jcs

OK let's play the ego game. I have a degree in Cognitive Science from UCSD with a specialization in artificial intelligence (and have written neural networks from scratch, worked on projects with GPU neural networks as well as machine vision). I've written complex real-time simulators including flight simulators using fluid dynamics to model airflow, custom energy-accurate integrators, and 100% custom from scratch impulse-based rigid-body physics. For the driving simulator I modeled the ground using cubic polynomials with a Newton gradient solver to intersect rays to compute tire-surface interaction with C1 continuity, allowing for very high speed accurate collision modeling (including physical bump mapping to simulate roughness when needed). The engine sound simulation was modeled on my Z06 Corvette (recorded from actual engine then modeled in real-time). The tire simulation is completely custom (a very hard problem to model realistically due to the nonlinear behavior of the elastic rotating tire), the network protocol is a custom UDP design which incorporates TCP design elements (reliable) with UDP elements (non-retransmitted elements such as position) for optimal network bandwidth utilization (used in the first XBox Live! game). Additionally, the physics simulation hid visible lag well over 500ms, including collisions. AI is used to drive the cars around the track and avoid collisions (the flight simulator also used AI to fly the aircraft, which could also perform post-stall maneuvers, long before real aircraft could do this in real life. The model uses a generalized rigid body moving through a fluid, which wasn't replicated by other developers until many years later). I wrote the first stereoscopic head-tracking multiplayer games for the PC (virtual reality), including 3D sound modeling, and showed John Carmack (now CTO of Oculus) how to do proper stereoscopic 3D when my company ported Quake to stereo3D for H3D Entertainment.
I'm an expert in real-time simulations and accurate modeling of real-world systems. I'm very familiar with the scientific method, as well as how to conduct single and double blind studies for psychology.
When I proved that the full frame look is a myth, people still argued and didn't believe it (some still don't believe it). I did the math, did the experiment, and showed the results for others to replicate. http://brightland.com/w/the-full-frame-look-is-a-myth-heres-how-to-prove-it-for-yourself/, 

 

There's another thread where someone created excellent 3D renderings, taking the lens out of the equation, and some folks still argued. So the point is even with the scientific method, math, real-world examples and simulations, people still argue because of their ego, just like you are doing right now. You are using ad hominem, which is an instant fail in debate. You've got to focus on the topic at hand, instead of attacking the other party, otherwise it becomes clear you have no valid argument in the debate.
From my background in mathematics, physical simulators, networked simulations, and artificial intelligence, I can see patterns in systems such as quantum physics when combined with concepts from psychology (as well as life experience) that make a good case for concepts like "thoughts make things", that working together and not fighting is more efficient in terms of energy, and am confident that a simulator which replicates large systems such as what is going on in the world today would coincide with economist Mark Blyth's analysis on the effects of Predatory Capitalism. The quantum experiments show that the universe reacts to perceivers in irrational ways. Einstein thought quantum entanglement was bunk, calling it 'spooky action at a distance'. He was wrong. These simple concepts should give one a powerful idea as to what the universe really is on a large scale. Due to the quantized nature of quantum physics (hence the name), some theorists believe the universe is a simulation running in a computer (and are creating tests to check the theory). Hinduism, Buddhism, Zen believe the universe and everything in it is God (which makes all of us elements of God). Alan Watts does a good job describing it (YouTube). And that there's a 'peek-a-boo' with our consciousnesses and the universe/God which correlates well with what we are learning about quantum physics.
Science can't explain everything (not even close), it's just a tool, and said tool has limitations. The human filter on reality also hides the true nature of reality, and the only way you can see this is through meditation and/or through compounds such as DMT (which shuts off the human filter 'software', especially 5MeO-DMT).
This post is intended to open your mind to possibilities beyond close-minded thinking. If you disagree and choose not to look deeper into what reality really is, that's cool. Many people enjoy living in illusion, The Matrix was a good example of this concept.

this is actually true ^^
im not smart enough to understand those videos, but the philosophical implications of quantum mechanics absolutely fascinates me

