Correct color, brightness, contrast, & tint in VHS c

[BrainStorm69]

Is there any objective way to get all these parameters "right" when capturing VHS, or is it just a matter of "what looks good to you?" I'm beginning to think this is the hardest part of capturing VHS. Please describe the method you use, and why you think it works.

[lordsmurf]

Restoration topic... moved..

Anyway, Elite Video BVP-4 Plus proc amp.
Sometimes, using ATI AIW card, the images controls.

Just a matter of "what looks good, and what looks correct". Usually, you only do this with bad sources or imperfect sources.

[edDV]

Is there any objective way to get all these parameters "right" when capturing VHS, or is it just a matter of "what looks good to you?" I'm beginning to think this is the hardest part of capturing VHS. Please describe the method you use, and why you think it works.

Yes there is, it envolves test signals, scopes and calibrated monitors. Correct system adjustment is the hardest part, and is important especially if the video is going to be viewed by somebody else on other equipment than yours.

I feel another rant coming ...
What forum are we in? OK restoration.

[rant]
In the professional broadcast world, video signal evaluation is done objectively using expensive test signal generators, special oscilloscope monitors and standardized procedures.
See: Tektronix Application Notes
http://www.tek.com/Measurement/cgi-bin/framed.pl?Document=/Measurement/App_Notes/index...Set=television

The home video hobbiest now has equipment capable of operating at near broadcast quality but typically lacks equipment or procedures for proper calibration. I think there is a need to develop some simple, objective calibration techniques that can be used by a home user for equipment setup without the need to buy additional special equipment.

The typical home video hobbiest now has at least a DVD player, a computer with VIVO (video in video out), a computer display, a surround audio system, a video monitor (TV) and various analog and digital recording devices. The overall goal would be to calibrate this environment for optimal, or at least adequate, audio-video recording and display. So lets get creative....

One untested idea:

The two most basic pieces of video test equipment are the test signal generator (color bars) and the waveform monitor which is a special type of oscilloscope. These are used to set levels.

For our purposes, a DVD player can operate as a fairly sophisticated test signal generator with NTSC, PAL, S-Video, Y-Pb-Pr, analog audio and digital audio outputs. The computer may offer composite, S-Video, audio and maybe DVI outputs but the signal quality will be more variable and less predictable. In theory, even the cheapest DVD players will give fairly uniform outputs. I may be wrong about typical DVD players outputting reasonably predictable levels. If they do, the next need is for standard test slides and audio tones that are readily available to all.

The THX Optimizer monitor calibration slide is a good start.
http://www.thx.com/mod/products/dvd/monitorPerform.html

It contains crosshatch, luminance stairstep, white level, color bars and a frequency sweep. The main purpose of this slide is monitor calibration, but it can also be used as a rough indicator of signal path performance. The advantage of this slide is that it is readily available (on rental or purchased DVD discs such as "Star Wars" or "Ice Age") in a calibrated form for DVD playback.

This slide doesn't do it all, the main disadvantages are a lack of calibration marks on the frequency sweep and a lack of "pluge" a simple black reference, but it's a start.

The first step is to hook the DVD player to the monitor (TV set) and follow THX monitor setup procedure.

Next record the slide from the DVD player to your VHS recorder and play that back through the proc amp to the monitor. This time adjust the proc amp for proper black, white, saturation and hue. Switch between the DVD player and the VCR and further adjust the proc amp for a match. Note the proc amp settings. These are your nominal settings.

Next capture the VCR-proc amp test slide to the computer. Play the computer file out to the monitor (TV set) and see how close it "Looks". Compare the computer output to the DVD player. That will give you an indicacation how close your computer inputs and outputs match. If ideal, the levels (as seen on the calibrated monitor) will match for the DVD player, computer and VCR-Proc amp.

If they don't match, the VCR or computer capture/output card may
need adjustment.

Note: never use the computer monitor for level setting. It is always wrong (e.g RGB color space, linear gamma, etc.). Use the TV set.

