What does dragon sensor mean for Scarlet?

[Lee Saxon]

Now I'm curious what's wrong with windowed 2K mode...

[PatrickFaith]

Now I'm curious what's wrong with windowed 2K mode...

Nuthing, I use it all the time, works perfectly as long as you black shade properly.

[Tom Gray]

TL;DR: The Dragon upgrade for Scarlet will cost more than the EPIC upgrade, will provide less power/framerates/rez compared to a Dragon equipped EPIC, yet at the end of the day, RED will always be there to accommodate you, by offering a discount on the new Dragon Scarlet, or by offering a reasonable upgrade path.



*To answer the OP's question, no one but Jim and Jarred know what a Dragon Scarlet will be capable of, but from what it looks like, it will be capable of 60fps 4K, but with a total investment close to $23K after upgrade. ($12K Scarlet Body, $11K upgrade price to get Dragon)

For that price, you could get an EPIC now, enjoy the high framerates, and then pay the cheaper $6,000 price quoted for an EPIC Dragon upgrade. You'll even get the upgrade quicker, since EPIC upgrades will be completed before they start working on Dragon Scarlets.




The Dragon upgrade, if they choose to offer it to Scarlet owners, will very likely bring some sort of 5K 24fps recording, but for more than the $6,000 price quoted for Epic owners. I'm guessing somewhere around $9K or $11K and that'll turn the Scarlet into an entirely new machine.

Keep in mind that the Scarlet has "about 25% the processing power of EPIC" according to RDC, and has a price to match. EPIC has extra horsepower that hasn't been taken advantage of, whereas the Scarlet was engineered from the start to be a TV/Indie/B Cam machine, utilizing all of the onboard processing power to churn out 4K 24fps. That's why features like 5K 24fps and/or the REDuser suggested "400 FPS burst mode" are harder to implement on Scarlet, simply because the power does not exist. It's like trying to get a Macbook Air to transcode R3Ds at the speed a Mac Pro or Z820 w/Rocket would—it's not technologically possible.

The REDCODE isn't the limitation, to my understanding, it's the power (ASICs) it takes to convert a RAW Bayer pattern image into REDCODE RAW, record that to SSD, provide a debayered monitoring path via HDSDI, as well as draw up the GUI on the EVF/LCD etc, etc. The EPIC has the power, Scarlet does not. That's why there's a $15K price premium on the EPIC. You're paying for framerates and power. This in turn makes the Dragon easier to install, because you don't have to upgrade as many components to get a usuable Dragon EPIC. Dragon just utilizes more of the onboard power.

The Scarlet when it was released before the Jan 1st price increase ($9,750) was substantially underpriced, IMHO. It worked, because it garnered a lot of press and generated a lot of new customers, but for that box in particular, it may not make sense to gut the whole thing and but Dragon capable ASICs in it.

Either way, RED will accommodate MX Scarlet owners, either by making the price of the Dragon Scarlet more than the cost to upgrade an MX Scarlet, or, by providing a discount to current Scarlet owners to buy a new Dragon Scarlet, or by some other means. It's in RED's DNA to keep the customer up-to-date. (or, at least have the option to. There are still quite a few Red One M's out there. )

To me, once being a business man, tells me this. Having two different chips sets made makes no sense.
chips get cheeper the more of one kind is ordered. I believe the scarlet and Epic are basically one in the same but is being detuned for obvious reasons and when the magic sensor is installed vavoom a new camera costs a lot and not much is really being done. IT'S JUST BUSINESS, nothing personal! Look at the sensor they have now, the same in both cameras... like I said, it's the way a good business is run. All the money is spent in R&D not in production.
Here is a better example. You walk into a Ford dealership and want to but a new truck, it's Jan, first. You want a discount and they tell you we only have 1500.00 to play with, look here is the invoice, this is what we payed for it and can't possibly sell it for that cheep, you leave. Well sure enough it's December the truck is still on the lot but now it's being sold for 15k less; do you really believe Ford is loosing 13.5k per car? If this were the case they would have needed to borrow money during the bailout but the didn't did they... think about it.. cheeper in numbers. They may have put chips in the Scarlet that didn't clock as fast but I'm sure will do much more than whats its doing now.

[Maik Müller Menendez]

Nuthing, I use it all the time, works perfectly as long as you black shade properly.

