View Full Version : ND filter versus faster shutter speed. Please educate me.
Nick Hope October 5th, 2007, 08:35 PM I do a lot of underwater shooting and most of the time in the past I've shot a lot with auto exposure and used the ND filters to reduce light as prompted by my VX2000 and now Z1.
Recently I've been doing a lot more manual exposure underwater macro work with lights and I find myself fairly randomly using either the Z1's built-in ND filters or a faster shutter speed (typically in the 1/100 to 1/300 range) to reduce light when for example the iris gets smaller than F6.8. I'm shooting 1080i50.
However I don't really understand the issues of when I should use ND filters and when I should use a faster shutter speed. Faster shutter speed is better for shooting rapid motion, right? Or is that just if you want to slow the rapid motion down in post?
I'm guessing that a faster shutter speed must in some way be a bad thing otherwise in auto mode the camera would speed up the shutter rather than maintaining 1/50 shutter and prompting for an ND filter.
Can anyone enlighten me or point me to a reference where I can learn about this?
Mikko Lopponen October 6th, 2007, 06:20 AM Faster shutter = less motion blur
Nick Hope October 6th, 2007, 06:31 AM Thanks Mikko, that makes sense.
When would I want motion blur and when would I not want it?
Tom Hardwick October 6th, 2007, 06:42 AM Faster shutter speeds for shooting rapid motion? Quite the contrary Nick. As Mikko points out, faster (shorter) shutter speeds cut down on motion blur, subject as well as camera. This has nothing to do with 'making the footage sharper' as a lot of people are led to believe, it generally just makes the footage look more jerky - staccato if you like.
If at all possible stick to your Z1's default shutter speed (1/50th sec PAL). You can halve that to 1/100th sec without it becomming too obvious on screen, but better to use the in-built ND filters to soak the light every time.
My slo-mo program looks a lot smoother if I stick to the default sh/spd too. It's because at 1/50th sec everything that happens in front of your lens is recorded, so if a fish swims past you the tape records A to B on one frame, then B to C on the next and so on.
Say you select 1/200th sec. Now you record A to B on the first frame and D to E on the next. The shutter was closed for 1/150th sec, and that portion of time wasn't captured.
Another thing against faster shutter speeds is that (with CCD chips such as you have) CCD smear becomes more of a problem the higher the speed. Best not to go there unless you want to print individual frames or stop the tape to analyse the golf swing.
tom.
Dylan Pank October 6th, 2007, 06:43 AM Nick, it's up to you. Fast shutter speed has a different look, motion is rendered as slightly stuttery (depending on how fast the shutter speed is) and is generally unnoticeable on very low motion or still frame shots.
In most cases if you want footage to look "normal" you'll shoot with a shutter speed as close to the default (50 for PAL) as possible.
One issue that might shy you away from fast shutter speeds is that the lack of motion blur stresses the HDV codec more if there is a lot of movement in the scene.
Nick Hope October 6th, 2007, 08:12 AM Thanks a lot Tom and Dylan, I understand a lot better now.
So similarly would most of the big-screen movies i watch be shot with a shutter speed of 1/24?
Thankfully most of the stuff I've shot recently at shutter speeds faster than 1/50 has been stuff without much movement, but I'll change my habit back to using the ND filters in preference.
How about degradation of the image due to the light passing through extra filters if the ND filters are used? Is it much of an issue with the Z1's internal filters?
Tom Hardwick October 6th, 2007, 09:53 AM So similarly would most of the big-screen movies i watch be shot with a shutter speed of 1/24?
No, not at all Nick. Big screen movies are shot on film, and if they're shot electronically then the DOP tries his utmost to make it look like film. Film is exposed to the light, then the shutter closes and a claw drags the next frame into position. At 24 fps each frame gets about 1/60th sec, though this can vary depending on the shutter sector angle (film cameras use a rotating focal plane shutter).
So film has this charisteristic staccato movement where more than half of what happens in front of the lens isn't recorded. We've all become used to it and are supposed to love it, but it purely an historical/mechanical encumberance that comes as part of the package.
Then along comes video with it's pro of total recall and resultant image smoothness and the die-hards winge and wring their hands at the 'video look' and miss film's cons, I know not why. With video you can pan at nearly any speed, with film you can't.
Don't worry about spoiling the image quality by using internal NDs. They're far less degrading than
a) using external filters or
b) varying the shutter speed.
But under water you're in an unreal world and I'd go so far as to say that almost anything goes. You show us a world that few of us enter, so use 1/25th to gain an extra stop, use a super wide-angle because we'll never see the barrel distortion and up the gamma on the timeline because it'll makes us gasp more.
tom.
