13th responses here...

[Chris Parker]

i hear what you are saying, but it hasn't hurt the RED One sales. it is a tradeoff, and i believe RED puts specs in the new cameras based on feedback from this forum sometimes. which is one of the reasons for their insanely loyal fan base.

just as many waited much longer for their RED ONE, many more will wait longer still for their new scarlet or epic.

the promise of something better from RED will keep many people from buying another company's product as well. something to consider.

 

[PapaBear]

After the November 13th announcement, the biggest problem RED Digital Cinema has remains the same: no matter how advanced the future cameras, as long as the big editing applications (from Avid, Quantel and Autodesk) remain unable to work with Native R3D footage, many people will choose to stay with simpler, faster and proven workflows, therefore shooting with other cameras than RED. It is amazing that Jim and Co. refuse to see and solve the main obstacle their product faces on the market.

 

[Michael Schrengohst]

"Personally, I wince a little when I read some kid on scarletuser talking about his plans to shoot with Scarlet this coming Spring or, "at worst," Summer. If Red delivers, this kid's excitement will be good for Red. But if this kid gets disappointed repeatedly by one projection after another that perpetually puts his Scarlet "just around the corner"? Hard to see how that ends up being good for Red."

Yes, didn't we go through this with the RED?
I think everyone survived? I don't think all the RED films
were realized. I am sure many were shot on something else.
Having a Scarlet or Epic won't change anything.
I have done several RED shoots with rented equipment.
I have been through several decades of "must have"
video equipment. I don't know where several
hundreds of thousands of dollars that I spent on
the "must have" equipment has ended up.

Most of it was probably sold for pennies on the dollar.
Some was donated/given away and I think I still
have the odd 3/4" deck and TBC sitting in my garage.

Scarlett & Epic will really move the technology ball
down the field. I will probably purchase some of these
new toys but in 20 years they will be sitting in my garage
next to my 3/4" deck.

 

[Stuart English]

Workflow

Workflow

After the November 13th announcement, the biggest problem RED Digital Cinema has remains the same: no matter how advanced the future cameras, as long as the big editing applications (from Avid, Quantel and Autodesk) remain unable to work with Native R3D footage, many people will choose to stay with simpler, faster and proven workflows, therefore shooting with other cameras than RED. It is amazing that Jim and Co. refuse to see and solve the main obstacle their product faces on the market.

I might agree with your analysis, but the conclusion you have reached is dead wrong.

Here's one public domain example as to why -

http://www.quantel.com/page.php?n=Working_with_RED_files

 

[Tom Lowe]

Tom, Hi,
I know that your preference is to produce "futureproof" content, then why even think about Scarlet?
Cheers,
cd

Well Scarlet FF35mm is a freakin 6K camera, so that's plenty future-proofing for me. It will downsample beautifully to "True 4K."

I want to agree with those who are asking for timelapse on Scarlet 35mm cams. Just a very basic intervalometer function is not very complicated. Either that, or a connection (stereo minijack?) so it can interface with an external intervalometer.

BTW, I assume that Scarlet FF35mm somehow takes stills without a mechanical shutter?

 

[Steve Gibby]

After the November 13th announcement, the biggest problem RED Digital Cinema has remains the same: no matter how advanced the future cameras, as long as the big editing applications (from Avid, Quantel and Autodesk) remain unable to work with Native R3D footage, many people will choose to stay with simpler, faster and proven workflows, therefore shooting with other cameras than RED. It is amazing that Jim and Co. refuse to see and solve the main obstacle their product faces on the market.

I went to the RED event at Fox Studios on the 13th. In part of the event Ted Schiliwitz gave an overview of the advancements in post workflow on the horizon, and representatives from various post application companies demonstrated what they have proposed. My friend Mike Curtis of HD for Indies has posted his initial summary of that info here:

Link: http://provideocoalition.com/index....ions_for_red_announced_from_apple_adobe_avid/

Post options are expanding rapidly...

 

[Cüneyt Kaya]

I might agree with your analysis, but the conclusion you have reached is dead wrong.

