Waveform and Vectorscope

[Eric Steelberg ASC]

Can anyone recommend good resources or tutorials for using a Waveform and Vectorscope in HD? I'm starting an HD feature and we can't afford a DT so I gotta do it all myself.

 

Is there anything like "Goodman's Guide: SDX-900" for the F900?

[Phil Rhodes]

Hi,

 

I've never understood why there's any confusion about this. Waveform is a graph of exposure composite of all the lines in the frame; vectorscope is an indication of saturation - the bigger the radial spike, the more saturated the image is becoming. Watch for clipping. Make sure everything's where you want it. What's the mystery?

 

Phil

[Tom Schlodder]

I'm finding myself somewhat baffled by the simplicity of the vectorscope as well :)

 

Well actually, it's with the YC Waveform chart.

 

I understand what I'm looking at, but translating that into actually tweaking the picture is a different story. Is there a common method to bringing the levels down to 100 IRE (or just below that point)?? For example: start by adjusting highlights, then midtones?

 

Turning on the video limiter clips the levels off at 100 but the resulting image is looks awful (which is no surprise.) Is there a rule of thumb as to when use the video limiter (such as: when the levels only need to be droped or raised by 5 IRE?)

 

I realize how incredibly simple-minded these questions must sound and I apologize for hurting your eyes and brain with them :)

 

Assume I know nothing and start explaining from there (I would think that there are many others like me out there that would appreciate a definative explaination) or refer me to a book you might know of.

 

Thank you very much in advance for the assistance.

[neal]

I understand what I'm looking at, but translating that into actually tweaking the picture is a different story.  Is there a common method to bringing the levels down to 100 IRE (or just below that point)??  For example: start by adjusting highlights, then midtones?

 

Turning on the video limiter clips the levels off at 100 but the resulting image is looks awful (which is no surprise.)  Is there a rule of thumb as to when use the video limiter (such as: when the levels only need to be droped or raised by 5 IRE?)

 

<{POST_SNAPBACK}>

 

The best way to bring the levels down below 100 is to change the lighting of a scene itself, not the signal level.

[Michael Nash]

Is there a common method to bringing the levels down to 100 IRE (or just below that point)??  For example: start by adjusting highlights, then midtones?

<{POST_SNAPBACK}>

 

Don't overthink it, just use common sense. A waveform is just another representation of the luminance you see on the screen. If something's over 100 IRE on the waveform monitor, you'll look at the picture monitor and go, "oh my gosh, those highlights are blown out!" What do you do? Use common sense: stop down your exposure; remove the highlighted object; or adjust the signal processing in camera to handle the highlights better. There's no magic or precise methodology to it. You're just looking at the luminance of the image represented as wavy lines on a graph, instead of as an image on a monitor.

[Nathan Milford]

Imagine having a constant, light meter reading of EVERYTHING that is in frame.

 

If you were using your meter and you were a stop over where you wanted to be on a particular highlight, you'd use a variac, net, dot, finger or whatever to reduce the light.

 

If you are looking at the waveform and you see a highlight 20 IRE over where you'd want it, you'd do the same.

 

It's pretty Mickey Mouse.

 

- nathan

[Leon Rodriguez]

There's a pretty good basic tutorial in the American Cinematogrpher Video Manual called Test Equipment in Digital Video Operations by Gerald McGinty.

 

Michael's right. Don't overthink it. These are just graphical reflections of the image itself which you will approve or not. But having said that, good reference points are important.

 

Get a good vectorscope chart to set your color to. The ChromaDuMonde 24 with Resolution Tie Bars from DSC Labs will do everything you need. It comes with instructions!

[Bill Totolo]

This is probably redundant but I would add that it would be helpful to send bars from camera to your waveform/vectorscope upon setting it up.

 

This way you have confidence in the colors you are seeing on the monitor and the exposures you are setting.

[Jason McKelvey]

Unless I'm using multiple cameras that I need to match, I don't use either. Look at a monitor... does it look great? Great! Does it look bad? Iris down/up, reduce/add chroma of R-G or whatever. Saturation clips will be obvious. If luminence clipping over 110 IRE, but you want it to look blown out and you're happy with what you SEE on the MONITOR, bring it to NTSC safe levels in post... don't use the cameras circuity. My reasoning is that the CCDs will react to white clip setting very different (unflatteringly) than a NLE software white clip adjustment. If anyone has a technical reason not to do that, I'd like to here it. If the blowout looks bad on the on set monitor but you still want the highlights to blow, then make changes to the knee point or slope. I'd love to see a DP set a camera to scopes without an on set monitor.

