KOMODO....

[Ash Worth]

Has to be FF, no question, that's where the market is going and why use the RF lenses if not for this?

 

[Misha Engel]

Has to be FF, no question, that's where the market is going and why use the RF lenses if not for this?

Looking at the picture in Jarred's post it isn't FF

It looks more like a 6k (5um pixel) sensor (Dragon 6k size) at most.

 

[jamesbridges]

You do not think it’s a Canon sensor?

 

[Blair S. Paulsen]

Walkarounding

Walkarounding

I can think of a bunch of reasons to stay at S35 vs FF or VV - especially if size, data rates, heat dissipation, etc are part of the design envelope. I love larger formats, but for a lot of production situations it can be more trouble than its worth, starting with shorter DoF. Let the Komodo be the ideal mix of size and capabilities that will make it a great tool in situations where you don't want a bigger footprint.

+1 on power draw/battery note. In that same vein, I'd love to have a well implemented "standby" mode to avoid missing a unique "moment" during boot up. I realize that the sensor might not be at the ideal temp in a quick start scenario, but perhaps that can be compensated for to some degree. IAC, I'm looking for an option other than:

1) keeping it running - burning batteries and generating heat
2) staring at the boot screen while I miss the shot of the day

Cheers - #19

 

[Mike P.]

Looking at the picture in Jarred's post it isn't FF

It looks more like a 6k (5um pixel) sensor (Dragon 6k size) at most.

$6k for 6k still has a nice ring to it.

The active area of that particular sensor does look quite a bit smaller than the gate... However, that gate looks it could accommodate FF35, so maybe there'll be an FF35 version?!

Also if that is in fact an RF mount, then it looks like it'll be a 3:2 sensor, which is the new hotness for anamorphic... And if the sensor is ~6kx4k, it'd give you ~8k Ana desqueeze, which is a nice middle ground.

 

[Steve Gibby]

Whatever Komodo is, I imagine it will be something very cool and ground breaking! Stay tuned folks...

 

[Mathias Erichsen]

Is it possible that Komodo is some kind of sensor box where you can have all different kinds of RED sensors.. could it possibly connect to a base (like DSMC2 and beyond) with a cord for high end recording AND also be a stand alone little camera with R3D Lite?
That could be pretty neat

 

[Misha Engel]

You do not think it’s a Canon sensor?

I have no idea if it's a canon sensor, I know that it is a RF mount and that the sensor is smaller than a canon full-frame sensor.

 

[Blair S. Paulsen]

Is it possible that Komodo is some kind of sensor box where you can have all different kinds of RED sensors.. could it possibly connect to a base (like DSMC2 and beyond) with a cord for high end recording AND also be a stand alone little camera with R3D Lite? That could be pretty neat

I think there are a lot of applications for small, high resolution camera heads connected to external infrastructure. For starters, I expect use of remote PTZ rigs that leverage modern robotics tech to become popular for a wide variety of live coverage use cases. Tokyo 2020 anyone?

If I were wagering, I'd bet on a S35-ish 8K sensor - that's not field swappable. Why? Anything bigger than S35 only increases the size of the unit and makes focus at longer focals more challenging. It's hard to imagine a design that would support sensor changes in the field without adding at least some size/weight/cost/complexity/etc. Now if you want to talk DSMC3...

Cheers - #19

 

[Mathias Erichsen]

It's hard to imagine a design that would support sensor changes in the field without adding at least some size/weight/cost/complexity/etc. Now if you want to talk DSMC3...

You could have a Komodo for each of your sensors. So no swapping of sensors just swap the Komodo... aaand maybe it could be a part of DSMC3 but also be quite functional with DSMC2 somehow...
Ok - just some wild speculations going on here

 

[Brandon Veen]

Going off the lead of Mathias Erichsen's post from Page 8 (overlaying the KOMODO over the EOS R), I found a video of the EOS R on a turntable at a closer angle, grabbed a frame for comparison, and threw some sensor lines on.

Red Line is the EOS R's 3:2 Sensor.
Orange is Helium S35.
Yellow is Monstro VV.

The image isn't perfectly lined up, but it's well within the difference between full frame and S35 (if anything the EOS R is being represented as slightly too big).
The bigger issue is that the angle still isn't perfectly matched up, so the EOS R sensor is going to be further to the left than the KOMODO's (not accounting the depth of the OLPF that might be making a difference as well).
That's why it's aligned based on the sensor cutout rather than the sensor edge itself.

With that in mind, it looks like the KOMODO's sensor cutout is closer to where the EOS R's FF cut is than where a S35 cut would be at, and additionally looks too tall and not wide enough to be the same size cut as a Monstro sensor.
If I had to guess, I'd say it's likely a FF 3:2 sensor. As we can flip the bottom of the mount and line it up and see the spacing is right.

