Additional Cooling Test - BMCC Results

Hello everyone,

Just thought I'd share some results I had while experimenting with additional cooling.



Just grabbed a simple peltier I had laying around, and decided to see if I could get the body to cool down enough to help the internal cooling perform better.

The body was cooled for around 20 minutes, before recording anything.

I did my best to keep everything constant in between the cooled and uncooled stages. There some slight inconsistencies in the framing, etc. but as far as exposure goes, they were both exposed under the same amount of light with the same settings.

UPDATE:
Link to DNG's - http://we.tl/O318BCT9ys

UnCooled - 1600 ASA - EV + 4 ( ACR ) 100% crop

Cooled - 1600 ASA - EV + 4 ( ACR ) 100% crop



So yes, its not very scientific but just goes to show how the image can clean up with proper cooling.
(The URSA's design starts to make perfect sense...)

I only have a BMCC, but I assume this would apply across the board with the other cameras. ( might with even help FPN problems? )

I'll have to go back an check but I believe its a 30W TEC, so its not even that powerful. ( might have to try with 60-100W TEC. )

Not very practical in any sense, but it would be "cool" to see an aftermarket solution.

Thats very good. My bmcc images is more noisy in warm weather and direct sunlight. I think i will consider
using icepacks, as some have suggested in the bmpc fpn debate. Cooling it with external fans seems a little
unpractical.

This is why I'm saying that BM should offer us a trade in for our BM4k's for the URSA..

With proper cooling I'm sure the FPN would be at a minimum..

Do you know PhaseCore cooling packs?

Google it, PhaseCore Elements, First Line Technology.

Interesting test, but the focus is different between the two.. maybe this is affecting your test results?

Excellent, well; noise must be important, but it opens a path ..

Were you pushing air toward the camera or pulling air away?

Interesting test, but the focus is different between the two.. maybe this is affecting your test results?

I don't believe it should make a difference. It happens with the sensor itself, I could have conducted it with body cap on and gotten comparable results.

Were you pushing air toward the camera or pulling air away?

The heat sink attached to the Peltier at the top has a fan pulling air away.

Also have a 120mm pc fan on the side blowing air into the ssd bay with the door open.

As stated it's a very crude test. Hoping to develop something a bit more formal, and potentially something functional that could actually be used in field.

Power is the big issue.

I'd suggest going much further with the test before we can see if this can be an idea. Like put the camera in 30C sunlight and let it shoot raw for 10 minute. Then do the same test with the camera cool to a reasonable temperature.

If it is a cooling issue I would think the results would be pretty obvious.

I wonder if the cooling fan connection has extra watts at it's disposal. So potentially a more powerful fan can be used in the current slot...for those users who don't mind a little extra noise.

I think in your setup, the fan blowing cool air into the SSD compartment is doing more to cool the sensor than your peltier. Moving around the sensor and SSD would have an immediate effect. However, I don't think its a good idea to blow air "into" that compartment. All sorts of contaminants could harm the internals of the camera. A second exhaust fan, that could clip to the SSD door mechanism, could be a good solution though!

Kickstarter campaing anyone?

EDIT: btw, icepacks are a bad idea (IMO) you run the severe risk of introducing condensation inside the camera body. Good way to kill your camera.

I thought that too. Maybe we could install a beefier fan. I personally don't mind the noise. Has anyone tried it?


Sent from my iPhone using Tapatalk

Cooling ANY camera will improve it's sensor performance.

http://www.jtwastronomy.com/products/ultimate.html

jb

Cooling ANY camera will improve it's sensor performance.

http://www.jtwastronomy.com/products/ultimate.html

jb

Very interesting, JB. Compelling image differences. A wonder what the steady state temperature goal for the Ursa sensor will be? Certainly not -15 C!

Very interesting, JB. Compelling image differences. A wonder what the steady state temperature goal for the Ursa sensor will be? Certainly not -15 C!

The trick is maintaining constant temperature.

That's when the mapping for FPN and dead pixels works best.....

It's what Arri do for example. They use a Peltier to maintain a constant sensor temperature. That's why they don't have a user "black shading" function.

The only other cameras I know of that use this approach are Blackmagic.

