PXW-Z150 low light issues - help?

[Cliff Totten]

Originally Posted by Paul Anderegg

One last note about those figures. Due to the loss of light sensitivity with increased pixel count, several manufactures have begun blatantly making up figures, FALSE FIGURES. These would be Panasonic and JVC. Both of these companies have introduced standard and "extended" modes to their line of cameras. The extended or "low light" modes add 6db of gain to the readout, so when the camera says 0db, it is actually at 6db. Because the figures are taken at 0db, they get a rating of f11@2000, when in fact their cameras are a full stop less sensitive. Remember, every 6db is a stop. The HPX370 and HM890 cameras are rated at f11@2000, but at 0db on extended mode where they receive this rating, they are noisy, and the image is identical to their standard setting at 6db of gain. Sony uses REAL image sensitivity, no games. When you see Sony say f12@20000, that is a real number, with no fake or hidden gain tricks applied.

I have owned and operated HPX and JVC camera with these modes, so this is from real world experience, not just guessed about.

Paul

But this is what happens in all cameras. When the voltage readings are measured at each pixel, (they are extremely low voltage readings) the circuit just does't take whatever it gets and calls that baseline "0db". Instead, that is all "amplified", deBayered and processed. The final result of all of that is what the camera companies will call "0db" for that model.

So, yes, all cameras have plenty of "pre-amp" work going on long before it ever reaches what the camera guy sees as "0db" in his settings. And of course, it gets even more complicated when you consider the pre-amplification needed to do a LOG curve read-out and mapping from those photosites on the sensor. LOG mapping requires additional gain processing before you ever see it on your screen/switch settings as "0db".

So, I think it's fair to say that "0db" can never be taken literally on any camera of any kind from any company. It really doesn't mean allot, it's sort of a false indicator for what is really happening behind the scenes.

CT

[Paul Anderegg]

Cliff, what I meant specifically for the JVC and Panasonic cameras, is that they have TWO DIFFERENT 0db modes, the ones they cal extended and high sensitivity add 6db underneath that 0db you see in the viewfinder. The s/n, such as 58db they claim in specs, is for their standard modes, so you have to deduct 3db from those s/n ratios for how clean a 0db image you will get at their claimed f11@2000. It's tricks like that that piss me off. Sony doesn't do this, if they say f12@2000 with a 62db s/n ratio, that is what you get, so you can add 18db of gain and know how clean your image will still be.

My work JVC HM890 stores separate gain settings for each mode. I have standard set to 0/9/18 and extended to 0/6/12. When I toggle between modes at high setting, the image stays exactly the same, meaning 18db at standard is as bright and noisy as 12db on extended. Might as well just add 18db of extended gain and put 0db on it, call the camera f32@2000 :-)

Paul

[Ron Evans]

My NX5U starts at -6db but becomes very noisy almost unusable at 12db. I too always assumed that this was really 18db and the noise then would make sense !!! If you are outdoors in manual the "set ND " often appears but of course in auto the camera will go to -3 or -6 before giving this message. A case where manual may not give as good an image if one doesn't realize that 0 is not really 0.

Ron Evans

[Paul Anderegg]

See page 9 starting with section 2.4 in the below PDF. Testing of the JVC HM650 and it's pretend f11@2000 extended mode. Tester shows that standard mode is only f7.5@2000, which is less sensitive than a 1993 Sony Betacam. JVC states in their brochure that standard sensitivity of the HM650 is F11@2000, which is a flat out lie. For reference, the PXW-X180 is f9@2000 with a real 60db s/n ratio, and I found it to be as bad as the X70 in low light.

https://tech.ebu.ch/docs/tech/tech3335_s08.pdf

JVC still will not provide signal to noise ratios for their cameras, and their brochures state just "excellent!" or some other nonsense.

Sorry to rant, but these stupid things make it hard for us to make educated camera purchase decisions. Short of buying every damn camera on the market to test, which I seem to actually do, there is no way to know what you will get in low light. About as annoying as drivers who do not signal with their blinker before cutting into your lane...grrr.

Paul

PS: Grrrrrrrrr!

[Pete Cofrancesco]

I'm glad I saw this thread. I also film theatrical performances and was considering this camera (z150) not for the 4k but bigger sensor for low light and maybe more dynamic range but now I see I avoided a costly mistake.