 

[Zack CC]

Going off the lead of Mathias Erichsen's post from Page 8 (overlaying the KOMODO over the EOS R), I found a video of the EOS R on a turntable at a closer angle, grabbed a frame for comparison, and threw some sensor lines on.

Red Line is the EOS R's 3:2 Sensor.
Orange is Helium S35.
Yellow is Monstro VV.

The image isn't perfectly lined up, but it's well within the difference between full frame and S35 (if anything the EOS R is being represented as slightly too big).
The bigger issue is that the angle still isn't perfectly matched up, so the EOS R sensor is going to be further to the left than the KOMODO's (not accounting the depth of the OLPF that might be making a difference as well).
That's why it's aligned based on the sensor cutout rather than the sensor edge itself.

With that in mind, it looks like the KOMODO's sensor cutout is closer to where the EOS R's FF cut is than where a S35 cut would be at, and additionally looks too tall and not wide enough to be the same size cut as a Monstro sensor.
If I had to guess, I'd say it's likely a FF 3:2 sensor. As we can flip the bottom of the mount and line it up and see the spacing is right.

Was waiting for someone to do that... Nice.

 

[Blair S. Paulsen]

If it does turn out to be FF, I wonder what windowing options it might support. I wonder if it would have sufficient processing power (and heat dissipation) for in camera downscaling?

Cheers - #19

 

[Jacek Zakowicz]

That would kill the Monstro- I could be wrong but I doubt it very much...

 

[Karim D. Ghantous]

You do not think it’s a Canon sensor?

RED would never use an inferior sensor in a cinema camera. What doesn't make sense is that it doesn't seem to be a stereo camera. Which is should be, you would think, if it's associated with Hydrogen. But, I like surprises. Sometimes I can guess the specs of a new RED camera (like I did with Raven IIRC), sometimes I can't.

 

[Aaron Lochert]

I'm surprised at the lack of conversation about where a new sensor fits into the existing lineup and what it means (if anything) for the unified body, mounts and lenses people already have.

I think people aren't talking about it because they're keeping their expectations in check. This camera is announced as not DSMC3 and not a replacement for DSMC2. They know how limiting physics can be, so they know that sacrifices are usually made in the process of miniaturizing something. Assuming it's a newer sensor than everything else, it's probably made some advancements despite those sacrifices, so I think a safe bet would be to assume it performs in the ballpark of everything else in the lineup.

I'll be happily surprised if it performs significantly better than anything we have from RED currently, but I think what everyone is most curious and excited about is the form factor and the types of shots they can now get within the RED ecosystem.

What I am curious about is how it fits into the inevitable DSMC3 or whatever it ends up being called.

Komodo being smaller makes me curious if DSMC3 will actually go larger. I actually think this would be a big benefit for them. More efficient cooling, more outputs, less modularity. Let Komodo be the sensor in a box you rig any which way you want. Let DSMC3 be the consumer-Ranger. IMO

 

[Jason Rivera]

My Two cents

My Two cents

Here is my prediction:

https://flic.kr/p/2gKMMZz

 

[Neil Abeynayake]

If it were just about numbers, the Alexa should have died 10 years ago. The Sony Venice will be ok, don’t worry.

Seems like you have not understood the post I made. A "disrupt in sales" means, projected sales quotas may not be reached. No matter how enterprising your thoughts might be, SONY keeps on changing workflow on every new generation of their camera bodies. That is seen as a deterrent in acquiring their cameras into existing pipelines by many folks. RED did not change anything radical to compromise their echo system yet. They simply added IPP2 into their DSMC2 camera lineup while maintaining the previous workflow. ARRI too had stayed with their workflow all along and added functionality as and when needed.

I don't know how much interaction you have with SONY professional group but we do work with them. According to one source at SONY, a certain Cine Alta camera ONLY sold 200 copies and they scrapped the program. Also, have you worked with the SONY Venice camera on a commercial or a feature project? We have and that workflow is NOT robust as you think. When you use a properly calibrated 4k or better reference monitor, you'd see image anomalies such as muddy picture details and crushed blacks when image elements are pushed to usable ranges.

So, before singing songs of praise around SONY, I recommend you to use them in a rigorous testing sequence.

 

[Evin Grant]

All I have to ask is that Komodo has a Sony Alpha mount option, I'm really lovin' my Batis and G-Master glass these days. Oh and AI-AF would be super appreciated as well.

 

[Brandon Veen]

That would kill the Monstro- I could be wrong but I doubt it very much...

