Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
12-04-2013, 06:48 AM   #61
Forum Member




Join Date: Jul 2013
Location: Oslo
Photos: Gallery | Albums
Posts: 63
H264 is a great codec for showing finished video but it's almost hopeless for editing as there's almost nothing to work with. I know this is already a problem in the video business with the Canon 5D. If you get the right footage it looks great but if you want to edit and color correct you have a problem.

12-04-2013, 07:13 AM   #62
Site Supporter
klkitchens's Avatar

Join Date: Oct 2011
Location: Allanta, GA
Photos: Gallery | Albums
Posts: 398
QuoteOriginally posted by Kunzite Quote
Is there anyone here making movies while moving the camera like in the second video?
Only Michael Bay.
12-04-2013, 10:57 AM   #63
Veteran Member




Join Date: Sep 2013
Location: Hamburg
Posts: 301
kadajawi wrote:
Is the h264 implementation really that bad?
I thought you have a K-5, don't you ? Here, we two can only speculate ...
How would it fare with the tree sample that I've uploaded recently?
Which trees on which movie do you mean ?
The one going down the steps ?

I can see it fall apart a bit cause I only used 60 Mbps instead of 80 Mbps
I get 25 Mbps in h264, any lower than that and it will look bad.
And that is with a much more sophisticated h264 encoder.
Ah ... you just talked about your transcoding of your K-5 MJPEG-footage.
Always use the highest rates possible of course - allways - because much gets
lost on YouTube anyway ... at least when you want a good look of your videos.
Otherwis we cannot discuss the quality practically ....

Pentax should have a setting that allows for 50 Mbps h264 videos minimum...
h264 is more efficient than MJPEG...
Yes. 50 MBit is minimum in broadcast anyway ... and in quality to !
I thought that the lack of SR is a bigger issue though...
That's something else we can't see being worked on yet as well ... true.
Sadly the K-5 isn't perfect either.. Then switching to video ... switching back and
forth between stills and video is just annoying.
True. With Canon 60D you can just push the trigger for photo as usual -
you just have
to change the exposure time then from 50 to your choice
And no focus peaking during video recording is just sad.
Well, do you mean some coloured contrast at the edges ?
A good sharp display is enough when you zoom in with the tele -
but only with parfocal lenses of course. Here the customers do ask
for autofocus which is not needed for prosumers anyway - so here you
do ask for a second path towards sharpness ... well, may be a good way ?

12-04-2013, 11:11 AM   #64
Veteran Member




Join Date: Sep 2013
Location: Hamburg
Posts: 301
QuoteOriginally posted by ArneTorp Quote
H264 is a great codec for showing finished video but it's almost hopeless for editing as there's almost nothing to work with. I know this is already a problem in the video business with the Canon 5D. If you get the right footage it looks great but if you want to edit and color correct you have a problem.
I heard that from a professional broadcast cutter as well, also he says those clips are
full of dirty artefacts ... I think this may be just a little snobbish even if true - because
if you look at the many shots used in broadcasting with enough quality left - this seems
to be quite a high expectation. And they all use Canon's full-frame-artists very often now.

But OK - Stations are differenciating here between daily specifications wanted - and footage
serving some special need for optical look ... and also small equipment in special situations.

12-04-2013, 11:38 AM   #65
osv
Pentaxian




Join Date: Mar 2007
Location: So Cal
Photos: Gallery
Posts: 2,080
QuoteOriginally posted by kadajawi Quote
I'm using the latest 10 bit version of x264, at the very highest setting. Constant quality rather than VBR, cause some scenes can easily get away at 2 Mbps, while others need 25 (usually I get 6-8, though that also depends on the lens. The 50mm usually gets smaller files). Only setting the bitrate doesn't make sense to me. I do a color denoise prior to encoding on all files (pretty much like what Lightroom does per default), and a mild denoise in general. If you have any suggestion on where to upload the file I uploaded to YouTube I will do so.
what i would do is to create a new youtube account, just for testing... put the text/graphics on every upload, and keep working with the encoding process and bitrate until those edges stay clean.

