PDA

View Full Version : Auto Calibration for sources



mahasf
07-30-2008, 06:33 PM
I do not hear a lot of talk about using the internal calibration tool (utility)for camera setup. If you have not tried it, you should...it works great!! I was demonstrating the tricaster to a friend who shades cameras on large production trucks, and he was amazed at what a great job this utiility did!!!!

Quiet1onTheSet
07-30-2008, 10:11 PM
I do not hear a lot of talk about using the internal calibration tool (utility)for camera setup. If you have not tried it, you should...it works great!! I was demonstrating the tricaster to a friend who shades cameras on large production trucks, and he was amazed at what a great job this utiility did!!!!Cool. :thumbsup: Specifically, what type of colorbars did you have generated from each video source, Alan -- 75% SMPTE? 100% SMPTE? or Full Field?

An inquiring mind would like to know. :D

Q1

mahasf
07-31-2008, 06:22 PM
75% from the cameras

Quiet1onTheSet
07-31-2008, 06:46 PM
75% from the camerasDemoed TriCaster to a friend who's up to his eyeballs in media production experience, and he too was utterly wowed at the ability of NewTek's TriCaster STUDIO product to match up his JVC HD100U cameras spot-on, and in little time at all!

Sheer amazement is on the faces of those who see it first-hand, right before their very eyes.

Q1
:agree:

amospro
08-03-2008, 10:00 PM
For the newbies, could you explain the process?

mahasf
08-09-2008, 07:10 AM
I place all of my cameras on color bars. I then go to the input section of the tricaster and restore all settings for each camera to default using the button--restore defaults, I then use the pulldown "Calibrate" to 75% bars for each camera, take the cameras off of bars and frame my cameras to a similar setup, and tweak.

TIMECARE
08-10-2008, 02:55 PM
Last Saturday we did a six camera shooting activity and we were banging with the best service providers on the island. We were entrusted with the video filming while one of our major competitors was entrusted with the setting up of the big screen to air the activity for the people seated down away from the action. Naturally we had to give them the composit video out. It was a most colourful night, including fireworks, and when we switched from one camera to other, the colours were perfect. They could not believe their eyes and we could noticed they had difficulty breathing comfortable. 3 cheers to tricaster auto calibrating.

GeekNews
11-19-2008, 01:55 PM
I place all of my cameras on color bars. I then go to the input section of the tricaster and restore all settings for each camera to default using the button--restore defaults, I then use the pulldown "Calibrate" to 75% bars for each camera, take the cameras off of bars and frame my cameras to a similar setup, and tweak.

Someone ought to create a video on this so us rookies get a clue..You lost me at "Frame my cameras to similar setup, and tweak"

Thanks in Advance Todd.

ted
11-19-2008, 02:44 PM
Mahasf, You brought up a good point that is often missed. I tried to calibrate once and didn't get the results I expected and remembered that I didn't reset the input settings. I can't say that you have to, but after resetting the input settings to default, I got much better calibration.

Geeknews, what he was saying is after you calibrate your cameras, and take them off color bars, try and frame all cameras similar, i.e. all cameras frame a medium shot of someone.
That will help you manually fine tune all cameras to match.
Hope that helps.

GeekNews
11-20-2008, 03:01 PM
Geeknews, what he was saying is after you calibrate your cameras, and take them off color bars, try and frame all cameras similar, i.e. all cameras frame a medium shot of someone.
That will help you manually fine tune all cameras to match.
Hope that helps.

Makes perfect sense thanks as soon as my box gets back from repair I am going to be experimenting with this. Its topics like this for the rookies like me that go ohhh man that makes so much sense..

Great Thread
Thanks
Todd.

brians0105
11-20-2008, 04:50 PM
... try and frame all cameras similar, i.e. all cameras frame a medium shot of someone.
That will help you manually fine tune all cameras to match.
Hope that helps.

When auto-calibrating and tweaking our 5 or 6 cameras for basketball, we frame all of our cameras on the logo @ center court. The logo includes blue, orange, black, and white... so it gives us a variety of colors. Plus, don't forget to look at the color of the court.

I adjust the colors and settings for camera one to please my eye, then match the rest of the cameras to look the same as camera 1.

Ourdoor games are even more fun (not)! A lot of our games started with the sun setting. We would often have to re-white balance and shade the cameras during commercials or half time. (But this if off-topic since it doesn't really have to do with auto-calibration.)

ditnoj
01-06-2009, 06:40 AM
Hello All!

I know this is an old post, but we're getting a new Tricaster this week so I'm doing some homework.
Am I understanding correctly that you all are starting the calibration process in the box and then tweaking on the cameras manually? Or are there tweaks you are making to individual inputs inside the box?

Thanks in advance!

D out ITNOJ

ted
01-06-2009, 10:34 AM
We always do the auto calibrate on TriCaster using the camera's color bars. Then we make adjustments on the individual inputs in TriCaster.
Make sure and do a white balance on the cameras before adjusting anything on TriCaster.
Was that your question?

ditnoj
01-06-2009, 06:08 PM
Yes, that was my question. Thanks!

pro---studio
02-25-2009, 03:55 PM
Sorry about my dumb question..

But it makes absolutely no sense to calibrate cameras with an generator based feed like color bars out of the camera.