This is a very crude way to adjust a system. A scope (waveform monitor) is needed to diagnose and pinpoint signal path performance. The latest generation of mid level editing packages (Premiere, FCP, Vegas 5, AVID, etc) have scopes that vary in performance and placement. These packages also generate test signals to the timeline. The basic tools are there to create a calibrated system.

More work is needed.
[/rant]

Assuming a calibrated system, then load the tape to be captured on the VCR and test capture it for evaluation on the edit program waveform monitor for levels. A pro tape will have color bars at the front and the entire tape is supposed to match the bars if quality control was maintained. Adjust the proc amp to match the bars and you are done. In the real world with home recordings, you will need to eyeball and match levels scene by scene with much trial and error.

Good luck.

[gshelley61]

That's basically what I've been using to set up my monitor... the THX video test patterns on some DVD's. There is a luma (contrast, video level) slate - the white boxes test; a black level (brightness) slate - the THX drop-shadow test; and the multi purpose slate edDV already mentioned. I get my monitor tweaked using that, which then serves as a baseline for me. When I apply color correction and black/white level adjustments, I use the luminance meter in my SignVideo Proc Amp to make sure I'm not pushing anything out of standard... and my eye and the montor to do the rest. If I'm capturing a laserdisc, I may have a look at a commercial DVD with similar material as a comparison. After some trial and error, practice and test captures, I settle on the what appears to be the best overall settings.

[BrainStorm69]

@LS - I'm not sure I'd really consider this restoration because for this I'm talking about even with a perfect source. That is to say, how do you know what you've captured accurately represents your source tape?

@edDV - nice post. I guess one of the main things I need to do is get a calibrated TV into my capture chain at the appropriate place - which I guess is at the point that it is receiving the same signal that my capture card is. I could do that, but then how do I know my capture card's proc amp settings are right to correctly capture that signal? Trial and error with a-b switching back and forth on the same calibrated TV between the original VHS tape and DVD I've created and eyeballing it?

My capture chain currently looks like this:

VCR--A/V Switch--Proc Amp--Capture Card

I guess I need to change it to:

VCR--Proc Amp--A/V Switch (has video monitor out)--Capture Card
..................................|
...........................Calibrated TV

At least then I would know that the capture card is getting the same signal that I've determined should be calibrated on the TV.

However, I've also thought of one other way which is crude but perhaps simpler. In the BT878 capture thread that FulciLives started, we compared outputted frames from our capture cards to a reference frame ripped straight from DVD. I was thinking that if I have both the VHS tape and the DVD of a movie (which I do of the Matrix), I could rip a frame (or several from different scenes) directly off the DVD and then by trial and error (and eyeballs) tweak things until the screen caps from my capture match the screen caps from the DVD. At least then I would know that the capture looks like the reference original. Thoughts??? Too crude???

@gshelley61 - You're part of the reason for this whole thread. When you originally posted those Titanic screen caps way back when, I thought they were amazing. I finally pulled out my Titanic VHS (which I found to my chagrin was the widescreen version) and in the last couple of nights tried to duplicate your caps. I found it very difficult (if not impossible) to accurately duplicate the color, tint, contrast and brightness of your caps. And then, in comparing my caps to yours, I wasn't sure when it was said and done that I even thought that your caps looked the way I'd want in the end. But of course that's on my computer screen, which of course isn't even the proper place to be doing the comparison in the first place, as edDV properly points out. Anyway, that's what actually prompted this thread.

I don't know, maybe I'm getting too hung up on achieving "perfection" when "close enough" should do.

[gshelley61]

Since I switched to capturing with a standalone DVD recorder, the process has become much easier. I monitor the output of the DVD recorder...

VCR-->DR-1000 Enhancer-->Proc Amp-->DVD Recorder-->Monitor

or, for composite sources...

