Fri 16-Aug-13 09:43 AM | edited Fri 16-Aug-13 09:50 AM by mklass
I didn't know it was contentious. A tempest is a teapot?
Never-the-less, they proved rule #1, don't use VR when you have the lens mounted on a tripod. So it's not surprising that they got the results that they got. Whether Thom Hogan agrees or not is immaterial.
The next question, when hand-holding is it better to "wait a few seconds" for the VR to stabilize, or get the shot before the bird flies away or after the action has stopped?
Per page #25 of Nikon manual for the 300 f2.8 G II the tester partially got it right. Do not use VR when lens is firmly locked down on a tripod. However, when the lens is not locked down, as with a gimbal or is attached to a ball head which is not locked down, VR maybe helpful.
My rule for VR is simply never turn it on unless you have a reason to do so. As such, I rarely ever turn VR on any of my lenses.
Who in his right mind would turn VR on, if the camera is on a stable tripod ?
I checked the response time of VR on my 70-200/2.8 VRI. At shutter speeds that are too low, you can clearly see the blur in the view finder. Upon activating VR, the image becomes steady almost immediately, not after waiting a "few seconds". My only lens with VR is the 70-200, and my default setting is "VR off". The only situation where I found VR very useful was when shooting from a rocking sailboat, where I use "VR active".
Birding with a D800 and 500VR on a free floating gimbal head I find that the "tripod mode" VR offered by this lens gives an advantage up to about 1/500s maybe 1/1000s.
I shoot with my free hand draped over the lens to limit vibration.
Just took about 40 shots this morning half with and half without "tripod mode" VR . Mostly f5.6 to f7.1 with shutter speed of 1/1000 and ISO from 400 to 1000. At first glance the VR shots seem better on average.
For lenses that do not have "tripod mode" VR I totally agree not to ever use VR on the tripod. I learned this the hard way with the 105VR macro. Also, when shooting with shutter speeds significantly faster than reciprocal of focal length I find the VR unhelpful.
Fri 16-Aug-13 04:30 PM | edited Sat 17-Aug-13 12:01 AM by dm1dave
"The next question, when hand-holding is it better to "wait a few seconds" for the VR to stabilize, or get the shot before the bird flies away or after the action has stopped?"
The best of both worlds - set the camera to CH mode - shoot a burst as soon as you have focus - The VR may not be stabilized in shot one but it will be by shot 3 or 4.
I use the AF-ON button for focus so I can keep VR engaged, shutter half pressed, separately from focus. When using VR I engage VR as soon as lift the camera. Most of the time VR has plenty of time to stabilize but if a bird may fly away quickly - I take the shot without worrying about VR stabilizing and then if the bird sticks around I will keep VR engaged and continue to shoot.
By the way; Wile I agree that VR is better after it has stabilized – it doesn’t take “a few seconds” as mentioned in the article. I would guess that it is usually locked in in less than one full second on most VR lenses. This will likely change a bit with each lens – the old 80-400 (Nikons first VR lens) really needed a full second or so to get good results.
I usually have VR on with anything under about 1/500s. If VR is already on and the shutter speed is between 500 and 1000 I will leave it on. I turn it off when I am over 1/1000s.
I used to keep it on all the time. Often I would be shooting birds - so I could go from a slow shutter speed when the bird is sitting in the shadows to a fast shutter speed after it takes flight and moves to full sun. It was easier to just leave it turned on then it was to remember to turn it on and off as needed.
I may have ended up with a few shots that were not as sharp as they could have been as a result of using VR at high shutter speeds but most of the time I think the loss of sharpness is negligible. It is similar to the difference in sharpness between shooting wide open and stopping down one stop on a high quality lens – you may see it if you are looking really close to find it but it has little impact on the full image with normal sharpening applied.
The possible damaging effects also seem to be inconsistent. Here is a shot at 1/4000s with VR on...
This image is incredibly sharp even at 100% and no additional sharpening applied. In fact if you look at my eagle gallery – most of the image dates prior to 2013 were likely shot with VR turned on.
<<Dave, at what shutter speed, if any, do you turn VR off?>>
Not Dave sorry, but my use of VR is if the focal length in mm equates to x1.5 in shutter speed, I turn VR off. In other words, on a 300mm VR lens I will turn VR off at 1/450. My reasoning is that in the pre stabilization days with e.g. a long prime lens the speed should always exceed x 1.5 of focal length to eradicate low speed camera shake.
The concern I have with your method is that it might limit the quality of your crops in post processing. Motion/vibration will have the same effect on a crop as it would on a longer focal length lens required to achieve the cropped FOV.
Despite Thom's speculations, I did some calculations awhile back based on the frequency of normal physiological tremor and VR sampling rates that suggested to me VR may still be helpful 80% of the time, and destructive 20% of the time at shutter speeds above 1/500. I've never done a controlled test though. Perhaps I will this weekend. In looking at some of my images, I have started to wonder if it might have the most harmful effect on bokeh.
Well I just had my first go at it. I tried with and without a monopod with and without OS (50-500) at 500 mm. I started at 1/125 and went up to 1/3000. It was difficult to tell the difference above 1/500. If I had to give an edge, I'd give it to VR. One thing is for sure, framing and focusing is much easier with VR than it is without it and I'm guessing that at least for me, the risk of shot degraded by missed focus without VR is greater than any negative effects VR might have - which I still haven't been able to spot anyway. I'll try it again (there is better light outside now) and post some crops - probably later when I'm housebound due to the rain predicted this afternoon.
