that you can run .avs trial scripts in media players. You can, but whether it works depends on what versions of certain Microsoft VC++ runtime files and which media player update you have installed, so don't trust it. Use VirtualDub to run .avs scripts. You'll note that there is no "Avisynth.exe" program item, because Avisynth doesn't have an interface. It's a scripted application, so all you need to create and run even a simple or complicated script is Notepad + Virtualdub. Avisynth is versatile; it can be used on a very simple or very sophisticated level depending on how far you want to take it. For the time being, keep it simple.

Thanks Sanlyn, I appreciate all your advice. I will investigate Avisynth further between now and your next post :) I can't believe we are on page 9 already-but I have learnt so much since my first post and bought some essential equipment along the way! Should I start a new thread?

Also I am hoping that TBC ON is all ok now but will await your analysis, should I be too concerned about what the analysis is for TBC OFF if I don't intend on using it switched off for captures? am not sure why the clips are so large as I always check compression before I capture and it all shows that HuffYUV is enabled, I am not sure how to confirm this is really on before I commit to a capture?

If you start a new capture project or tape, and come up with new questions or concerns, start a new thread.

You have more noise with TBC OFF, and you will need different levels settings than with TBC ON. This particular sample has bad and annoying top border line twitter with the TBC OFF (see comparison sample attached). The attached side-by-side sample shows that by using different levels for TBC OFF -vs- TBC ON, you achieved virtually identical histograms for both captures. In the sample, TBC OFF is the left half of of the frame.

You might have captured in huffyuv/YUY2, but you saved your sample in "fuill processing mode" in VirtualDub without changing the color depth and compression settings. By default, full processing mode saves video as uncompressed RGB unless you specify otherwise.

In capture mode, the compressor menu is in "Video" -> "Compression...".
The colorspace setting is in "Video" -> "set custom format...."

The Avisynth script I used with built-in commands to create the AVI for the attached sample:

Thanks Sanlyn, I am currently out so cannot fully digest the information. In your opinion, us TBC ON the best option ? Will it yield me results that are ok for capture and post processing? Will reply in full later :)

Set the line-TBC = ON.

Thanks for that,I know there are so many variables here but is my line TBC comparable to your line TBC function or is it wildly out? I guess I am looking for a benchmark to use when comparing my line TBC (and wider VCR) quality. Is the VCR and TBC up to scratch?

The TBC on my Panasonic AG-1980 and on my former JVC 7600's did not alter the contrast or gamma when turned on or off. Apparently your player has some sort of gain limiter that works at the same time as the tbc, or else it's simply a defect. Line-level TBC's don't ordinarily control contrast and gamma, they just control scanline timing and often have some level of noise reduction. According to your latest samples, whatever was distorting parts of the picture has been tamed. Different tape sources will behave differently, so if you see those distortions occur again you can bet it's not the TBC itself but some other power function in your VCR.

OK, I think I have found the culprit in the VCR chroma board and gone ahead and replaced it. Please could you re-review these clips (they look much better colour wise on the eye) and let me know your professional opinion ? Same routine - this one is TBC OFF

And this is TBC ON - enjoy :) Please note that I did not change any settings in the Proc Amp apart from actually reduce the contrast to 118

They look OK to me, although TBC OFF as previously noted has more noise and some bad twitter along the top border.

Hi Sanlyn, newsflash: I think I have identified a big issue between Vdub and the ATI card....more to follow soon!

There is an issue when first opening VDub in NTSC mode then switching to PAL_I

Please can you merge this into the previous post:

There is an issue between the ATI 9600, VDub v1.9.11 and the combination of opening VDub in NTSC mode. If the filter is subsequently switched to PAL_I, the picture looks ok but has patterning and rich colours making it look washed out with bright colours (I think this is also the root cause of the interference with TBC ON/OFF as I have been uploading recently). If Vdub is subsequently closed and restarted in PAL_I mode, the patterning disappears and a fairly respectable picture is available (as much as VHS can be). I discovered this by accident, is it a known issue?