To add my 2 cents judging from your pictures. It seems most of the noise/grain issue are due the 1 sensor not being able to render saturated color more specifically red which always has been difficult to reproduce. At the end of the day real world performance trumps all. It's tricky making purchasing decisions on specs and assuming imagine quality improves with each generation, larger sensor, more pixels etc

I think under less colorful performances it would perform fine. It be tricky for me to quantify how many performances would cause what you experienced. Though the benefits don't seem to warrant the risks for this application.

[Dave Patterson]

Does the PXW-Z150 has an ISO display option (instead of gain)? A friend of mine has this camera and wants to use his hand held meter with the camera, but can't find a way to set the camera to display ISO settings. I downloaded the manual and see no mention of an ISO display setting.

[Doug Jensen]

No, thankfully the Z150 designers did not give in to the silly ISO fad. ISO has no business on a video camera and just leads to confusion and incorrect exposure. And even on those cameras that do show an ISO, the number is almost never correct. Another complication is that the ISO value would fluctuate from gamma to gamma anyway, so the first step is to setup the camera with a specific Picture Profile -- and then rate it himself. Only then will he be able to determine the camera's true ISO number for that particular configuration. You should advise your friend to set his meter aside and use zebras instead. Everything he needs is bult right into the camera and far more accurate and faster than using a light meter. He should take that as good news - faster, easier, and more accurate. Put the meter down and step away.

[Paul Anderegg]

Along with ISO, they need to keep a 1/50 shutter selection OFF of the 60p codecs! They added that into one of the X70 firmware updates and now I have to jog through 1/50 to try and hit 1/30...SONY!!!!

Paul

[Andrew Smith]

Both Paul and Doug discussing geeky camera stuff in the same thread? This is my happy place.

Andrew

[Dave Patterson]

Thanks for taking time to reply, Doug.

I agree that rating a camera is crucial to understanding its sensor's sensitivity, and that the ISO response can vary with different gammas used. Such is the case with my SONY FS5 picture profiles.

This is the first time I have heard anyone advise against using a light meter. Many cinematographers swear by their light meter as a great tool for lighting and calculating ratios, and I see them used on set all the time.

Originally Posted by Doug Jensen

No, thankfully the Z150 designers did not give in to the silly ISO fad. ISO has no business on a video camera and just leads to confusion and incorrect exposure. And even on those cameras that do show an ISO, the number is almost never correct. Another complication is that the ISO value would fluctuate from gamma to gamma anyway, so the first step is to setup the camera with a specific Picture Profile -- and then rate it himself. Only then will he be able to determine the camera's true ISO number for that particular configuration. You should advise your friend to set his meter aside and use zebras instead. Everything he needs is bult right into the camera and far more accurate and faster than using a light meter. He should take that as good news - faster, easier, and more accurate. Put the meter down and step away.

[Paul Anderegg]

The 1" sensor in these cameras doesn't have enough dynamic range to sue a light meter. Light meters come from the film days, where latitude was several stops in either direction from proper exposure. How many 6db gain steps up or down do you think you can get away with on the Z150 and still have usable highlights or shadows, even in 10 bit?

Paul

[Doug Jensen]

Dave,

Let me clairfy. There is nothing wrong with using a light meter to set exposure with any camera that allows full manual exposure control. And that certainly includes the Z150. However, the operator absolutely must rate the camera themselves after they have painted the camera, and not rely on the bogus ISO numbers that might be displayed onboard the camera. In other words, if someone knows the true sensitivity of the camera, a light meter is perfectly fine. But the flip side is that it is also harder to use, takes more, time, and is more prone to causing an exposure mistake than using zebras. On the other hand, zebras are built into the camera and are fool proof for an experienced operator.

Could you you write a best-selling novel with a pencil or a typewriter if you wanted to? Yes. But most authors would choose a word processor today. We've got better technology in our cameras than light meters offer. Film is dead and we are all using electronic cameras now that offer tools not available to cinematographers 50 years ago.

[Dave Patterson]

Thanks, Doug. I learned a lot about my FS5 from renting your video series on that camera, so I appreciate your insight. I have favored using gain for a long time, as it simplifies the process when switching between various gammas on the FS5. I often run into multi-camera shoots where someone insists we match ISO settings on different cameras. I have explained that it isn't an accurate way to match cameras, but that usually falls on deaf ears.

cheers,
Dave

[Doug Jensen]

It sure does!

[Dave Patterson]

Not sure I'm clear on your reply, but setting different cameras to the same ISO has NOT been a reliable match in my experience.