Is the motion problem related to GOP vs I frame? I take it the external recorder did not help? Is there somthing that can be done in post to help? Does the FS700 have the same problem? Thank you for your time.
FS100 w/ Nanoflash Testing
- Thread starter Kholi
- Start date
Is the motion problem related to GOP vs I frame? I take it the external recorder did not help? Is there somthing that can be done in post to help? Does the FS700 have the same problem? Thank you for your time.
I don't think it's GOP related but possibly. Even before the GH2 hack (all the GOP changes), as long as you were shooting 1/40 it never bothered me. It does seem a little different now, but nothing crazy.
So perhaps? Someone with more expertise in that area may be able to help.
The external recorder didn't help, though. I shot everything to internal and external at the same time, both had the exact same motion characteristics. I thought that it may have been the pulldown in the external, actually, but no difference.
I don't think most people will notice it to be honest, so it's up to the shooter.
You and me both Kholi. I almost feel crazy to be so sensitive to such a minor detail of aesthetics, but it's a HUGE issue for me as well. Motion cadence similar to film is absolutely the single biggest "want" in terms of a camera feature when it comes to shooting something based in a drama / narrative context. An FS100, or many other true "video" cameras seem to be incapable of rendering motion in a film-like manner (exception given to those you already mentioned). I gravitate towards DSLR footage for this exact reason, which really isn't film-like either, but it at least approximates the feel close enough to be pleasing.
I can see your agony perfectly... wanting to settle down and use a "real" video camera for once, but that motion, it just won't adapt or conform to the feel you're looking for. I realize the majority of shooters just look at this as a non-sense issue, but I can fully comprehend why this is so agonizing.
Wish I understood the exact mechanics behind each cameras' shutter mechanism, and electronics / firmware well enough to pick out what differentiates a DSLR and regular large sensor video camera in terms of motion rendering. If we could at least determine the significant differences between the capture methods employed, it would take us down the right track in developing a new way to manipulate motion. There's gotta be a secret recipe, but until some kind of progress is made in this area, DSLRs like the GH2 continue to be the best bet for motion similar to film.
Last edited:
wannabeapilot
Well-known member
could you please link to an example of motion footage that your referring to?
both good and bad?
Well not that the FS100 is "bad" in any way, quite the contrary... it's just that 24p on the FS100 doesn't render in the same way as it does in other cameras. I love this camera, and even want to get one eventually for doing events and low light work / slow motion etc.
Here are 2 videos from the GH2 and FS100, both stylized with grading and some care taken to produce a cinematic result.
Both look great, I just find the motion cadence in the second to be a little off (although this is definitely a "better" example of FS100 footage in terms of pleasing motion). It all comes down to little more than personal preference, I'm just weird in the sense that I pick out silly things like the motion characteristics of video. If you don't see it, then good! You definitely don't want to be pestered by stuff like this : )
Last edited:
mraaron895
Well-known member
The biggest problem I have with those two video's is the lens choice and poor camera movement. Of course I'm more of a steady, soft move slider type of guy myself. But I will say that anything above 1/48 shutter starts to look like video. But any good camera operator should know this.
Funny enough, it's not really the issue for me. The camera could be completely steady on a tripod without movement and the motion captured on screen looks the same. It has nothing to do with how the camera moves or is being handled, but all to do with the way the motion of the different elements on screen are rendered.
It's like looking at any specific element of a production, perhaps color grading, and saying that if you only moved the camera a certain way it would change the look of the grade... which of course is silly; it's simply an innate characteristic of how the FS100 acquires images. To almost everyone this is an imperceivable issue and thus hard to describe accurately, very few people will ever pick up on it.
paulcurtis
Veteran
The motion cadence thing surfaces every now and again, usually without resolution.
I'm very curious because you'd think that it was a quantifiable technical issue rather than an aesthetic one. It should be possible to get to the bottom of it. I remember one good example was about detail in motion blurred movement, if there's too much detail it looks stroby. Is that the artefact you're seeing?
Digital sensors are pretty much identical in the way they work, so i could understand if you were simply comparing to film but you're also comparing to dSLRs and the GH2 in particular.
I really don't understand *what* these could be doing that the FS isn't - other than compression artefacts perhaps making motion more blurry because the compression isn't as good. Maybe?
It would be valuable to quantify what you find objectionable, enough to not use the FS in light of the positives, especially if you have any tests that show what you mean?
There was a thread on here where Frank found a difference in motion rendering he couldn't explain going 1080i50 instead...
thanks
Paul
I'm very curious because you'd think that it was a quantifiable technical issue rather than an aesthetic one. It should be possible to get to the bottom of it. I remember one good example was about detail in motion blurred movement, if there's too much detail it looks stroby. Is that the artefact you're seeing?
Digital sensors are pretty much identical in the way they work, so i could understand if you were simply comparing to film but you're also comparing to dSLRs and the GH2 in particular.
I really don't understand *what* these could be doing that the FS isn't - other than compression artefacts perhaps making motion more blurry because the compression isn't as good. Maybe?
It would be valuable to quantify what you find objectionable, enough to not use the FS in light of the positives, especially if you have any tests that show what you mean?
