Pentax K-3 Video Discussion

Tanzer · 11-02-2013, 06:37 AM

Well they (C/N) use in-lens stabilization.

panoguy · 11-02-2013, 08:06 AM

Originally posted by Tanzer

I'm still hopeful someone will soon comment further on the details of that implementation, such as AF speed, hunting, AF point indication, etc, as well as the quality/responsiveness of the focus peaking.

AF speed during video is slooooow, whether with screwdrive, SDM, or Sigma HSM lenses. Press and hold the AF button and it refocuses a tiny bit, pause, a tiny bit more, pause, a long sloooow refocus. Better to manually pull the focus (at least with this firmware 1.00)! Oddly enough, AF during live view (not recording) is speedy quick and has lots of CDAF options!

~~Focus peaking is active during video recording~~ (Not active, I was using a bad test subject), but is the same "white shimmer" in other Pentax bodies (or at least the same as my Q), which is not all that easy to see sometimes. The responsiveness of the peaking and the video display in general (on the back LCD) seems much faster than my K-5 or Q. Not much lag as you move around... but then I'm not a video expert.

One nice thing I don't remember from earlier bodies is that the live view display (in photo or video) updates the gain when you use Exp. Comp to show the approximate exposure results. Previously, I think the live view would always "gain up" to show details, but not the actual exposure.

Azzy · 11-02-2013, 08:27 AM

Originally posted by panoguy

AF speed during video is slooooow, whether with screwdrive, SDM, or Sigma HSM lenses. Press and hold the AF button and it refocuses a tiny bit, pause, a tiny bit more, pause, a long sloooow refocus. Better to manually pull the focus (at least with this firmware 1.00)! Oddly enough, AF during live view (not recording) is speedy quick and has lots of CDAF options!

Focus peaking is active during video recording, but is the same "white shimmer" in other Pentax bodies (or at least the same as my Q), which is not all that easy to see sometimes. The responsiveness of the peaking and the video display in general (on the back LCD) seems much faster than my K-5 or Q. Not much lag as you move around... but then I'm not a video expert.

One nice thing I don't remember from earlier bodies is that the live view display (in photo or video) updates the gain when you use Exp. Comp to show the approximate exposure results. Previously, I think the live view would always "gain up" to show details, but not the actual exposure.

weird I don't see focus peeking during video recording. It works before you press record but as soon as I start recording the focus peaking dissapeared (yes I've enabled it by pressing info)

panoguy · 11-02-2013, 08:37 AM

Originally posted by Azzy

weird I don't see focus peeking during video recording. It works before you press record but as soon as I start recording the focus peaking dissapeared (yes I've enabled it by pressing info)

Aha, you're right! I chose a bad subject for my peaking tests... a finely woven piece of fabric that happened to be laying next to my desk. (I don't really shoot much video) Apparently, I was seeing moiré on the back LCD as I changed focus during video recording (how's that for ironic?). When I tested it again just now by focusing on my keyboard, the peaking shimmer disappears when I press record, just as you've seen.

Sorry about getting folks hopes up! Let's lobby Ricoh for a Firmware 1.01 with mechanical SR and FP during video recording!

**7samurai** · 11-02-2013, 04:05 PM

Originally posted by Tanzer

Well they (C/N) use in-lens stabilization.

totally forgot about those....

**7samurai** · 11-02-2013, 04:48 PM

Originally posted by panoguy

Let's lobby Ricoh for a Firmware 1.01 with mechanical SR and FP during video recording!

I'm hoping they will fix it but I'm not optimistic. I was riding such a great "High" when they release the K-3. Now i'm totally depressed regarding the video. I know I can just buy a dedicated video recorder, but I was hoping they had at least caught up a little. This is K-01 video quality. No improvement at all. This might not be fixable in firmware if it's a heat issue with the SR when recording video. (Edit)

Pheo · 11-02-2013, 07:28 PM

It looks to me like they have optimised the SF during video to minimize the noise and disruption to picture.

I must admit I'm really not getting why everyone is so fussed about AF during video, for commercial uses etc it's not really needed IMO.