I will upload a clip shortly of this discovery for your analysis and expert opinion.

Assuming the above is correct, what do you think could cause this ?

Sanlyn, I am really hoping that this is the last TBC ON and OFF test with this configuration based on the issue I found with VDub and ATI 9600. I would really appreciate it if you could kindly cast your analysis over these (hopefully) last 2 clips that I will upload on this matter....this one is TBC OFF

And finally this is TBC ON

Latest capture looks good.

PAL_I is the PAL standard for the UK and Ireland. The source input properties are listed in the 3rd section of the latest settings guide, under "Video".. -> "Capture filters..", shown in the post at http://www.digitalfaq.com/forum/vide...html#post45238, as you've probably seen.

Thanks Sanlyn, is there any chance of running another quick histogram analysis on the TBC ON and OFF values to confirm empirically if this has changed materially?

Also, I have downloaded AviSynth but when trying to run the script you provided a few posts ago, I keep getting an error (despite changing the path to my drive). I go to 'file' then 'run script' in VDub but keep getting this error:

Error during script execution at line 1, column 10: Variable 'AviSource' not found
AviSource<!>("PATH")

Also thanks for the link to 'Capture filters' but I think the problem is deeper than selecting PAL_I as when switching from NTSC (default setting when opening VDub) then changing to PAL_I it causes patterning to appear on the video. The only way of stopping this from happening is to close down VDub then re-open with PAL_I as the new setting based on previous values prior to closing VDub....hope that makes sense?

You run Avisynth .avs scripts using "File.." -> "Open video file...", then locate the .avs file. VirtualDub sees AVisynth scripts as video files, because in fact the results of the script are "delivered" to VirtualDub as video. The run script item is for VirtualDub batch scripts.

If you close VirtualDub capture when "PAL_I" is selected, then the app should re-open to that same value. Maybe it's just a system quirk.

This is classic bad caps. :(

Note: OFF may be correct, ON correct.
But the inverse can also be true: ON correct, OFF not.

The sample of the guy standing in front of the monitors is terrible either way. One is hot, one is muddy and lossy. Neither is great. Color dynamic range is missing either way. The source itself may be at fault. And not just the physical tape, but any tape recorded by the source deck. This is more common with nth gen, but can happen on 1st with a crummy recorder.

The test samples and the image have transmogrified so many times over the past several days, I'd guess that the player will soon develop new problems. TBC board, power, chroma board, worked on so far....I would be looking for a shop rebuild or another player to test by now. The test with the guy in front of the TV consoles still has washed-out whites, but is useful for blacks, grays, and skin tons, and whether or not the tbc is actually doing anything (it appears to be working when ON). I say there's a problem with brights whether the tbc is on or not, and something other than this one test shot should be sampled. But I'm of the stronger opinion that users shouldn't feel they have to be married to a particular VCR. I've seen JVC/Philips with the same brightness problem and once owned one myself.

Good point: Units vary.

Hi Sanlyn, sorry for the delay but I have been systematically replacing caps on this machine - it is taking longer than planned but I will get there eventually in the meantime, could you kindly confirm which NOISE FILTER setting I should be using on the VCR - I have three: Noise Filter ON, Noise Filter OFF and EDIT ON. I personally find the setting that gives the most vibrant colours and sharpness is the EDIT ON but not sure whether it is the right setting to use. In this mode the manual sharpness control slider is disabled. Your thoughts would be appreciated.

Set the noise filter ON or OFF. Don't use auto features like EDIT ON, mainly because such features usually oversharpen, will artificially "enhance" the image, and offers no control over what it's doing. You've been studying captures long enough now to tell if that feature is worth using.

Thanks Sanlyn - I was reading up a little on this forum and came across this thread which applies to me. Unfortunately it hasn't been fully answered but perhaps the PAL version of the AG-1980 differs to that of the NTSC one? Is there anyone who can give a definitive answer on this as I feel that EDIT ON provides a superior picture....

