80 ire blackmagic 4k

Hi, I recently bought a Blackmagic 4k.

Is it normal for the iso200 to clip at 80ire, iso400 to clip at about 93iso and the iso800 to clip at about 108ire?

Also the IR colour shift is really quite bad at anything above ND6 is this normal? (Using Formatt 4x5.65 glass standard ND's)

Also is it correct the SDI can only output 'Film' look and not 'video' look?

This is common to most digital cameras. As we cut the amount of visible light hitting the sensor, the proportion of IR light increases since ND doesn't stop it. Some sensors are more sensitive to IR than others, so it might take more ND for other sensors to show it up. We see this all the time in the still camera world so it has to be compensated very regularly.

I'll let others discuss what is normal for the BMPC as I don't have one yet.

Also the IR colour shift is really quite bad at anything above ND6 is this normal? (Using Formatt 4x5.65 glass standard ND's)

The colour shift you are seeing is IR pollution. Digital sensors are more susceptible to IR than film is. Some cameras deal with it well but most will need IR filtration when going beyond 0.9 ND.

Most of us are using IRND filters or an IR cut filter.

How are you measuring your IRE clipping?

Hi, thanks for getting back to me, yeah I've read this in a few places, I did a side by side comparison with an FS100 which displayed very small amounts of colour shift even at ND1.2

Screengrab in Premiere BMPC - ISO200 80IRE clipping here:

In FCP ISO200 80IRE clipping

How are you measuring your IRE clipping?

Hi Adam, Ha that's a coincidence just been on your site and saw Keryn is your son. I worked with him in Essex on the 'Spoiler' feature

Hi Adam, Ha that's a coincidence just been on your site and saw Keryn is your son. I worked with him in Essex on the 'Spoiler' feature

Ah Keryn "Fizzy" Roberts!!

He sure did enjoy working on that film.

Hi, thanks for getting back to me, yeah I've read this in a few places, I did a side by side comparison with an FS100 which displayed very small amounts of colour shift even at ND1.2

This is worth a watch:

Screengrab in Premiere BMPC - ISO200 80IRE clipping here:

I'm sure someone more technical, like Tom, will reply regarding this...

Oh yeah thanks, I've seen that link. I've got a set of 4x5.65 Formatt ND's 3, 6, 9, 1.2, and believe Formatt sell plain (Not ND'ed) Hot Mirrors at 680 and 720 strength, which would you recommend? I think 680 cuts more of the spectrum than the 720

Thanks

Oh yeah thanks, I've seen that link. I've got a set of 4x5.65 Formatt ND's 3, 6, 9, 1.2, and believe Formatt sell plain (Not ND'ed) Hot Mirrors at 680 and 720 strength, which would you recommend? I think 680 cuts more of the spectrum than the 720

I've not used the Hot Mirror so can't really comment. I have their ProStop IRND in 0.3, 0.6, 0.9, 1.2, 1.8, 2.1 & 2.4.

When shooting with my Genustech Eclipse Fader ND I use this Hoya IR Cut filter:
http://amzn.to/1jXaSuq

Works really well.

I have much stronger green cast on my 4k than I do with my bmcc using the same tiffen 1.2irnd and the same lens. Weird.


Sent from my iPhone using Tapatalk

Is it true you can't stack IRND's? not that you'd need to with a full set though

How does it look in resolve?

Can you screen grab the resolve waveform to show clipping?

Is it true you can't stack IRND's? not that you'd need to with a full set though

As far as I know there are no issues stacking IRND. HotMirror on the other hand always needs to be the first filter the light hits.

Resolve 9 Lite: ISO200

Also is it correct the SDI can only output 'Film' look and not 'video' look?

Noticed this was missed.

The camera outputs whatever recording dynamic range you have selected. The viewing dynamic range only affects the built in screen.

I was wondering this too. My camera (BMCC EF) clips at around 80 giving me no information past that point.

If you are shooting in Film mode this would be because the data is in a log type curve. Protecting the highlights and shadows.

You the pull the contrast and saturation back in post to your taste.

If you are shooting in Film mode this would be because the data is in a log type curve. Protecting the highlights and shadows.

You the pull the contrast and saturation back in post to your taste.

Then why does ISO200 clip at about 80ire, ISO400 clip at about 93ire and ISO800 clip at about 108ire?

I thought log meant that it uses the full dynamic range of the sensor. This means my clipping should still only happen at 100 and gradually/ logarithmically step down to 0. The 20 steps between 80 and 100 should be filled up with info... no?

I tested all ISO's with relative T-stops on the lens (so all are lit the same) and ISO200 has less dynamic range, noticeably in the highlights, ISO400 has more, then ISO800 more still. Although ISO 800 is well noisy.

I'd have loved for this camera to work, image is leaps and bounds over the FS100 I usually use.

On all blackmagic cameras, ISO settings do not affect the actual image sampled from the sensor, there is no analog gain applied. So when you only change ISO, and nothing else, the image will always clip at the same brightness, only that that brightness will be mapped to different luminance values depending on the ISO setting. That means: When you select less than your cameras native ISO, you can't use the top bits of your ten bit color range, you are wasting part of your dynamic range! Still there are reasons to do so, as changing the ISO actually changes the transfer curve that maps sensor data to prores data and you might be able to supress some noise that way.

Oh right so its full dynamic range at the noisy Iso800 or less dynamic range at the clean iso200 with the BMPC.

So with the BMCC I think it's native ISO is 800, what happens to the IRE clipping at 1600iso and above?

Thanks
Jon

Just to toss this out there, with my BMPC 4K attached to my Macbook Pro via thunderbolt, Ultrascopes gave me a better visual for what's happening.

From ASA 400 to ASA 800, the histogram doesn't change the top level (at 80 IRE), but it does raise the low end quite a bit. It also affects the mid tones. To me, the higher end is squished in the histogram when raising from 400 to 800 since it's stopped there. From what I recall, going from 400 to 200 shifts the mid tones down more (possibly bring back some of the congested detail in the 80 range, but can't be sure).

Concerning IR pollution, see item #11 in: "BMPC-4K: Shooting tips from early users":
http://herefortheweather.wordpress.com/ ... rly-users/