Audio metering etc during record seems to work well. Need to try the audio pass through to headphones properly tomorrow

falconeye · 11-03-2013, 02:30 AM

Originally posted by Pheo

I must admit I'm really not getting why everyone is so fussed about AF during video, for commercial uses etc it's not really needed IMO.

K-3 will not be used for commercial video.

Video is a supporting role for dSLRs (except maybe the four cameras which are selected by commercial film makers). Supporting role as in family videos, event coverage, social media content. Where people don't bring their rigs, rails and focus pullers.

So, decent control over jello effect and sampling artefacts, decent stabilization, silent and snappy AF in video as well as peaking, are all requirements for this supporting role.

Of course, K-3 can still shine for its major role, photography

Class A · 11-03-2013, 02:52 AM

Originally posted by falconeye

So, decent control over jello effect and sampling artefacts, decent stabilization, silent and snappy AF in video as well as peaking, are all requirements for this supporting role.

I agree.

Do you have any idea why they may have went backwards compared to the SR implementation on the K-5?

The only reasons I can think of are

quieter operation, and
increasing battery life.

However, I feel that the price paid in terms of IQ is too high. Perhaps there are plans to improve the software-based SR over time.

**Mutters** · 11-03-2013, 03:23 AM

Originally posted by panoguy

Sorry about getting folks hopes up! Let's lobby Ricoh for a Firmware 1.01 with mechanical SR and FP during video recording!

I agree - now where have I seen that before.

I find it quite hilarious, and disturbing, that people are saying that no-one will use the K-3 for pro video.

Funny that. Must stop billing for my work then.

AB

Pheo · 11-03-2013, 03:38 AM

Originally posted by Mutters

I agree - now where have I seen that before.

I find it quite hilarious, and disturbing, that people are saying that no-one will use the K-3 for pro video.

Funny that. Must stop billing for my work then.

AB

That's what I was thinking!

**Mutters** · 11-03-2013, 03:48 AM

Originally posted by Pheo

That's what I was thinking!

Ha! Excellent. I was wondering how many are in the same boat.

Petition time for videographers and their K-3s.

AB

jaad75 · 11-03-2013, 04:59 AM

Originally posted by falconeye

silent and snappy AF in video as well as peaking, are all requirements for this supporting role.

Definitely not AF. Especially "snappy" AF.

falconeye · 11-03-2013, 07:47 AM

Originally posted by Class A

Do you have any idea why they may have went backwards compared to the SR implementation on the K-5?

There are many possible causes, e.g., management, software engineering team fluctuation, and one technical reason:

If they made the margins smaller, the max. amount the sensor is allowed to travel. Still image stabilization does not require as much margin as the K-5 has had.

However, I hear that LV still behaves as it did before, i.e., stabilizes the LV image. So, I don't know really.

**7samurai** · 11-03-2013, 07:48 AM

Originally posted by Class A

The only reasons I can think of are

quieter operation, and
increasing battery life.

I agree, it's strange that they would disable a K-5 feature for quieter operation and increase battery life.
External mic and a spare battery is not a big issue. And if it's an issue for some, then make Digital and Mechanical stabilization an option in the menu.

Of course we're just speculating here.

11-02-2013, 08:06 AM	#317
panoguy Pentaxian Join Date: Sep 2007 Location: Washington, D.C. Photos: Gallery \| Albums Posts: 3,327	Originally posted by Tanzer I'm still hopeful someone will soon comment further on the details of that implementation, such as AF speed, hunting, AF point indication, etc, as well as the quality/responsiveness of the focus peaking. AF speed during video is slooooow, whether with screwdrive, SDM, or Sigma HSM lenses. Press and hold the AF button and it refocuses a tiny bit, pause, a tiny bit more, pause, a long sloooow refocus. Better to manually pull the focus (at least with this firmware 1.00)! Oddly enough, AF during live view (not recording) is speedy quick and has lots of CDAF options! ~~Focus peaking is active during video recording~~ (Not active, I was using a bad test subject), but is the same "white shimmer" in other Pentax bodies (or at least the same as my Q), which is not all that easy to see sometimes. The responsiveness of the peaking and the video display in general (on the back LCD) seems much faster than my K-5 or Q. Not much lag as you move around... but then I'm not a video expert. One nice thing I don't remember from earlier bodies is that the live view display (in photo or video) updates the gain when you use Exp. Comp to show the approximate exposure results. Previously, I think the live view would always "gain up" to show details, but not the actual exposure. Last edited by panoguy; 11-02-2013 at 08:34 AM.

