[monochrome]

I don't see how 1/200 or 1/250 would make for a relevant improvement over 1/180. That's 1/3 to 1/2 EV.

Get into leaf shutter territory (medium format; NOT in the 645D) and you can have 1/500 to 1/1.000 easily. That seems more like it.

The benefit of 1/250 is primarily for backlit scenes with fill flash.

[Uluru]

What was done instead was to increase its reliability.

For a camera that will stick on the market for at least next 3 years, I find it unbelievable that they went all the way to redesign the shutter, and leave out perceptual flash sync speed barrier. Prospective users read it like that — nothing new will happen in the Pentax flash territory in the next 2-3 years. A big perceptual problem, that with the redesigned shutter shouldn't be there.
But is that entirely true?
At this stage, I think we're merely speculating. For example, latest firmware increased GR's shutter speed at a max aperture to 1/2500.

[goubejp]

I don't think they have redesigned the shutter. Increased im/s results from a faster Prime3 engine, and improvements on the Mirror box. I use the High speed flash when needed, anyway.

[Kunzite]

A clue towards a new shutter is the increased reliability (which I mentioned before). Or, maybe they found out the old MTBF was underestimated by a factor of 2
FSP could actually need shutter modifications, if the rearming is not fast enough.

AFAIK the SR lowers the effective sync speed a bit, so the shutter should actually be capable of 1/250 (in a non-stabilized body). Maybe they would be able to improve on this limitation, but using a faster shutter (as in: faster traveling curtains, resulting in a faster sync speed) is probably too expensive, for a $1300 body.

[Barry Pearson]

A clue towards a new shutter is the increased reliability (which I mentioned before). Or, maybe they found out the old MTBF was underestimated by a factor of 2
FSP could actually need shutter modifications, if the rearming is not fast enough.

If it was the mirror assembly that typically fails, rather than the shutter, the change in reliability could be down solely to the mirror assembly improvements.

[joergens.mi]

I don't think you need a faster shutter, but a shutter that is fully opened when the flash fires. Therefore the speed for the curtains must increase to keep the full view open at 1/250 seconds. I think the limitation is that 1/180 is the fastest speed at which the shutter is fully open. Higher speed only shifts a partly opened segment over the sensor plane.

If we assume 20% of the exposure time for opening and closing the curtains the speed of the curtains will be increased from 50km/h (14,5 m/sec) increased to 75km/h (20 m/sec) the masses of the curtains must accelerated and stopped without disturbing all other actions in the camera 8.3 frames per second no vibrations and bringing back the shutter to the start position. I think that'll be a completely new designed shutter - a complicated process with all the patents around.

[Uluru]

A clue towards a new shutter is the increased reliability (which I mentioned before). Or, maybe they found out the old MTBF was underestimated by a factor of 2
FSP could actually need shutter modifications, if the rearming is not fast enough.

AFAIK the SR lowers the effective sync speed a bit, so the shutter should actually be capable of 1/250 (in a non-stabilized body). Maybe they would be able to improve on this limitation, but using a faster shutter (as in: faster traveling curtains, resulting in a faster sync speed) is probably too expensive, for a $1300 body.

I don't know. I mean, I give up on presuming anything about "why Pentaxes won't work as others".
Sony also uses SR, and Olympus. Alpha 77 syncs at 1/250, and Oly EM1 at 1/320.
Their lower grade cameras, though, sync at lower speed. Old Alpha 65 synced at 1/160.

I remember reading somewhere many months ago, that Olympus deliberately omitted using a better shutter mechanism for the OMD5 to sync it to 1/320. They used a lower grade, saying that cost would not be justified, and that they are leaving it for the flagship (presumably EM1?). However, 1/320 is now available in their E-P5 too. It costs less than $1300.

And K-3 s is supposed to be some sort of flagship for Pentax, but the cost isn't justified at all? All that doesn't make any sense when comparing to others. And makes one wonder in what direction those Pentax cameras and the entire system are going to.

---

Last edited by Uluru; 10-13-2013 at 04:59 AM.

[IchabodCrane]

I don't know. I mean, I give up on presuming anything about "why Pentaxes won't work as others".
Sony also uses SR, and Olympus. Alpha 77 syncs at 1/250, and Oly EM1 at 1/320.
Their lower grade cameras, though, sync at lower speed. Old Alpha 65 synced at 1/160.