Nick Hope October 6th, 2007, 12:09 PM Thanks Tom. Really helpful insight. Just what I needed.
I've been shooting mostly macro recently recently with a flat port on the housing and a Century Optics +3.5 diopter screwed to the front of the Z1 and twin halogen underwater lights.
On a lot of my subjects I've been deliberately reducing the light with either ND filters or faster shutter in order to open the iris as much as possible (typically bringing it from say F6.8 to around F2) to try and shorten the depth of field to throw the background/foreground out of focus while keeping the subject (fish, crab, sea slug etc.) sharp to try and make the image more "film-like/cinematic/dramatic". I now know the faster shutter speed wasn't such a great idea, and I realise each case is different, but in general, using ND filters, do you think this is a good approach?
Mikko Lopponen October 6th, 2007, 02:59 PM Well I don't think its bad to raise the shutter. I've shot many different things with a high shutter and none of them look bad, infact I kinda like the more sharpened look. And yes, the individual images are very much sharper than the standard pal 1/50. For filmic quality try to keep the shutter at default, but if it's just for your personal pleasure and or filming sports then why not keep it faster? The staccato effect isn't that bad when we have 50/60 fps anyway.
Dylan Pank October 6th, 2007, 03:34 PM Nick, almost NONE of the films you see on the big screen feature a 1/24 shutter speed in the sense that it means in video.
They shoot at 24 frames a second yes, but the shutter is open for 1/48th of a second. This is because the vast majority of the movies at the cinema these days are still shot on 35mm motion picture film and the cameras feature a 180 degree shutter angle.
Of the rest that are shot on digital, generally they want to look as much like all those film-shot movies so simulate the 50%-shutter-to-frame rate ratio of 35mm movie film cameras.
A few low light night sequences of Collateral, Apocalypto or Miami Vice may have been shot at 1/24 but probably less than 10%, and other digital movies I've seen haven't used shutter speeds below 1/48 at all.
However the main difference is that theatrical movies are shot and shown (whether film or digital) in progressive format. The Z1 defaults to interlace.
Unless the camera is damaged, degradation of the image quality due to using the internal ND filters would be somewhere between negligible to the point of insignificance and non existent.
Tom Hardwick October 7th, 2007, 01:08 AM Mikko's right in that it's not 'bad' to shorten the shutter speed - it just gives a different look to movement (camera or subject). If you film a stationery subject with a stationery camera there's not the least bit of difference except that the higher shutter speed will mean you have to use wider apertures and maybe gain up - both of which will generally soften the image, not sharpen it.
Mikko's equating motion blur with unsharpness and of course in the still picture world this is very true. In movies it's not true in the slightest. Point your stationery camera skyward and film a plane as it traverses the frame. At a 50th sec each frame blurs into the next and the plane smoothly crosses the sky. At a 350th sec the plane makes 'jumps' to each new position on screen. It may be filmic, but it's painful to watch.
I've also seen footage where aircraft propeller blades and helicopter blades 'stand still' while the aircraft's in motion. High shutter speeds are at fault here. Mind you, the explosions and fight scenes in Band of Brothers and Private Ryan are greatly heightened buy the use of high shutter speeds; the 'jitteriness' adding to my jitteriness as I watch, horrified.
Nick, you're going about obtaining shallow dof in the right way - using biggish chips, wide apertures and close-up lenses. You don't say what focal lengths you're using, but of course the longer the better. Difficult to use telephoto under water as I know - I shot this on Super-8:
http://www.youtube.com/watch?v=NdI0hqv5tRY
tom.
Nick Hope October 7th, 2007, 02:47 AM Tom that's brilliant. When did you shoot it? Just for fun? Tell us more about how it came about.
Regarding focal length... With the Century diopter in place most of the time the front of the camera would be say 50-70cm from the subject. Manual focus control is not good on my housing, and it's difficult to judge focus as I'm just viewing the camera's top LCD through 2 panes of glass and some mucky seawater, so I was in auto-focus a lot of the time. But when I did switch to manual focus I seem to remember it was at infinity a lot of the time. I was often getting a shallow DOF even at smallish apertures like F6.8. My understanding of lens physics, and in particular diopters, is not good I'm afraid.
In due course I'll post some example material but in the meantime I have 20 hours of m2t files on my hard drive to wade through.
A couple more questions if you don't mind....
Do native 24P video cameras default to 1/48 shutter?
And why do wider apertures generally soften the image?