11-03-2013, 02:52 AM	#324
Class A Pentaxian Join Date: Aug 2008 Location: Wellington, New Zealand Posts: 11,251	Originally posted by falconeye So, decent control over jello effect and sampling artefacts, decent stabilization, silent and snappy AF in video as well as peaking, are all requirements for this supporting role. I agree. Do you have any idea why they may have went backwards compared to the SR implementation on the K-5? The only reasons I can think of are quieter operation, and increasing battery life. However, I feel that the price paid in terms of IQ is too high. Perhaps there are plans to improve the software-based SR over time.

11-03-2013, 07:48 AM	#330
7samurai Site Supporter Join Date: Apr 2010 Location: Rancho Cucamonga, CA Photos: Albums Posts: 474	Originally posted by Class A The only reasons I can think of are quieter operation, and increasing battery life. I agree, it's strange that they would disable a K-5 feature for quieter operation and increase battery life. External mic and a spare battery is not a big issue. And if it's an issue for some, then make Digital and Mechanical stabilization an option in the menu. Of course we're just speculating here.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Now that official K-3 specs are released, how good for video?	jake14mw	Video Recording and Processing	20	10-14-2013 12:56 PM
Pentax Optio WG-3 video download challenge	Eric Z.	Troubleshooting and Beginner Help	2	09-06-2013 06:33 PM
With HDMI out & 3.5mm stereo mic in the K-01 is better than the K-30 for video	jogiba	Pentax K-01	2	05-22-2012 01:52 PM
Spot one! robertstech.com discussion on K-01	hcc	Pentax K-01	18	02-08-2012 07:22 AM
Short Movie - Robot Warriors + Discussion on DSLR for Video	tiny	Video Recording and Processing	2	06-21-2010 12:29 PM

11-02-2013, 06:37 AM	#316
Tanzer Veteran Member Join Date: Mar 2012 Location: Perrineville, NJ Posts: 1,375	Well they (C/N) use in-lens stabilization.

11-02-2013, 08:27 AM	#318
Azzy Senior Member Join Date: Jan 2010 Photos: Albums Posts: 201	Originally posted by panoguy AF speed during video is slooooow, whether with screwdrive, SDM, or Sigma HSM lenses. Press and hold the AF button and it refocuses a tiny bit, pause, a tiny bit more, pause, a long sloooow refocus. Better to manually pull the focus (at least with this firmware 1.00)! Oddly enough, AF during live view (not recording) is speedy quick and has lots of CDAF options! Focus peaking is active during video recording, but is the same "white shimmer" in other Pentax bodies (or at least the same as my Q), which is not all that easy to see sometimes. The responsiveness of the peaking and the video display in general (on the back LCD) seems much faster than my K-5 or Q. Not much lag as you move around... but then I'm not a video expert. One nice thing I don't remember from earlier bodies is that the live view display (in photo or video) updates the gain when you use Exp. Comp to show the approximate exposure results. Previously, I think the live view would always "gain up" to show details, but not the actual exposure. weird I don't see focus peeking during video recording. It works before you press record but as soon as I start recording the focus peaking dissapeared (yes I've enabled it by pressing info)

11-02-2013, 08:37 AM	#319
panoguy Pentaxian Join Date: Sep 2007 Location: Washington, D.C. Photos: Gallery \| Albums Posts: 3,327	Originally posted by Azzy weird I don't see focus peeking during video recording. It works before you press record but as soon as I start recording the focus peaking dissapeared (yes I've enabled it by pressing info) Aha, you're right! I chose a bad subject for my peaking tests... a finely woven piece of fabric that happened to be laying next to my desk. (I don't really shoot much video) Apparently, I was seeing moiré on the back LCD as I changed focus during video recording (how's that for ironic?). When I tested it again just now by focusing on my keyboard, the peaking shimmer disappears when I press record, just as you've seen. Sorry about getting folks hopes up! Let's lobby Ricoh for a Firmware 1.01 with mechanical SR and FP during video recording!