Link here: http://www.digitalfaq.com/forum/video-capture/6500-panasonic-vcr-built.html

EDIT ON leaves you with no control over any sharpening artifacts, should you see them. If you believe EDIT ON is the better position, use it.

In fact, the "EDIT" mode disables all NR/sharpening circuits so it's supposed to give the rawest, purest (as in "straightest-from-the-head") image possible, without added post-processing,

Thanks Sanlyn and tester - would you therefore recommend that EDIT ON is the mode I should use for capturing?

I'd say it very much depends on each tape and recording conditions (quality of the original signal/source, amount of noise present on playback, dropouts, etc). Panasonic "Noise filters" do certainly, by default, substract a bit of detail and add a bit of oversharpening. I personally tend to use the edit mode (or NR OFF) if possible.

Then EDIT mode with the Phillips player is similar to EDIT mode in other JVC players, which is what I suspected.


If you don't want typical softening and smearing side effects from noise reduction and don't have problems with sharpening or other enhancements, use EDIT mode. As with most such players, EDIT is designed with "straight-through" playback for dubbing from tape to tape. With EDIT mode you'll learn to recognize and filter any source noise later in post processing, which is likely more sophisticated and less destructive than with most players' built-in features. You'll have to undertake some form of post processing anyway. With tapes that aren't so noise friendly you might decide that EDIT isn't doing enough cleanup during capture. To bypass your player's more aggressive built-in image processing, use EDIT mode.

Interesting thanks both - I definitely think the colours are more richer and vibrant when using EDIT mode but there are the grainy artefacts which is basically noise which I hope can be edited out in post processing....there seems to be more detail in the picture however - I guess it is about trading off clarity with noise?

All denoising methods pose a compromise, sometimes a little, sometimes more. My AG-1980 has decent grain removal during playback but sometimes with a rally dirty tape it appears to be over-smoothing. But AG-1980 DNR can't be turned off. So I play that tape on the 1980's lesser cousin, a PV-S4670 SVHS which has so little built-in denoising it effectively has none, thus it plays a tape "as-is" without modification. This leaves me with some post processing decisions to make and more work to do. But in the long run I think that for some tapes it's better than the side effects of less sophisticated built-in denoisers. You have to decide which compromise you want to work with. One rule of thumb is that whatever is removed from an image during capture can't be magically restored later.

I never understand this. Referring to JVC decks...

I don't see smearing with NORM or AUTO*.

Softening can exist, but rarely does. EDIT vs NORM/AUTO is about equal detail. If you need more detailer, that what detailers are for (example: SignVideo DR-1000). Or the Panasonic S-VHS, which is a tad sharper, especially if you crank the slider. It sharpens by default; you see tell-tale haloing.

My question is always this: Why buy a fancy VCR only to disable all the fancy stuff you paid for? :wink2:

Only use EDIT if it can be visually proven to be causing more harm than good. But that should be a minority of the time.

* Note: NORM is with calibration off. And it should be off. If you see AUTO, it's on. Turn if off! Calibration does some bizarre things to tapes, as it tries to pre-guess tracking. And it sucks at guessing. It leads to vertical jitter more often that not.

Capturing VHS, VHS-C, and Hi8. Transferring MiniDV over Firewire cable (the right one was hard as heck to find, by the way!)

1) VHS and VHS-C with adapter: JVC SR-V101US

2) for Hi8 archival, is capturing from a camera sufficient? I have the CCD-TRV68 (specs) which I'm sure is not a great camera. But all my tapes were recorded on an early model camera with no TBC, mono audio, and no S-Video. I assume I should use these playback settings: Edit mode and DNR disabled, TBC enabled

3) Do I need an external TBC? What sort of errors should I look for to answer this question, and what issues may the JVC VCR introduce?