The different pictures (colors, gamma, levels) are a result of not perfect calibrated internal processing of the camera. If you switch on the color bars, this internal processing is disabled. How can you match the cameras with a generator based source???

I would understand it if you frame all cameras on a reference color plate and do the matching then - but with the color bar generator?

Sorry but I've never heard of that method.

Maybe you could explain it to me and clarify my wrong thoughts.

Regards

Pro

ted
02-25-2009, 10:29 PM
The purpose of the calibration as I see it is to calibrate the signal from the far end of the cable all the way through the Tricaster. Thus matching for differences in types of inputs, timing the sources and getting a solid reference to color match. You have color correction to match the actual camera image.
White balance is different, but in practical terms, if you use the same white card and balance each camera from differenet angles, i.e. their set location, you can get undesired results. I was surprised the first time I had this happen. This would be one thing to be aware of if you do use the camera on a color chart.

I agree with you that it would be best to calibrate the actual camera image. If you brought all cameras side by side and Autocalibrated on the same card you would in theory have the best calibration.

I tried to find professional 100% color bar charts for months with no luck. I even tried talking one of the manufacturers at NAB into making one. I think it was Vertek. They were interested, but didn't know if there would be enough demand. If I see them this year I'll tell them you and I would buy one! :D

Remember, all this is JMHO. Based on 30 years of Broadcast experience, but that doesn't mean it's totally accurate. :D

pro---studio
02-26-2009, 05:16 AM
Hi ted,

thank you for answering my question.

It makes totally sense to compensate the quality loss of the (mostly) long cables to the Tricaster. Maybe it's because I work only with SDI connections that I don't had that in mind.

In professional broadcasting situations it is exactly like i described: Framing all cameras on a color plate and then using the CCUs to tweak them that they look equal.

I remember in the old days with valve based cameras (don't know if that is the right word) f.e. "Plumbicon" someone standing there with a little screwdriver (for 2 hours or so) doing a calibration straight in the camera. It was a really long procedure.

Regards

Pro

Quiet1onTheSet
02-26-2009, 12:25 PM
The purpose of the calibration as I see it is to calibrate the signal from the far end of the cable all the way through the Tricaster. Thus matching for differences in types of inputs, timing the sources and getting a solid reference to color match. You have color correction to match the actual camera image.
White balance is different, but in practical terms, if you use the same white card and balance each camera from differenet angles, i.e. their set location, you can get undesired results....

I agree with you that it would be best to calibrate the actual camera image. If you brought all cameras side by side and Autocalibrated on the same card you would in theory have the best calibration.

I suppose we'd also to well to alert any newbies on this subject, to the fact that it's sometimes tempting to discard the fact that you ought to ensure that all the video output cable from all the cameras inputting a signal to TriCaster, are of equal length, for better ease and accuracy in color phase/hue calibration.

Else, it's quite likely that some camera(s) might be utterly difficult to properly match up with other(s), due to an extreme phase incoherency caused by the excessive difference in signal delay, along various cable lengths.

Strive for excellence in camera matching at the outset (beginning with same-length camera cable); for in some cases, the undesirable result might present an insurmountable challenge to fix in post-production (editing).

Quiet1onTheSet
02-26-2009, 12:34 PM
Hi ted,

thank you for answering my question.



I remember in the old days with valve based cameras (don't know if that is the right word) f.e. "Plumbicon" someone standing there with a little screwdriver (for 2 hours or so) doing a calibration straight in the camera. It was a really long procedure.

Regards

ProOh, yeah, the vacuum tube cameras!

NRBC_Media
02-26-2009, 09:43 PM
DSC Labs makes color charts. I bought my last set from Filmtools.com for my day job. Still have to tweak some Thomson LDK20's and their chart works great.

However at the church, where my Tricaster resides, I just use a white card at stage center so all cameras can shoot, set Iris to same level on all cams by waveform, then AWB. So far my cams have been close enough for the waveform/vectorscope and my eye. As they age, it may change.

The Tricaster Calibrate feature is for cable compensation, ie luminance and chroma level, not necessarily related to WB or BB.

ted
02-26-2009, 09:58 PM
I suppose we'd also to well to alert any newbies on this subject, to the fact that it's sometimes tempting to discard the fact that you ought to [COLOR="Orange"] ensure that all the video output cable from all the cameras inputting a signal to TriCaster, are of equal length, for better ease and accuracy in color phase/hue calibration.

I "think" cable length differences are one of the things calibration takes care of as mentioned in my earlier post. But I could be wrong.

ted
02-26-2009, 10:09 PM
DSC Labs makes color charts. I bought my last set from Filmtools.com for my day job.

What we were talking about is an actual 100% Color Bar chart. One that exactly matches the 100% color bars from most cameras.
This would likely improve the calibration since you are actually using the "Cameras" electronics to calibrate from. At least in theory "according to Ted". :D I would love for NewTek to respond since they know the facts behind TriCasters calibration.

I didn't find anything like this on Filmtools.com.

NRBC_Media
02-27-2009, 11:14 AM
The charts I spoke of will do that.
Here is a link for a cheaper one.
http://www.filmtools.com/accu-chart-hdcr-high-definition-color-reference.html

**I am not affiliated with vendor or manufacturer.**

ted
02-27-2009, 09:41 PM
**I am not affiliated with vendor or manufacturer.**

It's obvious you are only trying to help with a resource. :thumbsup:

Now for the "however".
This isn't a 100% color bar chart as NewTek recomends from the source for auto cal.

