RED Komodo to start limited shipping soon for $6k

Mitch, totally agree.

What is with Panasonic leaving their utilitarian design roots, and putting bright red accents on everything these days? The DVX200 and the EVA1 are attention getting, but kind of in the bad way. Rivaling BM and Red for ugly camera design.
 
Last edited:
I suppose we can debate the answer - is it the people who hire you? Or the people who collectively hire them (the audience at home)?

I care more about the audience at home for the most part, but there are a lot of wannabe producers who insist on particular cameras.

But I also run into a lot of people who want Red because they think that's what determines the quality of their end result... most of those skip the preproduction phase, dive right into shooting, don't make any sort of plan for lighting or art direction, and rarely listen to anyone who tries to give them constructive advice.

In any case, the dominance of ARRI at the high-end of narrative production suggests that RED's branding is not affecting that sector.

The lifestyle branding probably has a lot more impact in the owner/operator market than in the high end, so it makes sense that the crazy colors only available in the custom order phase before Red ramps up its factory to manufacture Komodos for real. Those will all be the usual unobtrusive black. :)

That said, Red cameras are pretty much everywhere. Even on Arri shows, they often shoot VFX shots with Reds. I think it says a lot about how good a job Arri did with the Alexa that in spite of Sony, Red, and Panasonic having more advanced sensor technology than Arri, Arri is still #1 and Red is #2 in high end film production, Sony 3rd. It's in the non-Hollywood arena like Netflix partner productions that the competition is hotter.
 
I don't care what fancy decals the Komodo has, who has a camera with a 20 second boot time in this day and age?
 
I don't care what fancy decals the Komodo has, who has a camera with a 20 second boot time in this day and age?

I’ve read that it’s less and will continue to improve, but I don’t get that, either. I know guys with Gemini’s and it’s painful. Even my F55 boots in what’s the blink of an eye, in comparison(and I feel it’s slow). It seems like if they’re capable of creating cameras that can shoot at such high resolutions and frame rates and RAW internally, then a quick(at least reasonable) boot time should be do-able. Five seconds or less. But I’d rather see three or less in 2020.
 
But the slow boot time makes you THINK more about what you're shooting and be more intentional with your--- just kidding, yeah it sucks
 
I don't think slow boot time is that much of a pain because of the way I work, which is to just leave my camera on for most of the duration of a shoot. So often I see camera ops turning their camera off for a five minute break to save a minuscule five minutes on their FOUR HOUR battery which they have more than enough of to last the day. Then suddenly, "Time to roll!" "Hold on, I've got to wait 15 seconds for my camera to boot because I had wanted to save five minutes of battery on a FOUR HOUR battery." Silly to me. Who cares about saving five minutes of battery. It's not like I'm trying to keep the hour clock on the camera down for resale value, and most batteries I've used outlive the life of the camera. Starting with the fast paced wedding world, too many times I've missed moments waiting for my camera to boot up.

Even if the camera boots immediately like awaking an iPhone from sleeping, just the extra 1-2 seconds to get your fingers to turn the camera on is lost time in getting the shot. Of course if I'm doing a meal break or something more substantial I'll turn my camera off. I use 280w batteries on my Amira which last around four hours, and the BP-60 on the C300 Mark II lasts around four hours as well. Owning plenty of these to get through a full day shoot, which means if I leave the camera on ALL day for a 12 hour shoot, I only have to change batteries twice, and have plenty of warning so can choose to change them at opportune times, so there's really very little reason to be constantly turning my camera on and off throughout the day.
 
All else being equal, I'd rather have a shorter boot time. And even with a 4-hour battery, you have to change it some time. No one wants to be timing out their battery change for the least disruptive moment during continuous shooting. (Although at least Komodo has a dual-battery hot-swappable plate.)
 
I don't care what fancy decals the Komodo has, who has a camera with a 20 second boot time in this day and age?

On the DSMC2, they have a pre-record, so, as long as one has the batteries, the camera can be pre-recording the whole time. Sort of like going back in time.

So, just depends on the working style.


When i shot DSLR, i liked how quickly those cameras turned, but that mirror slap killed the moment in non-fiction, because the people around knew that i was about to record and they all changed their behavior once they heard the sound.

Would be nice to have instant on cameras though. I definitely hate the boot times on a Red with the 2TB mag. But it isn’t a deal breaker until a better and cheaper camera exists. At this time Arri has similar boot times. Sony Venice is quicker, but it costs $60K for a complete package. A Red is like $30K depending on which model and kit you get. Sometimes a lot less if one goes Dragon, and now Komodo.

