Your link shows up as a malicious website. Access is blocked by Kaspersky and Avira antivirus.

It's a false alert from what I can tell.

If I bypass the warning, I get a blank page. Will try it again later.

[EDIT] After 15 or 16 red malicious site warnings and 5 frozen browser tabs in 2 browsers, I finally got a standard browser download window. The main download page never painted completely.

The huff_Mt capture was decoded, but it's too dark and noisy to be a usable capture. It';s been deinterlaced for no particular reason and alternate frames discarded, so 50% of the original video is missing, motion is clumsy. Perhaps this was done for internet posting, but it's still not a useable image. Has oversharpening artifacts, lots of horizontal tape noise, and has noisy, wiggling edges (no line tbc?). It doesn't play properly in my VirtualDub -- no filters are mounted, but the output pane flickers. This looks like deinterlaced DV recompressed with muff_MT -- at least, that's what the artifacts look l like to me. Looks to be a copy of something, because even the black borders have the same noise as the images. I could be wrong.....

OOps. Never mind. Plays OK in Virtualdub. My antivius scan slowed things down a bit. My fault.

Very bad condition, as you probably know. After adding some luma gain in Avisynth, it now looks as if the source is a tape duplicate. Has been through some processing, such as deinterlacing and masking of the boittom border head switching noise that places the image off center, instad of cropping and adding balanced borders. Stiil, you might be able to rescue the clip, up to a point.

Below is frame 178 of the original sample, with borders removed. The image below that is a YUV histogram showing crushed darks, which doesn't include the black borders. There's no luminance above RGB 128, which is pretty dark. There is also a lot of luma flicker and some line timing distortion.

frame 178 (original, borders removed):
http://www.digitalfaq.com/forum/atta...1&d=1436892317

frame 178 (original, YUV levels histogram):
http://www.digitalfaq.com/forum/atta...1&d=1436892382

But with a lot of work you can get a clear but still fairly noisy video out of it:
http://www.digitalfaq.com/forum/atta...1&d=1436892440

The attached m2v version has no audio. The original audio track has an invalid PCM sampling rate. Still some bad ghosting and edge ringing.

This was only one test video, original from UYVY.
So a Rohcapture, without filtering.

Clear the film has mistake, because it was sent original in 25P and then was struck up on VHS.
Then Avisynth filter "PhaseShift" + SelectEven () helps there and QTGMC

The VHS tape is given to me here in such a way, totally has 3 same tapes.
Original the film with 16 mm of camera was taken up, stored away in 1969 and then in NTSC.
I have here the version PAL.
Film name "Woodstock" is hopefully to be written this permitted here.

Tomorrow, unfortunately, I do not have a better film quality and will look after it. With us it is 24:00 o'clock

Supplement_ audio is absent, now I notice only.
In the original it is embarrassing available.......

I suppose there has been already built with the norm change by NTSC to PAL big dung.
Then still progressively on interlaced on VHS.
Thus it looks, in 25P
Pity for the work

I wonder if the problem with Huffyuv files not decoding properly on another system is due to a mismatch of the interlaced decoding flag between the encoder and the decoder. The original version of Huffyuv uses a hardcoded value while later versions let you specify the number of lines that triggers the different coding method.

Note to admin: Perhaps good Idea for admin to change title of my original (this thread) to read: "Test file to see if Huffyuv/codec problems?". I uploaded sample file before realizing how messed-up it was, because it played ok on the capture PC!. I was intending to get RFI opinion, but thread wound up being a very good learning experience about Huffyuv/Lagarith instead. Once I get my Cap system working good again, (with playable files!) then I will start a new thread too see if still any RFI present on my captures:cool:

Some have noted the interlace parameter problem, but it doesn't seem to have fixed anything. Years back, I took old huffyuv caps from an old PC and couldn't read them in Virtualdub on a different machine, but they decoded properly when opened with Avisynth. At least, I think that's the way I worked around it. Its been so long ago, I don't even remember exactly how I did it, LOL!.