In fact it reads the following...
the chart is a neutral background of approximately 50% reflectance.
AND...
The chart ... is not intended to produce a display on a vectorscope corresponding to an electronically generated color bar signal.

So no, this isn't what we are talking about or is needed.
I think I might spend a few bucks and create our own, have a bunch printed professionally and sell them. :thumbsup:
I can't believe how much they charge for these things.

csandy
02-28-2009, 07:41 AM
Pro - I've had this thought myself - from shooting in college and throughout my career.

With consumer or "prosumer" cameras, you are right. It makes very little sense or difference if you use each individual camera's color bar generator.

Where it can make a small difference is when you use cameras of the same make and model, but for whatever reason have differing output signatures. Perhaps a not-do-good capacitor that are a few farads off, coloring due to wiring, or an impure conductor from different batches of metal. In modern electronics and signal processing, however - it is unlikely you will see much difference.

More helpful would be if you had a number different cameras with a mix in makes and models. Here you could get a good baseline reference of what each camera thinks is a SMPTE pattern and adjust accordingly.

Where the autocalibration feature IS very useful is with broadcast cameras. For example, I use 20-year-old dockable broadcast cameras that can adjust their output in the head, back, or CCU. Or a combination of all three. So you can color the output in several different ways - even with the same setting of white balance, pedestal, shutter, and filter.

Also, due to the age - the electronics can color the signal as well (yes, yes, yes - I can constantly adjust and calibrate all my cameras over time - but why bother?). When you can simply PRESS THE MAGIC BUTTON!

Yes, the bars change depending on the settings at the head-back-ccu, so making sure those line up are essential.

Now... what you desire is likely a giant SMPTE bar color poster that you can hold in front of all your cameras so you can calibrate off of the signal coming through the lens, so to speak.

Or you could just white balance.

csandy
02-28-2009, 08:37 AM
Here's a very good tutorial on how to use bars for your monitor.

Ted - the reason you likely didn't find SMPTE bars in paper form are because they were never designed to be used that way.

The tutorial link below should have a link to the obit of the guy who created the bars. It's a good history lesson. You're probably heard it before but forgot given your 30 years in the field!

http://www.videouniversity.com/tvbars2.htm

Colin

ted
02-28-2009, 10:47 PM
Ted - the reason you likely didn't find SMPTE bars in paper form are because they were never designed to be used that way.
Colin

But when Color Bars were invented, they didn't know about Tim Jenison or NewTek. :D

csandy
03-01-2009, 03:22 PM
How true!:agree:

PIZAZZ
03-01-2009, 04:33 PM
Simple solution we found in the past to assist in calibrating cameras with the VT/TriCaster, lay off a few DV tapes of bars and tone. Play that tape back in the camera in order to do the auto calibration. This was especially useful for the Canon XL1 back in the day because if you knew the secret to display bars, well they were not SMPTE standard bars.

From what I understand, the calibration in TriCaster and VT are more so to equalize the cable lengths not to adjust final video quality. You can have all the cameras calibrated to bars and still each camera look different because their white balance is off.

Order of Procedure for us:

Put cameras on bars
Calibrate each input in TC
Then focus cameras to stage or central area
White Balance each camera with something white large enough to fill the screen of the camera the furthest away.
COMPARE each camera for final quality test
Tweak Proc Amps if needed.

BTW a handy horizontal wipe helps greatly to sync cameras if you are having problems. Put your reference camera on top and the one you want to adjust on bottom. Wipe halfway. (H on a TriCaster keyboard shortcut)

After all of this is done, pray that someone doesn't drastically change your lighting.
If someone touches your lighting, beat them down with a large bat.

PIZAZZ
03-01-2009, 04:36 PM
I suppose we'd also to well to alert any newbies on this subject, to the fact that it's sometimes tempting to discard the fact that you ought to ensure that all the video output cable from all the cameras inputting a signal to TriCaster, are of equal length, for better ease and accuracy in color phase/hue calibration.

Else, it's quite likely that some camera(s) might be utterly difficult to properly match up with other(s), due to an extreme phase incoherency caused by the excessive difference in signal delay, along various cable lengths.

Strive for excellence in camera matching at the outset (beginning with same-length camera cable); for in some cases, the undesirable result might present an insurmountable challenge to fix in post-production (editing).

Peter's suggestion to always use the same length cable is not realistic.

The kits we built for the NCAA had 100, 250, and 500' runs of cable. It would be ridiculous to use 500' of cable if 100' works just the same.

The calibration tools/Proc Amp in TriCaster are there to assist us with the varying cable lengths. Copper is expensive. Shipping is worse.

ted
03-01-2009, 07:13 PM
Thanks Jef. I figured adjusting for cable length was part of the auto cal process. I also thought video level and color matching (from bars) was part of it too. But I've not found this to be the case.

I still wish someone from NewTek would enter here and give detailed info on Auto Cal from THEIR box. :D

NRBC_Media
03-05-2009, 10:41 AM
It's obvious you are only trying to help with a resource. :thumbsup:

Now for the "however".
This isn't a 100% color bar chart as NewTek recomends from the source for auto cal.

In fact it reads the following...
the chart is a neutral background of approximately 50% reflectance.
AND...
The chart ... is not intended to produce a display on a vectorscope corresponding to an electronically generated color bar signal.