JB

Very interesting, JB. Compelling image differences. A wonder what the steady state temperature goal for the Ursa sensor will be? Certainly not -15 C!

The trick is maintaining constant temperature.

That's when the mapping for FPN and dead pixels works best.....

It's what Arri do for example. They use a Peltier to maintain a constant sensor temperature. That's why they don't have a user "black shading" function.

The only other cameras I know of that use this approach are Blackmagic.

JB

Great insights, JB, as always. Thanks very much!

I wonder if the cooling fan connection has extra watts at it's disposal. So potentially a more powerful fan can be used in the current slot...for those users who don't mind a little extra noise.
.

+1 for a stronger/beefier fan if it is indeed that much efficient imagewise and work as it should, eletronics speaking. The good thing is that BMD make the fan itself user replaceable, so its a pretty solid idea to try to get a better cooling from there.

I remember once upon (a not too distant) time people were complaining about the (almost non-existent) fan noise. How things change lol..

How about water cooling a ssd block that fits inside the ssd slot and connecting the ssd externally? I bet if this could work the sensor would directly be cooled, I don't know if it would work but there's a ram that has water cooling, I think its hypernex H20 or something and I think if we used the same idea of the RAM's cooling structure it might work??

Just in case someone didn't know: BMCC is using COFAN blower fan, 5V, 0.08A, 2500RPM.
The first one in the list. Feel free to try anything else inside. I just don't really know if they will be properly powered?

Just tried connecting another fan, which is 5V, 0,25A and it didn't even start. So I guess there isn't much of a choice.

Title says it all..

( no where close to an URSA, but now I got water cooling. )

I'll first need to do some testing then I will share the results as soon as I can.

( from early stage testing, I found with this cooling the BMCC could actually be used for low light. )

So with heat no longer an issue.......I'm Ready for some 60p BMD....

apparently these people take suggestions for special purpose made cooling blocks so maybe you guys should work with them on something.

http://www.ekwb.com/

Red is recommending to black shade the Dragon at a specific working temperature, which is held pretty well under changing conditions with adaptive cooling by fan. I tested it at very low temperature without doing new shading and noise got really bad. So, it doesn't need to be cold, just needs to be calibrated for a specific temperature. Unfortunately, the cheaper models from BM do neither have the computing power (or memory?) to do a specific shading for it's working temperature and it's sensor nor do they have the sophisticated cooling of a Red. Please note: This all applies to FPN, dynamic noise definitely increases with temperature.

Some good reading here: http://www.red.com/learn/red-101/black- ... alibration

I found this:

could be fun adition to this cooling project.

Would be interesting to see BMD come up with a 60p upgrade kit for the bmcc and bmpc

*hint* *hint* ... hope you're listening bmd

I found this:

could be fun adition to this cooling project.

Next NAB, BMC running on free energy, instead of v-mounts

Looks impressive

Tests are currently under way.

I was actually hoping someone here has any info as to the sensor design on the BMCC.

Is the sensor vacuum sealed?

Looking online many scientific CMOS camera's ( which I believe share the same sensor/design. ) feature TE cooling and vacuum sealing.

http://www.andor.com/pdfs/literature/An ... ochure.pdf



Just curious as I dont want to harm the internal components with condensation.

When I first started this little project I had a few goals in mind. Reduce noise & FPN when shooting high ISO, eliminating hot pixels and removing “split sensor” horizontal line artifact that the BMCC exhibits.

After some time testing and researching further into the subject, I found I was only effectively able to do one of those things.

All are shot with same settings. ( ASA 1600 , 360° ) and both had a cool down/warm up time of about an hour.

Cooled temperatures are 12-17°c on the body and 27-35°c in the SSD compartment.

Uncooled temperatures are 30-35°c on the body and 50-60°c in the SSD compartment.









As the results show cooling the camera near completely eliminates hot pixels when shooting high ISO with low light.

Unfortunately the rest of the artifacts remain uncontrollable, at least by means of thermal influence.

Things like FPN, banding noise and the likes are a result of sensor design and hardware. These being out of control of users.