I agree, if it were 8K, 6K, or 5K.
But, if it were just 4K (4096x2730 or more likely 4278 x 2852), now that represents something that's been absent since RAVEN, and could exist healthily in the overall RED lineup.
There's actually a lot of precedent for why this is the most likely option, and really, we have to look through everything we've been told to try and prove why.

- It is not DSMC3
- It is not a replacement for DSMC2

These two lines are fairly important for how it comes to speculate about this thing, and important to keep in mind as we go through the other points.
Fair warning, I'm gonna take this as an opportunity to go fully speculative and not all of it is going to be on-topic for why it's likely a 36x24mm sensor.
And my points are being put out there open for anyone to jump in and debate.
With that said, let's start with the sensor.

- It is not 8k VV
- It is not a Dragon or a Helium or a Gemini or a Monstro Sensor

I know quite a few of us here remember RED's trouble with making Dragon VV cuts before Monstro was announced.
With that in mind, it's easy to see why RED would be avoiding something like that with KOMODO.
Additionally, RED historically hasn't gone the ARRI route of stitching two or three cuts together, and as that's off the table for KOMODO, we're left with two options:
1. A brand new sensor design made by their suppliers (same as the rest of the RED sensors).
2. Taking an 'existing' sensor and provide a unique OLPF and software processing to get it to feel like a RED.

Option 1 might seem like it's more likely knowing RED's history, AFAIK there aren't any other Cameras with cuts of RED sensors (excluding DXL), but once we factor the price point in mind, there's a strong case for Option 2.
And there's some precedent with Fujifilm, Sony, and Panavision:
- Sony provided Hasselblad, Fujifilm, Phase One, Pentax, the same medium format sensor, but they all had their differences that made them unique (lenses aside).
- Fujifilm threw their signature film simulations onto their medium format bodies to make it something different from their competitors.
- RED provided Panavision with Dragon VV sensors and Monstro Sensors for the DXL1/2.
- Panavision created their own OLPF and threw in LightIron color into the mix to make it different from RED.
- Sony created the 102 megapixel sensor for Fujifilm's GFX 100S with a pixel pitch of 3.76 microns.
- After the GFX 100S is released, Sony announces the A7R IV with a 61 megapixel sensor with a pixel pitch of 3.76 microns.
- The A7S series historically has been a 4278 x 2852 (12.2 Megapixel) 36x24 sensor.
- Sony provides sensors to other (competing) companies.
- Speculation 1: The A7R IV and Fujifilm GFX 100S share different cuts of the same sensor.
- Speculation 2: RED is working with Sony to create a 36x24 sensor, that may or may not also end up in an A7S III, which either will have a standard 4K sensor or go the raven route of 4.5K.
To set it apart, RED is going to follow Panavision's lead and have a special OLPF, like Fujifilm will do some software processing to get it to look the way they want it to, and since it's only 4K/4.5K, it's going to support normal R3D.

Jim's statement on H4VUser supports this: " We have completed the new and extraordinary sensor and redesigned the package. Its capability will vastly exceed the originally planned module. While it does not replace its big RED brothers, it will certainly be a complimentary camera for cinema grade images at the highest level at lower pricing."
The original module was planned to have R3D-lite. As Jim says the capability "vastly exceeds" what the planned module would do, it's easy to speculate that this thing will support normal R3D at 4K with a 36x24 sensor that's co-designed with Sony and given RED flavor.

It's likely part of the reason why interchangable OLPF's appear to be absent from KOMODO. The regular Standard/Skin Tone Highlight/Low Light ones just wouldn't have worked with the new sensor, and with a pixel pitch around 8 microns, they've felt a single OLPF is all the camera needs as it already will perform great in Low Light. This, I believe, is what we're looking at, and to me makes the most sense as to not cannibalize any sales of any of the DSMC2 Cameras. It's complementary. Not a replacement. If you want to shoot 5K and higher, you have to go to the realm of DSMC2/3.

- It does not use proprietary media

As we know, it uses CFast cards for Media and as Phil and Sareesh said on Pages 6 and 7, a lot of those top out at 130 MB/sec and 150 MB/sec.
Using RED's recording time calculator, the RED Raven at 4.5K full format, at 24 FPS at 3:1 compression has an average data rate of 122 MB/sec.
Gemini, at the same settings, has an average data rate of 135 MB/sec.
And, most importantly, at 4.5K 3:2, 24 fps, 3:1 Gemini averages at 151 MB/sec (4K 3:2 is 119 MB/sec).
There's no reason for KOMODO to use MINI-MAGs because it functionally doesn't need a faster data rate and CFAST allows for a smaller body.

PART 2 CONTINUED BELOW.