Here's one public domain example as to why -

http://www.quantel.com/page.php?n=Working_with_RED_files

is this based on the sdk?
or do they have the same integration like scratch, redcine or redalert

 

[Michael Brennan]

16Bit A/D

16Bit A/D

On paper the new 16bit A/D will improve tonality in shadows and offer greater scope to finesse colours in post.
Good work!

Question is how many bits in
a)recording and
b)post

are desirable to leverage the most from a 16bit A/D??

Mike Brennan

 

[Robin Balas]

On paper the new 16bit A/D will improve tonality in shadows and offer greater scope to finesse colours in post.
Good work!

Question is how many bits in
a)recording and
b)post

are desirable to leverage the most from a 16bit A/D??

Mike Brennan

A sidebar comment on that,
good medium format stills chips have 12 real bits of information in them according to the manufacturer (DALSA), still the MFDB makers use a 16-bit A/D and claim benefits.

Personally I saw nothing changed when it went from 14 to 16 bit so in my personal opinion oversampling beyond 2-3 bits is without discussion a waste of space, you are simply sampling noise better, not the data. So I wouldn't bother worrying about 16 bit or not.

As long as you get at least 12 bits of real data you are better off than most MFDB's or you are equal to them, which is providing incredibly good and smooth details and tonalities. To support those 12-bit you need a 14 bit A/D to provide oversampling to avoid quantization or round off issues when pushing the shadows hard.
If the Monstro lives up to the hype, which it very well could do as it still has some time to cook yet, it should sport a S/N ration of 84dB before you "need" a 16 bit A/D to describe the information with a comfortable oversampling. Will it do that? I hope so, but only time will tell (or Graeme Nattress could know some more on this for motion use, I only know stills equipment)

But what equipment can be used to do post? If you "cook" (i.e. RED Alert or RED Cine) and grade your files close to the end result before going into a NLE, you only need a 10-bit file to be very happy - so most NLE's will do it justice. But if you need to do heavy post or grade it you must re-"cook" your files to benefit from the full bit depth. Best example is heavy lifting of the shadows without artefacts, it is a huge difference doing that on a 16 bit file vs. a 10-bit one. Other stuff is ok on the 10-bit so it depends a lot.

I'm really exited about this new stuff, but 2010 is a really long wait...:cold:
MHO.

 

[Jannard]

There is a temptation to believe that all CMOS sensors are the same... and that you can interpret from one manufacturer to another based on a particular spec. This is an absolute fallacy. There are so many differences in pixel design and sensor design that one could not catalog them all. 12 bit sensors today are much better than 12-bit sensors yesterday. The Mysterium vs. Mysterium-X (both 12-bit) is a case in point. The Mysterium-X has 1 full stop less noise and greater DNR. It runs almost twice the speed and has about half the read-reset time. Monstro (16-bit) is a big leap again. All are CMOS sensors made by RED. RED's sensors differ from every other manufacturer's CMOS sensors in many ways. Just because one CMOS sensor does one thing doesn't mean ALL CMOS sensors do the same thing. The "secret sauce" inside sensor designs is widening between companies. It used to be "meat, salt, pepper and onions". Now, it is every ingredient in the pantry. It is amazing the different "dishes" that can be cooked.

Jm

 

[Peter McCully]

There is a temptation to believe that all CMOS sensors are the same... and that you can interpret from one manufacturer to another based on a particular spec. This is an absolute fallacy. There are so many differences in pixel design and sensor design that one could not catalog them all. 12 bit sensors today are much better than 12-bit sensors yesterday. The Mysterium vs. Mysterium-X (both 12-bit) is a case in point. The Mysterium-X has 1 full stop less noise and greater DNR. It runs almost twice the speed and has about half the read-reset time. Monstro (16-bit) is a big leap again. All are CMOS sensors made by RED. RED's sensors differ from every other manufacturer's CMOS sensors in many ways. Just because one CMOS sensor does one thing doesn't mean ALL CMOS sensors do the same thing. The "secret sauce" inside sensor designs is widening between companies. It used to be "meat, salt, pepper and onions". Now, it is every ingredient in the pantry. It is amazing the different "dishes" that can be cooked.

Jm

Brilliant. Thanks for telling us for a fact what I hope and dream the new sensors will be.

