The future...

[Gavin Greenwalt]

Gavin,

the quality of high ISO was logarithmically increased rapidly in the past couple of years and will continue this trend all in that way in the future

Yes but the important lesson is that there is no reason to output "two ISOs" of the same RAW image to generate an HDR image.

 

[Nils Ruinet]

Yes but the important lesson is that there is no reason to output "two ISOs" of the same RAW image to generate an HDR image.

Exactly. If it's already in the RAW, you don't need HDR, just use curves or windows to get the info you need.
I think what Sanjin is basically asking for is a new sensor with more native dynamic range where you could dig deeper into the blacks. That has nothing to do with the so called HDR, which involves multiple exposures to add more dynamic range than what the existing sensor can handle.

 

[Mark Warren]

That sounds awesome. I'm excited to hear whats coming.

However, there is an army of us that are anxious for the chance to buy a Scarlet. We're excited that there are more opportunities out there, but somewhat dismayed at talk about new opportunities when we have yet to have an opportunity to purchase a camera announced years ago. Forgive my frankness.

How close is Scarlet to EPIC-X release?

RED's marketing is usually to talk about what's coming in 2015, 2020, or 2025 to distract you from what you cannot have today. Be distracted and obey!

(As I've said before, the delay on Epic-X is actually good for me, for all sorts of sublime and ridiculous reasons, so... I can have a gentle dig!)

 

[Sanjin Jukic]

Yes but the important lesson is that there is no reason to output "two ISOs" of the same RAW image to generate an HDR image.

Super clean high ISO will definitely rule exposure in the future of digital still and cine cameras.

RED should pay a high priority to high ISO for development of higher dynamic range (together with a new A/D circuit) at its upcoming new sensor called Red Dragon.

One good example is Sony F3 that has a pretty good high ISO that also enormously helped Sony in development of F3 internal S-Log curve

that increased F3 DR from about 11 stops to 13.5 stops that all available at 4:4:4 dual/3G HD-SDI recording.

 

[Don Woods]

Its hard to believe that it has already been 5 years. I remember the first time Jim shared his idea and vision with us on DVXuser. Amazing how far this has gone and I can't wait to see what is coming.

 

[Phil Anderson]

I have to admit it really looks like they are progressing, and after a bumpy start like most new companies are really starting to roll. They are beginning to look like the company we all expected them to be.

Anyone care to speculate details what the "one lens issue" Jim refers to in regards to Scarlet means? A post-quake part shortage? A build issue? A firmware issue? A design issue? A two month delay? Six month delay?

 

[Wayne Morellini]

Super clean high ISO will definitely rule exposure in the future of digital still and cine cameras.

RED should pay a high priority to high ISO for development of higher dynamic range (together with a new A/D circuit) at its upcoming new sensor called Red Dragon.

One good example is Sony F3 that has a pretty good high ISO that also enormously helped Sony in development of F3 internal S-Log curve

that increased F3 DR from about 11 stops to 13.5 stops that all available at 4:4:4 dual/3G HD-SDI recording.

Sanjin, do you know of any camera's or sensors with super clean high iso? I thought I had located one around 2006 or 2008, from planet82 I think, with 120db signal to noise ratio, but that did not pan out. I am trying to find a Omron technology that had 154+ db dynamic range of 27 stops from around 2006 that went quite.

 

[Sanjin Jukic]

A super clean high ISO means very usable higher ISO speed that can make very similar image quality during exposure that you can get today with use of slower shutter speed (or longer exposure time).

Time when a slower shutter speed can be compensated with a higher ISO speed and also can be very nicely used to extend DR in digital motion acquisition is coming sooner that you would might think.

That's all.

 

[Wayne Morellini]

A super clean high ISO means very usable higher ISO speed that can make very similar image quality during exposure that you can get today with use of slower shutter speed (or longer exposure time).

Time when a slower shutter speed can be compensated with a higher ISO speed and also can be very nicely used to extend DR in digital motion acquisition is coming sooner that you would might think.

That's all.

So you don't. I've spent a few hours reviewing some information and can give you some links to find such sensors. You will also need high dynamic range aswell as high signal to noise ratio to best do this trick. I've come across a few other interesting things as well.

Red might be interested in the following for high dynamic range pixels. It needs refining of course, because it is security camera chip. I have one of the small stills cameras from the early century, and it works like a charm for the day. I had a continuous reset/drain pixel idea years ago, so this is not far off. As far as I am aware the current Red system used is a dual frame exposure as a retrofit after we suggested it, rather than an integrated pixel multiple reset combining circuit. This has the problem of producing multiple overlapping exposures of motion, where my scheme doesn't, though it would be interesting to see how the noise floor holds up with my system. The energy barrier idea is also an direction I was interested in.

www.melexis.com/Asset/Autobrite-Whitepaper-DownloadLink-5815.aspx

This one seems to describe what I was saying about needing hdr with h-iso at night. 154db hdr.
www.electronicsbus.com/hdr-image-sensor-night-vision-surveillance-security-systems/

Here is the competition from 2003, 170db over 27 stop dr with correct consistent color and down to 0.001 (star light) lux sensitivity, both figures Red would be proud to have (the low light sensitivity might not be conventional enough though).
www.industrial.omron.eu/en/news/news/cmos_image_sensor.html?archive=1&page=11
www.ims-chips.de/home.php?id=a3b5c2en&adm=
www.ims-chips.de/home.php?id=a3b15c1en&adm=
These people are doing some serious research to advance the industry and fhd chips.

Here is a list of known sensor companies:
www.image-sensors-world.blogspot.com/p/image-sensor-companies-list.html

Here is a good place for sensor technology news: Much more than you are going to find around here, or your average digital cinema production or stills sites.
www.Iimage-sensors-world.blogspot.c...-max=2012-01-01T00:00:00+02:00&max-results=50

This is lenless image sensor: Not really the same as my scheme I have mentioned, but a bit similar.
www.image-sensors-world.blogspot.com/2011/07/lens-less-furrier-domain-capture-camera.html

Transparent pixels:
www.image-sensors-world.blogspot.com/2011/07/samsung-applies-for-transparent-pixel.html
I was considering separate methologies for this.

25mp 5k 53fps global shutter 35mm fillfactory sensor, and 5mp and fullhd 2/3rds. The original home made digital cinema camera project that led to the project that inspired Jim to do Red was based on a advanced fillfactory sensor. I was hoping back then to attract a sugar daddy company to do a commercial camera and cheap mass market version.
www.onsemi.com/PowerSolutions/product.do?id=VITA25K
www.onsemi.com/PowerSolutions/product.do?id=VITA5000
www.onsemi.com/PowerSolutions/product.do?id=VITA2000
Now fillfactory tech now owned by onsemi and smalcamera tech now owned by melexis with 154db dr were both snapped up around the same time by cypress semiconductor, before being resold latter. So I don't know if the onsemi has picked up any of the 154db autobrite technology, the data sheets are not there ye.t

Anyway why mention it, it's interesting, and I allways thought if a well financed company can cherry pick these technologies and companies they could make a pretty kick ass sensor. The rate these companies are changing hands (and these are not all of them) the tech might be obtainable.. Of course Red knows half of this already.