So no, this isn't what we are talking about or is needed.
I think I might spend a few bucks and create our own, have a bunch printed professionally and sell them. :thumbsup:
I can't believe how much they charge for these things.

The gray backgound is 50% reflective, used for setting Gamma levels. Relectance has nothing to do with 100% bars. 100% or 75% bars refers to the chroma to burst level, not overall video level.

The chart is used to confirm your WB, Gamma, and Black Balance, via the waveform/vectorscope, or the half wipe method if you prefer. Auto cal only sets cable compensation. I use auto cal with 75% bars, as that is what my cameras put out. It sets the peak voltages for luminance and chroma, and I think I witnessed a phase correction on the scope, but that is only on the bars output of the camera. The image pickup will vary as discussed earlier.

You will never get a chart that will match generator bars in level, but you can "paint" the cameras to match each other with the chart, and display a correct vector matrix.

xleggs
03-19-2009, 08:16 PM
The calibrate feature is great on the tricaster.

I use it to compensate for the very different cables that my cameras have. Luckily, I'm using three of the exact make and model of camera (Canon XH-A1), but not everyone is. I never assume that my three cameras will output the same exact signal, either- there are noticeable variations in the cameras.

The calibrate gives the tricaster an idea of what to look for in a picture, but I still find myself painting in each camera. I prefer to do that on a real person sitting on the set.

Don't move your cameras away from their positions to white balance- they should stay in the environment in which they'll be shooting and balance on a white card where the main subject will be.

(My first job in TV involved shading RCA TK cameras in 1987, so I have an idea of how far we've come with technology.)

NRBC_Media
03-24-2009, 05:45 AM
I started with the TK-45 and TK-46's. Freelanced with a few TK-47's. They were really nice for painting. Then we switched to Ikegami, then Sony.

I changed stations and they have Thomson LDK's, I really do not like their manuals.

Wish RCA was still around.

celtics23
08-08-2012, 09:00 AM
I'm well aware this is an old post, but I'm dealing with all of the issues discussed here and have a few questions!

I'm using two Sony HDR-FX1 cameras with a Tricaster Pro for a 2-camera shoot. If I understand correctly, the extent of the usefulness of the 75% auto-calibration tool is that it helps the Tricaster equalize the incoming video signals from each source, given different cable lengths. Is this correct?

What is the professional way to match the color from two cameras in the Tricaster? Is it to first auto-calibrate the inputs, then white balance both cameras, and then point both cameras at the same color bar chart and adjust the inputs using the calibration tools in the Tricaster?

Regarding the calibration tools, in what order should the controls be adjusted? I don't want to play around with them and end up with each control in an absurd (high or low) position relative to the defaults.

Any help would be appreciated!

celtics23
08-14-2012, 12:19 PM
Someone respond, please!!!

Lee-AVP
08-14-2012, 12:43 PM
Our process is roughly this: Set a camera as we like it, then use the flash card to copy the scene file out to the other cameras. Our cams go out on rentals a lot, so getting all the menu options and such consistent across the shoot is important.

Then we white balance, then send color bars to the Tricaster and do an auto calibrate. I find this was helpful for analog transmission, not so much with SDI.

We do our final toning to the eye, from the Tricaster. Though, this is hardly ever needed if the cameras are properly white balanced to each other first.

The biggest thing, for us, is making sure the white balance is happening to the same color temperature preset. If not, sure, the whites will look the same, but everything else is drastically different as the camera interprets color with totally different parameters.

I, too, would like to know a little more about broadcast industry process for this.

celtics23
08-15-2012, 02:17 PM
We do our final toning to the eye, from the Tricaster. Though, this is hardly ever needed if the cameras are properly white balanced to each other first.

The biggest thing, for us, is making sure the white balance is happening to the same color temperature preset. If not, sure, the whites will look the same, but everything else is drastically different as the camera interprets color with totally different parameters.

I am using two Sony HDR-FX1 cameras connected to the Tricaster Pro. So, one thing is that you are suggesting white balancing the cameras BEFORE doing the auto-calibration using the color bars. I thought that the auto-calibration was to be done before.

Secondly, is it possible that there are differing settings on the camera that are resulting, given an identical lighting setup and pointing to the same white piece of paper, in differing color? I'm using all manual settings on the camera.

celtics23
08-15-2012, 03:10 PM
Let me say a bit more! I simply don't understand what the auto-calibration tool has to do with the image being outputted from the camera. I have run a test where I recorded the same setup in the Tricaster and onto dv tapes, and the color balance of camera 2 was off in the Tricaster version but not in the footage recorded onto dv tape. Therefore, I figure the color balance issue can be resolved in the Tricaster, and that fundamentally the cameras are color-balanced well.

celtics23
08-24-2012, 01:17 PM
I might as well just keep posting here (despite the conspicuous lack of responses hehe), since my questions are increasing in number.

What exactly does "painting a camera" refer to? This language is used in this thread and also in this product description for a DSC color chart:
http://www.filmtools.com/hd124.html

In general, does anyone know where I can learn more about proper technique for matching cameras (that is, the image) on multi-camera shoots?

Brian Mirrlees
08-25-2012, 12:01 AM
Think of the process for camera alignment/set-up as follows:

i) cameras "go on chart" for the critical alignment or check by the video engineer/operator. This is done under ideal conditions so as to achieve the best possible result...

