Black Magic PocketHD and 4K production camera

You know if 60p is a heat issue with the current design I wonder if they could get 48p just stable enough to work. Might be a good compromise for some people. 2x slow motion isn't as good as 2.5x but still slow motion. Personally I don't use slow motion very much so it isn't a big deal to me. If I did want slow motion I would want more than 60p anyway.
 
True North said:
About the only feature that would completely round out the BMPC is a higher framerate recording option for slo-mo. I think most would be more than fine with a 60 or 120fps 2K mode. In one of the posted videos the Blackmagic rep hinted that higher frame rates could be in the near future.
Click to expand...

I'd love higher frame rates too. It would make the BM4k almost perfect...at least on paper. What I've been reading is that the current hardware can't do it, even if recording to lower resolutions like 1080p. Apparently, it has something to do with heat management.

Samuel, just out of curiosity, do you know if the CMV12000 can be read-out faster, either by line-skipping, or windowing the sensor? That might potentially allow faster framerates without much extra heat being generated, at the cost of resolution. Or, does the whole chip have to be active at once?

True North said:
I am still blown away by how much camera you are getting for the price. Unbelievable really.
Click to expand...

No dissagreement there. Global shutter FTW.
 
Wesley Byram said:
Samuel, just out of curiosity, do you know if the CMV12000 can be read-out faster, either by line-skipping, or windowing the sensor? That might potentially allow faster framerates without much extra heat being generated, at the cost of resolution. Or, does the whole chip have to be active at once?
Click to expand...

The CMV12000 does 100 fps in 12-bit-mode 4096x2160, 200fps if you just want the central 4096x1080 (you can discard the sides of that before writing to disk if you wish).
No idea if it creates more heat this way, I haven't measured it (I believe this sensor limit is ACD-and-bandwidth-related).
I believe the heat issue is in the ASIC, not the sensor, but I could be wrong (because heat in the ASIC doesn't degrade IQ -it freezes the camera- whereas heat in the sensor does reduce IQ).
 
wow... in the hands-on bit with the pocket cam, looks like there's plenty of jello frame already. Curious if it'll be that extreme in the shipping version, and of course it's difficult to tell for certain from a video like this... not that it couldn't be controlled, but the stabilization necessary to do so would certainly take away from its svelteness.

Hope it's not like that... I'm seriously tempted otherwise.
 
mcbob said:
wow... in the hands-on bit with the pocket cam, looks like there's plenty of jello frame already. Curious if it'll be that extreme in the shipping version, and of course it's difficult to tell for certain from a video like this... not that it couldn't be controlled, but the stabilization necessary to do so would certainly take away from its svelteness.

Hope it's not like that... I'm seriously tempted otherwise.
Click to expand...
Me too!

You may already be aware but, for the record, the pocket cam is still rolling shutter whereas the 4KBMCC is global shutter.
 
mcbob said:
wow... in the hands-on bit with the pocket cam, looks like there's plenty of jello frame already. Curious if it'll be that extreme in the shipping version, and of course it's difficult to tell for certain from a video like this... not that it couldn't be controlled, but the stabilization necessary to do so would certainly take away from its svelteness.

Hope it's not like that... I'm seriously tempted otherwise.
Click to expand...

That's where the Lumix glass may help a bit if they can get image stabilization as part of the M43 electronic connection.
 
Also, there have been some cameras released recently that show more rolling shutter issues in the LCD than in the recorded footage (e.g. Nikon D5200). It's very weird and I don't understand how this can be, but there it is. The pocket BMC may be like that, since I would expect it to have slightly faster shutter than the original BMC.
 
John Brawley did mention that the RS seemed a little worse on the Pocket than on the regular BMCC. That might be because lenses are a little more telephoto on the Pocket vs. original BMC and because the Pocket is obviously going to be more susceptible to hand shake. Both of these amplify movement and call attention to any rolling-shutter issues, even if the sensor readout speed where the same on both cameras.

Now, a good question is if the readout speed is actually worse/better/same. If it's just a 1920x1080 version of the chip in the BMC, or in fact the very same chip windowed to 1920x1080, then we would expect the rolling-shutter performance to actually be a little better than on the BMC, because there are fewer pixels to be read. Fewer pixels=less time to read them=less rolling-shutter. I wonder, though, if Blackmagic "underclocked" the chip, slowing the readout, to control heat. The BMC-Pocket is pretty small. Not much room for heatsinks...
 