LD Player-->Detailer III-->Proc Amp-->DVD Recorder-->Monitor

The JVC DVD recorder basically captures exactly what is sent into it... in other words, it creates very nearly an exact duplicate of the source as far as color, contrast, sharpness, etc. (at the higher bitrate settings). By monitoring the output of the recorder, I simply adjust the processors to my liking. I also will pop in a commercial DVD of similar material in order to make comparisons between my settings and a professionally transferred DVD. I also play back my recorded DVD-RW's on my home theater setup and my daughter's bedroom DVD/TV to see how it looks on a couple of different systems.

[vhelp]

I personally never go by others "screen pics" or my own for that matter.
That is, I don't use them as gauges.
.
And the reason is obvsiou. Color space and codec. Our pc screens do
not duplicate in parallel what we *shoud* be seeing - as trueth.

There *is* no (IMO) right way. The correct way (as I do it for my DV
projects) is to not adjust the color setting at all. But, before I
go into a long and exhaustive discussion (ok, maybe I won't) I have
a question for you. Only because it could be an area that you did
not look into yet.

* What is your encoder that you are finding these issue to be bothersome
.. to you ??

* and, are you frameserving ??

* and are you using any Filter Plugins (how many) ??

Ok. too many questions

I don't have Titanic on VHS. I wanted the widescreen version, but I was
told that (those widescreen scenes) were only in certain areas. Perhaps
I was misinformed. Anyways. I did not get the tape.
.
Personally, I prefere a VHS tape that was Digatally Mastered as
these tend to have super-duper detail and quality, vs. one that wast
not Mastered. This is from my experience. Otherwise, I'd post a pic
or two, from my ADVC-100 or TRV22.
.
I'd do a DVD to VHS transfer, but as you know (or don't) when transfering
DVD to VHS, it suffers from the "black level" bug, and is too dark.
When captured (and although it may look good in your captured AVI) it's
way off the scale when you convert it. So, this would be an invalid
scenerio or reference But, I wish there was a way to transfer a
DVD to VHS w/out this bug. It could serve as a good reference or gauge
.
And, I don't have the Matrix on VHS either. I do have the Animatrix though :P
animation on tape (widescreen version too) Anyways.

In my process for VHS, I don't ahve any color issies (anymore) and my
chain is:

VCR (JVC S-VHS HR-S3910U) -> ADVC-100 -> canopus codec -> TMPGenc.

* note, commercial movies require my TRV22 in place of my advc. But
anything else that *I* record TO my vcr is ADVC all the way.

-vhelp

[BrainStorm69]

The JVC DVD recorder basically captures exactly what is sent into it... in other words, it creates very nearly an exact duplicate of the source as far as color, contrast, sharpness, etc. (at the higher bitrate settings).

How do you know that for sure?

* What is your encoder that you are finding these issue to be bothersome
.. to you ??

* and, are you frameserving ??

* and are you using any Filter Plugins (how many) ??

....

I'd do a DVD to VHS transfer, but as you know (or don't) when transfering
DVD to VHS, it suffers from the "black level" bug, and is too dark.
When captured (and although it may look good in your captured AVI) it's
way off the scale when you convert it.

*I'm using huffyuv. I'm not saying that it's necessarily bothersome to me, but I am wondering if it is accurate.

*I'm not frameserving.

*I'm not using any filter plug-ins.

I've not heard of the "black level" bug. Have any more information on it?

[lordsmurf]

The JVC DVD recorder basically captures exactly what is sent into it... in other words, it creates very nearly an exact duplicate of the source as far as color, contrast, sharpness, etc. (at the higher bitrate settings).

How do you know that for sure?

Experience.

[BSR]

hhhmmmm, only 1/4 a beer.

I would like to know more about the balck level bug also.
Was going to transfer some test images from DVD to VHS and then recapture to see where my process stood as far as color and cropping.

But this issue seems that it will render the color part inaffective.
Do you agree?