OK let's play the ego game. I have a degree in Cognitive Science from UCSD with a specialization in artificial intelligence (and have written neural networks from scratch, worked on projects with GPU neural networks as well as machine vision). I've written complex real-time simulators including flight simulators using fluid dynamics to model airflow, custom energy-accurate integrators, and 100% custom from scratch impulse-based rigid-body physics. For the driving simulator I modeled the ground using cubic polynomials with a Newton gradient solver to intersect rays to compute tire-surface interaction with C1 continuity, allowing for very high speed accurate collision modeling (including physical bump mapping to simulate roughness when needed). The engine sound simulation was modeled on my Z06 Corvette (recorded from actual engine then modeled in real-time). The tire simulation is completely custom (a very hard problem to model realistically due to the nonlinear behavior of the elastic rotating tire), the network protocol is a custom UDP design which incorporates TCP design elements (reliable) with UDP elements (non-retransmitted elements such as position) for optimal network bandwidth utilization (used in the first XBox Live! game). Additionally, the physics simulation hid visible lag well over 500ms, including collisions. AI is used to drive the cars around the track and avoid collisions (the flight simulator also used AI to fly the aircraft, which could also perform post-stall maneuvers, long before real aircraft could do this in real life. The model uses a generalized rigid body moving through a fluid, which wasn't replicated by other developers until many years later). I wrote the first stereoscopic head-tracking multiplayer games for the PC (virtual reality), including 3D sound modeling, and showed John Carmack (now CTO of Oculus) how to do proper stereoscopic 3D when my company ported Quake to stereo3D for H3D Entertainment.
I'm an expert in real-time simulations and accurate modeling of real-world systems. I'm very familiar with the scientific method, as well as how to conduct single and double blind studies for psychology.
When I proved that the full frame look is a myth, people still argued and didn't believe it (some still don't believe it). I did the math, did the experiment, and showed the results for others to replicate. http://brightland.com/w/the-full-frame-look-is-a-myth-heres-how-to-prove-it-for-yourself/, 

 

There's another thread where someone created excellent 3D renderings, taking the lens out of the equation, and some folks still argued. So the point is even with the scientific method, math, real-world examples and simulations, people still argue because of their ego, just like you are doing right now. You are using ad hominem, which is an instant fail in debate. You've got to focus on the topic at hand, instead of attacking the other party, otherwise it becomes clear you have no valid argument in the debate.
From my background in mathematics, physical simulators, networked simulations, and artificial intelligence, I can see patterns in systems such as quantum physics when combined with concepts from psychology (as well as life experience) that make a good case for concepts like "thoughts make things", that working together and not fighting is more efficient in terms of energy, and am confident that a simulator which replicates large systems such as what is going on in the world today would coincide with economist Mark Blyth's analysis on the effects of Predatory Capitalism. The quantum experiments show that the universe reacts to perceivers in irrational ways. Einstein thought quantum entanglement was bunk, calling it 'spooky action at a distance'. He was wrong. These simple concepts should give one a powerful idea as to what the universe really is on a large scale. Due to the quantized nature of quantum physics (hence the name), some theorists believe the universe is a simulation running in a computer (and are creating tests to check the theory). Hinduism, Buddhism, Zen believe the universe and everything in it is God (which makes all of us elements of God). Alan Watts does a good job describing it (YouTube). And that there's a 'peek-a-boo' with our consciousnesses and the universe/God which correlates well with what we are learning about quantum physics.
Science can't explain everything (not even close), it's just a tool, and said tool has limitations. The human filter on reality also hides the true nature of reality, and the only way you can see this is through meditation and/or through compounds such as DMT (which shuts off the human filter 'software', especially 5MeO-DMT).
This post is intended to open your mind to possibilities beyond close-minded thinking. If you disagree and choose not to look deeper into what reality really is, that's cool. Many people enjoy living in illusion, The Matrix was a good example of this concept.