Tom Hardwick October 7th, 2007, 02:56 AM You're brilliant too Nick :)
I've just watched your spotted-dick shark video on YouTube and greatly admired the utter smoothness of the tracking shots under water. Amazing! More later, I have to help a man on his timeline and it's my wedding anniversary today as well - oh, er.
tom.
Tom Hardwick October 8th, 2007, 05:52 AM OK, a close-up lens (in reality simply a positive dioptre single element - though the Century one may be a cemented doublet to justify its cost and to reduce aberrations), is designed to work with your camera's lens focused on infinity. So a +3 dioptre lens will mean infinity focus on the camera equates to 1/3rd metre, whatever focal length you're using.
So you're getting the results you're after by using telephoto focal lengths and the close-up attachment. Because you're focusing so close the d o f will be small and as you've found, apertures like f/6.8 might well be needed to have d o f field cover the animal you're filming.
Why do wider apertures soften the image? Because a lens is a lot of glass elements held together in an internally blacked cylindrical tube. Of course it's never totally black and in zoom lenses there's also other clutter in there to move the elements about and give OIS.
At wide apertures the light isn't concentrated down the centreline of the cylinder, but bounces around very close to the inner wall of the tube, so causing flare. Also your 12x zoom probably has 15 elements in line, and decentering of any element (inevitably allowed by production and manufacturing tolerances) means that wider apertures make the losses emanating from this more noticeable.
'Splash' was shot just for fun. I used an underwater Eumig Nautica Super-8 camera - a very simple affair that didn't even have exposure lockdown. For lots of the shots of me I simply swam holding the camera in front of me, facing backwards, and got the pool attendant to track alongside me and to wait under water as I dived in.
It was shot over 3 months of lunchtime swims back in the 80s, in Sweden, which helps explain the continuity errors. I just wanted to show the delights of weightlessness, that's all. My cutting ratio was my highest ever, as it's extremely difficult to swim and hold a camera steadish. I swim better than that now!
tom.
Nick Hope October 8th, 2007, 06:58 AM I thought I recognised your face. Nice stroke! I especially like that slomo opening shot.
Thanks again for the explanation. This is really great stuff for someone like me who missed out on some basic theory.
So according to your explanation, large irises (low F numbers) are to be avoided and according to Sony, small irises (large F numbers) are to be avoided too (p32 of the Z1 manual refers to diffraction and I have noticed this myself). So does that mean I should be staying in say the F3.1 to F6.8 range where possible?
I did some underwater macro tests with the Z1 when I got my Century diopter. They are here if anyone's interested:
http://wetpixel.com/forums/index.php?showtopic=20474
Tom Hardwick October 8th, 2007, 07:22 AM Thanks for the link - interesting thread Nick.
There's nothing special about 'video lenses' as Sony seems to infer - all lenses suffer from diffraction losses the smaller the aperture that's used. In theory all lenses are at their very best wide open, but in practice flare from the interior of the cylinder, vignetting of the frame and decentering of the elements ensures this is never so.
Our Z1s have a 'sweet spot' in the aperture scale at f/4. Any wider than that and the lens vignettes the corners of the frame, and the zebras can clearly show that wide open the edges of the frame are a good 1.5 stops under-exposed.
Of course for general movie making this doesn't matter a toss and can even be used to good effect - 'highlighting' the object in the centre of the frame quite nicely. If you use your camera for making panoramas you'll quickly notice that frame vignetting makes matching up sucessive shots very difficult.
Cameras that have automatic (and generally undocumented) internal ND generally stop the lens closing below f/4 these days, the ND taking over as the owner thinks he's stopping down. (Sony PDX10, A1 etc). This is to stop diffraction blurring the image, a problem that is exacerbated by short focal lengths.
Pinhole cameras (remember them?) needed a tiny hole to keep the image sharp(ish), but not so small as to have diffraction take precidence.
So with our tiny chips and short focal lengths (and even shorter focal lengths if you use a wide-angle converter) you must stay well clear of small apertures. In fact our Z1's menu offers you the option of not allowing you to stop down past f/5.6 or so. If you're shooting HDV, take this option. I'm much less afraid of using maximum aperture than I am of using minimum.
But never be afraid to experiment I say. Video is the greatest of teachers.
Opening slo-mo shot on Splash was a Canon 1014E at 60 fps.
tom.
Nick Hope October 8th, 2007, 07:36 AM I think you mean in the picture profile? You can limit the iris to no smaller than F11, F6.8 or F4 in there.
I have often shot with a picture profile limiting the iris to F6.8 and will do this religiously from now on.
I have my maximum gain in PP1 set to 12dB and I'm going to keep sharpness at 9 as well from now on having read comments from DSE and others about the over-sharpness of HDV cameras and noticing this myself in my recent footage shot with no picture profile set.
