# Why the A99 is slower than the A77 :-)

Started Sep 18, 2012 | Discussions thread
Re: Are you kidding right?

VirtualMirage wrote:

Here's a quote of what I wrote in a previous thread that gives an explanation to my reason/theory:

The distance a shutter needs to cover to make a complete cycle on a full frame sensor is longer than for a crop sensor or in crop mode. About 53% longer. This part is obvious.

So the time it takes for a a full frame sensor to meet 6 fps at said distance is faster than the time it would take for a shutter in crop mode to meet 6 fps for its distance.

Thus, if a shutter on a full frame sensor is clicking at 6 frames per second, for a crop sensor that same speed would account for a higher frame rate with no need to speed up the shutter. This is because the shutter has to cover less distance.

Here is some loose math in a very basic environment for ease of understanding and formulas (I know it is more complex than this). My math is assuming our "shutter" is just a single line/string, not a full frame covering multi-leaf shutter that it actually uses. Because of this assumption, the real life shutter speed would actually be much faster. But conversion in speed difference from a full frame to a crop frame should be about the same. Here it goes:

For a full frame sensor (35.8 x 23.8mm) shutter to meet 6 frames per second, it needs to travel at about 142.8 mm/s.

For a crop frame sensor (23.5 x 15.6mm) shutter to meet 6 frames per second, it needs to travel at about 93.6 mm/s. To meet 10 frames per second, it needs to travel at about 156 mm/s.

So assuming that only the cropped portion of the full frame needs to be covered by the shutter for a proper exposure, the speed of the full frame shutter does not need to change that much (if at all). So the camera could be "amping up" the shutter speed by less than .5 fps in full frame equivalent speeds to meet a 10fps crop speed. Or...since most manufacturers only document their frame rates in whole numbers, the actual full frame shutter speed could be slightly faster (maybe up to .5 fps) or the the crop mode frame rate could be slightly slower (maybe by .5 fps), or a combination of both.

With this theory in mind, I believe it is more of a shutter limitation and not a bandwidth limitation as to why crop mode can shoot at 10fps while full frame is limited to 6fps.

Paul

Sorry, but posting false information twice does not make it right, nor more right the second time around. I could not follow your logic a week ago, and it still escapes me today.

Your premise in the bold-italic quoted text is that even though the FF shutter travels over the entire FF distance at the same speed you get a faster frame rate with crop mode vs FF mode because the crop area is smaller. That is just plain wrong.

Frame rate is a measure of how many times per second the shutter mechanism executes a complete operating cycle. That includes traversing over the entire operating distance to expose the frame, and then back in preparation for the next frame. For the A99 in full frame this is 4000 rows of pixels that need to be exposed/covered but the shutter. If the shutter does 6 cycles per second, you get 6 fps, and that's all there is to it. Any single pixel only gets exposed 6 times per second.

Now by your logic, if we use a smaller portion of the sensor and keep the same traverse speed we get a higher frame rate? I mean, think about it, what sort of frame rate would you get if the crop area were only 10 rows of pixels instead of 4000? Would you get 4000 / 10 x 6 = 2400 frames per second? With the shutter traveling at the same linear speed??? I would agree that with a cropped sensor the image is "painted" in less time because it takes the shutter less time to cover the shorter distance, but the shutter still has to complete its full traverse over the FF and return.

You dismissed the suggestion a week ago that the frame rate could be increased if, using the same shutter traverse speed, the traverse distance were shorter; i.e., limited to the number of pixel rows of the crop area. But that is, in fact, the only way to make the mechanism cycle faster -- decrease the distance the shutter has to travel at its fixed speed, thereby allowing the camera to begin the shutter reset action sooner. (I may be wrong here, but I suspect shutters are spring loaded in the forward direction, and motor or solenoid driven in the reset phase of the cycle.) As was suggested, this would likely complicate the shutter design, and I suspect this is not implemented in practice.

If it were in fact implemented, or if you compare a FF camera with an APS-C camera where traverse speeds of the shutter blades are the same, then you will get higher frame rates with the smaller size sensor. Of course, the ratio of frame rates would not be linearly related to the ratio of the frame sizes because the traverse time during the exposure phase is only a portion of the total shutter cycle duration.

For a FF camera where a faster frame rate can be selected with a crop of the FF, the maximum frame rate is determined largely by the lesser of a) the maximum cycle speed of the shutter mechanism; and b) the maximum rate at which the data from the sensor can be read out to memory. I suspect the mechanicals of the A99 shutter can give you 10 fps max, and that is available in crop mode where less sensor data is being read to memory. When reading the FF sensor, frame rate is limited by the data readout speed, and the shutter is delayed slightly before each cycle to match the readout speed of 6 fps. But the shutter still travels over the sensor at the same rate.

So, shutter traverse speed is likely fixed by the spring tension; shutter release frequency is adjusted to match data readout rate (lower frequency of shutter tripping with FF); and of course "shutter speed" (frame exposure time) is set by the time delay between front and rear curtain, even if using electronic front curtain.

JF

JohnFrim's gear list:JohnFrim's gear list
Sony SLT-A57 Sony DT 18-55mm F3.5-5.6 SAM Sony DT 18-250mm F3.5-6.3 +3 more
Complain
Post ()
Keyboard shortcuts: