[rawr]

So far it looks to me (in sweeping through this 17 page thread) that the evidence for Movie SR being electronic, and the only SR option available on the K-3, is limited to assessing video examples from just one pre-production camera. And from another thread where K-3 image samples from the same site were discussed, I think the firmware involved was as early as v0.35.

Unless I have missed something, that's not much to go on in drawing conclusions about how the SR in K-3 video will work and look in a production camera.

---

Last edited by rawr; 10-21-2013 at 12:52 AM.

[repaap]

I tend to shoot with monopod anyway if i plan to shoot video. Even tripod. So how does this affect my shooting?

[Ayoh]

It affects your shooting regardless because traditionally with Pentax software SR the sensor image remains permanently cropped even if SR is off. As such you will lose fov on all wide angle lenses.

[Steve.Ledger]

Ayoh are you implying that with mechanical SR the sensor image isn't cropped?

[repaap]

you seem to know a whole lot about this camera, even that there is quite few examples out there with 1.00 firmware.

[Rondec]

Ayoh are you implying that with mechanical SR the sensor image isn't cropped?

That is true. If you shoot video with the K5, whether or not the SR is on, the image is not cropped, while on the K-01 and K30 it is always cropped, whether or not the SR is on. But the mechanical SR provides a much better appearing video result.

[Steve.Ledger]

So, tell me - how does the K-5 shoot 1080p or 720p if the sensor image is not cropped?

[StigVidar]

So, tell me - how does the K-5 shoot 1080p or 720p if the sensor image is not cropped?

K-5 does have approx. 5000 pixels on the long side. When shooting 1080p video, every 5th pixel is used for video and the remaing pixels is ignored.

[Steve.Ledger]

K-5 does have approx. 5000 pixels on the long side. When shooting 1080p video, every 5th pixel is used for video and the remaing pixels is ignored.

1080p refers to the short side.
And where do you get your precise information from?

[Class A]

The change to a bad implementation of software SR really does defy all logic.

I see where you are coming from, but at least it could be argued that the software SR is silent (whereas the hardware SR isn't).

What defies all logic to me is that they made the button layout on the D-BG5 different to that on the K-3. This is a joke, only its not funny. I get that they don't have the width on the grip to exactly replicate the camera layout, but instead of moving just one of the buttons, all three (AF, AE-L, green) move around relative to the rear wheel.

[Steve.Ledger]

The hardware SR is silent on the K-7. Are you saying it makes a noise in the K-5?

And talking about the grip button placement is off-topic I think.

[johnmflores]

Is there any real need to use SR [at all] in movie mode? In camera SR is always a compromise whether it's mechanical or not.

Here's an example of how SR is useful. The camera on the outside of the car is a Pentax Q. The camera on the inside of the car is a Panasonic LX-7, also with image stabilization.

Vimeo

The outside shots are nice and stable despite the movement and the road surface irregularities. Fast forward to 3:40 and look in the upper right corner by the logo and you will see the top of the car's B-pillar shift left and right as SR thinks that it needs to compensate for the bend in the road.

And here's an example of the K-01

Vimeo

This is handheld with the DA21 (at ISO3200 in low light BTW) and the K-01 and SR did an admirable job.

In both of these cases, SR (mechanical or digital or otherwise) saved my hide. This footage would have been all but unusable without SR, and I'll have to wait and see how the K-3 performs. The thing about video is that it is so exacting; one minor flaw at a critical moment can ruin a shot.

[StigVidar]

1080p refers to the short side.
And where do you get your precise information from?

Of course, my bad. Every 2 or 3 pixels then.

But what do you want to know, really? I have long experience with video from K-5 and K-01 and it is easy to see that the video has a narrower field of view on K-01 than K-5.

Both cameras do a sub-sampling on the sensor image when doing video, but K-01 do it from a smaller area than K-5.

[Steve.Ledger]

I have long experience with video from K-5 and K-01 and it is easy to see that the video has a narrower field of view on K-01 than K-5..

That's going to depend on your lens. 1920x1080 is 1920x1080

@johnmflores
I'm not entirely convinced tbh because we've been using GoPros on the outside of the cars my son and his mate race on bitumen and dirt tracks for a few years now. The GPs have no SR but the camera is mounted on a decent quality (Camzilla) twin suction mount. There is no crazy bumping and shaking at all. Perfectly acceptable footage.

You Tube

You Tube

[Rondec]

I see where you are coming from, but at least it could be argued that the software SR is silent (whereas the hardware SR isn't).

What defies all logic to me is that they made the button layout on the D-BG5 different to that on the K-3. This is a joke, only its not funny. I get that they don't have the width on the grip to exactly replicate the camera layout, but instead of moving just one of the buttons, all three (AF, AE-L, green) move around relative to the rear wheel.

The hardware SR does make noise, but anyone who is actually shooting film won't be using audio from the built in mic -- they'll either have an external mic (in which case you don't hear the SR noise) or a completely separate recording device. I don't know otherwise, except to say that to me the K-01 video footage with SR is almost unusable for anything, while the k5 is pretty decent, even with SR on.