F3 Testing Series

[Andy Shipsides]

Hey guys

I am doing a three part series on F3 tests.The first part is here: AbelCine Tests the Sony F3 ISO Rating | CineTechnica

Next test is on dynamic range in different gamma modes, which I will post sometime next week.

Andy

[Aaron Newsome]

Excellent Andy. I'm looking forward to it. Your posts on the blog are always top notch. Thanks for the efforts.

[Nate Weaver]

Originally Posted by Andy Shipsides

Next test is on dynamic range in different gamma modes, which I will post sometime next week.

Hey Andy, that's excellent.

Please take a look at whether black gamma raising actually pulls some shadow detail out (thereby helping latitude), or whether it just takes what's already there and moves it around.

Somebody on another board is thinking it doesn't help and is just making more visible what's already there (which is of debatable use) rather than a true improvement. I'm inclined to believe him, he seems like a smart cat.

[Alister Chapman]

Raising the black gamma which acts as a black stretch won't normally reveal otherwise invisible detail, but it will increase the number of data bits used to record the shadow areas, so it should allow you to do more in post before you get any degradation, however the difference will be small.

[Andy Shipsides]

I agree with Alister. Small difference, but worth looking in to. I'll make that part of the next test.

Andy

[Nate Weaver]

Originally Posted by Andy Shipsides

I agree with Alister. Small difference, but worth looking in to. I'll make that part of the next test.

Andy

Eh, if two people are saying it doesn't drag any real picture info up from under 0IRE, and from what I can tell, it's true as well, then I dunno if it's worth wasting any time on.

I was hoping there was a way to just get some more out of it without the S-Log bit coming next month.

Even if it was all that info recorded to 8bit, that's still useful to me because there are times when myself and others want that low-con Red look and we don't intend to crush it back in color-correction.

[Alister Chapman]

What's really needed is a way to open up the iris without blowing out the highlights, that's the only way you will really see deeper into shadows. This is exactly what s-log should do.

[Leonard Levy]

This brings up an interesting related question for me.

I've always wondered why the detente setting (0) for cap black on both Panasonic cameras like the HVX and DVX series and the Sony EX series cameras was actually higher than 0 IRE- somewhere between 5 an 10 IRE usually.

I always lowered the black settings accordingly on those cameras to get cap black to "0" under the theory that
1- it was more accurate if I didn't trust that grading would be any done well or at all, and also
2- under the assumption that other issues being equal, using the full contrast range of the signal was preferable to condensing information into a flatter picture - Thus it would be cleaner to record shadow detail
say between 0 and 30 IRE than between say 8 and 32.

However after reading this thread I realized that there are more data bits available if the shadows are slightly elevated.

I generally shoot for other people who I don't trust to do a decent job of grading so I try to give them the closest to perfect color and contrast I can on set. But for myself is it preferable to shoot with elevated blacks for later grading? What about the idea of using more contrast in the first place?

BTW I was going to use the word "more bandwidth" instead of more contrast but I realized I don't really understand what bandwidth is .