3) miniDV over Firewire: Scenalyzer or WinDV?

4) for archival & editing lossless format, is there any benefit to UTVideo or HuffYUV over MagicYUV?

5) Recommended resolutions: The guide here says:

352×480 = VHS, S-VHS, Hi8
640×480 = anything being captured as AVI intended for advanced editing in an NLE

So... which is it? I'm archiving but I do intend to edit, correct, etc as well.

Thanks in advance.

According to the published specs, the camera you mention has an internal tbc, likely at least a line-level tbc. Capture with the tbc enabled. You wll have to test to see if built-in dnr does what you want it to. Sometimes bult-0in dnr can overly soften or blur during motion. Built-in dnr in older gear is often rather primitive and can degrade images. Post-process apps have more sophisticated noise reduction. Many users of high-end vcr's often disable builtin-in dnr where possible and prefer to use after-capture software for less aggressive noise reduction. Tbc's aren't used during shooting. They're used for playback and capture.

External tbc's are usually frame-level rather than line-evel tbc's. A frame level tcb works on the frame timing signal, not on individual images or individual scanlines within frames. Uncorrected frame-level timing errors show up as dropped or inserted frames, frame hopping or judder, and off-sync audio. Signal noise can also generate errors that "look like" embedded copy protection errors with many capture setups -- these show up as image distortion, top-border "flashing", luminance variations, and other effects. Your camera's internal tbc might be good enough. If you have a dirty signal, you need an external tbc.

Both will work, but WinDv seems to be the overwhelming favorite.

For analog capture, people seem to have fewer problems with huffyuv. Fast and accurate, with low cpu usage during caps. Analog tape is captured as YUY2, which more closely resembles the YPbPr video storage used by analog tape. Huffyuv, Lagarith, and UT Video are in more universal use and are recognized by most media players. You can use MagicYUV, but it is optimized for high-resolution and 4K, is not as widespread in use as others, is not recognized by many media players, and many users have complaints with its use in some editors. If you use huffyuv, note that standard versions of huffyuv cannot compress YV12, used in much post-processing. Lagarith and UT Video are in wider use for post-processing. Your choice.

Unfortunately that guide is dated and was written for the CRT era, long before larger monitors were in use. See this post, and the post following: http://www.digitalfaq.com/forum/news...html#post50698

There is also an updated Virtualdub capture guide here, if that's what you're using: http://www.digitalfaq.com/forum/vide...-settings.html

Thank you, now I know what to look out for! :)

thanks. I will try MagicYUV first, if I run into troubles I can always switch to UT Video or Huffyuv. If I get a chance I will actually compare the 3 at vhs resolutions

Quick sanity test on my Custom Format settings for Virtualdub for VHS, VHS-C, and Hi8

PAL:
720x576
YUY2


NTSC:
720x480
YUY2

Yep, you are sane. 720x480/576 YUY2 is fine. :salute:

Forgive me for asking too many questions, I grew up being filmed by this equipment and never actually used it :)

Here are some very small (<150 frames each) samples I captured then converted to x264.

Any glaring signal errors I need to clean up before I begin capturing en masse? I don't know what is typical or expected analog behavior versus errors that need to be corrected.

"analog noise rec.mkv" is noise that appears on recently recorded tape, but doesn't seem to show up when playing back video that was recorded many years ago. Is that weird, or just a sign of the age of the tapes? Other file names should be self explanatory and some are just random scenes I picked out.

Thanks for any input.


https://www.dropbox.com/sh/0ywe1fm00...WPia8XdMa?dl=0

You need an external frame-level tbc.

OK, my turn...
JVC HR-S7800U (I should be arrested for what I paid)> TBC-1000 (my wife would arrest me if she knew what I paid)>Hauppage 850 USB>32bit PowerDirector using MAGIC/Lagareth/highBR MPEG2>Windows 10x64 with i7-4770.
Thank you for this great forum.

Hope I didn't jump in too early.