vbr is what gives you the small file sizes, and i think that youtube will accept it as an upload.

one thing that really kills the quality is making those multiple re-encoding passes through the media... it should come straight off of the editing timeline into the final compressed upload; the graphics, denoise, etc., is done all at once.

with lightroom you are working with raw files, so it's a non-destructive edit, but with h.264 video, everything you do as a separate process destroys the quality... only re-encode it once.

i really like the editing work on that parkour video, but it has major issues:
1) 25fps is too low of a framerate for anything, especially an action video like that, the stuttering was terrible... a major mistake.
2) it's only 720p, he should have given it the full frame size... quality suffers greatly when you reset the frame size during the encode.

in case you are curious, here are the stats for the the vimeo-encoded version of that video, i downloaded it:

ID : 1
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L3.1
Format settings, CABAC : Yes
Format settings, ReFrames : 4 frames
Codec ID : avc1
Codec ID/Info : Advanced Video Coding
Duration : 4mn 57s
Bit rate : 2 500 Kbps
Maximum bit rate : 9 643 Kbps
Width : 1 280 pixels
Height : 720 pixels
Display aspect ratio : 16:9
Frame rate mode : Constant
Frame rate : 25.000 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive
Bits/(Pixel*Frame) : 0.109
Stream size : 88.6 MiB (94%)
Writing library : x264 core 128 r2216 198a7ea
Encoding settings : cabac=1 / ref=3 / deblock=1:0:0 / analyse=0x3:0x113 / me=hex / subme=7 / psy=1 / psy_rd=1.00:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 / trellis=1 / 8x8dct=1 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 / threads=12 / lookahead_threads=2 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 / constrained_intra=0 / bframes=3 / b_pyramid=2 / b_adapt=1 / b_bias=0 / direct=1 / weightb=1 / open_gop=0 / weightp=2 / keyint=75 / keyint_min=25 / scenecut=40 / intra_refresh=0 / rc_lookahead=40 / rc=2pass / mbtree=1 / bitrate=2500 / ratetol=1.0 / qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / cplxblur=20.0 / qblur=0.5 / vbv_maxrate=3750 / vbv_bufsize=12500 / nal_hrd=none / ip_ratio=1.40 / aq=1:1.00
Encoded date : UTC 2012-10-10 00:51:58
Tagged date : UTC 2012-10-10 00:51:59
12-04-2013, 11:59 AM   #66
osv
Pentaxian




Join Date: Mar 2007
Location: So Cal
Photos: Gallery
Posts: 2,080
QuoteOriginally posted by ArneTorp Quote
H264 is a great codec for showing finished video but it's almost hopeless for editing as there's almost nothing to work with. I know this is already a problem in the video business with the Canon 5D. If you get the right footage it looks great but if you want to edit and color correct you have a problem.
the way that sort of thing has been handled for decades is to re-encode it to a lossless or semi-lossless format, and then edit it... but since h.264 is only 4.2.0(?), it's going to be crippled from the start.

the problems with dslr video quality have been around since the platform first came out; the bbc refused to broadcast the footage from those early canon dslr's, and even as late as a year ago, the bbc refused to accept the nikon d800 video on it's networks:

"800... The test procedures were as described and recommended by the EBU, in Tech 3335 (http://tech.ebu.ch/docs/tech/t.... Video performance is not really acceptable at 1080p, much less so at 720p. Even though the sensor has 36.8 million photo-sites, it achieves only about 1,355x764, little better than 1,280x720.
Many thanks
Andy Quested
Head of Technology BBC HD & 3D"
Nikon DSLRs 'Pass' BBC Broadcast Test, iPhone Used for Live BBC Interview
12-04-2013, 12:18 PM - 1 Like   #67
Pentaxian
ScooterMaxi Jim's Avatar

Join Date: Jul 2009
Location: Minnesota
Posts: 1,464
Pan back and forth quickly on any CMOS camera and you will get rolling shutter - fact of life. That is why instructions expressly indicate you need to turn off SR. Posting videos of misusing the camera proves nothing.