The weather here is fine with bright sunlight tomorrow unlike today which is similar to your day. I'll go to the bird reserve and see what I can do with my 50-500mm trying to replicate similar tests. My only problem is the birds tend to be quite distant and I learned long ago, following your advice, to get closer. Anyway, worth a try an I'll try with and without VR, handheld and tripod mounted.
That's why I'm using a hat with a fine weave pattern for this test. It is much more patient for a controlled test than birds. The only difference I'm seeing with my second test is due to noise at varying ISO levels. I set the camera to manual, but had auto iso on to stay within the camera EV range as I increased the shutter speed. Unfortunately I ended up with different ISO settings as the light changed, e.g., different noise levels that would mask any difference. so I am going to try a third time with manual and adjust iso manually as well. The sun is out so maybe I will have consistent light levels.
Sat 17-Aug-13 03:11 PM | edited Sat 17-Aug-13 03:50 PM by Chris Platt
Here was the target. D7000 and Sigma 50-500 on monopod. Distance about 45 feet. Use the link below to download the test results image file (6.4M). You'll have to download the file to view it in full resolution. I'm calling it a draw at shutter speeds above 1/500. I did this test a number of times and am not seeing any practical reason to avoid using OS/VR above 1/500.
Can't get the link to display so here is a text version:
Edit: for those that don't want to download the full test file, here are two 100% crops taken at 1/4000. Top image is OS-On and bottom is OS-Off. They were taken at ISO 1600, and I applied the same level of noise reduction to both which may have softened them slightly.
There's nothing wrong with that, excellent sharpness. I see that it was taken at 1/125 @f13 and as such would expect OS to be in play without any adverse effects on a monopod as Nikon and Sigma claim that the stabilization should be used in that situation.
Sat 17-Aug-13 04:11 PM | edited Sat 17-Aug-13 04:14 PM by Chris Platt
For some reason the site is not letting me post links, so I removed the dot after www so it would not be recognized as a URL. Try replacing the dot between "www" and "dropbox.com/s/xb25jfunmzpyupx/OS%20TEST2.jpg"
Some were a bit softer with it on, others with it off. There was such a slight difference that it could have been caused by other variables as well. Any of the very slight differences would have no practical significance at any reasonable viewing size/distance. My conclusion - if you're worried about it, turn it off. If you find it useful at higher shutter speeds to have a steady view for framing and focusing, leave it on and don't worry about it.
Thanks for taking the time to do this Chris. This is a great, practical test that put these differences between VR/OS or off at higher shutter speeds into perspective.
Your conclusions are consistent with my experience shooting birds in flight. If/when VR degrades IQ in certain circumstances the amount of degradation is negligible. There are so many other variables that are more likely to cause lost shots.
I try to remember to turn off VR at high shutter speeds just because it provides little advantage and it drains battery power.
I suggest you read the excellent article by Thom Hogan on this subject, www.bythom.com/nikon-vr.htm. One should not always believe what the Nikon Marketing Managers dream up. If one is worried about tripod vibrations, one ought to use a proper tripod. Besides, there are things like "mirror up" and "exposure delay".
I have read Thom’s articles and I stand by my statements based on my own experiences. If I am using support (I use a 5 series Gitzo) and my gimbal head or ballhead is not fully locked down and my shutter is below 1/500s I use VR.
I respect Thom, but I don’t take everything he says (or everything Nikon says) as gospel. Thom is looking for tiny differences that rarely matter when the final image is viewed as a whole.
I don’t see uber sharpness as the holy grail of image quality either. See my post above to see how I broke the “don’t use VR at over 1/500s” rule for years wile bringing home exceptionally sharp images of eagles pulling fish out of the water.
Thom even gives us some perspective in that article...
--------------------------------------- “...what does a spoiled-by-VR shot look like? Well, "spoiled" is perhaps too harsh a term. Sub-optimal is probably a better one.”
“Yes, there's some nitpicking going on here. VR not correcting right is a bit like tripod mount slop (fixed with a Really Right Stuff Long Lens Support) or ringing vibrations in the tripod legs (fixed by using the right legs for your equipment): you don't see it until it's gone, and even then usually only if you're pixel peeping”
I particularly like the “...even then usually only if you're pixel peeping...” line. I like nice sharp images as much as anyone but I am not a pixel peeper. My goal is not to create perfect pixels but to create a photograph to be viewed as a whole. Any flaw that requires viewing, on-screen at 100%, to identify will never be visible, no matter how large you print, when the image is viewed at a distance that allows you to see the entire image all at once.
There are countless things that can happen while shooting that can create “sub-optimal” results. The fact that VR “might” cause problems sometimes and in some situations is pretty low on the "list of things that I worry about wile shooting."
>> “If one is worried about tripod vibrations, one ought to use a proper tripod. Besides, there are things like "mirror up" and "exposure delay".
I use the most sturdy tripod available on every shoot - I also regularly shoot at a focal length between 400mm & 800mm, sometimes on a DX body.
Mirror-up and exposure-delay are not always practical when shooting a skittish or slow moving critter in low light with a 680mm focal length. At these super telephoto focal lengths shutter speeds of 1/100 – 1/500 are quite slow – they are fast enough to stop the motion of a walking animal or swimming duck but still susceptible to camera shake and vibrations introduced trough the photographer’s hands holding onto the camera or wind buffeting the lens.