camera comparison F3 - ALEXA - RED MX

Cinesin said:
Looking at the lat charts, I'd have thought the MX (same sensor as the Epic) would have been better, but it isn't at all. With F3 on S-log, it seems like it would easily surpass it. Not that this really matters that much but it's interesting how, for the money, you can go that far. I'd be using diffusion a lot anyway so the high freq. stuff means little to nothing to me.
Click to expand...
I think people are expecting too much from the epic. Its just a smaller MX.
 
Thanks for the test and thanks for posting it. I agree, the Alexa just has a great look. What caught my attention is the whites. The Alexa did not have "white" whites yet it makes the most appealing, filmic images. In contrast, the F3 has the most true colors of the three. In some ways this leads one to believe that "True" colors trend toward a video type look. Film being linked to narrative work is often make believe and certian looks lend themself to suspending reality in a way... Interesting.

One other thing, it is quite annoying to me that these top end professional cameras all have skew. I realise this often does not show due to proper shooting technoque but it bugs me that it is there right around the corner. What if your subject was an athlete, would the tilted verticals be up to the standard of the rest of the camera or production? I don't think so but the camera world has seemed to have accepted this. I think it is limiting.
 
Do you think film doesn't have a rolling shutter? A film camera doesn't have a shutter that opens and closes like an SLR. A global shutter is a video look. You shoot around it.
 
Film has a mostly-global shutter, with a little bit of rolling at the start and end of the exposure. I'm going to do a skew test with "The Skewer" on my S16 film camera while I'm doing other tests with it, and that'll demonstrate how film reacts.
 
Bassman2003 said:
One other thing, it is quite annoying to me that these top end professional cameras all have skew. I realise this often does not show due to proper shooting technoque but it bugs me that it is there right around the corner. What if your subject was an athlete, would the tilted verticals be up to the standard of the rest of the camera or production? I don't think so but the camera world has seemed to have accepted this. I think it is limiting.
Click to expand...

I think this is a good blog on the matter: http://blog.abelcine.com/2009/04/15/the-difference-between-a-global-shutter-and-a-rolling-shutter/

Specifically:

"First off, CMOS does not necessarily mean rolling shutter and CCD does not necessarily mean global shutter. Predominantly that is the case, but it is not an absolute. There are CCD chips with rolling shutter and CMOS chips with global shutters. Also, the notion of rolling shutter = bad / global shutter = good is not particularly true either. A rolling shutter can have a more organic, more — wait for it — FILMIC look. Note I said ‘can’ as there are no absolutes in this world. A rolling shutter can also dramatically increase the fill factor of a photosite, making for greater sensitivity and a higher signal to noise ratio."

***

Sensors, and for that matter, cameras, are about trade offs. All things being equal, I prefer a global shutter. But usually, all things don't end up being equal. I would assume Sony would give us a global shutter on the PMW-F3 if it could have been done at the same price point without sacrificing other qualities of the image sensor. Most likely that is not the case. The Abel Cine blog - which mentions the EX1 and EX3 by name, suggests that rolling shutter can help make photosites more sensitive. I don't know the technical details on this sort of thing, but to the degree that that is true - it may have a significant impact on why the PMW-F3 has such incredible light sensitivity and such an impressive signal to noise ratio.

To the degree that there are not comparable global shutter cameras on the market in my price point, I accept this. All cameras have limitations, and no camera is perfect for all jobs. Use the one that fits the application.
 
Last edited:
Barry_Green said:
Film has a mostly-global shutter, with a little bit of rolling at the start and end of the exposure. I'm going to do a skew test with "The Skewer" on my S16 film camera while I'm doing other tests with it, and that'll demonstrate how film reacts.
Click to expand...

Technically it has a completely rolling shutter in the form of a circular disk. However, for practical effect you notice is much more at the beginning and end of each exposure (due to the edges of the disk) as opposed to evenly throughout the exposure as with video.

Short answer is, film does have rolling shutter but its not the same. More like a parabolic wave form than a straight line.
 
Yes, but the point is -- through much of film's exposure, it is exposing as a fully global shutter. In a CMOS camera, it never is. Film behaves more like a global shutter than it does a rolling CMOS shutter.
 
Barry_Green said:
Yes, but the point is -- through much of film's exposure, it is exposing as a fully global shutter. In a CMOS camera, it never is. Film behaves more like a global shutter than it does a rolling CMOS shutter.
Click to expand...

So, I'm asking as someone who's always worked in digital, but with film, as the shutter angle decreases, wouldn't the potential for skew increase? Obviously, the roll off would be very different than a video CMOS skew, and not necessarily running from top to bottom either.

Is a rotary shutter always going to make one rotation per film frame in the gate, and will it always move at a constant speed, or is it possible to rotate the shutter at a faster speed than the film speed, but still have it only make one rotation per film frame in the gate?

In essence, could you assign the film to move at a rate of 24fps through the gate, but tell the rotary shutter to spin a complete spin in 1/48th of a second, then halt for 1/48th of a second, then repeat, etc - with a shutter angle of 180 degrees, for an effective exposure length of 1/96th of a second?


....Oh, and the other film skew conversation would be if we discussed out of phase shutter adjustments, such as in Saving Private Ryan.
 
The shutter makes one rotation per frame. If you were to shrink the shutter down to a ridiculously tiny size, such as, oh, 10 degrees, then yes it would act very very much like a CMOS rolling shutter. But, it also depends on the orientation of the mirror to the frame -- some film cameras slice down the frame like a CMOS, some slice sideways across the frame.
 
Barry and I are really in agreement here. I can only talk from my experience with my Eclair. Two things conspire to produce a rolling shutter effect in film. If you increase the frame rate to shoot slow mo the degree of rolling shutter increases at the edges. Especially since you may then make the shutter faster to avoid motion blur.

Anyway, you can get a rolling shutter skew on film, but at regular frame rates and regular 180 degree shutter angle, not much. When you do it tends to be at the extreme edges (parabolic wave) of the image so its not as noticeable as video doing it in the middle.
 
Last edited:
You must log in or register to reply here.
Back
Top