I did as much measurement as possible, but at the end of the day I was really just making guesses as to how the internal cooling behaved/responded to the extra cooling. It would be nice if the thermals were better documented, or at least explained a bit more in depth. It’s quite possible the additional cooling has less influence than I originally thought....or it’s possible I’m cooling too much and camera is actually counteracting it as a safety measure. (Counterproductive in my case.)

Are there some issues…yes. Power being one of them and also condensation. ( not to mention time to cool the camera. ) Getting closer to the sensor would also benefit. having to cool the whole body is overkill, but really the only way without totally dissembling/modding the camera ( not going to mess with that. )

I’ll probably play around with this some more, but this is (at least for now) the conclusion to the cooling test.

Hope some of you guys found it interesting or "cool"

The thing to understand is that the camera and sensor is actively cooled.

But peltier systems can just as easily heat as cool.

I believe that potentially what's happening inside the camera is that it's trying to maintain an optimum operating temperature. So if the camera is very very cold, it is possible the cooling system "heats' the sensor to this optimum operating temperature.

The reason it would do this is because that optimum operating temperature is the temperature all the factory BM calibration for each camera is done. In other words, for that calibration (dead pixles and FPN) to be at it's best, the sensor needs to be at the same temperature the calibration was done at.

Also, the "split sensor" issue is INHERENT to the sensor design. Cooling will have no effect on it. You'll never ever get rid of that. It's common on many sensors. My M8 does it as well when it's very very underexposed.

jb

The thing to understand is that the camera and sensor is actively cooled.

But peltier systems can just as easily heat as cool.

it was one of your earlier posts that made me aware of this, I was actually hoping that by effectively cooling the body ( heat sink in this case ) I could achieve a similar affect to what you would expect in a stacked/multi-stage peltier setup.

I believe that potentially what's happening inside the camera is that it's trying to maintain an optimum operating temperature. So if the camera is very very cold, it is possible the cooling system "heats' the sensor to this optimum operating temperature.

The reason it would do this is because that optimum operating temperature is the temperature all the factory BM calibration for each camera is done. In other words, for that calibration (dead pixles and FPN) to be at it's best, the sensor needs to be at the same temperature the calibration was done at.

If that is the case ( which I believe you are right about ) its too bad that BMD doesn't have a Black shading function for their cameras as do RED.

My current setup has the ability to cool roughly 10-15 below ambient. ( remains pretty constant I might add. ) Correct me if I'm wrong but If calibration/mapping was done with the lower temperature I can achieve would the effects of FPN/hot pixels be less? Or is it the same as with a higher temp & calibration? My instinct would say lower temp + calibration = cleaner image?

Also, the "split sensor" issue is INHERENT to the sensor design. Cooling will have no effect on it. You'll never ever get rid of that. It's common on many sensors. My M8 does it as well when it's very very underexposed.

Yeah its kind of unfortunate, I read about it over here regarding the dual amplifier design: http://www.andor.com/learning-academy/d ... amic-range

They noted: NOTE: Due to the splicing together of the low and high gains, the transition region between them is not seamless but has been optimised as far as possible.

Thanks again for your insights John, helpful as always

Yeah its kind of unfortunate, I read about it over here regarding the dual amplifier design: http://www.andor.com/learning-academy/d ... amic-range

They noted: NOTE: Due to the splicing together of the low and high gains, the transition region between them is not seamless but has been optimised as far as possible.

Thanks again for your insights John, helpful as always

It should be noted that with a very cleverly placed power window and fine tuned grade - you can easily hide the split.

I did so when creating some very low light shots - rated at 12,800 ASA on the BMCC.

https://vimeo.com/83725166/settings/privacy
PW: lowlight

ungraded:
http://tommajerski.com/publicimages/str ... _1.5.6.jpg


graded (pushed 4 stops)
http://tommajerski.com/publicimages/str ... _1.5.4.jpg


Lit with only a street lamp, 35mm 1.4

No special cooling used - just NR and a split fix grade.

also, worth noting that it was a moderately cool British evening time, but the camera was warm from filming all day long.

So if FPN (Pixed Pattern Noise) can be easily calibrated against (take 10 frames against some neutral cards), then cooling is a valuable option. Although would require some kind of post-processing app to de-noise.