I'm also wondering about a bit more colour. I notice a lot of people choosing around +2. What do you do?
Nick Hope October 20th, 2007, 08:47 AM OK, a close-up lens (in reality simply a positive dioptre single element - though the Century one may be a cemented doublet to justify its cost and to reduce aberrations), is designed to work with your camera's lens focused on infinity...
Tom, just coming back to this, does this mean I really shouldn't use the diopter on anything else except infinity? Should I even go as far as locking the focus on infinity and then just adjusting the other parameters (zoom, distance to subject etc.)?
Just wondering about this because I'm cutting my macro footage now and a lot of it is not quite sharp. Some of such footage may well have not been on infinity focus.
Tom Hardwick October 20th, 2007, 08:58 AM If you fit a close-up lens it simply means that infinity as marked on your lens is a lot closer. You can still use the camera's focusing quite happily, but of course adding the extra element(s) brings with it pretty nasty abberations. Generally the centre of the image should be ok, but it's the edges that suffer. Stopping down helps, but then again the more you stop down the more you head towards diffraction. You can't win.
So many manufacturers are making their lenses 'tele-macro'. The HC7 is a case in point - you have to use max telephoto (because of vignetting) but it does allow great closeups from a distance.
tom.
Nick Hope October 20th, 2007, 10:46 AM Thanks Tom. Well the achromatic Century diopter is definitely sharper around the edges the single-element diopter I was using before. But I'm only getting razor-sharp clarity and detail on the occasional shot. Many of the other macro clips shot with the Century are ever so slightly blurry, but still usable. The blurriness can't be seen on the camera's LCD. Only on my 24" monitor. A winning combination of lighting, iris, focus and distance from subject is pretty difficult to achieve (particularly underwater!).
I wish the exposure metadata could somehow be captured along with the footage, not just displayed on the camera during playback. Would make it much easier to learn from mistakes.
Attached is a still from some of the better underwater footage with the Century diopter. I guess that slug's about 2 inches long.
Greg Boston October 20th, 2007, 12:06 PM the achromatic Century diopter is definitely sharper around the edges the single-element diopter I was using before. But I'm only getting razor-sharp clarity and detail on the occasional shot. Many of the other macro clips shot with the Century are ever so slightly blurry
Nick, that's because using the diopter makes your DOF a lot more shallow. As you stated, you are underwater and are somewhat subjected to current moving you around, certainly enough to give you some soft images by moving just a little bit in relation to the subject. It's the nature of the beast.
I'll also second what Tom says about lenses and sweet spots. He mentioned you can tell that edges are underexposed vs. the center of the image and I have confirmed this while using a waveform monitor. You'll see the edges droop down on the scope until you close down to around F4, at which time the line becomes flat. As a general rule, the sweet spot kicks in around 2 stops down from wide open on most every zoom lens I have used.
-gb-
Tom Hardwick October 20th, 2007, 12:13 PM The 'everso slightly blurry' shots - these can't be blamed on dof problems or on very slight miss-focusing (i.e. it's sharper somewhere else in the shot)?
Greg - all lenses vignette the edges wide open - primes as well as zooms.
tom.
Nick Hope October 22nd, 2007, 06:02 AM The 'everso slightly blurry' shots - these can't be blamed on dof problems or on very slight miss-focusing (i.e. it's sharper somewhere else in the shot)?
Probably so. I think on this particular macro trip i went a bit overboard with opening up the iris to shorten the depth of field for a "film-like" effect. So I was often as wide as F2. Probably didn't need to do this as the diopter was shortening the DOF anyway.
(Will post some finished edited footage in due course)
Ryan Avery November 26th, 2007, 03:34 PM I thought I recognised your face. Nice stroke! I especially like that slomo opening shot.
Thanks again for the explanation. This is really great stuff for someone like me who missed out on some basic theory.
So according to your explanation, large irises (low F numbers) are to be avoided and according to Sony, small irises (large F numbers) are to be avoided too (p32 of the Z1 manual refers to diffraction and I have noticed this myself). So does that mean I should be staying in say the F3.1 to F6.8 range where possible?
I did some underwater macro tests with the Z1 when I got my Century diopter. They are here if anyone's interested:
http://wetpixel.com/forums/index.php?showtopic=20474
If anyone is interested, you can email me and I will send you an Excel file which will calculate minimum focusing distance with a diopter attached.
Send email with a BOLD title that say "SCHNEIDER CLOSE UP CONVERSION".
Thanks,
Ryan Avery
Regional Sales Representative
Schneider Optics
|
|