There was a thread on here where Frank found a difference in motion rendering he couldn't explain going 1080i50 instead...
thanks
Paul
Well it's somewhat of a lost cause I think; attempting to extract specific qualitative aspects of motion cadence among cameras like the FS100 and the GH2 or film. I wish I could find a way to express what I observe to be objectionable, but my only direct analogy would be what 24p film looks like being displayed @120hz 5:5 pulldown (not frame interpolation), used to help with telecine judder. It's an added smoothness that somehow distinguishes it from the normal cadence of film. There's probably a better way to describe it, but that's the best I can do.
Has nothing to do with motion blur, nothing to do with moving the camera "like in films", either. If you can't see it then you can't see it, cool. But it isn't the same, at all, and some cameras can't get there no matter WHAT you do.
It's very difficult to point it out to someone, honestly. To fully do so, you'd have to have the exact same image shot at the exact same time in the exact same frame (no shifts) and show them both at the same time, then back to back. Everything has to be exacts, otherwise those that can't already see it won't notice it, or will register the differences in the composition etc. not the motion.
FS100 24P (sony 24p in general) still has a slight "video" motion characteristic. It just can't get there and the only explanation I've heard from anyone is "this is how Sony likes their images."
Seeing a lot of Alexa footage, major motion pictures and just test footage, I can't tell the difference between that and film in regards to motion. It's perfect to me. But, sometimes I can see something odd in Epic footage, sometimes it looks right. GH2 looks right most of the time, but not exactly right in the way that Alexa is right. Canon is around the same as RED to me, give or take.
Edit here: and I still think the camera produces a very good image. Very very good. I just could not get it there in terms of motion (and those other things listed) so I wouldn't be comfortable using it for straight-up narrative unless it was just ultra lowlight material. But, I'm sure others can get better results out of it than I can and I'm still going to use it for my feature.
Last edited:
Bern Caughey
Veteran
Recently two individuals (one a big league ASC DP, & the other a Panasonic rep who used to be with Sony) have suggested using a 220* shutter angle to produce motion blur that better replicates film's.
Seems counterintuitive, & I haven't tested this theory yet, but will.
Seems counterintuitive, & I haven't tested this theory yet, but will.
Lammy
Well-known member
Thanks for the analysis on the FS100 w/Nanoflash dude. Especially boosting up on the shadows with the Nano there.
I was most interested in this camera last month as an alternative to the 5D3. I'm not gonna splash down cash or rent one anytime though. I'm also prepping for a feature too, but I'll be shooting in the summer and by then I know they'll be other alternatives around (possibly...)
Glad you're doing all the heads up testing here though, will look forward to re-visiting this thread when I need to ^_^
Keep it up!
I was most interested in this camera last month as an alternative to the 5D3. I'm not gonna splash down cash or rent one anytime though. I'm also prepping for a feature too, but I'll be shooting in the summer and by then I know they'll be other alternatives around (possibly...)
Glad you're doing all the heads up testing here though, will look forward to re-visiting this thread when I need to ^_^
Keep it up!
ColumbiaMan
New member
I'm with you guys on the motion issue. It bothers the heck out of me! If you guys would be so kind to advise a technical newb:
In order to alleviate this motion issue, what would one theoretically be gaining/losing if I boosted it up to 30fps with a shutter speed of 1/60th? Would this be even less of a film look?
In order to alleviate this motion issue, what would one theoretically be gaining/losing if I boosted it up to 30fps with a shutter speed of 1/60th? Would this be even less of a film look?
Kyle McConaghy
Well-known member
Hey, Columbia man... Definitely stick with 24fps for a filmic look. Some have suggested slowing your shutter speed down to 1/40 of a second. Haven't tried it yet. If you keep your detail below -2, that will also help.
What Columbia are you from/in?
What Columbia are you from/in?
wannabeapilot
Well-known member
Having only been able to do one comparison so far between the nanoflash at 80mbs with long gop recording, and the internal avchd at 24mbs.
I am very pleased to say that the nanoflash is definitely better than the internal recording
color rendition 4:2:2 vs 4:2:0 is not really different at all (at least not to me)
sharpness and visible detail is noticeably increased however
looking at a spiderweb in my nle and having the two clips one over the other (avchd 24mbs and nano at 80mbs)
Its surprising to see that a lot of the web has just been thrown away by the avchd compression
and looking at the exact same spot with the nanoflash footage and seeing the missing parts makes me very glad i decided to purchase the nano
I am very pleased to say that the nanoflash is definitely better than the internal recording
color rendition 4:2:2 vs 4:2:0 is not really different at all (at least not to me)
sharpness and visible detail is noticeably increased however
looking at a spiderweb in my nle and having the two clips one over the other (avchd 24mbs and nano at 80mbs)
Its surprising to see that a lot of the web has just been thrown away by the avchd compression
and looking at the exact same spot with the nanoflash footage and seeing the missing parts makes me very glad i decided to purchase the nano
I'm looking at s35mm film right now, going through it frame by frame. On my next monitor is RED material. Nope. No difference in motion. What is this magical "videomotion" that you guys see? Is it only on the FS100? Because stuff shot on that looks pretty good to me too. I just saw the AF100 film The Raid in a theater, looked pretty good motion wise too.
If there are no examples and people can't give a reason, then it falls into the audiophile tomfoolery category. "You just can't hear it!" "hmm, ok..."