Given that DNG is an open format, this should be relatively straight-forwards to code together.

I presume that for the sake of testing, we should have a test like:

* Point the camera at a neutral card under flicker-free lighting
* Defocus
* Take 10 frames
* Measure the standard deviation per pixel over the 10 frames, take the average of that
* Do this for a range of body temperatures

If this first test demonstrates that the sdev is reduced at lower temperatures (i.e. random noise is reduced), then next test would be like:

* Do test 1 for a range of lighting environments (including blackout)
* See if it's possible to use the blackout frames to predict the noise at the other brightness levels (i.e. a FPN cancellation function). Then the filming process is easy:

* Drop the camera temperture and hold it
* Put the cap on and take 10 frames
* Record at will
* Use post-process app to denoise the recording (using the 10 black frames as reference)
(final step : wish that there was a Magic Lantern project for BlackMagic cameras so that you can do this all in camera at record time)

Black Brushless DC Cooling Blower Fan 5015S 12V 0.1-0.3A 50x15mm
EUR 7,37on Ebay

Hate to bring this old thread back from the dead.

Just thought I'd share some new insight and findings related to the cooling in the BMCC.

Recently I decided to delve deeper into the idea of beefing up the internal cooling of the camera by increasing the power available to the peltier.

All this came with a better understanding of the heat exchange/management system of the camera.
( shoutout to Davi Silveira for being so helpful and informative in sharing his knowledge in his camera teardown. )

The sensor is cooled by what appears to be 2 stacked ( multi-stage peltiers ) which then transfers heat via flat cooling pipes to the body of the camera. ( that's why the camera warms up, including camera's like the pocket and production. )

( more about the flat cooling pipes http://www.amecthermasol.co.uk/product4 ... =30&fg=148 )

(the diagram is not drawn to scale or proportion, but should provide an idea of the layout )


Another great source of heat appears to be the FPGA Processor in the camera, which is only cooled via a small heatsink. For those who have long wondered by their SSD's came out blistering hot, now you know why. ( the processor is directly behind the SSD bay )

I initially believed the plan of juicing up the peltier would require a total dis-mantling of the camera, but to my pleasant surprise the connector for the internally powered peltier was right within reach once you remove the the fan from the bottom. ( the cable with the red and black wire )

Upon testing I found a some interesting details about the peltier.
1. Its very low power. ( voltmeter read: 2.5v @0.35-0.4A ~ 1W)
2. It does not appear to be controlled and is temperature independent. It just powers on and stays on. ( this was contrary to popular belief that it is set to maintain a certain temperature. )

I have successfully been able to power the peltier externally with higher power. ( 5.6v @.9A ~ 5W )
from preliminary results the extra power does have an immediate impact cleaning up the image, with almost 100% no hot pixels.

I will post proper test results soon.

Hope someone found this the slightest bit interesting/informative.

Thanks Csaba!
The question is, what happens when you are using more power for the peltier in the long run?
If it cools down too much, condensation will happen and eventually freezes, forming an ice cap.

And the higher power put into the peltier will be transformed into additional heat and has to be dissipated too.

Hope someone found this the slightest bit interesting/informative.

Fascinating!
Thanks for sharing.

Thanks Csaba!
The question is, what happens when you are using more power for the peltier in the long run?
If it cools down too much, condensation will happen and eventually freezes, forming an ice cap.

And the higher power put into the peltier will be transformed into additional heat and has to be dissipated too.

Those are excellent points, not one's I haven't considered myself.

Condensation is the biggest risk, I don't think the peltier runs cold enough for ice in my current climate, but condensation has formed on the body of the camera with my peltier/water cooler setup.

The extra heat load is being handled from the mentioned peltier/water cooler setup. ( scroll up to see what I'm talking about. )

Csaba,

Hey, nice resurrection! I'd say in having ran the camera outside of it's body and having first hand (literally) experience that 90% of the heat is coming from the processor. I can't describe to you how hot that thing gets, it's shocking. I actually am thinking of getting rid of my huge heatsink and going with something more modest in order to fit a better heatsink and/or fan to cool down that processor. I think the design of the camera is so compact that most of the sensor heat is really the processors heat transferring over. I'll send you some pictures tonight, sorry I got busy last night. The processor spills heat from the heatsink and the backside of the PCB where it's located. A crazy amount of heat!