This isn't to excuse the fact that the K-0130/50 video will show jello even with gentle side-to-side movement with SR on, and generally perform better with SR off. Yes, we need to know if the situation has improved at all - and I understand that is the point of the thread.
12-04-2013, 12:41 PM   #68
Pentaxian




Join Date: Sep 2010
Posts: 2,609
Yes, I only have a K-5.


I meant the one going down the stairs, with tons of trees all around. The one that fell apart so horribly on YouTube.


I might try senselessly increasing the bitrate of my video so that YouTube thinks it deserves higher quality, but I can only say what I uploaded looked great.


Again, 6 Mbps can look great... if you give it plenty of time and power to encode. If the encoder is able to make use of all the very sophisticated features of h264, then the files will be tiny and the quality high. But that is just not possible in camera.


Yes, more contrast at the edges. I've tried it on the a57, and it works GREAT. Having a sharp display is nice... but... we don't have that. If the screen had FullHD resolution, and enough size so that it actually makes sense, then yes, it'd be fine. But something can be totally sharp on the screen, or even in 720p, but once you look at it in 1080p it will be out of focus. Focus peaking tells you exactly which part of the video is in focus, without having to squint or anything. It's easily visible. Yes, zooming in, focusing, zooming out works... unless you want to shoot zoomed in, or you/your motive is moving, you want to focus on something else, etc.


@ScooterMaxi Jim: My point is that with the SR system in the K-5, and some care, you can surpress lots of jello, and you get reasonably stable video, without any gear to stabilize the camera (and if you have that gear it will be even better). The K-5 SR system has few downsides... you have to be a bit more careful with pans (something Pentax could fix), and you MIGHT pick up some noises when it is quiet. Maybe shorter battery life too. That's it, that's all I can think of. It works, as I think my videos have shown. And if the K-5 could do it, and the SR system of the K-3 is even better... why didn't Pentax let us use it during video shooting? The software implementations are nowhere near as good... and instructions rightfully indicate you should turn off SR. Because electronic SR just doesn't work well. Mechanical SR does work though, and Pentax should have used it. Btw., the K-01/30/50 show jello even with no movement with SR on...


These are my settings btw., taken from the shot that was so problematic on YouTube.


ID : 1
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High 10@L5.1
Format settings, CABAC : Yes
Format settings, ReFrames : 16 frames
Codec ID : avc1
Codec ID/Info : Advanced Video Coding
Duration : 42s 0ms
Bit rate : 24.7 Mbps
Maximum bit rate : 36.7 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate mode : Constant
Frame rate : 25.000 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 10 bits
Scan type : Progressive
Bits/(Pixel*Frame) : 0.476
Stream size : 124 MiB (100%)
Writing library : x264 core 140 r2377 1ca7bb9
Encoding settings : cabac=1 / ref=16 / deblock=1:-1:-1 / analyse=0x3:0x133 / me=tesa / subme=11 / psy=1 / psy_rd=1.00:0.15 / mixed_ref=1 / me_range=24 / chroma_me=1 / trellis=2 / 8x8dct=1 / cqm=0 / deadzone=21,11 / fast_pskip=0 / chroma_qp_offset=-3 / threads=6 / lookahead_threads=1 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 / constrained_intra=0 / bframes=16 / b_pyramid=2 / b_adapt=2 / b_bias=0 / direct=3 / weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=25 / scenecut=40 / intra_refresh=0 / rc_lookahead=60 / rc=crf / mbtree=1 / crf=20.5 / qcomp=0.65 / qpmin=0 / qpmax=81 / qpstep=4 / ip_ratio=1.40 / aq=2:1.00


As for the parkour video... I think he shot it at 1080p, which explains the 25p. Also, it gives it a filmic look, doesn't it? I'm one of those who actually prefered The Hobbit in 24p rather than 48p...

