Avatar Sequels no longer on Red?

[Hrvoje Simic]

Um...huh? I know it's not generally your style but could you explain what you mean in more detail. Perhaps start by explaining, in detail, what you meant about having high hopes for Gemini @1600 - you know, the process. Maybe I misunderstood what you meant.

Then, maybe you could explain, in detail, what you thought I meant by Gemini @ 3200 - you know, the process. Maybe you misunderstood what I meant.

Analog gain electronically boosts the signal prior to quantization and defines what gets captured. Digital equivalent boosts what got captured. So two different routes with differing advantages and disadvantages.

Analog (real) gain push cuts off the DR in highlights and digital ISO equivalent push to 1600 just redistributes the captured. So...if the sensor is clean and saturated enough with 1600 ISO equivalent exposure it should offer more highlight range than rated at- and exposed to- ISO800 in normal mode, or 3200 in low light mode.

So if all works out...with Gemini, generally speaking, one could choose between three possible approaches of exposure:

1) ISO800 rating in normal mode - for cleanest image and best density
2) ISO1600 rating in normal mode - for more balanced DR distribution, better highlight range and better low light performance
3) ISO3200 rating in low light mode - for sensitivity priority and even better low light performance

The same logic applies to Venice with ISO 500 and 1000 in normal gain and ISO2500 with analog gain boost, and with cameras using this dual gain principle as well.

 

[Brian Boyer]

Analog gain electronically boosts the signal prior to quantization and defines what gets captured. Digital equivalent boosts what got captured. So two different routes with differing advantages and disadvantages.

Analog (real) gain push cuts off the DR in highlights and digital ISO equivalent push to 1600 just redistributes the captured. So...if the sensor is clean and saturated enough with 1600 ISO equivalent exposure it should offer more highlight range than rated at- and exposed to- ISO800 in normal mode, or 3200 in low light mode.

So if all works out...with Gemini, generally speaking, one could choose between three possible approaches of exposure:

1) ISO800 rating in normal mode - for cleanest image and best density
2) ISO1600 rating in normal mode - for more balanced DR distribution, better highlight range and better low light performance
3) ISO3200 rating in low light mode - for sensitivity priority and even better low light performance

The same logic applies to Venice with ISO 500 and 1000 in normal gain and ISO2500 with analog gain boost, and with cameras using this dual gain principle as well.

Okay, cool, that's what I thought. We're on the same page. Except, I was talking about Gemini rated @ ISO 3200 in normal mode, not the low light mode, which is why your response was confusing.

I'm assuming that's possible. I was under the impression low light mode had to be manually chosen and wasn't something that automatically kicked in at a set ISO.

To be clear, when I say rated I mean with a light meter and the camera set for ISO 3200 and all the corresponding lens and ND settings. Wouldn't ISO 3200 in that context provide the same benefit in terms of highlight capture (1 stop) over ISO 1600 as ISO 1600 has over ISO 800?

 

[Hrvoje Simic]

To be clear, when I say rated I mean with a light meter and the camera set for ISO 3200 and all the corresponding lens and ND settings. Wouldn't ISO 3200 in that context provide the same benefit in terms of highlight capture (1 stop) over ISO 1600 as ISO 1600 has over ISO 800?

Sure, same logic. Two stops more at the top if exposed based on that signal distribution.
But there is no free lunch here. Two stop push brings more tradeoffs. As one pushes the material, density gets reduced, which reduces tonal subtleties and overall image richness. Which also reduces the potential of post production.

Due to online gear propaganda sites playing journalism it has become a trend to judge high ISO push performance predominantly based on the absence of noise. Which is okay for online blogosphere pics viewed on portable devices but not really for a material which needs proper post and appearance on the big screen and brings a DoP with some specific requests.

 

[Brian Boyer]

Sure, same logic. Two stops more at the top if exposed based on that signal distribution.
But there is no free lunch here. Two stop push brings more tradeoffs. As one pushes the material, density gets reduced, which reduces tonal subtleties and overall image richness. Which also reduces the potential of post production.

Due to online gear propaganda sites playing journalism it has become a trend to judge high ISO push performance predominantly based on the absence of noise. Which is okay for online blogosphere pics viewed on portable devices but not really for a material which needs proper post and appearance on the big screen and brings a DoP with some specific requests.

Trade-offs are almost always inevitable, which is why I indicated I'd like to see how Gemini @ ISO 3200 fares. It may not be as bad as the numbers would make it seem in certain situations.

I recently recalled in a different thread how RED trotted out MTF graphs of Epic M-X and Alexa before Epic was released. They touted how thick the M-X graph was compared to Alexa as if it were a surefire indicator of real-world, overall image quality. Well, we all know who won that battle.

Over the last 9 years of reading Reduser on a regular basis, I've seen plenty of noise-free, artifact-free, "thick" negatives that, to be blunt, look like ass.

While those technical issues do indeed exist, too many people fixate on avoiding them when they should really be focusing on achieving consistently better looking imagery through improved taste and technique.

So, if increased highlight capture, which I argue is associated with the perception of better looking images, can enhance that perception for an already aesthetically pleasing image, somewhat reduced density, tonal subtleties and overall image richness may go unnoticed or be a small price to pay.

As I've said elsewhere, the image that looks better is better. The math can sometimes be misleading.