For instance, light that is used to illuminate the line-up chart is of a single, true colour temperature. You don't want to be mixing light from tungsten or HMI or flourescent sources. As well, you want to ensure that the light is not being "dimmed" as it hits the chart.

This is critical to your line-up, tungsten light will have a nominal colour temperature of 3200 degrees Kelvin. If the AC power is being reduced to the fixture or being dimmed, it will have the affect of lowering the colour temperature of the tungsten filament. It will appear to be "warmer" on the chart, and the camera will see this as being "plus RED". You will then be looking at a false image on the scope/monitors and will remove more RED (gain, gamma, ped) than is actually required, if at all.

Your eyes and the camera will easily see a difference in 300 deg. K to 400 deg. K.

You can demonstrate this yourself if you are in a studio which has Tungsten light fixtures and a dimming system for light circuits. When the dimmer circuit is set full at 100 percent the light will appear to be "white" (3200 deg. K). As the voltage is reduced to the filament when it is dimmed, the filament will start to show a warmer range first yellow, then and orange range, then a red range. This change in colour range is the actual shift down the specturm in colour temperature.

(N.B. Don't look at an intense light source for a pro-longed period of time... you can cause permanent damage to your eyes by doing so. If you need to look at light fixtures / sources at full intensity levels, use a "gaffer's glass" - a dense protective viewing filter.)

Similarly, if you are working with HMI light instruments, you can alter the nominal colour temperature of the HMI lamp by dimming. Many of the electronic ballast units for HMI fixtures have a built in dimmer which reduces the voltage to the HMI lamp. HMI lamps will have a nominal colour temp. of 5500 degrees K. or 5600 degrees K. and these lamps are best used to match or supplement natural daylight conditions. But when they are dimmed, the colour temperature will shift down the spectrum and start to approach a green range. Again, this will cause havoc on your line-up as the camera will see an overly greenish look on the chart, and you will incorrectly be reducing green in the camera, or adding RED and BLUE levels to over compensate for the false GREEN.

You will typically use a 3200K Tungsten or Incandescent light source in a controlled situation like a studio, or theatre setting. Most ENG/EFP light kits are also equipped with Tungsten light fixtures - not all, but most.

We encounter daylight conditions for the majority of outdoor, and location shoots. Typical of this would be sports events, in the stadium, on the golf course, in the ball park etc. and therefore HMI light instruments are used for punch or supplemental lighting as their nominal 5500K colour temperature is the best match for natural daylight.

So, enough of the Grade 11 physics lesson on light... back to the line-up.

The whole objective of the camera line-up process is to yield the best possible electronic image that the camera is capable of making. This is accomplished by using internal test reference signals in the camera (BARS, CAL signal) to check and align the electronic performance of the camera. The video operator uses the camera and its electronic signals in conjunction with a colour match or reference monitor and a waveform monitor and vectorscope.

Once you have checked the electronic performance of the camera video path, it is time for the camera to go on chart. This would be done with a grey scale / log reflectance chart as a minimum.

The popular DSC series of charts have features of a gray scale chart, colour chip chart and back focus chart etc. all combined into one test chart system... thus the product name CAMALIGN COMBO.

Ironically, the best way to get a good colour video image from a camera, is to first make sure that you are producing a good black and white video image. Use the gray scale portion of the chart to accomplish this. You want to minimize as best as possible extraneous colour sub-carrier from the RED, GREEN, and BLUE channels of the camera. Viewing this on the waveform monitor which is operating in FLAT mode (both luminance and chrominance) you want to remove or minimize the amount of unnecessary colour in the ranges of BLACK (pedestal) GRAY (gamma) and WHITE (gain).
You can cross check your adjustments on the vectorscope. Here the objective it to have minimal "noise" displayed on the scope. and to center the colour space as uniformly as possible at the intersection point of the X and Y axis of the vectorscope graticule grid (0 x, 0 y).

At this stage, the video operator can also use different charts or the combined features of one chart to adjust the gamma level and cross-over, the knee and white clip, level dependency and enhancement. Black stretch, etc.

Traditionally, this used to have to be repeated for each camera. Today, many newer models of cameras have features where a master set-up file can be created on one camera and then transferred and stored into different cameras of the same model/series. Check your camera manual to see if you have this feature.

ii) Camera "painting" - at this stage typically you are looking at the cameras in the real life working condition either on set in the studio or in the field shooting the ball park, race track, football field etc.

This is where is gets subjective - the video operator is looking for the best possible match between all cameras shooting the same content on the set or in the scene. The video operator want the performance of the cameras to be uniform from one to the other. This is so the viewer will not be distracted unnecessarily by things like the news reader having a brown jacket on camera 1, but it looks more of a reddish brown on camera 2. Or why the nurse's uniform in the afternoon soap opera is white on camera 1 and 2, but has a bluish tint on camera 3.

Even though the camera were all aligned on the same chart and in ideal conditions and therefore should perform the same, there will always be some degree of subjective adjustment after the fact.