12-04-2013, 01:41 PM   #69
Pentaxian




Join Date: Oct 2009
Location: Far North Qld
Posts: 3,250
Original Poster
QuoteOriginally posted by ScooterMaxi Jim Quote
Yes, we need to know if the situation has improved at all - and I understand that is the point of the thread.
Thank you..
12-04-2013, 01:45 PM   #70
Veteran Member




Join Date: Feb 2013
Photos: Gallery
Posts: 917
QuoteOriginally posted by ScooterMaxi Jim Quote
Yes, we need to know if the situation has improved at all
it has. but not sure to what extent it has become useful yet...
12-06-2013, 05:21 PM   #71
Pentaxian
ScooterMaxi Jim's Avatar

Join Date: Jul 2009
Location: Minnesota
Posts: 1,464
QuoteOriginally posted by kadajawi Quote
Yes, I only have a K-5.


@ScooterMaxi Jim: My point is that with the SR system in the K-5, and some care, you can surpress lots of jello, and you get reasonably stable video, without any gear to stabilize the camera (and if you have that gear it will be even better). The K-5 SR system has few downsides... you have to be a bit more careful with pans (something Pentax could fix), and you MIGHT pick up some noises when it is quiet. Maybe shorter battery life too. That's it, that's all I can think of. It works, as I think my videos have shown. And if the K-5 could do it, and the SR system of the K-3 is even better... why didn't Pentax let us use it during video shooting? The software implementations are nowhere near as good... and instructions rightfully indicate you should turn off SR. Because electronic SR just doesn't work well. Mechanical SR does work though, and Pentax should have used it. Btw., the K-01/30/50 show jello even with no movement with SR on...
No doubt about it - frustrating situation. I think the K3 more or less took the best from the two cameras, and improved video - including the headphone jack (pretty basic stuff). The K-01 has the same battery as the K5 - good life and no overheating. The K-01 has terrible compression compared to the K5; K-01 has no high end and lots of AGC noise in silent passages. You have a mixed bad no matter which way you go; the K3 had to get better - and did it appears.
12-06-2013, 08:53 PM   #72
Pentaxian




Join Date: Sep 2010
Posts: 2,609
QuoteOriginally posted by ScooterMaxi Jim Quote
No doubt about it - frustrating situation. I think the K3 more or less took the best from the two cameras, and improved video - including the headphone jack (pretty basic stuff). The K-01 has the same battery as the K5 - good life and no overheating. The K-01 has terrible compression compared to the K5; K-01 has no high end and lots of AGC noise in silent passages. You have a mixed bad no matter which way you go; the K3 had to get better - and did it appears.
I don't think the K-3 took much from the K-5. Certainly not the superior bitrates, or the mechanical SR. If it had say 50 Mbps h264, perhaps 80 Mbps at 4:4:4 for those interested in grading), and the mechanical SR from the K-5, perhaps in an improved version that recognizes panning, then the K-3 would be on the forefront of video DSLRs and might win videographers from other brands. Those are probably two rather simple additions... and if the adjustable aliasing filter would work during video it would be perfect. (Ok, there is still pixel binning). But Pentax would have a good chance with those additions.

Here's hoping that Ricoh will do a better job at adding features after the release...
12-07-2013, 09:54 AM   #73
Pentaxian
ScooterMaxi Jim's Avatar

Join Date: Jul 2009
Location: Minnesota
Posts: 1,464
I'm not sure any of that can be accomplished through a FW change. It is quite clear that Pentax designs its processing engine with stills in mind, first and foremost.

You are right that little was taken from the K5 (you certainly would rather have compressed format - if only executed less aggressively, you'd want focus peaking even if only before you press the shutter button, you'd want to be able to adjust mic in levels, resistance to overheating, etc.). What it does take from the K5 in terms of performance (not hardware) is slightly better resistance to rolling shutter - but mechanical SR obviously would have been a better solution to take from the K5.

I'd be hard pressed to choose between the K-01 and K5 for video - as I really do like the video quality of the latter. However, I'd have to go with the K-01 overall. Not that I'm going to get a K-3 for video (or at all, most likely), but the K-3 choice would easiest of all.
12-07-2013, 09:18 PM   #74
Pentaxian




Join Date: Sep 2010
Posts: 2,609
QuoteOriginally posted by ScooterMaxi Jim Quote
I'm not sure any of that can be accomplished through a FW change. It is quite clear that Pentax designs its processing engine with stills in mind, first and foremost.

