[Eruditass]

I made some videos of burst mode on the K-x. I found out that JPEG burst mode buffer dependent on both the ISO and what is in the scene, while RAW and RAW+ are limited by write speed of the SDHC interface and doesn't change. Observe:

K-x JPEG ISO200:
YouTube - K x Burst JPEG ISO200

K-x JPEG ISO800:
YouTube - K x Burst JPEG ISO800

K-x JPEG ISO1600:
YouTube - K x Burst JPEG ISO1600

K-x JPEG ISO12800:
YouTube - K x Burst JPEG ISO12800

K-x RAW:
YouTube - K x Burst RAW

K-x RAW+:
YouTube - K x Burst RAW+

The pictures were taken of some kitchen glassware. When I put the lens cap on, at ISO200, I got excess of 40 shots (I'm not trying to kill my shutter life). At ISO12800, I still got 8 or so. So scene limits at lower ISO, hardware processor/amplifier limits at higher ISO. Around 1600 is where it starts to drop below 17 frames.

Noise Reduction: was left on default - starts at 800, and at medium.

File Size: They were all pretty much the same size, around 5 MB - I deleted them but I just took an ISO 200 and ISO 12800 shot in Av mode, same exposure:

200: 5174KB
12800: 5374KB

Now, the buffer sizes between the two came out to be 8 vs 30+. I don't think that played a role.

---

Last edited by Eruditass; 11-06-2009 at 02:01 PM.

[Eruditass]

Oh and I've gotten over 2000 photos at this point on this charge on my duracell white tops. Btw, many batteries are capable of this if you burst all day long instead of spreading it out over weeks.

Poor K-x shutter.. at least it is rated more than other DSLRs in this range.. 100K?

[er1kksen]

Wow, 100K? That's what the 40D I just replaced with a K-x was rated to. Nice to know I'm not downgrading in that respect. I wouldn't worry about running the shutter down.

[dragonfly]

This is full size, right? How does it perform if you choose a smaller file size (reduced # of pixels)?

[rawr]

Yep, 100K, according to the K-x press release:

The K-x also features a more reliable shutter unit, which passed a demanding action test of more than 100,000 shutter releases.

100K plus.

Let's see at 5fps, that means Pentax would have done stuff like run their test camera(s) on a test bench continuously or intermittently for a total of about 333 minutes, or 5 and a half hours, without failure (100,000 actuations/5 ps = 20,000 seconds total duration, then /60). Sounds pretty robust.

I wouldn't buy one of those test cameras 2nd hand though.

[Eruditass]

This is full size, right? How does it perform if you choose a smaller file size (reduced # of pixels)?

Yep, full size, 3 star.

Reducing the file size helps very slightly. At 2 MP and 1 star, RAW and RAW+ gain only 1 or 2 shots each, though the speed afterwards is a slight bit faster.

JPEG at ISO 12800 went up to 14 or so. From 12 MP, lowering to 10 and 6 gained a couple shots at ISO 12800. Lowering the star quality gained 1 or 2 over the lowering of the resolution while at 12 MP.

Ok, no more for my poor shutter. These are on 1/2 dead batteries but I did plug in new ones with minimal differences.

edit: these tests were with shadow comp all the way up. The OP tests were with it off. I don't think there is a big difference with that, but highlight comp effectively doubles the ISO so take that into account. No testing with Lens correction - I don't feel like finding a DA lens to put on.

[er1kksen]

As long as I can get the full 4.7 fps for about 5 frames in RAW, we're good. Is that how it works?

[Eruditass]

Yep, check out the video. The fps after it fills isn't as horrendously slow as previous models IIRC

[Eruditass]

Did some bursts with NR off, this makes ISO 1600 and below have buffers of somewhere like 30+. However, as I had read.. 3200+ ISO is actually done through software and the buffer size is around 8 even with NR off