[vhelp]

I'm going to have to do some HUFFY tests of my own. I have a
hunch that I'm wherling around in my head. (It does have
something to with color loss, as in your case)

The "black level" issue w/ DVD -to- VHS (dvd recorded to VHS)
is there. Some time ago, I found a link that described this
phenomina, but I can't find it.
.
There is an issue w/ this D to V, but I haven't completely
figure out any further details. Other than, for you all (those
that are interested further) to test for yourself.

My sencond theory (hunch) is w/ respect to D to V being that
your DVD player may have a BL setting, and eliminate this
issue, and those w/ such (if it works) will be able to use
*there* D to V for further testing (or, following this thread)

You still haven't mentioned your encoder. Anyways.

I beleive that TMPG may have an advantage over CCE (assuming
that you ARE using CCE here) Only because w/ TMPG, you can
turn on/off CODEC use, where in CCE, you are *restricted*
to a given cocec. That is, YUY2 source. But, I'm not sure
about CCE's true outcome. But, if you are using CCE, and
you have issues w/ color matching (your post) then my comment
stands, w/ regards to TMPG.
.
I realize that you said its not that big a deal (in so many words)
but still, it must be bother you, to get that much *closer*
to the original source (detail and all)

I also belive that (specially for DVD projects) your encoding
project should not be lowered to 352 x 480 (even though its
DVD standards)
.
There are too many tainting of OP's (mine included) about size
beging everything.. (ie, and 352 x 480 is better in some cases)
I disagree. I beleive that 720 x 480 (or, 704 x 480) is best
used with *everything* when your target goal is DVD. We want
*every* detail. Specially for future TV sets, which will no
doubt, be expecting greater quality or resolution. I do
realize that users want to fit *completely* a given source's
contents (ie, movie) on one CD (DVD-R) and a 352 x 480 is the
choice that other here recommend. And, that may be ok in
some cases, but if users are looking for *quality* then, this
suggestion is not a answer I would recommend. I would always
recommend 720 x 480 for DVD, where quality is the goal.
.
I especially feel this 720 x 480 should especially apply for
VHS sources.. mostly because the detail is already less than
that of Cable/TV/Satellite/DVD etc. Lowering it for 352 x 480
or to fit on a disk is not truely giving any serisous benefits.

So, where am I getting at, when I introduce size in this
topic ? Well, I feel that to some degree, its tainting the
color detail to some degree w/ VHS. VHS has lesser color
detail that other most sources. Lowering res adds to this
phenomina (imo) so I thought I'd throw my thoughts about this
in here

And, what gshelley61 wrote..

The JVC DVD recorder basically captures exactly what is sent into it... in other words, it creates very nearly an exact duplicate of the source as far as color, contrast, sharp

..it has a little trueth/support to what I just mentioned above.

The more bitrate, the more detail. So true. The less bitrate,
the the less detail (you see more blocks, resulting in loss of
detail) ...

So, I also believe that *specially* when we are testing issues
out (color detail, for instance) we should use good bitrates
in our test encodes, if only to put *doubt* out of our minds.
.
I believe in high bitrates. Because it gives maximum detail (and
enjoying results)

Anyways.

Another thing I'd like to note here..

Is about external devices (filters, enhancers, etc) I want to
eliminate this all together. (Unless a given source requires
it in the area or name of restoration purposes. But
for normal every day capturing encoding and authoring purposes,
we should not become cripled/dependable on external sources.
I stopped using my SIMA SCC device. I need to be free from
such things that (IMO) upset the equation. This color detail
issue has me thinking. Start with the very bare bones basic,
to eliminate any doubts. TBC, Enhancers, etc. I have now
eliminated. And, now I have very strong convictions that
I have approached the 95+ % mark in reproducing the sources
orignal look (detail, etc) by this move.
.
Well, something to think about.

Ok. So, I'm gonna do some testing to.. and see what I come up
with.

-vhelp

[BrainStorm69]

Let me know what you find out.

With respect to my encoder, I'm talking about the avi capture, before the mpeg encoding process. Although now I guess there's another variable to consider for the final finished product

[FulciLives]

I capture AVI format with a PCI capture card so my capture is YUY2 to begin with and then I use AviSynth because I like the capabilites of it (and the filters for it like Convolution3D) and it keeps everything YUY2 and then I encode with CCE and it is YUY2.