You are right that little was taken from the K5 (you certainly would rather have compressed format - if only executed less aggressively, you'd want focus peaking even if only before you press the shutter button, you'd want to be able to adjust mic in levels, resistance to overheating, etc.). What it does take from the K5 in terms of performance (not hardware) is slightly better resistance to rolling shutter - but mechanical SR obviously would have been a better solution to take from the K5.

I'd be hard pressed to choose between the K-01 and K5 for video - as I really do like the video quality of the latter. However, I'd have to go with the K-01 overall. Not that I'm going to get a K-3 for video (or at all, most likely), but the K-3 choice would easiest of all.
Focus peaking before pressing the shutter is a bit useless, isn't it? I might as well use AF for that... what I need is assistance when I have to focus manually. Especially at smaller apertures/with wide angle lenses where a lot is more or less in focus you'll run into issues in getting the focus just right. I find it easier to focus using the screen when I have a shallow DoF... cause I can see the area not in focus being clearly not in focus, and I suppose I can get away with it being not perfectly right. With a large DoF you can see that the focus point was behind the motiv... and it disturbs me, even the motiv is sharper than I would have gotten it with a shallow DoF. But there it is the sharpest point in the scene...


Between the K-01 and K-5 I'd easily go for the K-5. Better video, better ergonomics, better stills. K-5 and K-3 would be more difficult... I'd like it for stills (that AF...), but video is too much of a let down, so it'd probably be a K-5.
12-07-2013, 11:51 PM   #75
Pentaxian
ScooterMaxi Jim's Avatar

Join Date: Jul 2009
Location: Minnesota
Posts: 1,464
QuoteOriginally posted by kadajawi Quote
Focus peaking before pressing the shutter is a bit useless, isn't it? I might as well use AF for that... what I need is assistance when I have to focus manually. Especially at smaller apertures/with wide angle lenses where a lot is more or less in focus you'll run into issues in getting the focus just right. I find it easier to focus using the screen when I have a shallow DoF... cause I can see the area not in focus being clearly not in focus, and I suppose I can get away with it being not perfectly right. With a large DoF you can see that the focus point was behind the motiv... and it disturbs me, even the motiv is sharper than I would have gotten it with a shallow DoF. But there it is the sharpest point in the scene...


Between the K-01 and K-5 I'd easily go for the K-5. Better video, better ergonomics, better stills. K-5 and K-3 would be more difficult... I'd like it for stills (that AF...), but video is too much of a let down, so it'd probably be a K-5.
I guess it would come down to the kind of shooting you do, and even more so the lenses available to you. My better, faster lenses that are properly set up for video are all MF - and high quality lenses bring out more with focus peaking than a lot of consumer stuff. With my MF-oriented lens arsenal (on the Pentax side), the K-30 and K-01 are superior to the K5 in terms of getting stills keepers. I would much prefer the K5 for video action, but that is a rare need for me. Not sure that if I even had fast, AF lenses suitable for video that I would be shooting AF - unless it was an unusual action scene, and even then it would take a lot of practice to adjust. There are professionals using the K-01 for video work - a bit surprising, but an indication that it isn't all that bad. In my opinion, neither the K5 nor the K-01 are up to broadcast quality, but both would be adequate for industrial video work (low end pro).
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
dslr, jello, k-01, k-3, k3, pentax, pentax k-3, youtube
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
A tale of 3 sensors K-3/K-01/K-7 Painter Pentax K-3 21 11-16-2013 09:38 PM
Pentax K-3 v/s the competition Deedee Pentax K-3 3 10-08-2013 11:35 AM
New Firmware for K-500, K-50, K-5 II, K-5 IIs, K-5, K-30, K-01, K-r, and 645D is out Adam Pentax News and Rumors 60 09-16-2013 05:46 AM
K-01 v K-05 shutter sound rob_k20d Pentax K-01 12 08-05-2013 05:02 AM
40mm xs v 40mm ltd on K-01 texasgirl Pentax K-01 9 03-15-2012 12:10 PM



All times are GMT -7. The time now is 10:11 PM. | See also: NikonForums.com, part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top