Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

5K is alive...

It might have, that's what I thought also. Frankie was working very quick, and I guess they also had the 300mm prototype?

I love the test case, it totally looks like a crank case with a supercharger behind it, though the lines and size remind you of a transmission.
 
I want the breadbox-cam...theft proof! :)
 
uslatin said:
...
I love the test case, it totally looks like a crank case with a supercharger behind it, though the lines and size remind you of a transmission.
Click to expand...

mopar :)
 
Total mopar lines! Hehe... I had a few mopar parts in a dodge of mine. I love the filled in big block power look of a good Dodge. However Scarlet is more of a ZR1 on slicks.

cool0036.gif
 
Jannard said:
We have successfully tested the 1st S35 Mysterium-X sensors in "Frankie 2". The images look incredible right out of the gate. We will be posting shots of the "Frankie 2" rig and frame grabs soon. Motion a bit later.

Wooooohooooo!

Jim
Click to expand...

Congrats Jim, we can't wait to see what you crank out next.

Happy Holidays!

Cheers,

Tim Fassnacht
 
Is it just me, or did the image get darker since yesterday?
 
Is it a sign that the other evening my wife and I decided that "Frankie" would be a great name for our next daughter?! :huh:
 
I love Frankie 2, a handsome brute!

Jim is definitely very brave and very generous, I can’t think of another company that would openly share photographs of things being prototyped. These kinds of things are generally VERY sensitive; I know that I hate showing things in development as I always want to put my very best foot forward, and would almost rather die than show work in progress until it is absolutely up to spec and finished. As a designer and SW and HW architect I feel very privileged to be allowed to come along for the ride and to get a peek on how things are going.

Jim your unprecedented generosity and openness is of course always greatly appreciated,

Cheers v. much

Eric

P.S. has anyone started an open/unlocked thread about the compressed vs. uncompressed debate?

.
 
I don't think it is prudent to open that thread again. Not many understand exactly how our sensor, DSP and compression engine works so there is a lot of random speculation about what we can, cannot or "should" be able to do. Since we are not in the position to explain it... :)

Jim
 
Joshua Brown said:
But Reduser is built on speculation and vaporware!

-Josh
Click to expand...

Speculation and vaporware are both usually much different than misconception.

Jim
 
Jannard said:
I don't think it is prudent to open that thread again. Not many understand exactly how our sensor, DSP and compression engine works so there is a lot of random speculation about what we can, cannot or "should" be able to do. Since we are not in the position to explain it... :)

Jim
Click to expand...

I completely agree that for general cinematic applications that compression is the only sensible way to go;

however (on a selfish note) I was personally interested in the RAW vs. Compression issue, as this could affect some of the work we are trying to do. We are trying to design various new types of 3d extraction systems from stereo footage [S3d]. Red one is definitely the best game in town for this. However, using conventional auto-correlation and pattern matching algorithms to extract explicit 3d models from stereo sequences one finds that compressed imagery tends to fool or snag these algorithms resulting in false stereo matches and false geometry (even if the imagery seems visually flawless to the human eye) . [we are working on algorithms to get around this problem but we have not tested any RED stereo footage as yet, would be nice to see if the RED schemes go better for auto-correlation and 3d feature extraction than more conventional compression schemes].

For VFX and computer game applications development, a set up of minimum four pairs of stereo cameras to record full 3d 360 of dynamic sequences in a controlled environment would be very useful, and hence we are very keen on the FF35 (Scarlet). For our purposes we would only record very short sequences of the order of five seconds or so, and we could also choreograph some of the action to produce sets of “pre“-key framed sequences. The higher resolution imagery from the FF35 (scarlet) and Epics would be invaluable for our work and others besides. The option of RAW for a low number of frames would be very nice indeed. However I understand that the compression scheme is an integral element to the sensor and early stage signal processing and hence uncompressed is not very practical and is counter to the whole system design.

The software we created uses a new spatial paradigm to very elegantly and efficiently represent complex 3d scenes from stereographic imagery producing extremely clean and life like models and is (conservatively 400x) more efficient when rendered than convention VR models of the same complexity. We see the FF35s as a real life line for us whatever the scheme of data/imagery is produced and are looking forward to building a lot of high quality but affordable advanced software and hardware 3d products around the new breed of RED cameras.