TMPGEnc Plus uses the RGB colorspace.

So if anything one should get a "closer to original look" in terms of colorspace with a total YUY2 chain than having to change colorspace by using TMPGEnc Plus (VirtualDub also is an RGB environment).

- John "FulciLives" Coleman

[vhelp]

Sorry guys. I was having trouble this evening w/ my VCR and Cables.

I finally got things sorted out, but I'm pushed to the limit, and my head is
about to explode. However, I was able to fiddle a little w/ a previous
recording of my DVD movie, "The Fifth Element". This is a 2.35:1 aspect ratio
movie, recorded to vhs in letterbox mode on my AD-1500 player setup.

AD-1500 -> VCR (JVC S-VHS HR-S3910U)

Because of the "black level" issue w/ this setup, I had a pretty hard time
with a few capture w/ my ADVC-100 device. Afterwards, I found out that I
had forgotten about my IRE setting, and switched it back to 0 (which is
the lighter side of things) And, after another re-capture w/ the advc,
I found a slightly better turnout in quality. Still though, the color
seems to be washed out a bit. And so, it doesn't seem like a good gauge
(dvd to vhs, then capture from vhs) But, anything else recorded to vhs
is very ok. Now, for another gauge.

I did a little playing around w/ CCE vs. TMPG (just short scenarios) and
I found that CCE was fine w/ a Canopus saved AVI ( fourCC string "cdvc" )
and fed inside CCE. Go figure. Unfortunately, I couldn't IVTC the DV
source w/out having to re-compress it to DV again - which would void
the scenario/test. So, I continued to stay w/ TMPG, because you can do
pretty much everything.

@ BrainStorm69

I had lots of trouble gauging the color (as you did/have) but in the end,
I was able to get a fair output (from AVI to MPEG-2) though the final
source looked washed out a bit. Still, within guidelines, but washed out
none the less. But, what I did like in my process (and results) were that
in dark scenes (where the ship takes off, before Coby is put to sleep)
the scene was very fluid, and not "blotchy looking".. washed out or not.
.
I think it was the codec and the pre-prep of TMPG for DV encoding that
made the difference. Still though.., it wasn't a fair scenario due to
the taintiveness in the issue w/ the "black level".

I wish I had Analog Cable, and HBO or some movie channel that we
could use (gauge) and capture the same programs/movies for testing purposes.
But I don't. My only source is (as you know by now) Antenna.

Oh yeah.., I did try my hand at my two tests (one above, and huffy) and
my result w/ huffy was not a good one. I used Smallville as my content.
I also recorded it to vhs tape (SP mode) for future reference.
.
After a few test w/ huffy and vhs, I used my ADVC-100 (cause it does a
fantastic job of vhs tapes recorded by my vcr) and I found that it did a
much better job w/ min final encodes vs. huffy.

I'm still testing, but I'm pretty darn bushed out for the time being.

I could try and U/L a small piece of my results.. (for curiousity sake) ?

Cheers,
-vhelp

[vhelp]

IMO, I think its virtually impossible to gauge by eye
the color level of your captured AVI source against tv sets.

The best pre-prep of color setting that can be done is with
TMPG's color filter box. When you turn on [x] Overlay
*there* is where you should judge. But, unfortunately, you can't
even do this, becuase the overlay on most cards are not the same
as that of TV set tubes. However, its in *this* overlay mode,
that you should adjust the color under, because it will *turn up*
the color level to that closest to the tv set's. When you turn
on the overly in TMPG (to adjust the color levels) you are in
effect, getting an accurate and final value for each color
platform you choose. (ie, Settings; YUV; CMYK etc) I use the
CMYK, because it offers the closest to natural colors. However,
only the ATI-Rage Fury Pro cards can best utilize the Overlay
feature that TMPG features (and we see more detail than most
other graphics cards do)

So, in short, you are adjusting the color at the wrong end of
the chain. Do it at the end, where you're about to encode it,
inside TMPG (and under overlay mode)

-vhelp

[BrainStorm69]

So, in short, you are adjusting the color at the wrong end of the chain. Do it at the end, where you're about to encode it,
inside TMPG (and under overlay mode)

Another interesting thought...but isn't that going to add quite a bit of time to the encoding process?