 

J

[Tenolian Bell]

I know some guys who were shooting a music video on the F-900. They wanted to saturate the colors like your standard MV. So with the paint box and the monitor they played with the colors until they got what they wanted. Without a vector scope however they had no way of telling if they'd stretched the colors beyond color gamut.

 

Once in post and editing the MV, the colors were distorted, bled, aliased, and had red moire. Because they'd stretched it beyond color gamut and didn't realize it.

 

They tried to do some HD color correction but still could not fix the problem, the colors were just gone.

[Tenolian Bell]

Recently I was helping out a friend and worked as an electric on his short film which was being shot on the Varicam.

 

I'm not sure of the DP's previous experience with HD, but they were shooting with a monitor only no waveform or vectorscope.

 

I think the DP is rather new to shooting, I believe he works as an electric most of the time. Because he didn't seem to have a great deal of confidence in his eye, he really depended wholly on the monitor and was constantly relighting everything.

 

The director want a pretty low key, dark feel for the lighting. The DP took it literally and crushed the blacks as we were shooting, really just by not lighitng the shadow areas. I became concerned has to how deep he was letting the blacks go because he was leaving no way to retreave the shadows in post.

 

I suggested to him that if we were looking at this scene on a waveform monitor the large mass of black behind the actors at times would be below 0 IRE, and detail in that would be unrecoverable in post. I don't think he totally knew what I was talking about and said he didn't want any detail recovered.

 

Even when shooting film if I want the blacks to be completely crushed I don't crush them with exposure, I crush them in post.

 

The project is now in post and I heard the director wanted to do a 35mm print blow up, but there may be a problem with it because the shadow areas are too underexposed.

[Phil Rhodes]

Hi,

 

Thing is, if he was lighting to the monitor (which is not incorrec) and if he liked what he saw, then there's nothing wrong with the way he's shooting other than it's potentially risky if you later decide you don't want it that crushed. If he's happy with the blacks as they were on the monitor, the fault lies with the filmout technology. Personally I'd expect problems in this area due to the inability of film to record any light that isn't blasted into it with the force of several express trains, but even then you should be able to pedestal it all up until the black point on the video matches the black point on the film. Then all you have to worry about is the stock and processing, which I am informed in real-world situations will be within half a stop by the time you've got to a print.

 

Phil

[Jason McKelvey]

Without a vector scope however they had no way of telling if they'd stretched the colors beyond color gamut.

Once in post and editing the MV, the colors were distorted, bled, aliased, and had red moire. Because they'd stretched it beyond color gamut and didn't realize it.

<{POST_SNAPBACK}>

 

Please help me understand why the colors were "distorted, bled,..." etc. on the monitor in post, but didn't show up on the on set monitor???? Sounds like it wasn't a matter of needing a waveform/vectorscope per say, but needed a better production monitor.

 

Jason

[kalunga]

The Los Angeles Final Cut Pro User Group (LAFCPUG) reviewed a DVD based tutorial by Intelligent Assistance, entitled:

 

Practical Color Correction For FCP 3 and 4, OS X

 

(60 minutes of Tutorials, 70 searchable 'How2s' and Advice pages)

 

 

Unfortunately it is MAC only, and I belive only available through the LAFCPUG store and costs $35.00, but the intersting thing about it is that there are dedicated sections on how to read waveform and vectorscope monitors with examples. There is also a section explaining what can actually be corrected in post and what can not.

 

 

http://www.lafcpug.org/reviews/review_cc_ia.html

 

best of luck

[Tenolian Bell]

I don't know. I wasn't there on set when they'd shot it. I just saw what the video looked like as they tried to edit. The DP and director claimed it looked fine on monitor on set.

 

They know about calibrating a monitor, and said they'd done that. But the state of the monitor on set I can't attest for.

 

After seeing their reults I do believe without a doubt it would be most essential to have a vector scope when doing extreme color saturation in camera. The danger of the end result is completely useless video.

[Phil Rhodes]

Hi,

 

Hmm, all I need now is for someone to commission me to produce some Premiere tutorials...

 

Phil

[Felipe Perez-Burchard]

Another good book that will explain understanding waveform, vectorscopes and histograms is Color Correction for digital video by Jamie Fowler and Steve Hulfish.

 

-felipe.

[Bill Totolo]

Unless I'm using multiple cameras that I need to match, I don't use either.  Look at a monitor... does it look great?  Great! 