The audio bitrate on the first sample was non-standard and gave two of my mpeg encoders an error:
For my histograms I used PC levels. As your EDIUS image shows, converting PC levels to TV levels brings the blacks up to RGB16. However, the EDIUS vectorscope and waveform still show clipped (compressed) darks at RGB 16. Whether dark detail is crushed at RGB 0 or at RGB 16 is not the issue. The issue is that darks are compressed into a narrow range. The vectorscope and waveform also show useable data is cut off below the midrange. In other words, the tape is too dark.
http://www.digitalfaq.com/forum/atta...1&d=1436978119

YUV Levels histograms show the different black levels for PC and TV.
left image = PC levels show clipping at RGB 16 and below, and suppressed midtones.
right image = TV levels clipping above RGB 16, which is raised from black to dark gray. Very little midtone data. PC levels raises the black border to RGB 16, but data is still dark gray with little detail. Usually the borders are RGB 0 anyway rather than RGB 16. Most of the darks in the image are obscured and at the same dark level as the borders.
http://www.digitalfaq.com/forum/atta...1&d=1436978264

ColorFinesse was set up in After Effects to examine the same video at TV 16-235 levels for PAL BT.601. A screen capture of the ColorFinesse color system setup is attached as AE Color Finesse - video system setup.png. The image below demonstrates how the video would display as TV RGB. There are 4 charts: Luma waveform (upper left), YC waveform (upper right), YRGB waveform (lower left), RGB Histogram (lower right). All show data compressed at or below RGB 16, which displays at "TV Black" on TV. All charts show cutoff below the midtones.
http://www.digitalfaq.com/forum/atta...1&d=1436978704

The histograms below show the effect of adjusting gamma and using a basic contrast mask to elevate gamma in the range at and below RGB 128. 16-bit dithering was used with the SmoothAdjust plugin to level out gaps in the RGB range and to heighten brights above RGB 128. The histograms show increased detail and output in YUV levels (left) and in RGB TV display (right). There is still some super-black data in the display, but that area contains no details anyway and is mostly RGB 0 black border.
http://www.digitalfaq.com/forum/atta...1&d=1436978925

The newer samples have several problems. Sample #3 is oversaturated and looks almost like animne. I still don't know why users deinterlace video unless it's for web or PC-only display.

And thank you, Goldwingfahrer, for the clip with audio. Attached is a version of the earlier HuffMT.m2v...but now with sound!
:)

I have provided test with mpeg2 first, with the Procoder, see Screen.


Yes, I know, I can play there in edius [6.08 + 7.50 with the curve YUV something, today I have also made, however, does not become so great.
On 3 VHS of tapes and today on a new package with 2 VHS of tapes [part 1 +2] and also on the Blu-ray-Stram [m2ts Codec WVC-1] the Intro is exactly equally black.

But on the Blu-ray at least the audio is quite good.

I have already noticed this if I look at the Stream, see Screen 2 in VDub.
RVB is RGB [is a French programme, green =Verte]

Okay, this was only one teststream.
Do not forget, tape VHS, is original the film in NTSC 23.976 and was encoded very badly to Pal.
now I have 5 springs here [inclusive of the Blu-ray] and in the kalibibrierten TV + in the calibrated Eizo SX and also [interlaced] in the professional studio controlling monitor I see that the colours are oversaturated and the picture very badly.
Unfortunately, I have no better spring, must still continue to search.

And still what... I am a fan from interlaced, no matter whether in mpeg2 Stream or in MP4 Kontainer as a MBAFF [MBAFF = Interlaced] mostly TFF.

Analogously Captures are here always TFF when I see, however, that a film original in 25p was sent and then improperly badly on VHS it was covered up, this tries to consider with Avisynth and PhaseShift.

However, I see... has not succeeded me here.