The C500ii is an interesting camera, and overall i really like it, but i didn’t get the same image quality as the Gemini. They were close, but the Gemini had half to a whole stop better DR, colours seemed better through the range, and the out of camera look felt more “movie-esque” straight from camera. The Canon looked great, but had a slightly more video edge to it. Either camera is great, and the fast boot time on the Canon is a big plus. However, the Red’s having the pre-record makes up for the slow boot times in some situations. And i suppose Arri is just above reprove.
 
Last edited:
I imagine a future where a camera is always recording when it is on, as, "I forgot to hit record" is a big mistake that can and does happen to anyone and everyone at some point, so this mode would make sure that would not happen. So, there would still be a record button, but it would be like meta data where that footage is set aside when the official recording is done would be duly noted as such, and then the footage when not officially recording would be a backup of sorts that you'd dig into another folder for if you ever need to use it.

This would not be practical until the cost of media storage is down enough that the additional media costs of using this mode would be negligible (which could be a long time from now, though for cameras shooting more compressed codecs perhaps not so much), and there would also be an option to turn off the constant recording when wanting it off.
 
Last edited:
In 2K prores LT that is doable. I think it is just over 100Mbps.
 
All else being equal, I'd rather have a shorter boot time. And even with a 4-hour battery, you have to change it some time. No one wants to be timing out their battery change for the least disruptive moment during continuous shooting. (Although at least Komodo has a dual-battery hot-swappable plate.)

Advantage of the BP batteries on Komodo is you can hot swap one and keep rolling on the other.

Bolex boots in about 20 seconds but can take several minutes to reach operating temperature. There is a slight difference in image brightness between the left and right sides of the sensor before it reaches thermal equlibrium. Between the internal battery and a beefy Switek external one I have about 8 hours total on time, so I never turn it off. The Switek trickle charges the internal battery when connected, so I only use it on the tripod. Internal battery is good for 3+ hours, twice the recording time for the 512GB SSD.

How long does it take your phone or your laptop to boot? Cameras are essentially micro super computers these days.
 
Last edited:
I imagine a future where a camera is always recording when it is on, as, "I forgot to hit record" is a big mistake that can and does happen to anyone and everyone at some point, so this mode would make sure that would not happen. So, there would still be a record button, but it would be like meta data where that footage is set aside when the official recording is done would be duly noted as such, and then the footage when not officially recording would be a backup of sorts that you'd dig into another folder for if you ever need to use it.

This would not be practical until the cost of media storage is down enough that the additional media costs of using this mode would be negligible (which could be a long time from now, though for cameras shooting more compressed codecs perhaps not so much), and there would also be an option to turn off the constant recording when wanting it off.

My 2700 VariCam has essentially that function. It can be set to continuously record until the media is full(LOOP RECORD) then it will just begin to write over the oldest material. And shooting 24P on 64GB cards with five card slots that's somewhere around 10-12-15 hours of uninterrupted recording(can't remember the exact run time per card in 24, but somewhere north of 2 to 2 1/2 hours or more per card). So, for an average production day of 8-10 hours(sans any slow-mo), you could just load up five cards and roll from the first take till you wrapped. Just hope that timecode was jammed and the script girl had great notes and ID'd the good takes.
 
My 2700 VariCam has essentially that function. It can be set to continuously record until the media is full(LOOP RECORD) then it will just begin to write over the oldest material. And shooting 24P on 64GB cards with five card slots that's somewhere around 10-12-15 hours of uninterrupted recording(can't remember the exact run time per card in 24, but somewhere north of 2 to 2 1/2 hours or more per card). So, for an average production day of 8-10 hours(sans any slow-mo), you could just load up five cards and roll from the first take till you wrapped. Just hope that timecode was jammed and the script girl had great notes and ID'd the good takes.

That would be cool if the REC button was more of a flag or marker. Then at the end of the take, the scripty could wirelessly add the "good" take metadata directly to the last take on the camera. Then DIT would be even more important, as they would be exporting only the actual takes, but possibly repairing anything missed.
 
My 2700 VariCam has essentially that function. It can be set to continuously record until the media is full(LOOP RECORD) then it will just begin to write over the oldest material. And shooting 24P on 64GB cards with five card slots that's somewhere around 10-12-15 hours of uninterrupted recording(can't remember the exact run time per card in 24, but somewhere north of 2 to 2 1/2 hours or more per card). So, for an average production day of 8-10 hours(sans any slow-mo), you could just load up five cards and roll from the first take till you wrapped. Just hope that timecode was jammed and the script girl had great notes and ID'd the good takes.

2700 was a great camera, I used to rent it and shoot with it quite often although I never had casue to use that function. It's funny, I think it was a better camera than the 3700 in a lot of ways and it was a lot cheaper.
 