I agree, i was a bit worried about 2k before i got my Scarlet, but after doing some experiments i'm happy with the results. We have just to be more careful when shooting 2k and 1k, because we loose the ability to downscale, which washes away some noise and sharpens the image. That saves some real post-processing, but it's a bit of cheating. ;)

But, if your 2k-footage looks "terrible" than you're doing something wrong, probably with your 4k-footage too. This might be not obviously visible yet, because most of our 4k-footage gets always downscaled to 2k/HD, but later when everyone here got his 4k-monitoring-solution, the guys with bad 2k-footage might not be happy with their 4k-footage.

Remember, 2k-footage is just the same thing as a 2k-center-crop of a 5k-image, noise-level is the same, compression is the same. So if your 2k-images are noisy and not sharp, then your 4k-footage will look similar on 4k-screens.

[Brent G Miller Jr]

2K at 60 fps should definitely not look terrible. I shoot 120 at 2k on the Red One all the time and it looks amazing. Have you tested it yourself? How do you know it's terrible?

Because seeing the 2k @ 60 footage compared to downconverted 4k to 2k footage on Scarlet absolutely looks bad. You can see the difference right away. I would easily dish out another $6k for a Scarlet upgrade to just give me 4k @ 60. I really don't think it's asking for a lot.

[Brent G Miller Jr]

I agree, i was a bit worried about 2k before i got my Scarlet, but after doing some experiments i'm happy with the results. We have just to be more careful when shooting 2k and 1k, because we loose the ability to downscale, which washes away some noise and sharpens the image. That saves some real post-processing, but it's a bit of cheating. ;)

But, if your 2k-footage looks "terrible" than you're doing something wrong, probably with your 4k-footage too. This might be not obviously visible yet, because most of our 4k-footage gets always downscaled to 2k/HD, but later when everyone here got his 4k-monitoring-solution, the guys with bad 2k-footage might not be happy with their 4k-footage.

Remember, 2k-footage is just the same thing as a 2k-center-crop of a 5k-image, noise-level is the same, compression is the same. So if your 2k-images are noisy and not sharp, then your 4k-footage will look similar on 4k-screens.

On paper, I would aggree 100% but I have yet to see one single video showcasing Scarlet's 2k @ 60 footage that actually looked "good" so either no one out there knows how to shoot or Scarlet's 2k @ 60 looks less than desirable... Everything is so soft and grainy, it's like shooting on an entirely different camera...

https://vimeo.com/36255092
https://vimeo.com/42664365
https://vimeo.com/38761180

But the real proof is here....

https://vimeo.com/34875819

...It has mixed 4k @ 24 and 2k @ 60 footage and the difference when it cuts between the two is MASSIVE. It makes the 2k footage look absolutely terrible whereas viewing compressed in 720p on vimeo, both should look virtually identical in terms of overall image quality. but it doesn't.

FYI: I played with a Scarlet at NAB and LOVE the 4k. We want to order some of these but without a "good looking" 60p option, it's hard to take that plunge when considering cameras like the C500...

[Lee Saxon]

You can't compare native 2K to downsampled 4K. There'd have to be a major problem with the 4K material for the native 2K to stand up. That's not a legit comparison.

Shoot in 4K, CROP to 2K, and compare. That's how you find out if there's anything wrong with native 2K.

Perhaps even more relevant would be to compare native 2K mode to 2K-only cameras.

[Brent G Miller Jr]

You can't compare native 2K to downsampled 4K. There'd have to be a major problem with the 4K material for the native 2K to stand up. That's not a legit comparison.

Shoot in 4K, CROP to 2K, and compare. That's how you find out if there's anything wrong with native 2K

Granted, but I still would have assumed the 2k would look better than it does. It's so soft and grainy that, dare I say, that some other cameras shooting at native 2k look sharper/cleaner/better...

[Jon Shaw]

This was shot yesterday, 2k 60fps, no noise reduction, just a slight curve and WB adjustment. Obviously doesn't look as good as 4k 60fps but when you see the 2K file there isn't much noise



Password: ginclear

[PatrickFaith]

This was shot yesterday, 2k 60fps, no noise reduction, just a slight curve and WB adjustment. Obviously doesn't look as good as 4k 60fps but when you see the 2K file there isn't much noise



Password: ginclear

Nice, looks great with no artifacts... (getting the dynamic range on fire can be tricky)!
I also like the concept of the shot.