Abel Cine Reviews FS700

Ehm, again Andy misreads the Dynamic Range chart like with the FS100. Can someone count the stops please and tell me what you see, because I don´t see 11,5 stops...

And it looks like Cine 3 yields the most dynamic range. The last stop almost vanishes in the dark BUT the brightes part is only at 100 IRE, so when you put that back to 109 IRE it could be more overall than cine 1 and 2.
Cine 4, no idea.
 
Last edited:
I do find the dynamic range testing a little dubious. I mean there are usable stops and there are theoretical stops. In the super brights you may have 3 stops but they're not actually useable, you can't bring them down. I've personally measured about 9 stops of usable range (for me personally) stuff that can be moved around. I'm not disputing 11.5 stops but just questioning where they are and how they are squeezed into that 8 bit container.

cheers
paul
 
Actually what is interesting is the highlight handling in the candle light - the difference between the 100 and 700 is remarkable, this is the yellowing highlight shown very obviously. I wonder what PP Andy was using on the 100 to get that. Looks like no PP at all (PP Off)

cheers
paul
 
One thing i notice is that the skin tones look better on the FS700 or is that just me? Also the detail is far better... on the 100 the skin almost seem to blur out or get caught up in banding. But on the 700 we get that nice HD cripsyness. Here are two unaltered grabs from the video on AbelCine to show the coloration difference. Hope you don't mind.

View attachment 53430View attachment 53431

The first image is from the 700 and the 2nd is from the 100.

View attachment 53432View attachment 53433
 
Last edited:
I just got done reading the internal reflection thread as I'm thinking about getting an FS700. The FIRST thing I noticed on the candle light test is how much stronger the internal reflection of the candle flame is on the FS100. It's plain as day and dancing like James Brown. On the FS700, you can see it, but it's much fainter; which is obviously a good thing.

With that said, it looks like on the FS100, there is only 1 internal reflection of the candle, but on the FS700, it looks like there are multiple reflections of the candle, but they could almost pass as a lens flare.
 
paulcurtis said:
I do find the dynamic range testing a little dubious. I mean there are usable stops and there are theoretical stops. In the super brights you may have 3 stops but they're not actually useable, you can't bring them down. I've personally measured about 9 stops of usable range (for me personally) stuff that can be moved around. I'm not disputing 11.5 stops but just questioning where they are and how they are squeezed into that 8 bit container.

cheers
paul
Click to expand...


I find that interesting. In this video,

https://vimeo.com/groups/fs700/videos/41973197

the very last thing the guy says (@ 5:22) is that it has 13 stops. Is 13 even remotely possible on this sensor?
 
Tim Nitsch said:
I find that interesting. In this video,

https://vimeo.com/groups/fs700/videos/41973197

the very last thing the guy says (@ 5:22) is that it has 13 stops. Is 13 even remotely possible on this sensor?
Click to expand...

It's not so much the sensor as the 8 bit container that's the limiting factor, after all the sensor is cut from the same or similar cloth to the F3. When calling out range it's an issue of what you feel you can objectively see in the shadows before it becomes too noisy. The real world and scopes can be different.

Simplistically, say 8 bit can handle 8 stops with 32 levels of brightness for each stop, from 1 to 255. So, for all cameras, to get more then you need a curve to squeeze the range so that the brightest parts take up less data (as light is non linear, as are our eyes) but 13 stops would be 2^13 contrast, to 1 to 8192. Somehow that would need to be squeezed into 1-255, so how many steps for each stop? That's what i mean about useable range. Because out of those 13 stops even if you could discern a difference a whole stop might be in the range of 253 to 255 (for example, so representing a stop with 2 light levels). Frankly you'd be hard pushed to see that. In the real world i believe the contrast with the sun out would be about 20000 to 1, that puts some perspective on battling with ranges

Everything is about how to best squeeze a large range into a small container using the limited PP tools we have.

In an ideal world sony would allow us to create our own LUTS to translate sensor data into that 8 bit container. But that's isn't likely to happen.

cheers
paul
 
You must log in or register to reply here.
Back
Top