Live sports remotes present their own challenges to the video operators... largely due to changing light conditions like on a partly cloudy/sunny day.
Not only to they have to contend with constant shading adjustments (iris and ped/black) due to changing levels of cloud cover, but they routinely have to change the overall colour look of the camera to a degree when natural lighting conditions are altered and heavy amounts of cloud move in. This is because the cloud is starting to block the sunlight, and the affect is that the colour temperature of the natural daylight will start to increase and becomes "cooler" looking when it starts to approach 6000K, 6500K, 7000K etc, ranges of colour temperature. So the trick is to warmup the look of the cameras. Or the reverse... the Baseball game or Golf coverage began on a heavily overcast afternoon, so the cameras were all painted to a "warm" look. Later in telecast, the cloud cover began lifting, so the cameras had to be backed off from the warm look, and returned to the normal look or "painting" for when they have more sunlight.

celtics23
08-25-2012, 11:36 AM
Brian,

Thanks for the info. Overall, I am still trying to figure out how all of this fits into the context of using a Tricaster (Pro, in this case) for a multi-camera shoot. Once you have adjusted color bar signals from each camera in the Tricaster, and the cameras go "on chart," do you do the adjusting of each input in the Tricaster itself or on the camera? Also, if you do not have a hardware vectorscope and waveform monitor, is the evaluation of the image coming from each camera done on the Tricaster's waveform and vectorscope? (And are these scopes broadcast-accurate?)

On that note, when you are at the last step in the process, i.e. painting the cameras, what monitor do you look at to evaluate the image (The Tricaster live output, a tv monitor connected to the Tricaster's output, or somewhere else)?

jcupp
08-25-2012, 03:57 PM
You use the calibrated broadcast monitor. :) Or as close to it as you can get. How to calibrate a monitor. (http://spareroommedia.com/video/monitor_setup.html)

celtics23
08-25-2012, 06:16 PM
I did the following procedure with the two studio cameras and the Tricaster Pro:

1. White balance both cameras to the same white card, using their respective lights and the exposure that I use on the talents.

2. With the cameras on but covered by the lens cap, set brightness and contrast so that black on the waveform in the Tricaster is at 7.5.

3. Using the vectorscope in the Tricaster, adjust V and U offset so that the grey blob is right in the middle, where X and Y meet.

4. Send color bars through both cameras and confirm that the vectorscopes match (they pretty much matched).

5. View, through the tv we have connected to the live output monitor, a split screen of two identical t-shirts with multiple colors (I don't yet have a proper color chart!), once being shot by each camera, and then the talent, and then record both of these.

6. View the shot footage in Adobe Premiere, the editor I use.

Now, the situation is that the images I got on the live output when comparing the two talents, and the two t-shirts, looked pretty identical, but in my editor, one of the cameras had a slightly greenish hue. Independent of whether the computer monitor on which I was viewing Premiere, why would this be? Shouldn't the images, while perhaps looking different than they did on the TV, still look just as SIMILAR to each other?

To pain the cameras to match each other, i.e. the last step in the whole process, which proc amp controls should be adjusted first (hue, saturation, etc.)?

Brian Mirrlees
08-25-2012, 10:07 PM
Which output are you using on the Sony HDR-FX1 to reach the Tricaster Pro?

Analog composite video?

Component video (Y/Pb/Pr)?

S-VIDEO out (Y/C)?

Are you using the same method for each camera video feeding to the Tricaster?

What is the length and type of cable you are using on each camera?

celtics23
08-26-2012, 08:50 AM
Brian,

I'd need to wait until tomorrow to find out the specific brand and length of cable, but we use component cables from both HDR-FX1s. However, and I know this is a red flag, for one of them, we have x cable connected to y cable connected to Sony's component to camera cable, while with the other one, we have just the y cable connected to Sony's cable. I might add that y cable is fairly flimsy. Is it imperative to have identical length and brand of cables for each source? Do you have any recommendations for good component cables?

It's a shame that the Sony cable is incredibly flimsy.

celtics23
08-27-2012, 09:46 AM
The component cables we are using are:

For Camera 1, from the Tricaster (with BNC adapters) a 10-foot long version of this cable:
http://www.amazon.com/Steren-Python-Component-Video-Cable/dp/B000EA1LXQ

Connected to one of the these:
http://www.svideo.com/rgb12extension.html

Connected to the Sony cable that connects to the camera's port.

For Camera 2, we are using just the second cable (connected to the Tricaster with BNC adapters) connected to the proprietary Sony cable.

Any suggestions? Is it better to use BNC cables with an adapter at the end to connect to the Sony cable, or to use component cables with the BNC adapter to connect to the Tricaster?

What is the difference in quality, if any, between the following two cables, for a camera to Tricaster distance of only 25-40 feet?

http://www.bhphotovideo.com/c/product/529408-REG/Monster_Cable_127640_400cv_Advanced_Performance_Co mponent.html

http://www.bhphotovideo.com/c/product/529419-REG/Monster_Cable_128066_700cv_Ultra_High_Performance. html

Brian Mirrlees
08-27-2012, 11:30 AM
Considering replacing your current cabling with a better quality 75 ohm video coax in a multichannel snake with BNCs. You could use a couple of 50 foot mults, one per camera. This will allow you to continue working in your component workflow for connecting the cameras to the Tricaster.

Suitable products are available from a variety of manufacturers... Belden, Canare, Clarke, Gepco, etc.

Below is an example of the Canare 3 coax mult cable available from major supplier... customer has to terminate the ends or they can do it for you for an additional cost.

http://www.markertek.com/Cables/Bulk-Wire-Cable/Bulk-Video-Cable/Canare-Corporation-Of-America/V3-5C.xhtml

celtics23
08-27-2012, 04:42 PM
Excellent recommendation Brian, thanks.

