Warm Cards - www.warmcards.com

[Bill Davis]

I'm clearly not doing a good job of communication here.

I'm not advocating any sophisticated color grading in post. For most video projects that's overkill.

And I'm definately NOT not talking about scene by scene correction here...

A "pushed" warm-card or cool-card color balance is a GLOBAL change that affects the totality of EVERY frame thereafter.

My contention is that with a modern NLE - there is simply NO advantage to doing that in the field.

As long as every frame is consistent as to camera white balance - the post white balance tools are so incredibly easy and simple to operate in an NLE, that the effort to dig out and use so called "warm cards" or any other artificial color alteration method is as silly as using a hand crank to start an automobile.

If you want ALL your footage a touch warmer or cooler - just park your playhead on a still, call up the NLE's color corrector - correct to taste, then copy that setting and paste it to the rest of your timeline.

It's maybe a 15-second UNDOABLE process.

Look, in this brave new world where all our video images are streams of numbers, my opinion is that futzing with color in the field is a waste of time. ANYTHING you can do in the field regarding color correction, you can re-create simply and easily in post - because you're computer will have WAY MORE processing power than whatever white balance electronics can be shoehorned into your camera body. So why not use the better tool?

Take the minute you'd use for warm card use and apply that to getting a better EXPOSURE. Or re-thinking your framing. Or re-considering the worn out icebreker you want to do to get to know the nice lady/gent doing craft services --- you know - something you CAN'T CHANGE as easily after the fact in post.

FWIW.

[Jack Walker]

Originally Posted by Bill Davis

My contention is that with a modern NLE - there is simply NO advantage to doing that in the field.

There are two reasons I can think of off hand:
1. Time. Depending on how long the video is, there's an additional half hour to several hours of rendering time when making global changes in the computer.
2. With native HDV isn't it advisable to re-encode as little of the footage as possible to maintain quality?

[Matt Davis]

Originally Posted by Bill Davis

I'm clearly not doing a good job of communication here.

I think you are, we've got similar points of view. We're both advocating spending time doing a proper white balance, but sometimes, one may need to push the reds a little.

If you don't push the reds at acquisition time (i.e. shooting) you'll have to take what you've got and pull it at post production.

Any 'filter' REMOVES data in post. A filter or effect in the camera should be about optimising what gets put onto tape.

Formats such as DV and HDV (8 bit 4:2:0/4:1:1) lack a lot of chrominance detail. Boy, it needs all the help it can get.

Therefore, IMHO, with these low end formats sometimes it's best to record something that's close to what you want in post, because to yank what you'd otherwise get to where you want it to be will stretch the limited data too far. I'm not saying Warm Cards should an 'effect', just a useful kick in the right direction like (maybe) a ProMist or Daylight could be.

And please one day just try using the lowest version on a cloudy day or in shade some time, and compare with even an adjusted preset WB. More pleasing, more rounded colour balance...

But the nurse says I must rest now...

[Benjamin Hill]

Originally Posted by Bill Davis

Look, in this brave new world where all our video images are streams of numbers, my opinion is that futzing with color in the field is a waste of time. ANYTHING you can do in the field regarding color correction, you can re-create simply and easily in post - because you're computer will have WAY MORE processing power than whatever white balance electronics can be shoehorned into your camera body. So why not use the better tool?

Well, "better" is subjective, but people with a lot more credibility than I have argued the exact opposite of what you're saying here. Scott Billups has a write-up in his book "Digital Moviemaking" about color timing and how doing this in-camera vs. in post can actually be much higher quality; he includes histograms to make his point in a very scientific way for those who are skeptical- but I didn't believe it until I tried it.

It really makes sense, because before your video is compressed and recorded to tape, you are theoretically working with unlimited information (variables like sensor size, glass and operator skills notwithstanding). When your signal is compressed down to HD, HDV or DV etc. then you are actually losing information that can't be restored. So color correcting your footage in post in the NLE certainly isn't the same thing, unless you are recording full uncompressed video which most of us aren't. For people who know the look they want and know what they're doing, warm cards are a smart and fast way to get great-looking images from harsh formats like DV. They've done it well in in "Always Sunny in Philadelphia" on FX, so it's like any effect or technique, it's all about how it's used.

[Bill Davis]

>Any 'filter' REMOVES data in post. A filter or effect in the camera should be >about optimising what gets put onto tape.
>
>Formats such as DV and HDV (8 bit 4:2:0/4:1:1) lack a lot of chrominance >detail. Boy, it needs all the help it can get.


Perhaps I don't understand DV storage data correctly, but I've always understood that that digital video typically stores pixel color values as RGV values in a range of 0-255 each for the three color values.

If a "filter" merely takes all the red pixels in a scene that have a value of 200, 40, 40 - and shifts them to perhaps 220,20,20 - how does that "filter" the data?