Last edited:
Samuel H said:
The CMV12000 does 100 fps in 12-bit-mode 4096x2160, 200fps if you just want the central 4096x1080 (you can discard the sides of that before writing to disk if you wish).
No idea if it creates more heat this way, I haven't measured it (I believe this sensor limit is ACD-and-bandwidth-related).
I believe the heat issue is in the ASIC, not the sensor, but I could be wrong (because heat in the ASIC doesn't degrade IQ -it freezes the camera- whereas heat in the sensor does reduce IQ).
Click to expand...
Well if the CIS2051 were the sensor in the pocket cam, it could do 200 fps at 1,920 x 1,080 in rolling-shutter mode and 100 fps in global-shutter mode -- but the camera would probably melt.
 
Wesley Byram said:
Now, a good question is if the readout speed is actually worse/better/same. If it's just a 1920x1080 version of the chip in the BMC, or in fact the very same chip windowed to 1920x1080, then we would expect the rolling-shutter performance to actually be a little better than on the BMC, because there are fewer pixels to be read. Fewer pixels=less time to read them=less rolling-shutter. I wonder, though, if Blackmagic "underclocked" the chip, slowing the readout, to control heat. The BMC-Pocket is pretty small. Not much room for heatsinks...
Click to expand...

Tzedekh said:
Well if the CIS2051 were the sensor in the pocket cam, it could do 200 fps at 1,920 x 1,080 in rolling-shutter mode and 100 fps in global-shutter mode -- but the camera would probably melt.
Click to expand...

both spot on, and still I'm not sure about the CMV12000: the docummentation talks about 2-3W power consumption (depending on where you read it; I haven't measured it, but that should be at full speed). The CIS2051 says 20W

so:
* they may have underclocked the sensor for the pocket cam, and this could explain the jello (the other explanation would be that jello is in the LCD but not in the footage, as in the D5200)
* I still doubt sensor heat is the issue for not allowing >30 fps on the BMC-4K, ASIC limitations seem more likely
 
Last edited:
Heat.. . . . . No other photo/video camera of this size can run without overheating, that I know of, to be really dependable as a video camera. Why would this one be different?
 
Wesley Byram said:
I just now watched this. The Pocket looks very drool-worthy, but...holy cow look at the jello. He's barely moving the camera and the rolling-shutter is very prominent. The finalized, shipping version isn't going to record footage that looks like that...right? I mean, it can be the smallest, cheapest, highest DR, raw recording camera in the world (which it is), but if a movie camera requires that nothing in the frame actually moves very much, then what use is it?
Click to expand...

Can someone give the exact time when they see the worst rolling shutter?
 
Wesley Byram said:
I just now watched this. The Pocket looks very drool-worthy, but...holy cow look at the jello. He's barely moving the camera and the rolling-shutter is very prominent. The finalized, shipping version isn't going to record footage that looks like that...right? I mean, it can be the smallest, cheapest, highest DR, raw recording camera in the world (which it is), but if a movie camera requires that nothing in the frame actually moves very much, then what use is it?
Click to expand...


So I guess you forgot that you can put this on a tripod.......

People are impossible to please SMH..

Use this on a tripod, gorilla pod, skater dolly, heck it's so small you can prop it up against anything. Work to the strengths of a tool.....
 
philiplipetz said:
Can someone give the exact time when they see the worst rolling shutter?
Click to expand...

Check out around 10:40. I recommend watching it in full screen at HD res to see it most clearly. Now, bear in mind that this is handheld video of the back of the BMC-Pocket LCD, which is itself handheld, so that might be making it look worse than it is. It's also a pre-pro camera, so the jello might not be so bad in the shipping product. It could also be, as Samuel says, a trick of the Pocket's LCD, and not something so present in the recorded footage. I'm not writing this cam off yet.
 
J.F.R. said:
So I guess you forgot that you can put this on a tripod.......

People are impossible to please SMH..

Use this on a tripod, gorilla pod, skater dolly, heck it's so small you can prop it up against anything. Work to the strengths of a tool.....
Click to expand...

Well, yeah that would make it better, but then if you need to carry around a rig or a tripod all the time, it becomes a lot less "pocketable," doesn't it? Besides, stabilization only takes out the handheld shakes. Passing cars, for example, are still going to lean an awful lot.

Of course, this is all assuming that the rolling-shutter is really as bad as it looks in this video. I'm not going to complain without seeing some real footage from a production camera.
 
Last edited:
You must log in or register to reply here.
Back
Top