[vhelp]

You mean, if you use the color filter that TMPG provides ??

No, not really. But, that largely depends on the codec you are
using, and the color filter(s) you choose, during the encoding stage.

But, if you have a well tuned AVI and CODEC, than
in most cases, color settings won't be needed. But,
it seems that VHS sources do

In my test sample, I use CMYK, and only used the last seting as -3,
and that was it for color settings.

Don't forget, my source is DV. I really like it, because I'm starting
to beleive that it's 4:1:1 color space is not such a bad space.
the quality I obtain from this format is very good. So, I think that
this 4:1:1 bug (though it does exist) is over hyped too. Actually,
you can eliminate the bug, inside TMPG and w/out using any 411
filter like vdub or AVIsynth

Like I said, I want to eliminate as many in-between steps as possible.
That means frameservers and external devices.

Note, you never want to you the HSV filter in TMPG. It's snales speed.

-vhelp

[vhelp]

Ok. Here are some pics from my process, demonstrating the results
after a pre record from a DVD movie, which has the "black level"
bug issue.

The pics are from a recording I made from my AD-1500 DVD player to my
VCR on standard VHS tape in SP mode and captured w/ my ADVC-100 device.

(I used my own style of cropping and Aspect Ratio 'izing to 2.35:1 AR)

I encoded it to 16x9 format, and ran it through vidubMOD and copy to
clipboard the window image, and imported into paintshop pro, saved it to
a .PNG file (lossless) No resize or other nonsense was done to it. (It's
bad enough, so theres no need to make things worse)





These pics are from the MPEG-2 I encoded, using the above setup.
Oh, and the 4:1:1 bug is gone (thanks TMPG) .. see the red light

-vhelp

[vhelp]

Before I hit the hay (bed) I thought I'd throw out a thought that had
accured to me.

I have the AD-500 (much older dvd player) remember ?
Anyways.

This particular model had its video source on the light side vs. my
AD-1500 's dark side. I'm wondering if I perform the same DVD to VHS
transfer, would I have this same "black level" bug, using the same
process I did above. hmm... I'll have to try. But it's too late
and I have to go ta bed. I'll try tomorrow maybe. This is very
interesting to me now

Cheers,
-vhelp

[vhelp]

I haven't ben around here to post my finds w/ my AD-500 player
and other tests, but..

I finally did some tests, and I found that the AD-500 does play
the video in a different "black level" plain. However, when I
captured it w/ my ADVC-100 it still have the weird color problem
and also, it was "noisy". Very strange. I did have my setup a
bit weird to begin with, and wires going from other ends of the
room, and that could have tainted the final capture result. I'll
have to do another test run. But I wanted to report back the
result of this planned event. I'm sorry my theory did not pan
out

I have the full screen version of "The Fifth Element" I think
that what I could do is set the dvd players in fullscren mode
(if that's possible) and capture like that and compare against
my full screen vhs version. I don't know. I'm still working out
my *own* set of frustrations.

I also think that I should continue the test scenarios, but under
the same attributes - ie, 4:3 apsect ratio of the tv. The capture
is based off of this aspect.

Anyways, I thought I'd keep (bump) this topic w/ my findings thus
far.

-vhelp

[vhelp]

Just an update to my own tests.

I did a test scenario, this time w/ my TRV22 and to my surprise, the (so
called) "black level" issue was not as evident as in my ADVC-100 captures.
.
On that note, I'm going to run an experiment w/ my ATI-Rage Fury Pro
capture card, and see how it fairs up w/ this phenomina. I'm going to try
various codecs, and then its built-in YV12 captures (you can only do this
w/ ATI's MMC app, unfortunately)

Until then.
-vhelp

[FulciLives]

Just an update to my own tests.