Anyways that's what I've found, I know that cooling the sensor is number one for a cleaner picture. I hope you succeed in your external peltier power system. It's very tricky

that 90% of the heat is coming from the processor.

Processor either works or doesn't. It cannot work badly and give more or less noise.

In all this story I just don't understand the following: It's BMCC, not BMPCC. It has a fan. The fan should apply a stream of cold air to the processor radiator and throw it's heat away before it has a chance to heat a body, heatpipes and a cryocooler. But it does not. Why? And BTW there is an exhaust fan but where is an intake hole?

Anatoly, BMD uses materials, aluminum and now magnesium, partially because they are good conductors of heat. The camera's cooling systems are designed to draw heat away from the sensor and processor to heat the camera bodies which then radiates the heat to the air. Having a fan to blow cool air sounds good, but two fans might still be required then.


Sent from my iPad using Tapatalk

The Traveling Wilburys had a list of things that "He loves" from their song "Dirty World", and a Big refrigerator was smack in the middle of the list! In this case, a big refrigerator really cleans up the "dirty world" or image in this case! A bit random I know, but with all this talk of cooling, it just jumped into my mind!

Hey Csaba i'm totally glued to what you're doing with the BMCC !!! That looks fantastic !!! Hope you'll reach your goal.

I'm waiting for your tests !!!

Looking forward to hearing more about this!

Hate to bring this old thread back from the dead.

Just thought I'd share some new insight and findings related to the cooling in the BMCC.

Recently I decided to delve deeper into the idea of beefing up the internal cooling of the camera by increasing the power available to the peltier.

All this came with a better understanding of the heat exchange/management system of the camera.
( shoutout to Davi Silveira for being so helpful and informative in sharing his knowledge in his camera teardown. )

The sensor is cooled by what appears to be 2 stacked ( multi-stage peltiers ) which then transfers heat via flat cooling pipes to the body of the camera. ( that's why the camera warms up, including camera's like the pocket and production. )

( more about the flat cooling pipes http://www.amecthermasol.co.uk/product4 ... =30&fg=148 )

(the diagram is not drawn to scale or proportion, but should provide an idea of the layout )


Another great source of heat appears to be the FPGA Processor in the camera, which is only cooled via a small heatsink. For those who have long wondered by their SSD's came out blistering hot, now you know why. ( the processor is directly behind the SSD bay )

I initially believed the plan of juicing up the peltier would require a total dis-mantling of the camera, but to my pleasant surprise the connector for the internally powered peltier was right within reach once you remove the the fan from the bottom. ( the cable with the red and black wire )

Upon testing I found a some interesting details about the peltier.
1. Its very low power. ( voltmeter read: 2.5v @0.35-0.4A ~ 1W)
2. It does not appear to be controlled and is temperature independent. It just powers on and stays on. ( this was contrary to popular belief that it is set to maintain a certain temperature. )

I have successfully been able to power the peltier externally with higher power. ( 5.6v @.9A ~ 5W )
from preliminary results the extra power does have an immediate impact cleaning up the image, with almost 100% no hot pixels.

I will post proper test results soon.

Hope someone found this the slightest bit interesting/informative.

Maybe replacing the internal FPGA heatsink with a more efficient one would decrease the temperature of the FPGA. Also creating an air intake into the faceplate would benefit too. Here is an interesting pdf about Highly efficient cooling in low-airspeed environments:

http://pdf.cloud.opensystemsmedia.com/embedded-computing.com/Cool.Jul07.pdf
http://www.coolinnovations.com/technology/technical-articles/

Time for an update. I sort of left my promise of test examples unfulfilled, with that in mind I don’t have any DNG’s or samples from my previous test.

This is going to be sort of a conclusion to this subject, and my opportunity to address some points.

DISCLAIMER: I’m not an engineer or have much of any experience or qualifications in this field, however I believe the underlying points of these statements hold true.

As I have stated before and for anyone who understands principles of an image sensor, cooling will improve overall image quality. Low-light or Daylight, (less so on a well exposed image) But results vary on different sensors.