11-02-2013, 04:05 PM	#320
7samurai Site Supporter Join Date: Apr 2010 Location: Rancho Cucamonga, CA Photos: Albums Posts: 474	Originally posted by Tanzer Well they (C/N) use in-lens stabilization. totally forgot about those....

11-02-2013, 04:48 PM	#321
7samurai Site Supporter Join Date: Apr 2010 Location: Rancho Cucamonga, CA Photos: Albums Posts: 474	Originally posted by panoguy Let's lobby Ricoh for a Firmware 1.01 with mechanical SR and FP during video recording! I'm hoping they will fix it but I'm not optimistic. I was riding such a great "High" when they release the K-3. Now i'm totally depressed regarding the video. I know I can just buy a dedicated video recorder, but I was hoping they had at least caught up a little. This is K-01 video quality. No improvement at all. This might not be fixable in firmware if it's a heat issue with the SR when recording video. (Edit)

11-02-2013, 07:28 PM	#322
Pheo Veteran Member Join Date: Nov 2006 Location: Brighton, United Kingdom Photos: Gallery Posts: 311	It looks to me like they have optimised the SF during video to minimize the noise and disruption to picture. I must admit I'm really not getting why everyone is so fussed about AF during video, for commercial uses etc it's not really needed IMO. Audio metering etc during record seems to work well. Need to try the audio pass through to headphones properly tomorrow

11-03-2013, 02:30 AM	#323
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	Originally posted by Pheo I must admit I'm really not getting why everyone is so fussed about AF during video, for commercial uses etc it's not really needed IMO. K-3 will not be used for commercial video. Video is a supporting role for dSLRs (except maybe the four cameras which are selected by commercial film makers). Supporting role as in family videos, event coverage, social media content. Where people don't bring their rigs, rails and focus pullers. So, decent control over jello effect and sampling artefacts, decent stabilization, silent and snappy AF in video as well as peaking, are all requirements for this supporting role. Of course, K-3 can still shine for its major role, photography

11-03-2013, 03:23 AM	#325
Mutters Loyal Site Supporter Join Date: Aug 2007 Location: Melbourne, Australia Photos: Albums Posts: 262	Originally posted by panoguy Sorry about getting folks hopes up! Let's lobby Ricoh for a Firmware 1.01 with mechanical SR and FP during video recording! I agree - now where have I seen that before. I find it quite hilarious, and disturbing, that people are saying that no-one will use the K-3 for pro video. Funny that. Must stop billing for my work then. AB

11-03-2013, 03:38 AM	#326
Pheo Veteran Member Join Date: Nov 2006 Location: Brighton, United Kingdom Photos: Gallery Posts: 311	Originally posted by Mutters I agree - now where have I seen that before. I find it quite hilarious, and disturbing, that people are saying that no-one will use the K-3 for pro video. Funny that. Must stop billing for my work then. AB That's what I was thinking!

11-03-2013, 03:48 AM	#327
Mutters Loyal Site Supporter Join Date: Aug 2007 Location: Melbourne, Australia Photos: Albums Posts: 262	Originally posted by Pheo That's what I was thinking! Ha! Excellent. I was wondering how many are in the same boat. Petition time for videographers and their K-3s. AB

Thread Tools	Search this Thread
	Search this Thread: Advanced Search

11-03-2013, 04:59 AM	#328
jaad75 Veteran Member Join Date: Sep 2007 Location: Cracow Posts: 457	Originally posted by falconeye silent and snappy AF in video as well as peaking, are all requirements for this supporting role. Definitely not AF. Especially "snappy" AF.

11-03-2013, 07:47 AM	#329
falconeye Veteran Member Join Date: Jan 2008 Location: Munich, Alps, Germany Photos: Gallery Posts: 6,871	Originally posted by Class A Do you have any idea why they may have went backwards compared to the SR implementation on the K-5? There are many possible causes, e.g., management, software engineering team fluctuation, and one technical reason: If they made the margins smaller, the max. amount the sensor is allowed to travel. Still image stabilization does not require as much margin as the K-5 has had. However, I hear that LV still behaves as it did before, i.e., stabilizes the LV image. So, I don't know really.