TrueColor HD200/250 in progress

[Paolo Ciccone]

Hi all.

Just wanted to confirm that I received a HD250, thanks to JVC for the loaner, and I have started calibrating the camera today. I can tell you for sure that this model behaves quite a bit differently, from the color matrix point of view, than the HD100 so no surprise that the old scene files didn't fit. It's a whole new game.

Fortunately I have now a full size ChromaDuMonde chart, thanks to DSC Labs, and the added color precision shows on the VectorScope.
Just for kicks I tested my old HD100 with TrueColor V3 with the ChromaDuMonde and I'm happy to report that the colors in fact fall in the right spots.

Anyway, I will be working the rest of the next two weeks to design and verify TrueColor HD250, I'll post here when it will be ready.

Take care.

[Tim Dashwood]

I found that the actual matrix controls did not seem to change too much. The gamma and color gain seem to be the most different.
In fact, I decided to throw TC3 up on the scopes while I was creating HD200/250 versions of my own scene files and was able to match it to the HD100 by bringing the gamma level down to MIN and COLOR GAIN to -2. The vectorscope was similar and I think the only small adjustment I made was R Gain 4 and R ROT 3.
Other than that it seemed to match almost perfectly.

Out of curiosity I also created a "DSC CDM calibrated" setting using the proper methods as described by DSC, but found it very difficult to maintain the hue of the violet blues in the lower right corner through to a NTSC downconvert. This is of course one of the most difficult hues to accurately recreate in NTSC, so I spent a few hours trying to overcome the challenge.
I ended up creating a new scene file that I believe can achieve accurate reproduction of colour and tone on through to a NTSC downconvert into 601 colour space. I'll post it soon when I post the other HD200/250 versions. On the vectorscope there doesn't seem to be a trade-off, but there are usually no free rides in camera calibration.
I'm curious to see how close our results are.

[Paolo Ciccone]

Hmm, the camera that I have has quite a different behavior. As usual, when I do this, I start from scratch, setting highlights and blacks values in the right place, setting the gamma, removing noise etc.
After all was said and done, the color matrix was scewed toward the red. While I remember this beeing the case for the HD100, the HD250 seems to be even more so.

More details later.

[Tim Dashwood]

I did find that the preset white balances were skewed to red a little. I manually white balanced on the DSC chart before beginning. Even the 3200K manual white balance vs. the 3200K preset showed a shift on the vectorscope toward skin tones. Odd, but obviously more "pleasing" out of the box.
I'm not suggesting to avoid presets, but you may want to decide if your scene file will be used with manual WB or preset WB before calibrating - or maybe make two versions for the exacting users. I designed all of mine based on manual WB this time around, with the intent being that HD100 series and HD200 series could be used in a multicam setting.

[Paolo Ciccone]

Yes, I white balance using the DSC chart, before starting working on the colors. Interestingly enough, the same exact scene with the exact same lights balanced at 3200K with the HD100 and at 2800K with the HD250. I didn't want to post it before because I want to investigate what is the reason for this. I don't know of it's the camera or anything else that I might have missed but it's interesting that you found a similar behavior.
More about this later. I have a Sony F-350 as well, I see what is the result with that camera.

[Tim Dashwood]

Originally Posted by Paolo Ciccone

Yes, I white balance using the DSC chart, before starting working on the colors. Interestingly enough, the same exact scene with the exact same lights balanced at 3200K with the HD100 and at 2800K with the HD250. I didn't want to post it before because I want to investigate what is the reason for this. I don't know of it's the camera or anything else that I might have missed but it's interesting that you found a similar behavior.
More about this later. I have a Sony F-350 as well, I see what is the result with that camera.

You should try using the exact same lens to eliminate a color cast variable. That's what I did. I own the 13x3.5, but specifically used the 16x5.5 for calibration because that's what 95% of the users own. I also taped down the aperture ring so that I could compare sensitivity. You'll be surprised what happens!

Also, make sure that the scene file you use to do the white balance on both cameras is set to NORMAL on all matrix and WB PAINT settings.