 

[Joseph Ward]

There is a temptation to believe that all CMOS sensors are the same... and that you can interpret from one manufacturer to another based on a particular spec. This is an absolute fallacy. There are so many differences in pixel design and sensor design that one could not catalog them all. 12 bit sensors today are much better than 12-bit sensors yesterday. The Mysterium vs. Mysterium-X (both 12-bit) is a case in point. The Mysterium-X has 1 full stop less noise and greater DNR. It runs almost twice the speed and has about half the read-reset time. Monstro (16-bit) is a big leap again. All are CMOS sensors made by RED. RED's sensors differ from every other manufacturer's CMOS sensors in many ways. Just because one CMOS sensor does one thing doesn't mean ALL CMOS sensors do the same thing. The "secret sauce" inside sensor designs is widening between companies. It used to be "meat, salt, pepper and onions". Now, it is every ingredient in the pantry. It is amazing the different "dishes" that can be cooked.

Jim

Jim.

Years ago, I found CMOS to be ''better'' theoretically compared to CCD for me, but forgot why. I was happy that you have taken CMOS to the next level and proved it. I know they are both good of course.
I was wondering why you choose CMOS over CCD? Just would like some insight of the history behind Red and your decision making. Thanks. :detective2:

 

[Steve Lewis]

The "secret sauce" inside sensor designs is widening between companies. It used to be "meat, salt, pepper and onions". Now, it is every ingredient in the pantry. It is amazing the different "dishes" that can be cooked.

Awesome! I'll go with the 16-bit Peking Duck, thank you.

 

[C.H.Haskell]

There is a temptation to believe that all CMOS sensors are the same... and that you can interpret from one manufacturer to another based on a particular spec. This is an absolute fallacy. There are so many differences in pixel design and sensor design that one could not catalog them all. 12 bit sensors today are much better than 12-bit sensors yesterday. The Mysterium vs. Mysterium-X (both 12-bit) is a case in point. The Mysterium-X has 1 full stop less noise and greater DNR. It runs almost twice the speed and has about half the read-reset time. Monstro (16-bit) is a big leap again. All are CMOS sensors made by RED. RED's sensors differ from every other manufacturer's CMOS sensors in many ways. Just because one CMOS sensor does one thing doesn't mean ALL CMOS sensors do the same thing. The "secret sauce" inside sensor designs is widening between companies. It used to be "meat, salt, pepper and onions". Now, it is every ingredient in the pantry. It is amazing the different "dishes" that can be cooked.

Jm

Post of the day. I am hungry.

 

[Seth Larney]

Post of the day. I am hungry.

Amen to that!

 

[Ketch Rossi]

There is a temptation to believe that all CMOS sensors are the same... and that you can interpret from one manufacturer to another based on a particular spec. This is an absolute fallacy. There are so many differences in pixel design and sensor design that one could not catalog them all. 12 bit sensors today are much better than 12-bit sensors yesterday. The Mysterium vs. Mysterium-X (both 12-bit) is a case in point. The Mysterium-X has 1 full stop less noise and greater DNR. It runs almost twice the speed and has about half the read-reset time. Monstro (16-bit) is a big leap again. All are CMOS sensors made by RED. RED's sensors differ from every other manufacturer's CMOS sensors in many ways. Just because one CMOS sensor does one thing doesn't mean ALL CMOS sensors do the same thing. The "secret sauce" inside sensor designs is widening between companies. It used to be "meat, salt, pepper and onions". Now, it is every ingredient in the pantry. It is amazing the different "dishes" that can be cooked.

Jm

Now you start talking like me... people tell me that I should stop talking about Lasagna, Tirami Su' and Creation Lobster sauces.

BTW, still recovering from the announcements and trying to get my head around of it all and which cameras will really be the best path for my project and my company, then I remind myself of the RED ONE and all I can think is, KETCH buy some new RED ones now, talk less, shoot more... you get the picture:innocent:

Any way by the experience I had so far following this crazy out of the world company (if you can called that) I know that NAB will have some serious SHIT* to show for and that what is now will no be then, so no reason to get overly exited now, must leave some Heart attacks for later, hehe.

ciao

 

[conrad gaunt]

A sidebar comment on that,
good medium format stills chips have 12 real bits of information in them according to the manufacturer (DALSA), still the MFDB makers use a 16-bit A/D and claim benefits.

