NikonGear'23
Gear Talk => Lens Talk => Topic started by: pkol on January 27, 2017, 00:58:47
-
Hi,
I received my copy of the 300mm f/4 PF VR last Friday. Since then the only opportunity I had to shoot outside was on Saturday so eager to try the new toy I went to the Copenhagen Women's March unfortunately I was only able to get there at 16:00.
I really enjoy using the lens, it's light, VR is amazing and the image quality is great... However when I was going through the photos I noticed that some showed what it seemed double contours... I was well aware of the issues with VR when the lens was released but I thought that Nikon fixed them a long time ago. By the way my lens SN is 2247xx, so well outside the troubled lenses (according to Nikon service Notice). I also noticed that the shots where the double contours appeared were at around 1/100s.
Here are some photos quickly processed in Lightroom, taken with D810 + 300mm f/4 PF
-
Since I noticed the double contour issue mentioned in the post above I decided to do some tests. For all the test shots the subject is 3m away from the camera (D810).
Now I'm a total novice regarding lens test, so I will explain what I've done and would like to get some guidance from the community regarding the tests and whether my copy of the lens needs to be serviced.
Here are 5 shots all handheld, in Shutter Priority, Auto WB and Auto ISO with VR Normal ON.
While I think the shots @ 1/20 and 1/160 are perfectly fine, I find a slight deterioration @ 1/80 and clearly double contours @ 1/100 and 1/125.
-
Next I decided to compare the shot @ 1/100, still handheld with VR Normal On, but now using:
1 Single release mode,
2 Single release mode and flash (SB-700), and
3 Mirror Up (MUP)
Here I find that using flash or MUP enhances substantially the IQ and eliminates the double contour issue.
-
Finally decided to use a tripod to test shooting @ 1/100 which one of the most problematic speeds with VR on.
So for the comparison below all the shots were taken with a tripod @ 1/100 and f/4 using a cable release and :
1 VR Off, MUP + EFCS
2 VR Normal On, MUP + EFCS
3 VR Off, Single Release
4 VR normal On, Single Release
In terms of IQ I think the order is 1>2>4>3. Here is the comparison:
-
Double contours mean something has moved during the exposure. Typically caused by too long shutter speed in conjunction with inadequate support for the camera/lens or a camera not held steady enough.
Do take the manufacturer's specification of how long shutter speed VR will allow with a big lump of salt. These numbers are obtained under the most optimal conditions. In practice in the field, VR is not enough if the shutter speeds get too long-. You probably would need 1/250 sec or faster for a 300 mm lens.
-
Although I have no experience with longer lenses I always read that the mirror shock can be problematic at certain shutter speeds.
I can see that between 1/80 and 1/125 I cannot get sharp photos handheld using VR. It seems that this is due to the mirror shock since I can consistently get sharp photos (without double contours) using Single release + VR On + MUP. Obviously this is not very practical and clearly not a solution in the field.
These tests left me with a few questions:
1 Why are the handheld photos (with VR on) below 1/60s consistently sharp and between 1/80s and 1/125s showing double contours?
2 Using flash @ 1/100s with VR On handheld does not show double contours. Why is that?
3 It seems to me that the issue of double contours is caused by mirror shock. Is that right? and if so why don't I see that on my Nikon 70-200 f/4?
4 I most likely am doing something wrong with these tests (this is the first time I try to test a lens in this manner). Any guidance would be greatly appreciated.
5 Most importantly, do you think my lens should be returned or serviced?
It would be great to hear your thoughts about these questions.
Many thanks in advance for your time.
Kind regards,
Paulo
-
Double contours mean something has moved during the exposure. Typically caused by too long shutter speed in conjunction with inadequate support for the camera/lens or a camera not held steady enough.
Do take the manufacturer's specification of how long shutter speed VR will allow with a big lump of salt. These numbers are obtained under the most optimal conditions. In practice in the field, VR is not enough if the shutter speeds get too long-. You probably would need 1/250 sec or faster for a 300 mm lens.
Many thanks for you reply Bjorn. I understand your point, but my question is how can I get consistently good results (no double contours) at shutter speeds slower than 1/60s. That is what I'm baffled with?
I also know that this is somewhat a mute point as one would need to use, as you point out, much faster shutter speed to handheld a 300mm. But the question remains in my mind and there must be an explanation.
-
Tripod, or at least a monopod. Plus practice a lot. You need to be able to treat the camera in a relaxed, not tense, manner, even when the going gets touch and involuntary camera shake is easily introduced. VR might help, but is always a factor of unknown magnitude and importance. Don't rely solely on VR.
That being said, I sometimes am able to get very sharp images with the 300 PF at ridiculously slow speeds, in some cases down to 1/6 sec. Still I try not to do too much hand-held shooting with this lens.
-
Nothing is wrong with your lens, Paulo. Doubles caused by wrong correlation between mirror box of D800-810 series, and VR-ed lenses. I have or had the same situations with many VR-ed lenses, especially with my 80-400G. There is some kind of engineering mistake, less so on the last D810. I set all of my zooms to ACTIVE position and left this setting forever. It is helping somewhat, just try to avoid 1/125 +/- 30%. Beautiful camera is that D810, with the best sensor on the market. Yeah, we can't always win... Good luck! LZ
-
I have one of the earlier copies (202xxx) and after the fix I still have (minor) problems at 1/80 - 1/160
It helps when you use a grip on your D810 (or even a bracket).
-
A benefit of these new cameras like the D810 is that it is easy to just turn up the ISO to avoid slow or troubled shutter speeds.