I did a test scenario, this time w/ my TRV22 and to my surprise, the (so
called) "black level" issue was not as evident as in my ADVC-100 captures.
.
On that note, I'm going to run an experiment w/ my ATI-Rage Fury Pro
capture card, and see how it fairs up w/ this phenomina. I'm going to try
various codecs, and then its built-in YV12 captures (you can only do this
w/ ATI's MMC app, unfortunately)

Until then.
-vhelp

I am going to solve your "black level" issue.

It really is not that hard but you make it complicated by refusing to stick to the standard proceedure.

Each capture device will be different but I know you like your Canopus ADVC-100 so let us go with that capture device.

First of all you refuse to set it up correctly. If the source is 7.5 IRE BLACK such as NTSC (other than Japanese NTSC) then set it to 7.5 IRE setting. If the source is 0.0 IRE BLACK (all PAL and Japanese NTSC) then set it to 0.0 IRE setting.

The Canopus ADVC-100 has the Canopus codec inside as it does the conversion.

So you must use the Canopus DV codec on your computer. Load the captured DV AVI into VirtualDubMod (or whatever flavor of VirtualDub you like) because this is one way to check what DV codec is in use. Do whatever you have to do (might have to remove some codecs) so that the Canopus DV codec is being used.

Now I know you are found of using TMPGEnc Plus as your MPEG encoder. I am NOT found of this encoder because of the strange behavior it has when it comes to luminance levels BUT it actually works REALLY WELL with the Canopus DV codec.

Load the DV AVI directly into TMPGEnc Plus. For this "test" do not use AviSynth and do not frameserve through VirtualDub etc.

Now in TMPGEnc Plus (do this before loading the AVI file) go to OPTION across the top and under that click on ENVIRONMENTAL SETTING ... to bring up the ENVIRONMENTAL SETTING window.

On the GENERAL TAB make sure you check mark "Interpolate YUV data from 4:1:1 to 4:4:4"

For the option under that "Set equation for color space" select "Basic YCbCr (Default)"

Now when you load the AVI and are setting things up make sure you DO NOT check mark "Output YUV data as Basic YCbCr not CCIR601"

If that does't work I will shoot myself in the head! :P

Now I realize that loading the DV AVI directly into TMGEnc Plus may not always be what you want to do. You might want to do some simple editing or apply a noise filter or perform an IVTC. If you have to do any of that DO NOT use VirtualDub. DO NOT EVER use VirtualDub. It is an RGB environment and will just muck things up. You should be able to use AviSynth just make sure you only use YUY2 filters (no colorspace change) and despite the popular belief you do NOT want to use the ConvertToRGB24() option at the end of the script. You might actually have to use the ConvertToYUY2() option but you might not. I admit I am not 100% sure how AviSynth might "throw" off the luminance. I don't think it will though.

Try it and then either praise for me being correct or curse me for being wrong

I admit that this method is something new to me which is why I never mentioned it before but it should work perfect for keeping the properties of the source (such as the luninance levels) looking the same down to the final MPEG-2 DVD spec encode.

- John "FulciLives" Coleman

[vhelp]

Hi fulci

I am stuborn to change (specially when I *knew* it won't work) but
I'll give the 7.5 IRE thing a go. Mind you, the source will be
much more darker. I know. It will. But, I'll try it (just for the
heck of it)

