C100: Zebra 2 at 100% and ISO 320 problem

Mercenary

Member
Does anyone else have this problem.

If I set my Zebras to zebra 2 and 100%, if I set my ISO to 320 the zebras don't work. Nothing, turn it all up to make it as bright as possible, point into a light and still no zebras. But change the ISO to anything else or take the zebras down to 95% and suddenly everything functions as expected?

This happens on all our C100's with the latest and previous firmware.
 
Here is my non-technical diagnosis:

If you have zebras set at 100% the white color of the zebras are exactly the same value as the burnt out area that is 100 percent. So in effect the zebras become invisible. They are there, you can't see them because you have white zebras over a white area. That is my educated guess. My history is 90% Sony & Panasonic, this is my first foray into the Canon World. However, white against white is invisible no matter who manufactures the camera, correct?

That's why I set mine at 95% (on all the cameras I have ever owned) so I know when to stop opening up iris, especially if it's for broadcast. I want a WARNING I'm getting too hot and if you set your second set at 100% (mainly for exteriors) by the time zebras kick in it's Too Late. I never met a shooter who sets at 100%.

My two cents...
 
Last edited:
Yes, Wide DR clips before 100% at ISO 320. With WideDR, it is best not to shoot below ISO 640 if you want to have a fully useable dynamic range available.
 
Yes, Wide DR clips before 100% at ISO 320. With WideDR, it is best not to shoot below ISO 640 if you want to have a fully useable dynamic range available.

My understanding is to utilize Wide DR's latitude of 12 stop range most effectively the ISO should be at 850. SO I keep it there as my "default" ISO setting. Is that what others are doing?
 
My understanding is to utilize Wide DR's latitude of 12 stop range most effectively the ISO should be at 850. SO I keep it there as my "default" ISO setting. Is that what others are doing?

Where did you read this info about wdr gamma? Only canon-log gamma needs ISO850 for the best DR (800pct or 72dB or 4000:1 contrast). If you set ISO to 320, you will get the same white peak from the chip but in the lower level of modulation (look at waveform monitor on LCD).

Only normal gamma has the knee ON (it is about 200-250pct of DR with knee), cine1 and wdr have no knee (even if you set knee ON in CP), cine1 has about 300pct of DR (300pct or 63.5dB or 1500:1 contrast), wdr is a little better, but it is not exactly specified.

But canon-log has 800pct of DR (72dB) in +2.5dB gain (ISO850), if you set for cine1 -6dB (ISO 320) then you will get the same white peak (at superwhite on waveform), because -6 +2.5=8.5dB; 300pct is 63.5dB + 8.5dB = 72dB - WOW!!. I tried this for cine1 and canon-log and this works, the reason is that the amplifier (ISO/GAIN) is before A/D converter (in analogue domain), if you set lower gain then A/D has headroom for another +8.5dB above 300pct of cine1 gamma. There is no clipping (with the same signal), but black is more pressed in case of cine1, you get different black and middles.

For wdr gamma I found that I need about ISO320-400 for the same result (signal from the chip without clipping, with the same exposure of the chip as cine1 or canon-log) in the first test of my new C100 one year ago. So i don't think that ISO850 is the best way for wdr. But wdr is different in highlights, so the result depend on the exposure of middles (different result in a day or at night).

The differences between gammas are not in DR (if you set different ISOs as I wrote above, the white peak is without clipping for the same exposure of the chip, the same signal from the chip), the differences are in the distribution of samples in black, middles and highlights. Cine1 with -6dB has the same DR as canon-log in +2.5dB (but there is pressed black, middles are a much higher), wdr is a very close to cine1 in the black and middles, but wdr has much more pressed highlights (and this highlights are desaturated, you need bring the saturation level back in highlights), canon-log has the best distribution of samples in the whole DR (it is the best start point for color correction at post, if you have true 8bit 422 I-Frame only recording from HDMI). This is the main difference.

I hope this help.
 
Last edited:
I like the much better saturation smoothing in transition from middle to the highlights with wdr gamma (over exposure places), cine1 has this transition in saturation levels worse (on my lamp shade it is really much more worse, but I make some desaturation of highlights at post and then it is much better). WF monitor helps with correct exposure, but also C100's LCD is a very good.

The Zebra 2 is only basic info about overexposure, WF monitor and LCD helps much more to you, this needs some time for learning.
 
Sooo, is the no zebras at 320 normal then :)

We use the WF all the time and use the zebras as a quick guide for whats hot. Especially when using the eyepiece which doesn't show the WF and is used when your out in bright sunshine.
 
I tend to work at ISO 850 or 640 for wdr, but I have not done any real testing. In fact, I was having a hard time following Pave1's analysis. I record to AVCHD and I would rather err towards crushing my blacks in post and maybe pulling down my superwhites to get some more definition in the highlights if the whites are clipping. At ISO 320, I am putting most of my DR below 50% and I am usually pulling up my middles in post in order to get a better image from poorly lit event coverage. What I don't understand is why Canon chose to record superwhites over 100% rather than just compress the entire image down to 0-100%. Can anybody explain that?
 
I have done some testing with WDR at 320, 640, and 850, and very much prefer the deep blacks I get at 320. Whites clip at 100, but not 110, it seems to me. Use what you like. For me, the shadows are typically more important than the widest possible DR gamut.

ALWAYS black balance before starting the day with this camera. Really helps.
 
...In fact, I was having a hard time following Pave1's analysis. I record to AVCHD and I would rather err towards crushing my blacks in post and maybe pulling down my superwhites to get some more definition in the highlights if the whites are clipping...

Canon C100 has some special issue - the ISO/GAIN amplifier is before A/D converter (not in the signal procesor).

If you set the scene's white peak to be ok on the chip (with iris, ND filtres, shutter speed)= correct exposure of the chip, then still you can get the A/D converter clipping from the wrong amplifier level (ISO/GAIN), even if you be below 100pct of modulation level on WF monitor .-)).

The same signal from the chip differ with ISO/GAIN level, after linear A/D conversion is also the gamma curve applied, the digital signal then differ in black, middles, highlights with different gamma.

The rest I wrote before .-) .

I like the wdr desaturation in highlights, but with higher ISO there is more noise in 25i (even if you have the same shutter speed 1/50 for both 25p and 25i mode). In 25p mode it is much better. But with the cine1 you can get a very similar result (with a little more pressed black and with desaturation highlights at post), it has one more advantage - wdr and canon-log with 8bit scale result sometimes in banding of the smooth gradient (on the sky, light going to the shadow on the wall, etc).

This was the main reason for me to use cine1 as the main gamma curve with C100. I also often shot at 25i, it is the second reason for using cine1 gamma (due to noise level in wdr at 25i mode, no ABB can help with this). I like wdr gamma only with ISO320 (25i) in sunny days with a lot of highlights, cine1 is my basic in interiors.

So my war is 25i vs 25p, cine1 vs wdr noise, wdr and canon-log banding or better DR?

I found that cine1 gamma is for me the best curve for some 8bit 420 AVCHD C100 Canon, but sometimes I try wdr or even canon-log, but this need Ninja recording (log) or 25p mode (wdr). This is "my problem" with C100, but I like this cam .-) .
 
Last edited:
Back
Top