Page 1 of 1

FPN & NOISE vs camera TEMPERATURE

PostPosted: Mon Nov 17, 2014 5:08 pm
by AndyGandy
Hello BMC users!

A pair of days ago I have read a next topic on bmcuser.com:

http://www.bmcuser.com/showthread.php?1 ... mera-WoW!!!

It is related to fpn and noise after changing firmware. I had made some test and _no success_ BUT found interesting depencency ratio of noise vs camera working temperature.
I want to share it with you because I found it very exiting!

First of all let me show you my screengrabs when I tested all all of noise and FPN things.
I need to say about this pictures first or about my test methodology.

All test was completed on 1.9.8 firmware. I have recorded 4K RAW dng files with various ISO and various camera temperatures. After that I dropped the dng files into Photoshop and when ACR opened I increased the Exposure to +3 setting to better watch the result. I know that in ACR applies by default some sharpening and color noise reduction but I just leaved it as is (Color NR = 25) in all cases.
Than I opened the dngs in Photoshop with Adobe RGB color profile applied and saved as 16 bit TIFF files. Than files was cropped for comparison and showed to you with 100% zoom (I mean 1:1, no zoom).

So here the test results:

1. Camera just turned on with the room temperature (about +24°C):

Image

Don't forget! I have raised exposure in ACR to +3 to look like that.

2. Camera worked about a hour. I have no idea what temperature it was but by the hand it feels like very warm:

Image

As you can see the more camera temperature the more noise (and FPN) it generates. So I decided cool it down to possible deggrees.
As I see at blackmagicdesign.com operating temperature of the camera allowed is 0°C to +40°C (+32°F to +104°F):
https://www.blackmagicdesign.com/produc ... techspecs/
So I decided to go stright to almost 0°C. Luckily weather in our region is pretty match my requirements. I mean it cold and dry. So I leaved the cam at the cold unheated room for about a hour and half. After that time it was very cold by the hand feeling. I think it was about +3°C or +5°C. I recorded the RAW again and ingested it in Photoshop.

MY JAWS ARE DROPPED!!!

Look yourself:

Image

Here the dng files from cold shooting:

https://www.dropbox.com/sh/nsd3tddy2ft4 ... oT9-a?dl=0

I never saw anything clearer like these! It just crazy cleanest image! You can shoot at ISO800 with that temperatures. And it looks better than ISO200 in normal room temperatures! I attached the comparision pics:

Image

TIME TO BUY SOME WATER OR LIQUID HYDROGEN COOLING SYSTEM FOR BMPC4K!!!

P.S. Night shootout is coming... "Waiting for the night" © dm.

WARNING!!! Don't forget that cooled down technics need to be warmed about a hour or two to work in normal temperatures because you can get the humidity condensate dropped to electronic parts!!!

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Mon Nov 17, 2014 5:26 pm
by rick.lang
Thanks, Andy. I was just going to post a link to your thread from the other forum! Now you have shown these samples, it would be most interesting for someone with the liquid-cooled URSA to attempt a similar test. The liquid-cooled URSA should keep the internal operating temperature of the sensor at a fairly consistent temperature, but would be good if we knew how effectively that is achieved and if the noise is consistently lower. Shooting with a camera near freezing temperatures may not have a general application, but if URSA's design temperature is consistently quite low, that bodes well for the noise in those images.


Rick Lang
Sent using Tapatalk HD

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Mon Nov 17, 2014 5:52 pm
by ChrisWarren
Great test!

This is a somewhat related question I've been meaning to ask the URSA Users/Owners...

How is 800iso? Is it better than or on par with BMPC 800iso?

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Mon Nov 17, 2014 6:02 pm
by cabraldemello
Fantastic! Thank you!

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Mon Nov 17, 2014 6:26 pm
by danielschweinert
very interesting! Cooling the camera externally has been already tested but it didn't worked very well:

http://forum.blackmagicdesign.com/viewtopic.php?f=2&t=24688

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Mon Nov 17, 2014 7:31 pm
by Mac Jaeger
Your findings are no general surprise, all sensors show less noise when cooled properly - that's why we use heavy peltier cooling on our astrophotography cameras.

But your test pictures still surpassed my expectations about just how much this sensor could benefit from better cooling! That makes me hope that my upcomming winter shootings will look good even though i'll have to do some of the shots at iso 1600...

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Mon Nov 17, 2014 8:43 pm
by Aaron Scheiner
After the last thread on this subject I ordered a big peltier unit off eBay. As soon as it arrives I'll sandwich it between my 4K camera's housing and a Dell Poweredge CPU heatsink :D .

Great test Andy :) .

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Mon Nov 17, 2014 9:02 pm
by AndyGandy
The shooting results are not very fine. I can say I can use ISO800 in my BMPC when cooled down but can't give a stop more to the footage. And its pretty dark even at 800 ISO:

Image

Some dng:
https://www.dropbox.com/sh/j3pn7i1scmqn ... TG-_a?dl=0

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Mon Nov 17, 2014 10:42 pm
by cabraldemello
Aaron Scheiner wrote:After the last thread on this subject I ordered a big peltier unit off eBay. As soon as it arrives I'll sandwich it between my 4K camera's housing and a Dell Poweredge CPU heatsink :D .

Great test Andy :) .


Do you mind giving details on it?

Could you please post a link to that peltier unit on Ebay and describe how you'd use it with the Dell?

Also, how noisy would that combination be?

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Tue Nov 18, 2014 12:18 am
by brent k
What would be nice is if you kept the BMCC/PC body, but removed the SSD, and rear screen, beefed up the internal cooling for up to 60p, and moved the control screen/recorder to an external device that functions exactly like an Atomos Shogun. That would be a killer camera, kind of like what Red does, but better in that the all in one control/record unit takes dual generic SSDs.

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Tue Nov 18, 2014 1:07 am
by CaptainHook
Hi, are you able to post a DNG from your camera? Thanks. :)

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Tue Nov 18, 2014 4:23 pm
by FrankApollonio
I'm curious how this would test with iso 1600

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Wed Nov 19, 2014 8:32 am
by Aaron Scheiner
@cabraldemello : I bought mine here, but there are many retailers on eBay that sell the same item, often with various combinations of surrounding kits.

One kit I saw while searching used liquid cooling (and as such no fan noise).

I suspect it would rapidly turn into a bulky setup, as you'd need a heatsink/water cooling block and the device itself needs to be powered by something decent.

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Wed Nov 19, 2014 5:33 pm
by Granit Neziri
I recently bought a BMPC4k myself and the first thing i noticed was the FPN, its quite bad on my camera. However, after doing some tests i have now solved the problem by letting the camera warm up for about 10minutes. By doing so i can use 400iso in low light with pretty much no FPN at all.

Re: FPN & NOISE vs camera TEMPERATURE

PostPosted: Thu Nov 20, 2014 3:38 am
by cabraldemello
Aaron Scheiner wrote:@cabraldemello : I bought mine here, but there are many retailers on eBay that sell the same item, often with various combinations of surrounding kits.

One kit I saw while searching used liquid cooling (and as such no fan noise).

I suspect it would rapidly turn into a bulky setup, as you'd need a heatsink/water cooling block and the device itself needs to be powered by something decent.


Many thanks, Aaron!