Thanks for him type with Color finesse, I have AE in the Adobe package CS6.
Work with it, however, not enough there my Stream always in UYVY or YUY2 are given and picture rushing in
Y canal have. Adobe anus effect can work only in the RGB.

... my translation hopefully well comes across....

interesting analysis tool. I noted some entries you show in red:

** Number of Audio Streams: The audio is AC3. The use of MPEG-1 Layer 2 obsolete in many countries, not used commercially in the U.S. for prime releases. It is not high quality audio compression.

**Horizontal Size = 704: PAL and NTSC 4:3 captured from VHS occupy only 704 horizontal pixels in a 720x480 capture frame. 704x480 is valid for DVD and BluRay. However, BluRay 16:9 DAR requires full width 720x480 for standard definition.

** Top Field First: encoded as progressive, so no field order is specified.
** DCT Type (Frame): The encoding is progressive, frame-based, no pulldown.

Yes, it is shocking the way some people process video, even if the source is old fashioned VHS. One can only do the best they can muster with the tools available to repair.

Well...I'm not sure what the "anus" effect means. I do noise removal and basic levels determination in YUV in Avisynth. Most Avisynth filters are YV12, a few work with YUY2. NLE's are not good repair tools -- after all, they are primarily editors, not total restoration tools. They process color well, but better when I use Avisynth to set the colorspace. Premiere, After Effects, and other expensive NLE's have color filters for YUV and RGB.

The 03.avi sample was difficult to correct. It has less noise than the earlier video. But the color is terrible, as you say. The first step was levels. saturation, and gamma correction in YUV (Avisynth). Chroma exceeds RGB 0-255 at both ends of the spectrum, in both YUV and RGB. Wide chroma and luma contrast can work with film, but the film spectrum is too wide for digital video. In the image below, the high saturation makes the colors "glow" and "bloom". The colors and levels are innaccurate, unrealistic, and unconvincing.
http://www.digitalfaq.com/forum/atta...1&d=1437014328

http://www.digitalfaq.com/forum/atta...1&d=1437014448

Levels correction and noise removal were run in Avisynth. This was followed by moderate color and gamma refinement in RGB (VirtualDub). YUV to RGB conversion was run in Avisynth. RGB return to YV12 was also run in Avisynth.
video Encoder: TMPGEnc Plus 2.5.
Video/audio remux to Dolby AC3: TMPGEnc MPEG Editor v3.

not everything what is there red it is recognised as wrong.

Yes, I know this. Various programmes inflate simply on 720.
Here, besides, I had with Adobe Encore of early problems, it could start with 704 x 576 nothing.
I have distributed TMPGEnc old and also the newer version, from the Authoringtool as well as from the Codierer, but on 3 PCs.
My favorite tool for the DVD production is still Ulead a power tools 2.

Yes, with Avisynth. I learnt here a lot in the Doom9 / Gleitz forum of user LigH, scharfis_brain, Didée and other.

Yes, YV12... I knows most filters work in this plan are.

here to the view control I can send under edius interlaced stream from timeline about the Canopus NX map by YUV or by Y/C in the tubes 50 hertz of studio controlling monitor.
Edius I already have since the first version, at that time still with Storm DV, the analogous by the way by Y/C in UYVY capturen was able... nice old time.

Yes, in Avisynth it would be the best....., unfortunately I have there a picture only in VDub on a progressive screen and not by the controlling monitor with 50 hertz interlaced.
I must deal there sometimes more with Avisynth.

Perhaps, they show me sometimes the Script you have provided, then I get on, I am not a beginner in VDub and Avisynth.

They should look sometimes at the picture like it is stored on the Blu-ray... I soon catch in to cry...:D

I used the Avisynth script shown below, followed by two Virtualdub filters. The script was opened in VirtualDub and the VirtualDub filters were applied to Avisynth input. The VirtualDub output can be saved to YV12 (I used Lagarith lossless compressor for YV12 output).