And as to chrominance detail, isn't that a fixed aspect of the color encoding step? No matter WHAT white balance you set, doesn't the color information always get "Mapped" to one of those 255X3 steps?

Are you saying that shifting white balance at the encoding stage somehow expands or shifts the tonality of the colors that the RGB data can describe and reproduce? That seems counter-intuitive to me.

Then again, I'm not a digital video engineer, so I'd welcome someone who knows how this works to set us all straight.

If I have time over the holidays, maybe I'll drop a note to Adam Wilt and see if this RGB value shift via white balance idea makes technical sense. But right now it seems pretty suspicious to my lay mind.

[Daniel Browning]

Originally Posted by Bill Davis

>Any 'filter' REMOVES data in post. A filter or effect in the camera should be >about optimising what gets put onto tape.
>
>Formats such as DV and HDV (8 bit 4:2:0/4:1:1) lack a lot of chrominance >detail. Boy, it needs all the help it can get.

Perhaps I don't understand DV storage data correctly, but I've always understood that that digital video typically stores pixel color values as RGV values in a range of 0-255 each for the three color values.

You are not properly accounting for the effect of compression. Having 0-255 for the three color values of each pixel would take 1,488 mbps, but HDV is just 25 mbps. Applying filters to such highly compressed data *does* result in loss, although you and your audience may not notice it because of your circumstances.

Originally Posted by Bill Davis

It's as silly to my brain as that "should I overexpose or underexpose all my shots" thread.

Expose properly - Always!

Overexposing a low-light scene by three stops and darkening it in post will yield far less noise than one exposed for 18% grey.

[Matt Davis]

Originally Posted by Bill Davis

Perhaps I don't understand DV storage data correctly, but I've always understood that that digital video typically stores pixel color values as RGV values in a range of 0-255 each for the three color values.

DV and HDV use a YUV colour space - ack, yes, there's supposed to be teeny little 1s in there somewhere to signify something about being logarithmic, but what you describe is uncompressed RGB video at 8 bits per channel (10 bits per channel yields - correct me if wrong - 1024 steps per channel). And then there's something about not using all 256 steps for video - so really only using 16 to 235 or thereabouts...

It's at this point I rapidly start gesticulating towards such luminaries as Adam Wilt and Graeme Nattress who have published some great articles on this stuff that explains far better and with rather more authority:

http://www.adamwilt.com/DV.html

http://www.lafcpug.org/Tutorials/bas...ma_sample.html

As the late John Peel once said: "It may not blow your mind, but it will breathe somewhat heavily on it".

[Peter Moretti]

Originally Posted by Bill Davis

I'm clearly not doing a good job of communication here.

I'm not advocating any sophisticated color grading in post. For most video projects that's overkill.

And I'm definately NOT not talking about scene by scene correction here...

A "pushed" warm-card or cool-card color balance is a GLOBAL change that affects the totality of EVERY frame thereafter.

My contention is that with a modern NLE - there is simply NO advantage to doing that in the field.

As long as every frame is consistent as to camera white balance - the post white balance tools are so incredibly easy and simple to operate in an NLE, that the effort to dig out and use so called "warm cards" or any other artificial color alteration method is as silly as using a hand crank to start an automobile.

If you want ALL your footage a touch warmer or cooler - just park your playhead on a still, call up the NLE's color corrector - correct to taste, then copy that setting and paste it to the rest of your timeline.

It's maybe a 15-second UNDOABLE process.

Look, in this brave new world where all our video images are streams of numbers, my opinion is that futzing with color in the field is a waste of time. ANYTHING you can do in the field regarding color correction, you can re-create simply and easily in post - because you're computer will have WAY MORE processing power than whatever white balance electronics can be shoehorned into your camera body. So why not use the better tool?

Take the minute you'd use for warm card use and apply that to getting a better EXPOSURE. Or re-thinking your framing. Or re-considering the worn out icebreker you want to do to get to know the nice lady/gent doing craft services --- you know - something you CAN'T CHANGE as easily after the fact in post.

FWIW.

What NLE are you using? I ask this in all seriousness, as I'm still learning Avid and I don't know how to do what you're saying as a change that takes seconds.

[Heiko Saele]

From my experience with DV/DVCAM and FinalCut Pro, no quick NLE adjustment to color has the same quality as an in-camera correction.
Apart from the visual quality, there are a lot of traps, the best example would be an overexposed part of the sky - the whole frame is exposed correctly, there's just one part of the sky that blows out. It doesn't matter in the original frame (because it's white), but try to color-grade that frame and every change you make to the the whites will color that part of the sky accordingly. Suddenly you have a yellow/orange spot in the sky. You'd have to mask the sky to avoid this behaviour - not possible if it's 5 p.m. and the piece has to be on air in half an hour.