Here's a semi-unrelated question: I adjusted the brightness of the image from each of the two cameras (with the lens cap on) using the proc amp controls in the Tricaster, setting it to 7.5 on the waveform. After doing this, I'm still getting parts of the background of my set to come up as below 7.5 IRE on the waveform. Why is this?

Brian Mirrlees
08-27-2012, 11:27 PM
I would suggest that you try the following:

1) Set both camera to BARS mode, and then either use AUTO CAL or manually adjust the level or the camera bar signal at the Tricaster.
If you adjust manually, follow this order:

i) BLACK LEVEL or SET-UP/PED
ii) VIDEO LEVEL or GAIN
iii) CHROMA LEVEL or SATURATION (colour)
iv) HUE or CHROMA PHASE (colour)

When complete, switch both cameras back to CAM (BARS output off)

2) Check the settings on each camera via the MENU:

i) Switch ND off (I assume you are shooting indoors)
ii) Null the COLOUR LEVEL and COLOUR PHASE for each camera, setting 0
iii) Make sure AE SHIFT, and WB SHIFT are also set to 0.
iv) Make sure the AGC / HIGH GAIN mode is set to 0db.
v) Make sure the at the SHUTTER is not activated.

3) Now shoot the your WHITE BALANCE card on both cameras, ideally positioned side by side, and under the same light source/condition.

4) Manually adjust the IRIS + / - so that the white card is properly exposed. Use the ZEBRA mode set to 100, the card will be properly exposed to 100 units when you reduce the IRIS so that the diagonal highlight lines or the ZEBRA are gone from the Viewfinder or LCD screen display. You can slowly adjust the IRIS up and down slightly in order to better find this point.

5) Activate the AWB - Auto White Balance - for both cameras

At this point you can check the output of both cameras at the Tricaster.

Ideally, the white card should appear uniform on the screen as you cut / take between the two camera sources. Also, the level should be uniform on the waveform display, and the the vectorscope. The white card should appear as a small circle, centered equally at the X and Y axis of the display.

Refrain from making any adjustment to the performance of camera video at the Tricaster. If you see differences between the two cameras, double check the video path by switching the cameras back to BARS. Again, the BARS signal from each camera should appear equal between both cameras.

If the BARS signal is appearing the same, but the WHITE card appears differ between one or both cameras, you should adjust for this difference at the camera. Use the WB SHIFT adjustment to make final adjustment.

Now you can have both cameras shoot the "set" under normal set lighting conditions. Readjust the IRIS / exposure as required.

Continue to visually evaluate the look of both cameras at the Tricaster. And again, refrain from adjusting for camera quality at the Tricaster. Use the Tricaster AUTO CAL or PROC adjustments in conjunction with the camera BARS signal to ensure that the video signal path is calibrated for proper set-up. video, chroma level and hue spec.

Once you are happy with the match / look of the cameras shooting talent on the set, you can store this set-up into one of the camera memories. You would do this for each camera.


Not sure if you have the operator manual for the HDR-FX1. You should refer to pages 34 thru 36.

Here is a link to an url where you can access the manual if you require it.

http://www.manualowl.com/m/Sony/HDR-FX1/Manual/245

celtics23
08-28-2012, 12:29 PM
Thanks for the advice on the proper procedure. A couple questions:

1. If after auto-calibrating on the Tricaster using the camera's color bars the waveform shows a lower than 7.5 black and a higher-than 100 white, should I adjust brightness and/or contrast in the proc amp controls to put those in the right limits? Also, should the proc amp controls be used to adjust brightness and contrast if, when shining on the white card with the right exposure, it appears above 100 IRE?

2. The issue we are having in the studio is that we have different lights, with slightly different color temperatures, on talent 1 (camera 1) and talent 2 (camera 2). In fact, there are two issues. One is that the color temp doesn't match PERFECTLY between the key and fill lights on each of the two talents, and secondly, that the color temp is unequal between the talents. Assuming I don't change the lighting setup by buying identically-colored lights, should I white balance each camera to its respective lighting condition? If I white-balance cameras 1 and 2 to the same lighting conditions (e.g. camera 1's lights) I know for a fact the white balance will be off, as expected, when I move camera 2 back to its condition.

I realize that the ideal situation is to have lights with the same color temp, but it sounds like the best way to do this is to white balance separately and then adjust the WB shift in the picture profile in each camera to get them exactly right. Am I wrong?

Brian Mirrlees
08-28-2012, 02:37 PM
Thanks for the advice on the proper procedure. A couple questions:

1. If after auto-calibrating on the Tricaster using the camera's color bars the waveform shows a lower than 7.5 black and a higher-than 100 white, should I adjust brightness and/or contrast in the proc amp controls to put those in the right limits? Also, should the proc amp controls be used to adjust brightness and contrast if, when shining on the white card with the right exposure, it appears above 100 IRE?


Yes, you should be able to do this... adjust for the black level first, then adjust for the video level second... but only relative to the BARS signal from the camera.

Question... by how much do the 7.5 black, and 100 white levels differ.? Can you try connecting the COMPOSITE video output from the camera to the Tricaster composite inputs and check if the level is the same or different.