J

<{POST_SNAPBACK}>

 

 

Well how are you going to confirm that what the camera is sending to the monitor is accurate unless you have a tool to measure the signal? Hence, a waveform/vectorscope.

[Jason McKelvey]

Well how are you going to confirm that what the camera is sending to the monitor is accurate unless you have a tool to measure the signal? Hence, a waveform/vectorscope.

 

How are you going to confirm what you are sending the waveform/vectorscope is accurate? I've seen that messed up before too, either by bad cables, wrong termination or wrong calibration of the WF/VCS itself.

 

I understand the use of a waveform/vectorscope and it's value on a set. I just wanted to point out that too many times, you end up with a bland, steril picture that is 100% NTSC correct, with accurate color representation, everything in the gamut, white clip where it's "suppose" to be... yuck. As a shooter, I don't like to let the editor have all the fun. Using a vectorscope to find a specific color that you see on the monitor is a time saver, but I still say shoot to a good CALIBRATED monitor.

 

Jason

[Bill Totolo]

How are you going to confirm what you are sending the waveform/vectorscope is accurate?  I've seen that messed up before too, either by bad cables, wrong termination or wrong calibration of the WF/VCS itself.

 

I

Jason

<{POST_SNAPBACK}>

Well, light meters have been used incorrectly and have had thier share of problems also. If you know how to use your tools and you trust your eye you develop a sense when something isn't right.

 

In the end no technology is infallible, you can only do the best you can with what tools work for you.

[Eric Steelberg ASC]

Well thanks for all the responses I guess. I wish I would have gotten them closer to July 19th...when I posted this and could have used the info before I shot the film.

 

In any event I quickly picked it up as it's the essentially the same using it on set as it is in all those telecine sessions. The vectorscope I could give two sh*ts about, but it was the waveform which was most important to check the exposure. I couldn't trust any monitor I saw...two of them were 'calibrated' yet didn't match.

 

By the way I highly recommend the Astro 3003(?) to anyone shooting video. It's an onboard HD capapble LCD monitor with integrated waveform/v-scope.

[Tenolian Bell]

Thing is, if he was lighting to the monitor (which is not incorrec) and if he liked what he saw, then there's nothing wrong with the way he's shooting other than it's potentially risky if you later decide you don't want it that crushed.

 

From expereince I've learned its best to keep your options open during post. Crushing blacks isn't necissarily the end of the world, but can be a pain in the arse later on for various reasons. The director may want to recover some shadow detail for story or asthetic reasons. And crushed blacks can cause technical problems later on because, essentially that means there is no information in that part of the frame, whether its tape or negative.

 

 

If he's happy with the blacks as they were on the monitor, the fault lies with the filmout technology.

 

 

Why would it be a filmout technology problem when there is no information in the shadows for the filmout to record?

[Elhanan Matos]

Thing is, if he was lighting to the monitor (which is not incorrec) and if he liked what he saw, then there's nothing wrong with the way he's shooting other than it's potentially risky if you later decide you don't want it that crushed.

 

Now that's only correct if the monitor is properly calibrated, and under the correct lighting conditions. Having a vectorscope on set is very important if you don't know how to calibrate your monitor or if your not watching it under the correct lighting conditions.

[Phil Rhodes]

Hi,

 

> Why would it be a filmout technology problem when there is no information in

> the shadows for the filmout to record?

 

It wouldn't, assuming that the technology is faithfully recreating what was seen on the monitor. However the original question implied that there was some additional problem caused by the crushed blacks - I would imagine that the problem was due to the blacks being crushed even more by the film, possibly to unacceptable levels. That would be a filmout problem.

 

Phil

[Tenolian Bell]

The problem they are facing is that their is no information in the shadows. So thier is very little to use in that part of the frame for the film out. The film out isn't causing the problem, the problem is lack of exposure during shooting.

 

What the DP didn't realize or perhaps understand is that video works in db and the IRE scale. Looking at a monitor you cannot tell if a dark area is at 10, 7.5, or 0 IRE, nor on a monitor can you tell if a highlight is at 75, 90, or 100 IRE.

 

Of course you do not want to light a flat uninteresting picture just to be NTSC safe. But that is where skill and mastry of your craft comes in. The DP has understands light, understands grey scale, exposure, and control of contrast.

 

If he/she has a grasp of these tools then a room can be lit up bright to the eye with the desired contrast built into the exposure. In post contrast of the whole image can be brought down, shadows and highlights will fall into place because the DP built the desired ratio into the exposure.