[Paolo Ciccone]

Yep, that was my guess too. The HD100 stock lens on the HD250 warmed up the scene. The Fujinon 18x lens that was provided with the camera is defintely "cooler" :)

[Sean Adair]

Excited, exuberant shout! You guys can only imagine how much I was anticipating this exchange, checking a couple of times a day for tidbits on how to approach a better base scene file for the 200 series.
It's all very exciting, and I plan to try some of the steps mentioned THIS morning within my limited test environment.

Some of the questions that come up for me:
What is the light "sensitivity" comparision between 100/200 series?

As "removing noise" is an important function of these settings, have you found any fundamental difference in picture noise? Perhaps the stock settings, or jumbled adjustments from tyros like me were responsible for some noisy results (I'm also tuned into the prompts like a large field of the same midtone color). Do you the think the settings you are working on now will help make the camera more robust in noise provoking situations?

Re the preset vs manual WB. Is the difference from the 100 series due to the manual preset being calibrated for a "cooler" lens? Or that with the 18x, the (same) manual setting ends up with a cooler look?
If either of you have an opportunity, I'd be very curious to know how the 17x I own fits into this color spectrum on the charts.
BUt by doing a manual white balance, we are able to eliminate the color shifts of the lenses as a factor, correct? I'm sure this get even more variable with lens adapters for 16mm & 35mm still cameras. I know my Nikon lenses have quite a bit of variation from each other. There must be contrast variation as well which would factor into the settings.

I do find that with much of my shooting I want to use presets. Often I'm moving quickly within indoor or outdoor environments, and want to have a stable setting. It may be a bit warm indoors, but it's looked attractive compared to the manual set. I'll compensate for shadow vs full sun in post if needed, but feel better off than manually setting for one of these.

I really appreciate you guys sharing your conversation as you move towards finalizing scene files. It really helps to understand the mechanics behind the settings, and allow me to get some more confidence in adjusting settings for variables that warrant compensation.

[Paolo Ciccone]

Hi Sean.

Originally Posted by Sean Adair

Some of the questions that come up for me:
What is the light "sensitivity" comparision between 100/200 series?

Too early to know exactly although I don't expect the cameras to be different. I will do accurate tests using the same lens on both cameras but given that the chip size is the same I expect results that are very close.

As "removing noise" is an important function of these settings, have you found any fundamental difference in picture noise?

The noise problems that many people report are intrinsic of the digital compression and sensor accuracy of digital acquisition. With the HD100/250 the problem is compounded by the HDV compression. Dark backgrounds of seemingly uniform color present a big challenge. Avoid them! I will post detailed information with before/after comparison about this in relation to the detail settings in the camera.

Perhaps the stock settings, or jumbled adjustments from tyros like me were responsible for some noisy results

I found the stock scenes in the HD250 to be considerably less sensitive to light than the DSC-calibrated configuration. The white chips register at a much lower level and the blacks are washed out. So, yes, the standard scenes will require much more light than what the camera is actually capable and in low-light condition this can add a lot of noise to your footage.
TrueColor, as always, will give you the widest latitude the camera can achieve.

Re the preset vs manual WB. Is the difference from the 100 series due to the manual preset being calibrated for a "cooler" lens? Or that with the 18x, the (same) manual setting ends up with a cooler look?

I adjusted the WB in both cases manually using the DSC chart which is spectrometer-tested for neutrality. With exactly the same light and same chart the two lenses/cameras registered a difference of at least 300K. I will have more information later.

If either of you have an opportunity, I'd be very curious to know how the 17x I own fits into this color spectrum on the charts.

I don't have that lens. Your best bet would be to wait for the TrueColor to be finished, install it in your camera and then do some tests by using the DSC chart. The Camalign FrontBox HD is affordable and you can use it to verify the bias of every lens that you will be using. I get new uses from that chart every day.

BUt by doing a manual white balance, we are able to eliminate the color shifts of the lenses as a factor, correct?