Personally I saw nothing changed when it went from 14 to 16 bit so in my personal opinion oversampling beyond 2-3 bits is without discussion a waste of space, you are simply sampling noise better, not the data. So I wouldn't bother worrying about 16 bit or not.

As long as you get at least 12 bits of real data you are better off than most MFDB's or you are equal to them, which is providing incredibly good and smooth details and tonalities. To support those 12-bit you need a 14 bit A/D to provide oversampling to avoid quantization or round off issues when pushing the shadows hard.
If the Monstro lives up to the hype, which it very well could do as it still has some time to cook yet, it should sport a S/N ration of 84dB before you "need" a 16 bit A/D to describe the information with a comfortable oversampling. Will it do that? I hope so, but only time will tell (or Graeme Nattress could know some more on this for motion use, I only know stills equipment)

But what equipment can be used to do post? If you "cook" (i.e. RED Alert or RED Cine) and grade your files close to the end result before going into a NLE, you only need a 10-bit file to be very happy - so most NLE's will do it justice. But if you need to do heavy post or grade it you must re-"cook" your files to benefit from the full bit depth. Best example is heavy lifting of the shadows without artefacts, it is a huge difference doing that on a 16 bit file vs. a 10-bit one. Other stuff is ok on the 10-bit so it depends a lot.

I'm really exited about this new stuff, but 2010 is a really long wait...:cold:
MHO.

erm, not sure how you'd get thirteen stops of dynamic range from a 12bit colour sample, can't be done, unless you use a log sensor. Since sensors generally aren't log, you need to oversample from a higher bit depth. My maths tells me you'd need at least a 13bit sensor, but frankly I wouldn't buy a camera with 13bits, it ain't lucky, so 14 would always be better
Sure you're sampling noise, thats what light is, but just because you can't see it on your 10bit/8bit/6bit (apple lcds) doesn't mean its not there or useful.
I'd use 32bit TIFFs if I had a 16bit sensor because I'm a bit anal about such things, and curves can crush data away. Log ain't useful in my opinion unless your trying to save space or doing a film out.
i think a 64bit OS will be needed to handle the memory requirements of even 9k footage, let alone 28k@16bit.
I also think it'd be good, since R3Ds are spatial compression and not temporal, to be able to output an unaltered, but trimmed R3D from redcine to save space, especially as the data rates increase. I don't see why it would be difficult if the source footage is of the same resolution and it would make long term archiving (still a major issue in my opinion) much easier.

 

[Marc Jonkers]

First of all : congratulations to Jim and the Red team !!!
The idea of “make your own Scarlet and or Epic” is just revolutionary !
This will open a lot of possibilities not only for Red but also for a lot of manufacturers (of lenses, ….)
I think November 13th will be a day to remember in video history!

The only thing I am afraid of is will Red make this true in such a short time schedule ?
Red is not bringing 1 or 2 products to market but a whole range of brains, lenses, battery’s , recording modules, I/O modules, …
Also a lot of new software has to be written for the devices, new red raw, new red cine, …..
All those things have to be developed /tested /produced / distributed etc in only a few months…
It seems they will have to work very hard at Red to make this come true.

As for sample footage, my guess is at the end of this year or beginning January …..

And for the fixed lens Scarlet : my guess is we will get a working model for NAB with all specs announced by then and for the price I think it will be 3.5 or max 4K for a fully working model.

 

[Sidney L. Plaut]

video history - video... what does this strange word mean, wasn´t that something they used a long time ago? like the cameras with 3 film rolls to make color?

Jim will be remembered as the man who killed video. not that I am going to miss it - remember filming outside in a forest or green grass and everything muddled together into a gooey green mass?

 

[Matt Dufilho]

So Jim,

What is the next announcement on the docket (accessory pricing, lenses, etc) and when can we expect it?

Don't you love how insatiably impatient we all are?


Thanks, again, for all that you and the team are doing.