Forgive my ignorance, why is it so hard to build a high resolution sensor (10-22mp) with a frame rate of 30fps with decent DR (cinematic quality) with a true electronic global shutter (as opposed to a rolling shutter or twin RS). For 3d extraction from stereo from multiple cameras this would be the best way to get all the pieces to fit together properly. Rolling shutters tend to produce spatially distorted 3d fragments that don’t fit together properly. For us a global shutter is the holy grail, but for our applications industrial cameras with global shutters generally look horrible and are impractical but the stuff we like to do from medium and large format stills are visually quite stunning.

How difficult is it to build a high quality high res globally shuttered camera or how far away are “we” from ever doing this? I.e. a special globally shuttered camera for high end technical work?

Any input Jim or other wise would be greatly appreciated.
 
Jannard said:
Speculation and vaporware are both usually much different than misconception.

Jim
Click to expand...

Correct. I was referring to the speculation about your sensors, DSP, and compression engine given the inability to release any actual specifics about it. As mentioned above:
Not many understand exactly how our sensor, DSP and compression engine works so there is a lot of random speculation about what we can, cannot or "should" be able to do.
Click to expand...

Again, very excited to be more present for this cameras round of pre-production release info!

-Josh
 
Eric/VERAXYZ said:
Forgive my ignorance, why is it so hard to build a high resolution sensor (10-22mp) with a frame rate of 30fps with decent DR (cinematic quality) with a true electronic global shutter (as opposed to a rolling shutter or twin RS). For 3d extraction from stereo from multiple cameras this would be the best way to get all the pieces to fit together properly. Rolling shutters tend to produce spatially distorted 3d fragments that don’t fit together properly. For us a global shutter is the holy grail, but for our applications industrial cameras with global shutters generally look horrible and are impractical but the stuff we like to do from medium and large format stills are visually quite stunning.

How difficult is it to build a high quality high res globally shuttered camera or how far away are “we” from ever doing this? I.e. a special globally shuttered camera for high end technical work?

Any input Jim or other wise would be greatly appreciated.
Click to expand...

In no particular order...

Still cameras do not have a global shutter.

If you shoot a non moving object, even a slow read-reset sensor will look great.

There are widely varying degrees of read-reset times for a rolling shutter. Some as fast as a film camera.

A global shutter will "feel" different than a rapid-reset sensor or a film camera when used for motion.

Assuming that we all get used to a global shutter look (less "filmic" and more sterile), technology will take us there is a few years. It has only been very recently that we can have a high speed, high dynamic range, good light sensitivity, and a rapid-reset sensor.

As we move forward, you are likely to see all aspects of the sensor improve. We are spending a ton of time and research money on this. It is one of the main reasons why we believe in a sensor upgrade, starting with the RED ONE, in all our cameras without having to buy a completely new camera every year.

It is impossible to find what you are looking for today. Sometime in the future, when we have delivered all you are asking for today (assuming you really want it once you have it), you will have a whole new set of wishes. That's how this whole deal works.

Why can't car companies just build a sedan that has 600 horsepower and gets 40MPG? Answer: They can. Just not for a few years.

Jim
 
Jannard said:
It is one of the main reasons why we believe in a sensor upgrade, starting with the RED ONE, in all our cameras without having to buy a completely new camera every year.
Jim
Click to expand...

Hello

Sorry if I misread this post or if this has been officially addressed elsewhere before but does this mean the brain of Epic-X could one day be upgradable to a monstro S35 sensor ?
or will the 'sensor upgrade' path for this camera (to get more DR) will only be a change of brain for a FF35 ?

Opening the door to the idea of a sensor upgrade for Epix-X would make a lot of people happy and more confident into buying the "1st generation Epic" knowing it could be upgraded (like RED ONE) if needed without having to sell the old brain to get new one... Maybe it is not possible because all the electronic in the brain would need to be upgraded along with the sensor and it would make no sense with the new modular program.

Thanks

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