That is correct but when the camera register a cooler color temperature than it actually is, you know that something is up. You might end up in a situation where you can't do an accurate WB so you decide to use one of the presets. If the camera assumes a 5600K WB point but the lens has a blueish bias you'll end up with an imbalance in your footage. Nothingt that cannot be fixed in color timing but it will require an extra step and more color information could be lost.

I'm sure this get even more variable with lens adapters for 16mm & 35mm still cameras. I know my Nikon lenses have quite a bit of variation from each other.

But with still photos you either use film, which has much more latitude than HDV, or RAW which again has a lot more color information than HDV. In fact the WB in RAW is stored as metadata and not as part of the pixel information. That's why you can move it around so much.
With digital cinema you have less data and the danger of "rounding-off" the color bits by using too much color correction is always present. That's why I spend so much time in making sure that the cameras that I use are calibrated for the most neutral look. This gives me as much color information as possible. If your WB is set right at the shot, you have less work to do and less risk of data loss.

Take care.

[Jeffrey Butler]

Well, I've been shooting with my HD250 for about two weeks now. I've definitely found some really sweet spots with the camera, to the point where I'm blown away how clean it is. I've got the 17x lens. But I've also found some scenes where I'm largely horrified at the results....I'm gathering some shots and working through all the settings as well as just re-learning how to shoot with this camera....I hope to contribute to this thread some how

The lens is really, really nice, and I'm overall happy with the camera. I don't mind saying that 1280x720 is just plain small. Popular, but small....for such a big camera! =D

All that said, I'm very much looking forward to your scene files and some detailed tweaking - it's clear the camera, with some minor tweaks shoots much, much better video that it does out the box, so with some major tweaks, I expect major improvements.

[Dave Beaty]

Thanks for taking the time and energy to produce these scene files. Without the proper test charts , scopes and know how, for most of us, it's a frustrating experience and we end up using factory defaults.

We have been shooting with the HD250s since November. We have noticed them to be a bit more noisy than the HD100? Have you found that also?

One question I have for you about your scenes, can you test the DNR feature. I am curious about the sacrifices we are making when we use it. Perhaps the best approach is to use the wider latitudes and less detail setting and leaving DNR turned off.

The other observation with the shutter is that at 30p and shutter 1/60, it seems to turn off. The (S) in the viewfinder goes off. We have also done some productions with 30p and shutter 1/30 which have been wonderful. As long as there isn't to much fast motion, the added exposure and softness is very pleasing. But I am unsure how this effects noise and gamma. It appears to improve the situation.

Dave Beaty

[Paolo Ciccone]

Originally Posted by Dave Beaty

We have been shooting with the HD250s since November. We have noticed them to be a bit more noisy than the HD100? Have you found that also?

I have the camera as a loaner from JVC so I didn't have the time to do a comparative tests between the HD100 and HD200.

One question I have for you about your scenes, can you test the DNR feature. I am curious about the sacrifices we are making when we use it. Perhaps the best approach is to use the wider latitudes and less detail setting and leaving DNR turned off.

The effect of DNR is one of the things that I plan to test. Regardless, TrueColor will provide a wider latitude than the stock scene so you will have less need to rely on the DNR.

Take care.

[Carl Martin]

All of you working on scene files for us here at DVI are very nice to do so. I for one appreciate all the hard work all of you are doing.

Thanks for taking the time out of I'm quite sure very busy schedules to help the rest of us enjoy creative scene files for our cams.

You all are awesome.

Thanks,

Carl

[Paolo Ciccone]

Thank you Carl, much appreciated

[Scott R. Smith]

I am happy to find this thread started on the HD250's. I had found Paolo's True Color settings for the HD110 after may first big shoot with the 250's was a bit disappointing. I adjusted the 250's to Paolo's TrueColor3 setting for the 110, and my second shoot was already a big improvement over the factory preset for the 250. I adjusted a couple of items slightly different than TC3, but as I don't have a decent DSC chart, is was mostly an intuitive adjustment, by eye.

Paolo and Tim, I look forward to seeing your final setting choices for 250TrueColor.