I imagine a future where a camera is always recording when it is on, as, "I forgot to hit record" is a big mistake that can and does happen to anyone and everyone at some point, so this mode would make sure that would not happen. So, there would still be a record button, but it would be like meta data where that footage is set aside when the official recording is done would be duly noted as such, and then the footage when not officially recording would be a backup of sorts that you'd dig into another folder for if you ever need to use it.

This would not be practical until the cost of media storage is down enough that the additional media costs of using this mode would be negligible (which could be a long time from now, though for cameras shooting more compressed codecs perhaps not so much), and there would also be an option to turn off the constant recording when wanting it off.

The problem with continuous recording isn't power or media, it's the ingest and management of all that crap footage just to get to the content of any value. Reality shows already suffer from ungodly amounts of content pouring into editorial.
 
AI will eventually take care of the ingesting and management (and editing).

For reality, in particular, it has to be so much right now because manual labor is still involved in creating stories. Producers don't always know in which direction(s) they want to go.

In the future, machines will create multiple storylines and humans will pick which ones they like best (and maybe clean up the final product a little).

Physical media will probably be mostly wiped out too (and only used for the most demanding data).
 
AI will eventually take care of the ingesting and management (and editing).

Maybe this is true, but I'm not so sure. The first gig I had was assisting at Break Thru Films documentary production company in Manhattan while they were editing Joan Rivers: A Piece of Work, which won the Sundance award for best doc editing.

The editor watched every second of footage they ever shot in its entirety. She just took notes during that first pass and made no edits or selects. Grasping all the available material was an integral part of her process.

You could argue that there's a value proposition or a fast-turnaround advantage to outsourcing creative decisions to AI. Maybe that's true, but it's a creatively and socially complex craft. If you want to maximize quality, you'll put a human to the task.
 
I mean...that's of course the right way to do it, the best way...with your heart and mind.

But don't be naive. lol

Trust that the AI will be just as good once it learns the purpose. It could provide you so many different edits, and so much faster.

And people would never know and may have liked all of them if they saw them (because in reality there are even multiple cuts made by humans but one gets chosen, or there is an "Alternate Ending" or a "Director's Cut").
 
I mean...that's of course the right way to do it, the best way...with your heart and mind.

But don't be naive. lol

Trust that the AI will be just as good once it learns the purpose. It could provide you so many different edits, and so much faster.

Pardon me for being so naive.

But it's not clear to me that AI will develop the social skills necessary for creating a complex narrative from intricate social interaction.

Will they be able to interpret shade and sarcasm? Maybe. But even in the world of Star Trek where there are general intelligence androids, they struggle with emotional interpretation and expression.

I read that "If the human brain were a computer, it could perform 38 thousand trillion operations per second."

Sure, an actual computer has perfect memory and near-instantaneous recall with mostly accurate logic circuits or whatever. But humans are highly evolved for complex social interactions. Matching them is no small feat.

Or consider the fact that Moore's Law may be coming to an end and computer speed may not be accelerating at the same rate as it was. There may be a limit to what can be achieved with classical computers.

Yes, quantum computing is lurking in the wings. But it may not be useful for every type of problem.

There are things that computers are way better at doing than humans. But may not the opposite be true?

(That being said, one of my favorite sci-fi stories in one of my favorite volumes is the very funny "Trurl's Electronic Bard" in The Cyberiad by Stanislaw Lem. It's about a computer AI trained to make poetry who is exiled by poets after it puts them out of business. They try to destroy it, but it wins them over and changes their hearts with eloquent verses as they approach it.)
 
IMO, you're overthinking it.

We're just talking about putting clips together. And machines will be able to eventually learn EVERYTHING and they will provide you dozens of edits for your human reviewal. Maybe even hundreds of cuts if you choose to have that many created. All in hours or even minutes.

Humans are socially awkward too. Not everything is smooth. You will adjust AI's work in post. And teach the machine about feelings.

Will it ever create something exactly how you're thinking it? No, of course not because that is impossible unless the machine is reading your brain.

Do you know how many human editors conflict in the bay? Director? Producers? Executives? That is like human AI!

___

Nonetheless, humans make a lot of stuff that is constantly criticized and disliked anyway, so at some point does it really matter who is making it? It's a hit or miss.

More complicated editing like narrative will slowly develop while AI commercial work will be adapted instantly, overnight. Like, seriously...you can now randomly throw some clips on a timeline from a product video and it could look like a commercial.

I think you're underestimating this simple computer process by overestimating the particular task.

As a guy who loves Star Trek, you're really against robots and change. lol
 
Back
Top