Exposure, Black Levels and O7Q

P

paulcurtis

Veteran
Had some time with the O7Q.

An observation. At the moment i'm just looking at the monitor and on board tools. Not sure if this is just me so...

Running 2K RAW, but all blacks seem very elevated all the time. If i lens cap the camera i'm still seeing a base line black of around 16%, the internal LCD has this at the far left. Changing profile doesn't radically change this and there's nothing you can tweak on the Slog profile that affects it either.

The false colour tools would never show purple or blue for example (without adjustment)

To my mind, linear RAW should run from 0 to 4096, there shouldn't be this idea of legalised blacks and 16% is higher than that anyway.

So i wonder what the signal path is? Is the 700 outputting 0 for black? My concern is that 16% is a lot to loose in terms of available range, especially for 12 bits.

Is this just an O7Q display thing?

It's quite difficult to tell from a DNG absolute values of black because the files come in with a black level set anyway.

Is everyone else the same?

cheers
Paul
 
Bring the image into Resolve and see what it shows you on the scopes.

I do know from my F55 that Sony seems to place blacks slightly higher then where you expect it would sit. Perhaps Sony has done this across the board.
 
Just to expand on what i mean a bit more, here's an rough iPhone photo of the waveform. The scene is just dusk outside, you can see that the whole dynamic range isn't that great. nothing clipping not even a little. My concern is that the range appears to be from 16% to 70% rather than 0% to 70. If this is written to a 12 bit container then that 16% is wasting a lot of space in this case. Space that ought to be for minimising graduations etc,.

(And yes i know i can increase ISO and raise the highlights, this is more an example)

o7waveform.jpg

Also from a basic monitoring point of view, i have no idea where black is clipping unless i alter all the false colour tools. It just seems not what i would expect.

cheers
Paul
 
Last edited:
starcentral said:
Bring the image into Resolve and see what it shows you on the scopes.

I do know from my F55 that Sony seems to place blacks slightly higher then where you expect it would sit. Perhaps Sony has done this across the board.
Click to expand...

I will do, but i suspect the DNG is flagged as having the black level elevated which means it would come in as black. That doesn't answer the question of whether there is range detail missing. I know log curves elevate the lower blacks but this seems *too* much and it's a general offset to the whole image

thanks
Paul
 
I would black balance the camera and shoot something with extreme range so that you can see where black and white is actually clipping.

I don't use the 7Q but perhaps there is a levels range selection settings somewhere on it? It almost seems like its inappropriately scaling your image? This could also be due to a setting on the SDI output (ie. YCrBr vs RGB) - if it even exists on the FS700.
 
Last edited:
Dennis,

Highlights are fine, you can't get blacker than a lens cap. The SDI is outputting RAW for this, hence the confusion about where black should be in this 'data' orientated situation.

Just wondering whether it's like this out from the 700 or whether the O7Q is elevating the blacks. Only way to compare with be with the Sony R5.

cheers
Paul
 
Paul,
Interesting question. I put my O7Q into multi-stream Quad mode and plugged my EX1 SDI port into SDI port B on the O7Q. The O7Q WFM distinctly shows Black for the FS700 around IRE16, while the EX1 shows IRE0 on the O7Q.

Another interesting little experiment is to monitor the Black(Lens cap on) on the WFM whilst changing the ISO on the FS700. The noise floor stays the same at IRE16, while the noise shows ever increasing levels of "black" as one raises the ISO, as one would expect.
 
cuervo said:
Paul,
Interesting question. I put my O7Q into multi-stream Quad mode and plugged my EX1 SDI port into SDI port B on the O7Q. The O7Q WFM distinctly shows Black for the FS700 around IRE16, while the EX1 shows IRE0 on the O7Q.

Another interesting little experiment is to monitor the Black(Lens cap on) on the WFM whilst changing the ISO on the FS700. The noise floor stays the same at IRE16, while the noise shows ever increasing levels of "black" as one raises the ISO, as one would expect.
Click to expand...

Well 16% on the waveform isn't the normal 'black level' is it? That's supposed to be 7.5% (code value 16 out of 255). I'll try the iso thing!

cheers
Paul
 
paulcurtis said:
Well 16% on the waveform isn't the normal 'black level' is it? That's supposed to be 7.5% (code value 16 out of 255). I'll try the iso thing!

cheers
Paul
Click to expand...
It sure isn't.....

edit: But, I see that switching the O7Q to DPX mode and the FS700 to HD, drops that Black to around zero for the hypergamma modes, controllable by the Black level. Using the Slog2 setting in HD, the O7Q shows a positive value of Black on the WFM, not adjustable, as you probably know. It would appear that "Black" in RAW mode is some positive value. I don't think there is an error in the CD WFM. This is the way the signal comes from the FS700 in RAW mode.
 
Last edited:
Highly likely this is the signal out of the 700, but i would like to know for sure.

The reason i'm concerned is that we're dealing with 12 bit linear (4096 values). This is barely enough for a linear signal, the F5/F55 is 16 bit (F5 14bit) by comparison. Now if 16% of the signal is not used that leaves 3400 isn values which may not be a big deal but to me it seems a enormous waste when you don't have that much to start with and no explanation as to why it would be like this. Especially when it's all skewed to highlight range.

It will also be nice to get LUTs on the O7Q to monitor with this and knock it into a better range for viewing

cheers
Paul
 
This has to do with the nature of the RAW signal out of the FS700, and it is actually something that we are addressing with the new firmware.
 
Mitch Gross said:
This has to do with the nature of the RAW signal out of the FS700, and it is actually something that we are addressing with the new firmware.
Click to expand...

That's good to know. In the Sony Slog white paper 0 should be at code value of 90, i'm pretty sure Slog2 should be that this too. So in a linear workflow then these 700 files wouldn't reverse LUT without adjustments. I've also found some RAW files with the 700+R5 although not ideal, do suggest values below 16%, i was going to look for more samples but if you guys are obviously way ahead on the ball so i'll probably pass on that.

Are you able to tell us whether this would result in a simple rescale within the O7Q or whether the change would mean additional detail is available?

Many thanks!
Paul
 
I'm not 100% sure, but I think it is just a remapping of the signal. You already have all the info available in the image.
 
You must log in or register to reply here.
Back
Top