Also, if you have a contact at a local broadcast station or production house, ask if you could bring your cameras in to check the BAR output level on an external waveform and vectorscope... This would help isolate the level problem even further, maybe the levels out of the camera are low/high.... just a thought for further troubleshooting.

If you are satisfied with the level of BARS from camera to Tricaster, either using component or composite outputs, and you still have have a high or peak white on the white card, then you should adjust the IRIS on the camera to reduce the light level accordingly (again, an easy check is to use the ZEBRA in the camera LCD/Viewfinder).


2. The issue we are having in the studio is that we have different lights, with slightly different color temperatures, on talent 1 (camera 1) and talent 2 (camera 2). In fact, there are two issues. One is that the color temp doesn't match PERFECTLY between the key and fill lights on each of the two talents, and secondly, that the color temp is unequal between the talents. Assuming I don't change the lighting setup by buying identically-colored lights, should I white balance each camera to its respective lighting condition? If I white-balance cameras 1 and 2 to the same lighting conditions (e.g. camera 1's lights) I know for a fact the white balance will be off, as expected, when I move camera 2 back to its condition.

I realize that the ideal situation is to have lights with the same color temp, but it sounds like the best way to do this is to white balance separately and then adjust the WB shift in the picture profile in each camera to get them exactly right. Am I wrong?

It seems by your description that you are working in a situation with mixed or different light... and I would suggest this is the root of you problem.

Are you sure that you have two light sources of different colour temperature, like tungsten or fluorescent light? Or are you confusing colour temperature with COLOUR MEDIA gel (Rosco, Lee, GAM) used on light fixtures to change the colour to a wash of colour used for dramatic effect?

celtics23
08-28-2012, 03:10 PM
I've attached a screenshot of what today's show looked like. Now, bear in mind that I'm waiting on an extra bulb to come in for the fill light on camera 2 (on the right, the guy with the headphones on),which should reduce the shadow on his right cheek and shoulder. Also, we are working on lowering the brightness of the monitor behind camera 1 (the monitor with the logo), and on trying to reduce the horrible glare on his desk, which is coming from his key light.

The key light on subject 1 (again, on the left) is a Coollights 5600K fluorescent tube fixture. The fill is an intefit CFL unit with bulbs that claim to be 5200K (with the white diffuser which might raise the temp a bit). His backlight is a lowel omni with a full daylight gel.

The key light on subject 2 is a supposedly 5400K and his fill is an interfit CFL unit with bulbs that very mysteriously say that are between 5000-5500K (!!). We don't yet have a backlight on subject 2.

Apart from any other observations you might make about the shot comparison, would changing the tubes and CFLs of the lights to identical-color temperature and high cri lamps work to reduce these issues, or is this a fluorescent light issue that can't be resolved even with high cri, "high quality" lamps?

Also, on a side note, I've been able to find kino-flo quality tubes online, but is it possible to get production-quality CFL bulbs with a definite color temperature, and if so, where?

Brian Mirrlees
08-28-2012, 08:57 PM
Honestly, the pictures look pretty good judging by the screenshot you provided.

I understand about the replacement lamp for the CAM 2 side fill fixture... I agree with you that, that should solve the problem with the shadow.

For the reflection on the desk top, if you don't already have flags (like from a Matthews grip kit) you can easily improvise with using some "foam core" board - used for graphic art displays. We use it all the time, cuts very easily with a utility knife. Available in black and white surfaces. Cheaper too than a flag set.

The brightness on the left monitor (logo loop) didn't bother me...

Good call on using the 5500K range of light since you are working with live monitor screens on the set. Looks like you just have to make some tweaks on the larger LG screen.

Another trick to try, is to route colour bars to the on set screens, and then shoot the screen with one of the matched cameras. Make adjustments to the on set monitor while looking at the feed of the camera on PGM out. Remember, you aren't trying to make the on set screens look good to the eye, but rather to the camera. Make sense?

The only other thing I could suggest is to apply some make-up / powder to the on camera talent. Nothing major, but a little powder will help to even out the skin tone, and take off some of the shine from natural body oil and perspiration. If you do choose to try a make-up powder, make sure you use a product designed for television/video work. It should have a proper TV base. The problem with using regular make-up cosmetics is that they tend to have a yellow, or greenish base colour. This can make the talent look somewhat sickly looking on camera...

I have no experience with the CFL lamps that you are using. I do work regularly with Videssence, Kino-Flo, and ARRI fluorescent fixtures, but the design and nature or the lamps is a bit different from the CFLs that you are working with.

However, I would advise on standardizing / using one type of lamp and to use one with a colour temperature which is in the range of 5500K to 5600K colour temperature.

celtics23
08-29-2012, 07:58 AM
I've been suggesting to both talents to put video make-up on both talents, but it hasn't yet happened hehe.

I'm currently waiting for kino-flo to respond to an email I sent about True Match CFL bulbs (95 CRI, 5500K) that I found on their website but cannot yet find at any vendor. http://www.kinoflo.com/Kino%20Flo%20lamps/True%20Match/True%20Match.htm

I think these would be ideal.

Thanks for all the suggestions and help Brian.

Brian Mirrlees
08-29-2012, 12:00 PM
If you are still interested in purchasing one of the DSC Front Box charts, you should ckeck out the DSC website and their August newsletter. They have a bundled sale available until Sept. 15th.