The source is VHS ( remember ?? ) A dvd recorded to VHS (as a test)
and then captured from w/ various capture cards. This "black level"
issue started ( thanks to me ) when I discovered that I could not
obtain the same level of color space (reproduction) from a VHS capture
that had a DVD movie recorded to it (from a home dvd player)
And, I was on about it, because I thought that it would serve as a
good test w/ respect to brainstorm69 's issue w/ getting the right
color from *his* vhs recording. I hit up against a bump, and blaimed
it on the "black level" bug I read elsewhere's on the internet. After
that, I've ben on my own trip (search) for the resolution to this.
It became *personal* to me.
.
Then, I coudn't get my DV (ADVC-100 nor TRV22) to produce the same
quality as a "DVD rip *and* encode" process. The results were always
washed out or really blochy look colors. This begain driving me crazy.
My DVD rips always had looked like the original DVD VOBs. (when I do
them *MY* way) And, my DVD -to- DV (cam/ADVC) too, came out the same.
But in my trials and tribs, I got some things screwed up in my system.
And I lost my abibilities. Well, I'm happy to say, I have them back.
Everything works as they once did before. Now back to VHS again...
.
I don't know for sure (cause I didn't ask) if brainstorm69 's color
issue is from the same cause (dvd to vhs, then captured) That was
*mine* (again, as a test.. to help out w/ the color issue) What I
was going to do was demonstrate how it was not necessary to adjust
the colors for VHS sources. There's how the snag started. Now you
know. But, in his case, it might not be such. I mean, his tapes
might just be recordings from TV; Satellite; Cable; Antenna; etc.
I just never asked (I don't think) And, in *that* case, you don't
have to adjust the colors. I have no issues. I don't adjust the
colors w/ my VHS -to- ADVC-100 capture and encode. With DV, and the
proper codec setup, no adjusting of colors is ever needed.. using
TMPG of course.

@ brainstorm69

I'm sorry about the confusion. Really.

I think I should have asked the question.. "what was recorded on
these VHS tapes ??"
.
And then go from there. (I just wanted to help)

I have a question or two, for Fulci, but I'll have to start another
thread on this..

-vhelp

[racerxnet]

This is a great post concerning the calibration of the PC monitor and TV. I would bet that 95% of consumers just plug and play without giving much thought to output characteristics. I use my Tektronix scope extensively for this purpose before I even start the capture process. I posted several links to the test patterns previously and hope they are of use to someone else. The Pluge and color bars are output to the TV via the Rex software or scope, helping set the brightness and contrast. Its been great at the OFFA site for upgraded firmware on a variety of DVD players where the added features truely help in playback quality. In particular, the IRE setting for the Apex 660 player and the hard drive capability. A work of art when one has the desire to improve the capabilities of the player.

http://www.epanorama.net/links/videotest.html
http://www.adamwilt.com/DV.html#EmailListsLink
http://www.bores.com/courses/intro/basics/1_antia.htm
http://www.adamwilt.com/DVvsMJPEG.html
http://scanline.ca/ycbcr/
http://www.cs.tut.fi/~leopold/Ld/VideoFormats.html
http://www.smpte.org/


MAK

[gshelley61]

Those are really informative links!

Here are a few test patterns just for fun...


SMPTE Bars



EIA resolution from the late 50's



Carmen Miranda



Indian Head



BBC

[edDV]

Luv those SMPTE bars, ther is so much information represented in those colors.

[Epicurus8a]

Thanks to racerxnet & gshelley61 for the links & posts. I haven't seen that Carmen Miranda chart in many, many, years!

[edDV]

Epicurus8a, do you remember whether it was Miranda or Link that served beer from a BNC connector at IBC? As I recall, it was component beer.

[racerxnet]

I guess this fits in on the restoration of old DVD players that most feel are beyond the useable life span. Anyhow, here is the link to OFFA on modded firmware for a variety of players. Not only is it better, but more flexable in its setup. You do have to register to download the files.

http://groups.yahoo.com/group/OneFirmwareForAll/

MAK

[BrainStorm69]

Damn, I spend so much time in the capture forum that I almost forgot about this thread. Thanks to all who have been working on it in my absence

@vhelp - this thread of mine did not come out of DVD taped to VHS. It actually just came from my own musings about how to know whether the color I end up with accurately represents what was on the original source.

I'm going to have to check out some of those links above when I have a little more time.