[Reportage]

Software as in just by changing a value will allow change of fps value and Hardware as in there are voltage regulators in place that does not give the shutter mechanism the power needed to run at possibly higher fps?

[RonHendriks1966]

No actual idea. My guess will be that the camera manufactuar buys a shutter that fits the camera. A faster shutter (top speed of 1/2000 or 1/4000 or 1/8000 or 1/16000 of a second) get more pricier and the same goes for higher fps or longer lifetime (100.000 or 300.000 actualisations).

[Marc Sabatella]

No way would they impose an arbitrary software limitation. While it's possible that entirely new firmware could be written that would speed things up slightly, it's got to be mostly hardware - and probably the mirror assembly more so than the shutter.

[kristoffon]

Both

From one side the shutter and mirror assemblies take a certain time to work. Faster ones can be built but will cost more.

On the software side you need powerful processors to turn 14 MP worth of data into JPEG files. Just take note of how long lightroom takes to do a full size raw conversion and realize the camera does that each time you click. And if RAW lots of RAM are needed to store the data it is written to the SD card.

[Lowell Goudge]

it is simply hardware.

the mirror speed and damping mechanism, plus the shutter itself are the principle limits.

as you speed things up, the stresses go up tremendously, and the lifetime falls.

so if you want a faster frame rate, it will cost a lot more.

[Adam]

It's limited by the mirror and the sensor/processor.

[stevebrot]

Consider everything that must happen on each exposure cycle and you will see why it is unlikely that the fps is an arbitrary software setting.

1. Close lens aperture (conceivably from full open to full closed)
2. Raise mirror
3. Release first shutter curtain
4. Release second shutter curtain
5. Capture the image to buffer
6. lower mirror
7. Open lens aperture
8. Return both shutter curtains to ready position

The bottleneck is one of three places, 1) the lens aperture, 2) the mirror, or 3) the image processor. My bet is on #3. My reasoning is that high fps were routine for professional bodies on film cameras.


Steve

(Thinking in horror about the iris of vintage glass k-mount glass at high fps...)

[Ex Finn.]

I would say that keeping the mirror and shutter intact is the limiting factor, double the speed, quadruple the forces. (correct me if I am wrong).
For a $763.00 body the K-7 5+ frames per second is pretty good IMO.

[alohadave]

Software as in just by changing a value will allow change of fps value and Hardware as in there are voltage regulators in place that does not give the shutter mechanism the power needed to run at possibly higher fps?

Both.

The shutter has a maximum speed that it can operate at. It's related to the flash sync speed, because the longer the sync speed, the fewer actuations per second the shutter can operate.

For software, the main limit is the buffer size. The bigger the buffer, the more pictures you can take without the camera needing to pause to write the pictures to the card.

[shuie]

If the buffer needed to be larger why couldn't a high speed grip be offered with a additional RAM buffer in the grip?, maybe even offering an addiional processer and SD chip? Just a thought.

Shu

[alohadave]

If the buffer needed to be larger why couldn't a high speed grip be offered with a additional RAM buffer in the grip?, maybe even offering an addiional processer and SD chip? Just a thought.

Shu

In theory, if the camera was designed to do that, then sure you could add additional buffer.

Cameras aren't designed like PCs are. They are closed systems with very limited expansion options, if any.

[ManuH]

The shutter has a maximum speed that it can operate at. It's related to the flash sync speed, because the longer the sync speed, the fewer actuations per second the shutter can operate.

I don't think so. The shutter travels the frame in 1/180s. So in theory you could have max. 180 fps! Thus the shutter is not the bottleneck. Someone on dpr posted a link to the Samsung sensor specifications. It says that a 2 channel readout version is able to give 3fps and a 4 channels more than 5fps. So the sensor readout speed seems to be the real bottleneck.

[Lowell Goudge]

For those discussing the shutter note that sync speed has 3 components. Leading curtain time, trailing curtain time and open time. Logically leading and trailing curtain times are the same then add 1-2 ms for the max power flash duration. Let's say flash duration is 2 mS that leaves 1.5 ms each curtain to be at 1/200 overall

This means the shutter coils easily cycle fast enough

The limit is mirror and aperture control

If you could tolerate blanking out the viewfinder at high rate, leave the mirror up and aperture closed down

From memory the green button metering time is about one second which is largely aperture closing time

[Lowell Goudge]

As an aside if the aperture control was totally in the lens with no camera linkage it would be a lot faster

This is the price of compatibility backwards

[stevebrot]

As an aside if the aperture control was totally in the lens with no camera linkage it would be a lot faster

This is the price of compatibility backwards

In theory, the camera could decouple the mirror and aperture travel between exposures. Shutters might also be designed to operate in both directions so that there is no need to "cock" the shutter between exposures.



Steve