OK let's play the ego game. I have a degree in Cognitive Science from UCSD with a specialization in artificial intelligence (and have written neural networks from scratch, worked on projects with GPU neural networks as well as machine vision). I've written complex real-time simulators including flight simulators using fluid dynamics to model airflow, custom energy-accurate integrators, and 100% custom from scratch impulse-based rigid-body physics. For the driving simulator I modeled the ground using cubic polynomials with a Newton gradient solver to intersect rays to compute tire-surface interaction with C1 continuity, allowing for very high speed accurate collision modeling (including physical bump mapping to simulate roughness when needed). The engine sound simulation was modeled on my Z06 Corvette (recorded from actual engine then modeled in real-time). The tire simulation is completely custom (a very hard problem to model realistically due to the nonlinear behavior of the elastic rotating tire), the network protocol is a custom UDP design which incorporates TCP design elements (reliable) with UDP elements (non-retransmitted elements such as position) for optimal network bandwidth utilization (used in the first XBox Live! game). Additionally, the physics simulation hid visible lag well over 500ms, including collisions. AI is used to drive the cars around the track and avoid collisions (the flight simulator also used AI to fly the aircraft, which could also perform post-stall maneuvers, long before real aircraft could do this in real life. The model uses a generalized rigid body moving through a fluid, which wasn't replicated by other developers until many years later). I wrote the first stereoscopic head-tracking multiplayer games for the PC (virtual reality), including 3D sound modeling, and showed John Carmack (now CTO of Oculus) how to do proper stereoscopic 3D when my company ported Quake to stereo3D for H3D Entertainment.
I'm an expert in real-time simulations and accurate modeling of real-world systems. I'm very familiar with the scientific method, as well as how to conduct single and double blind studies for psychology.
When I proved that the full frame look is a myth, people still argued and didn't believe it (some still don't believe it). I did the math, did the experiment, and showed the results for others to replicate. http://brightland.com/w/the-full-frame-look-is-a-myth-heres-how-to-prove-it-for-yourself/, 

 

There's another thread where someone created excellent 3D renderings, taking the lens out of the equation, and some folks still argued. So the point is even with the scientific method, math, real-world examples and simulations, people still argue because of their ego, just like you are doing right now. You are using ad hominem, which is an instant fail in debate. You've got to focus on the topic at hand, instead of attacking the other party, otherwise it becomes clear you have no valid argument in the debate.
From my background in mathematics, physical simulators, networked simulations, and artificial intelligence, I can see patterns in systems such as quantum physics when combined with concepts from psychology (as well as life experience) that make a good case for concepts like "thoughts make things", that working together and not fighting is more efficient in terms of energy, and am confident that a simulator which replicates large systems such as what is going on in the world today would coincide with economist Mark Blyth's analysis on the effects of Predatory Capitalism. The quantum experiments show that the universe reacts to perceivers in irrational ways. Einstein thought quantum entanglement was bunk, calling it 'spooky action at a distance'. He was wrong. These simple concepts should give one a powerful idea as to what the universe really is on a large scale. Due to the quantized nature of quantum physics (hence the name), some theorists believe the universe is a simulation running in a computer (and are creating tests to check the theory). Hinduism, Buddhism, Zen believe the universe and everything in it is God (which makes all of us elements of God). Alan Watts does a good job describing it (YouTube). And that there's a 'peek-a-boo' with our consciousnesses and the universe/God which correlates well with what we are learning about quantum physics.
Science can't explain everything (not even close), it's just a tool, and said tool has limitations. The human filter on reality also hides the true nature of reality, and the only way you can see this is through meditation and/or through compounds such as DMT (which shuts off the human filter 'software', especially 5MeO-DMT).
This post is intended to open your mind to possibilities beyond close-minded thinking. If you disagree and choose not to look deeper into what reality really is, that's cool. Many people enjoy living in illusion, The Matrix was a good example of this concept.

Hey Shawn, the 1DX II 4K is soft. It's ok, it looks great! Even the GH4 4K has more detail, however the 1DX II has much better color control and the image looks more organic. To get full detail 4K we need an 8K sensor and good downsampling to prevent aliasing (Nyquist)...

The Universe is a kind of reactive computer. What you think it will create, you being part of it. 
 

Hey Shawn, the 1DX II 4K is soft. It's ok, it looks great! Even the GH4 4K has more detail, however the 1DX II has much better color control and the image looks more organic. To get full detail 4K we need an 8K sensor and good downsampling to prevent aliasing (Nyquist)...

How do we heal ourselves and the world?
 