In the script, QTGMC is used in progressive mode ("InputType=1")

The Avisynth filters used:
- SmoothAdjust v2.62 http://latoninf.free.fr/d9/SA/SmoothAdjust-v2.62.7z
The doom9 thread for SmoothAdjust. The filter includes the functions SmoothTweak and SmoothLevels: http://forum.doom9.org/showthread.php?t=154971
- FixChromaBleeding http://avisynth.nl/index.php/FixChromaBleeding, which requires ChromaShift http://avisynth.nl/index.php/ChromaShift.
- HQdering http://avisynth.nl/index.php/HQDering (the web page also has links for Deen and MaskTools2).
- QTGMC http://forum.doom9.org/showthread.php?t=156028.
- AddGrainC (is included with QTGMC support files package)

VirtualDub filters used:
- ColorMill v2.1 http://fdump.narod.ru/rgb.htm
- Gradation Curves http://members.chello.at/nagiller/vdub/index.html.

I have attached a VirtualDub .vcf filter settings file for the settings I used. If you have not used .vcf files previously: a .vcf is a simple text file that can be opened with Windows Notepad for reading (but not a good idea to change the text!). To load the filters and settings into VirtualDub, Use "File.." -> "Load processing settings...", and navigate to the location of the .vcf file. Most users keep the .vcf file in the same folder with their video project. Don't place .vcf files in the VirtualDub plugins folders.

The two VirtualDub filters mentioned must be in your VirtualDub plugin folder.

Everything can be tweaked and refined as you wish. I ran out of time for perfection.

Cool... big thanks.
When do you sleep, actually?

FixChromaBleeding and HQdering is not to be found any more on the aforesaid side, however, it has still found in my archive.

The first step is done, if the "ts" file has though not taken separate the file gecapturte anew [YUY2] with VDub to Lagarith YV12 coding.

Then the next step is with VDub. I have there 3 versions. 1,9.11 + 1.10.4 + 1.10.5...
must decide me sometimes for a version.

Every night. :)

Damnation! :smack:
It's been a while since I downloaded from those links. It appears they have moved. For those who don't have the missing plugins, I have attached For_Avisynth.zip. It contains Deen, FixChromaBleeding, ChromaShift, and HQdering. The version of Deen I have been using is the 2005 version.

For those who don't have the VirtualDub filters, I have attached For_VirtualDub.zip. It contains Gradation Curves and ColorMill.

Use 1.9.11. I still see occasional complaints about the other versions.

The script and VirtualDub filters used for 03.ts can't be used for the darker clip. They have completely different problems. I will forage for the filters I used for the dark clip and post later. It may take a day or two, I will be traveling for family affairs for the next couple of days and won't have my computer.
:depressed:

DirectShowSource I take for over 10 years not really more, there were earlier a lot of problems with it. One never knew which decoder there was taken, even if one had adapted the "Merit", there always remained a risk.
Mostly demuxe I the Streams and work on only the picture contents by means of Avisynth.
Besides, the Capturestreams are given in YUY2 [Canopus NX] or in Lagarith, or come directly by SDI to the PC.
There reaches "AviSource"
Now for "Ts" I would have used rather LWLibav.

In addition as the 3rd filter I have still inserted ChromaShift in the filter chain VDub.
Blue colour of the shirt of the reporter is something too far on top.

With TGMC and the settings I get no so good result like with the above Script.... Cool.
Only the picture rushing something is mostly to be eliminated, it must not be completely, however, the Encoder draw mostly the picture still a little bit softy.

I did a few test encode-decodes and I wasn't able to trigger the problem this way. For example, encoding a Huffyuv-MT file with "lines" set to 576 and decoding it with "lines" set to 288 was no problem. However, ticking the box that says "Ignore is-interlaced flag" for the decode side did corrupt the image.

Attachment 4970
Attachment 4971