I remember reading somewhere many months ago, that Olympus deliberately omitted using a better shutter mechanism for the OMD5 to sync it to 1/320. They used a lower grade, saying that cost would not be justified, and that they are leaving it for the flagship (presumably EM1?). However, 1/320 is now available in their E-P5 too. It costs less than $1300.

And K-3 s is supposed to be some sort of flagship for Pentax, but the cost isn't justified at all? All that doesn't make any sense when comparing to others. And makes one wonder in what direction those Pentax cameras and the entire system are going to.

They can only bite off so much at one time and I think most observers would conclude they already bit off a lot going from the K-5 series to the K-3. We shouldn't assume sync speed won't get revised in another hardware revision somewhere down the road.

[Kunzite]

And K-3 s is supposed to be some sort of flagship for Pentax, but the cost isn't justified at all? All that doesn't make any sense when comparing to others. And makes one wonder in what direction those Pentax cameras and the entire system are going to.

The K-3 has so many improvements, complaints that they didn't include X or Y and because of this they're doomed don't make any sense. They had to draw a line somewhere and say: "OK, that's all we'll include in the product"; otherwise there would be no product, ever.
We know where they're going by recognizing the K-3 as a revolutionary product.

[joergens.mi]

if you had a look at all the changes

• AA-Filter
• 27 Point Autofukos
• 86000 Point White Balance
• High Speed 8.3
• Autofocus
• ...

All that cost a lot of nonrecurring development cost, and at a time there is only one Budget and there is a timeline. Ricoh has prioritized other than the flash community - maybe unluckily.

[thibs]

They can only bite off so much at one time and I think most observers would conclude they already bit off a lot going from the K-5 series to the K-3. We shouldn't assume sync speed won't get revised in another hardware revision somewhere down the road.

The thing with SR is that since the sensor moves, you need to take that into account for the flash sync.
Hence, if the sensor wasn't moving it is probable the sync would be higher (between 1/200 and 1/250).

This of course doesn't change a thing as to the general rule for flash sync: faster shutter means faster flash sync.
I know older Sony bodies had different sync speed if steadyshot was on or off. What is the situation now, I dunno.

As for Oly cams, I dunno the details but remember the sensor is smaller so easier to have higher flash sync yet with sensor stab.

[joergens.mi]

Why must I take Sr into account, sensor and shutter are different parts and there is no need for aligning both of them except for starting the shutter when stabilised. The shutter itself must be big enough to see the complete available sensor.

[cfraz]

I don't know. I mean, I give up on presuming anything about "why Pentaxes won't work as others".
Sony also uses SR, and Olympus. Alpha 77 syncs at 1/250, and Oly EM1 at 1/320.

One often cited difference in the way the "Pentax K-5 didn't work as other" cameras was it's quiet operation.

A possible reason sync speed of 1/180 sec wasn't improved is that the design criteria included the notably quiet shutter/mirror actuation of the K-5 series. Note that they did rework the mirror damping to cope with the higher frame rate, presumably resulting in less vibration and noise from the mirror. Perhaps a shutter with 1/250 sync speed with it's faster moving curtains could not meet the noise criterion.

If I had a choice between 1/2 stop faster shutter sync speed (which would come into play only on rare occasions for me) and a K-5 like noise level (which happens every exposure), I would take the quiet camera.

[Kunzite]

The complete sensor's surface, in all possible positions (as it will be moved around by the SR mechanism).
With a very precise synchronization, it should be possible to release the second curtain earlier - so that it would reach the sensor's edge right after the first curtain fully uncovered it. I'm not sure it's feasible, though, nor that it would make any measurable difference.
Even so, there's another problem, probably the most important: synchronization with the flash; its pulse must be precisely timed for that window, and perhaps that's not possible with the current P-TTL. There are examples where a higher speed sync can be attained only with certain flashes.

Faster traveling curtains would surely raise problems with the shutter vibrations and noise. Olympus cameras, for instance, have shutter blur issues (and firmware workarounds); and our K-7 was analyzed for similar issues.

[joergens.mi]

The noise of a faster system I've forgotten. An I'm with you cfraz I like a quiet camera like the K-5. and prefer low noise against higher sync speed

The shutter must be fully open when the flash fires. and the sync times from pressing the trigger to fire the flash and the Shutter full open must be highly snchronised. to set the time to the second curtain is an easy job than.

The first curtain must be fully down and the second must stay still (there is some time jitter for this moment because its mechanik and it must work from -10 to +40°C. )
The flash must be fired that the full light is near to that moment.
Flash ist of
The second curtin must run down.

That isn't so easy when all must fit completely within 4 MilliSeconds