Normally I'd agree regarding external USB drives, however USB 3 is really fast, and the WD Ultra 2TB drives are rated at 110MB/s read (and get about that for real: https://www.youtube.com/watch?v=z8p03IjCclA ). 4K C300 II files are around 440Mbps or ~55MB/s which gives us around 2x overhead which matches my experience as well. 1DX II 4K60p is ~800Mbps or 100MB/s and those work as well too. Transitions might cause a burst slowdown, however since reading is faster than the file bitrate the system can catch up in most cases (my real world experience). For something more complex, sure, an SSD or HD RAID 0 is needed. Our 4K complexity comes from green screen, compositing, visual effects, large images, etc. (large images can also choke/expose PP CC bugs). I would expect the 3.5GB/s (3.5 GIGABYTE per sec!) NVMe PCIe drive to blow away just about any multidrive solution except perhaps 2 NVMe drives in RAID 0 ;). There is no seek overhead, it's RAM and PCIe is faster than SATA (latency), and it's going to be faster than multiple spinning disks with seek latency (when total bandwidth is the same or more). Loading apps in a blink is pretty cool (and does save time- it adds up), and with 2T there's enough room for editing a full project 100% on the drive (our episodes are less than 30 minutes finished). Once released the project can be moved to a 3.5 HD archive (4 drive dockers are handy- can fill drives and put in a case then take offline and store). 4 Samsung 850 Pro would be ~2.2GB/s  RAID 0, still getting smoked by the single 960 Pro at 3.5GB/s (except in total storage).
 
It depends on the 4K material: GH4 tends to be easier, A7S2 harder, and C300 II and 1DX II 60p the hardest. Perhaps also consider trying more RAM (e.g. 64GB) before getting a new machine. WD Black is plenty fast unless you're compositing a bunch of 4K videos in the same frame (transitions will be fine).
 

Our 2010 2.93GHz 12-core MacPro with GTX 980ti couldn't edit complex 4K footage smoothly in Premiere Pro CC (in OSX El Capitan or Win10x64Pro). FCP X for the most part can handle medium complexity 4K (chroma key, multiple 4K streams at once, very high resolution stills (Ken Burns etc.). However, I'm quite a bit faster in PP CC, and I need to research plugins to replicate what I can do with Lumetri (if such a plugin exists). It was time to upgrade the video editing machine, and it seemed like building a Hackintosh would create a far faster machine than anything Apple currently has to offer.
The goal was to balance high performance with low noise on air cooling only (tricky!). Here's the machine:
Corsair Carbide Series 330R Blackout Edition Ultra-Silent Mid-Tower Case ASUS X99-DELUXE II LGA 2011-v3 ATX Motherboard (ASUS has fancier boards however this appears to be just as fast) Intel Core i7-6950X 3.0 GHz Ten-Core LGA 2011-v3 Extreme Edition Processor (have been buying or building dual CPU Xeons for over 10 years. The single chip high clock 10 core looked to be the best fit for video editing vs. lower clock more core Xeons) Corsair Vengeance LPX 64GB (4x16GB) DDR4 3466 (PC4-27700) C16 with Airflow, Black CMK64GX4M4B3466C16 Noctua Dual Tower CPU Cooler for Intel LGA 2011-0/LGA 2011-3 Square ILM/1156/1155/1150 and AMD AM2/AM2+/AM3/3+,FM1/2 NH-D9L Noctua AAO Frame Design, SSO2 Bearing Premium Quality Quite Fan NF-A9 PWM (Extra fan for CPU cooler) ASUS GeForce GTX 1080 8GB ROG STRIX Graphics Card (STRIX-GTX1080-A8G-GAMING) (purchased over eVGA (normally favorite) due to heat issues with ACX 3 and the VRM chip) Corsair ML120 Pro, 120mm Premium Magnetic Levitation Cooling Fan CO-9050040-WW (to replace rear case fan) Corsair ML140, 140mm Premium Magnetic Levitation Fan (2-Pack) (to replace front fan and add additional fan to cool hard drives) Samsung 2TB 960 PRO M.2 Internal SSD (fastest drive available (~3.5GB/s), for OS, Apps, temp projects) WD 4TB Black 7200 rpm SATA III 3.5" Internal HDD (long term storage for video projects) EVGA SuperNOVA 1000 P2 1000W Power Supply (higher quality PSU's increase component life and reliability) Microsoft Windows 10 Pro (64-bit, OEM DVD) (lower cost than download!) SanDisk 16GB Extreme USB 3.0 Flash Drive (downloaded Win10 from Microsoft and used key from DVD to create USB installer) I was able to get it stable at 3.96GHz (in turbo/boost) with core temps max around 65-69C (all fans running 100%, fan speed temp ramping tuned in bios). Memory was only stable at 2666 (so much for 3466 rating). It will run at 4-4.1GHz, however some cores temp spike to 80C, so is probably too much for this air cooling set up.
The latest NVidia drivers are still buggy with Premiere- the "black screen bug with Ultra key" is still present. I installed the latest CUDA development video driver figuring it might be more stable than the game-oriented drivers (black screen bug hasn't appeared in latest green screen project yet). 4K editing in software mode is pretty smooth (full res preview)! So if the GPU driver has severe bugs during a project I can switch to software mode and still work reasonably fast.
Windows 10's font rendering looks much crisper than El Capitan. Window rendering and overall GUI speed is also much faster- feels like you are working faster. Once Windows 10 checked out with Premiere Pro I dug a bit deeper into what it would take to install OSX (would test on a hard drive then replace with SSD). Here's the closest I found for this motherboard (uses apparently the same bios). Doable however a lot of time hunting down kext (kernel extensions) and trial and error. So for the time being the new machine is Windows only- I can access both machines through a USB switch for keyboard/mouse (monitors have dual inputs).
At idle this machine is nearly silent, much quieter than the 2010 MacPro. It's good to see that after 6 years there's been some progress in CPUs and memory (GPUs have had much more performance growth). The new NVMe PCI SSD drives are insanely fast, however 4K editing straight from a USB3 port and CFast 2.0 works fine (as does editing from a hard drive over SATA- not the bottleneck).
Hopefully this info will be helpful for anyone looking to build a fast and quiet 4K editing box.
 

Our 2010 2.93GHz 12-core MacPro with GTX 980ti couldn't edit complex 4K footage smoothly in Premiere Pro CC (in OSX El Capitan or Win10x64Pro). FCP X for the most part can handle medium complexity 4K (chroma key, multiple 4K streams at once, very high resolution stills (Ken Burns etc.). However, I'm quite a bit faster in PP CC, and I need to research plugins to replicate what I can do with Lumetri (if such a plugin exists). It was time to upgrade the video editing machine, and it seemed like building a Hackintosh would create a far faster machine than anything Apple currently has to offer.
The goal was to balance high performance with low noise on air cooling only (tricky!). Here's the machine:
Corsair Carbide Series 330R Blackout Edition Ultra-Silent Mid-Tower Case ASUS X99-DELUXE II LGA 2011-v3 ATX Motherboard (ASUS has fancier boards however this appears to be just as fast) Intel Core i7-6950X 3.0 GHz Ten-Core LGA 2011-v3 Extreme Edition Processor (have been buying or building dual CPU Xeons for over 10 years. The single chip high clock 10 core looked to be the best fit for video editing vs. lower clock more core Xeons) Corsair Vengeance LPX 64GB (4x16GB) DDR4 3466 (PC4-27700) C16 with Airflow, Black CMK64GX4M4B3466C16 Noctua Dual Tower CPU Cooler for Intel LGA 2011-0/LGA 2011-3 Square ILM/1156/1155/1150 and AMD AM2/AM2+/AM3/3+,FM1/2 NH-D9L Noctua AAO Frame Design, SSO2 Bearing Premium Quality Quite Fan NF-A9 PWM (Extra fan for CPU cooler) ASUS GeForce GTX 1080 8GB ROG STRIX Graphics Card (STRIX-GTX1080-A8G-GAMING) (purchased over eVGA (normally favorite) due to heat issues with ACX 3 and the VRM chip) Corsair ML120 Pro, 120mm Premium Magnetic Levitation Cooling Fan CO-9050040-WW (to replace rear case fan) Corsair ML140, 140mm Premium Magnetic Levitation Fan (2-Pack) (to replace front fan and add additional fan to cool hard drives) Samsung 2TB 960 PRO M.2 Internal SSD (fastest drive available (~3.5GB/s), for OS, Apps, temp projects) WD 4TB Black 7200 rpm SATA III 3.5" Internal HDD (long term storage for video projects) EVGA SuperNOVA 1000 P2 1000W Power Supply (higher quality PSU's increase component life and reliability) Microsoft Windows 10 Pro (64-bit, OEM DVD) (lower cost than download!) SanDisk 16GB Extreme USB 3.0 Flash Drive (downloaded Win10 from Microsoft and used key from DVD to create USB installer) I was able to get it stable at 3.96GHz (in turbo/boost) with core temps max around 65-69C (all fans running 100%, fan speed temp ramping tuned in bios). Memory was only stable at 2666 (so much for 3466 rating). It will run at 4-4.1GHz, however some cores temp spike to 80C, so is probably too much for this air cooling set up.
The latest NVidia drivers are still buggy with Premiere- the "black screen bug with Ultra key" is still present. I installed the latest CUDA development video driver figuring it might be more stable than the game-oriented drivers (black screen bug hasn't appeared in latest green screen project yet). 4K editing in software mode is pretty smooth (full res preview)! So if the GPU driver has severe bugs during a project I can switch to software mode and still work reasonably fast.
Windows 10's font rendering looks much crisper than El Capitan. Window rendering and overall GUI speed is also much faster- feels like you are working faster. Once Windows 10 checked out with Premiere Pro I dug a bit deeper into what it would take to install OSX (would test on a hard drive then replace with SSD). Here's the closest I found for this motherboard (uses apparently the same bios). Doable however a lot of time hunting down kext (kernel extensions) and trial and error. So for the time being the new machine is Windows only- I can access both machines through a USB switch for keyboard/mouse (monitors have dual inputs).
At idle this machine is nearly silent, much quieter than the 2010 MacPro. It's good to see that after 6 years there's been some progress in CPUs and memory (GPUs have had much more performance growth). The new NVMe PCI SSD drives are insanely fast, however 4K editing straight from a USB3 port and CFast 2.0 works fine (as does editing from a hard drive over SATA- not the bottleneck).
Hopefully this info will be helpful for anyone looking to build a fast and quiet 4K editing box.
 