For me cooling did the best job at removing Hot pixels. Sort of makes sense…cooling “HOT” pixels.
Why I didn’t see more improvement beyond that comes down to way the 2.5K and the soon to be 4.6k sensor works.

These sensors achieve their excellent dynamic range using special circuity built onto the sensor that samples each column of pixels at both a Low and High gain, both streams of 11 bit data are then merged to create a 16 bit raw image. The sensors are also readout in halves, that’s what leads to the split sensor artifact.

More about that here: http://www.andor.com/learning-academy/d ... amic-range


Split sensor artifact shown below.


This is important to the way the sensors GAIN or ASA is then controlled by the user. As has been extensively been discussed in the past, the 2.5K (& pocket ) sensor both share a base sensitivity of 800 ASA. In camera adjustment of sensitivity is all digital gain, as opposed to traditional analog gain which is standard in most camera systems. (Not exactly true since there is some better mapping for the shadows that occurs when using 1600 ASA, but for sake of argument it is all digital gain.)

For this reason the Cinema Camera and Pocket Camera will never be as robust as something like an a7s. As you are essentially always shooting 800 and simply stretching the information in camera or in post. This to me is most noticeable in a more recent test I did, where I cooled the camera to near freezing. When looking at near black parts of an image no matter how much I raised the exposure, there was nothing to recover. Analog gain will always be superior vs digital gain, at least in regards to resulting in better SNR.

( and yes there is a radiator hanging out my window, because you can bet I'm going to embrace a Canadian winter to do cooling tests also that isn't clipped highlights under the water cooler, its ice forming on the peltier/cold plate)



While it is not certain what specific GAIN configuration BMD is using on the sensor, the sensor itself the “CIS 2521f” does have various gain modes. It may be possible for BMD to employ a gain configuration that may be better suited to low light shooting when desirable. This does have a reduction in dynamic range and bit depth because it would essentially be derived from a single channel of the 2x 11 bit stream. It may be possible that BMD could explore this, although very unlikely at this point If they haven’t already done it.



So with this in mind, don’t shoot with these cameras like you would an a7s. It’s not going to work as well, no matter how cool the sensor is. That’s what all this has taught me. So don’t go cooling your cameras with LN2 or putting them in the freezer. It won’t help you shoot in moonlight.

Also as another address to camera thermals, I frequently see sensor heat as a potential limit to High Frame Rate recording. I want to throw my thoughts into this one, because although I agree with this point I don’t believe it’s as severe as it’s put out to be. Heat is generated by the sensor, but to a degree far less than something like the FPGA in these cameras. If we look at what available information we have about the sensors, they operate typically at least than a few watts. Taken directly from the 2.5k datasheet



“Power consumption < 2W at 100 fps”

less than 2 watts in the grand scheme of the entire thermal power design of the camera is very insignificant when it comes to its heat contribution. Your cam gets hot far more from the Peltier, the SSD, the battery and mostly the FPGA.

So to say the “The cam can’t handle it because of sensor heat." is not likely to be as great of a factor as one might think.

I’m not saying the cam can therefore handle HFR the way it is, however to me is suggests is has a lot more to do with the processing pipeline or insufficient power.

With that little rant over, some closing words.




Nothing can change shooting with light, use lots of it. No cooling can substitute a lack of lighting. That’s what I have taken from all this.

The more you know!

Having said this...I'm not done...yet...

less than 2 watts in the grand scheme of the entire thermal power design of the camera is very insignificant when it comes to its heat contribution. Your cam gets hot far more from the Peltier, the SSD, the battery and mostly the FPGA.

So to say the “The cam can’t handle it because of sensor heat." is not likely to be as great of a factor as one might think.

I’m not saying the cam can therefore handle HFR the way it is, however to me is suggests is has a lot more to do with the processing pipeline or insufficient power.

Exactly.

It's not the sensor itself, it's the combined total of all the workflow in there to make it all work at such high data rates and bit depths.

And the peltier get's hot because it's MAINTAINING a sensor temperature that gives a more consistent performance.

Fascinating stuff !

JB

It looks like they have that great metal case. Any way to thermally couple the hottest chips to it?