Yep, Canon STM lenses are preferable for situations that mic is close to the camera. 
You can also hear the noise from one of PB AF tests (11m20s):
 

All of our Canon L lenses and the 50 1.4 work fine. However they all make varying amounts of noise (50 1.4 is loudest). Canon STM lenses are silent (limited selection: http://explore-lenses.usa.canon.com/stm-lens-stepper-motor.html ). When close mic-ing to the subject, typically not an issue, however if using a camera mounted mic, it can be an issue.

That is remarkable... how many hits do you get with these videos? Also, is she your wife?

Ages 3 and up.
 

For a variety of reasons, mathematical and physical, most 4K cameras don't really produce actual, clean, alias-free 4K of resolution. The C300 Mark I and C100 Mark I/II produce true 2K using 4K sensors. The C300 II does not produce true 4K (measured at 3300x1750) nor does the 1DX II (looks similar to C300 II 4K). To produce real, alias-free 4K we need to sample at 2x the resolution (per Nyquist) or 8K. Thus 8K cameras can finally give us full 4K resolution for playback. This looks amazing on a 4K monitor (skintones are nice too!): 
 

i love that video - hahah   great to see both personalities out there and the littles dogs too!  (For me a bigfoot is only the inner fear, never let it outgrow your love)
Keep on shinning guys !

Good stuff!

@tomekk asked for a 1DX II test to see if it suffers from a problem reported for the 1DC. 1DX II 4K24p recorded for 20 minutes in a 75F office with lens cap on- didn't get very hot. Ramped from ISO 200 to 12800, Lumetri exposure set to 2.5 stops to show noise pattern:
 

Hey Dan the dark edges on the hair are probably from the green screen keyer (Ultra in PP CC). Will look into that issue for future shoots. Thanks for pointing it out.
Thanks FilmMan. C300 II is set up with 'Alexa' settings, using an Alexa LUT in Lumetri, and various quick tweaks with shadows, highlights, and saturation. Sound- one channel is the Schoeps CMC6 MK41 and the other is the Audix SCX-1 HC, both straight into the camera (with limiter enabled). These mics sound very similar despite the difference in cost. While the limiters and AD converters in Sound Devices recorders are better, the C300 II does a pretty good job.

@tomekk asked for a 1DX II test to see if it suffers from a problem reported for the 1DC. 1DX II 4K24p recorded for 20 minutes in a 75F office with lens cap on- didn't get very hot. Ramped from ISO 200 to 12800, Lumetri exposure set to 2.5 stops to show noise pattern:
 

Thx jcs, I appreciate it.

@tomekk asked for a 1DX II test to see if it suffers from a problem reported for the 1DC. 1DX II 4K24p recorded for 20 minutes in a 75F office with lens cap on- didn't get very hot. Ramped from ISO 200 to 12800, Lumetri exposure set to 2.5 stops to show noise pattern:
 

Jcs, the video looks good! The C300 II is a high quality beast.   Skin tones look really good.  Sound is really good.  Well done.  By the way,  it is an important topic which affects many people.  Cheers.

In case you were wondering...