Another work around is to buy the battery grip (and the large battery from the D4) this will add more weight to the camera and stabilise the shot.
Back ground double lines are a VR risk.
I think your images from lovely Copenhagen are very fine!
You could always try my lens,,, Seems you are in CPH,,,
-
There is some consensus out there that adding a tripod collar to the 300mm PF (used to improve how the lens is gripped and increase inertia) can improve VR performance on troubled bodies.
-
Yes, that would also help - Good point!
-
Many thanks for all your comments and suggestions.
While I do understand that VR does detract from IQ and should only be used sparsely I am still dumbfound regarding the VR performance within a narrow band of shutter speeds. i will have to test the lens on my D750, but anedoctely it seems to be affected in a similar manner. I might be wrong about this and will have to test it to be sure.
It seems to me that most of you think that my lens VR behaviour is normal for this lens. I wanted to make sure that I did not get a "lemon" at least with regards to VR related issues.
I was hoping that, given the knowledge in the NG community, I would get some answers or at least some pointers towards answering these questions:
1 Why are the handheld photos (with VR on) below 1/60s consistently sharp and between 1/80s and 1/125s showing double contours?
2 Using flash @ 1/100s with VR On handheld does not show double contours. Why is that?
3 It seems to me that the issue of double contours is caused by mirror shock. Is that right? and if so why don't I see that on my Nikon 70-200 f/4?
These questions are still echoing in my head and I do not have the experience/knowledge to address them. It is very possible that the test I did were not adequate to provide any usefull information towards unconvering possible answers.
Once again many thanks for your time.
Regards,
Paulo
-
I think your images from lovely Copenhagen are very fine!
You could always try my lens,,, Seems you are in CPH,,,
Many thanks for your offer. I moved to Copenhagen in the summer of 2015 and am enjoying my time here.
It would be great to meet you.
Regards,
Paulo
-
Your welcome :) just send me a PM or start a topic here:
NikonGear » Travelogues » Future NG Events - and Location Reveries
-
VR actively participates in a feedback loop. Thus it is no surprise it sometimes can exacerbate issues (overshooting comes to mind).
I have seen lots of double contours with VR ON using 1/500 or fast with the long Nikkors such as AFS VR 400/2.8, 500/4, 600/4 (all of the latest generation) with these lenses mounted on a very stable tripod.
Only use VR as a very last resort and most problems will go away. My experience for what it's worth.
-
I was hoping that, given the knowledge in the NG community, I would get some answers or at least some pointers towards answering these questions:
1 Why are the handheld photos (with VR on) below 1/60s consistently sharp and between 1/80s and 1/125s showing double contours?
2 Using flash @ 1/100s with VR On handheld does not show double contours. Why is that?
3 It seems to me that the issue of double contours is caused by mirror shock. Is that right? and if so why don't I see that on my Nikon 70-200 f/4?
Paulo,
at first sight, this looks like resonance to me: https://en.wikipedia.org/wiki/Resonance. It seems there is an interaction between the vibration of the body at this shutter speed and the natural frequencies of the VR system. This behaviour is often spotted in cars when driving at a specific speed on a road with a ryhtmic surface pattern. You get a loud vibration sound which disappear if you change speed or road section.
Adding a grip or using a tripod instead of handholding changes the mass and configuration of the system, changing thus the vibration mode. A different result in these circumstances is a strong indicator of a resonance issue.
It is very hard to damp complex mechanical systems at all frequencies as parameters are interelated and often have opposite effects. The art of silencing a car in all circumstances, for example, is a specialised business with entire companies only dedicated to vibration analysis and dampening.
Specifically on your questions:
question 1: because the impulse produced between 1/80s and 1/125s has the right frequency to induce the resonance
question 2: most probably because the duration of the flash is shorter than 1/4th of the resonance frequency. It is short enough to not record the full amplitude of the vibration at resonance
question 3: because this body plus lens configuration has different mass and vibration modes and because the VR of the 70-200 has different natural frequencies. The global system resonates then at different frequencies, which are either not produced with this lens/body configuration or are produced at a shutter frequency outside of the range you use
-
Bjorn,
I find your experience priceless and am grateful to you for sharing it here in this website!
If I understand you correctly, the VR technology feedback loop causes in some circumstances the movement to be exacerbated rather then minimised. This seems to be more prominent at around 1/100s, in my case.
I have now shot close to 100 frames at 1/100s and the double contours manifest itself in ~98% of the time, when handheld, VR on and in single shuter release. However adding the SB-700 or using MUP instead of S (release mode) I cannot see double contours - that puzzles me.
Because of that I'm inclined to think that the mirror slap must be interfering with the VR feedback loop specifically at shutter speeds around 1/100s. Which also explains the results posted earlier with a tripod and cable release. In fact I'm thinking that the mirror vibration of the D810 at around 1/100s interferes with the VR implementation on the D810.
Could it be that the mirror in my D810 is not properly calibrated and the issues only manifest itself at these shutter speeds? This might be a question coming from my ignorance...
-
Paulo,
at first sight, this looks like resonance to me: https://en.wikipedia.org/wiki/Resonance. It seems there is an interaction between the vibration of the body at this shutter speed and the natural frequencies of the VR system. This behaviour is often spotted in cars when driving at a specific speed on a road with a ryhtmic surface pattern. You get a loud vibration sound which disappear if you change speed or road section.
Adding a grip or using a tripod instead of handholding changes the mass and configuration of the system, changing thus the vibration mode. A different result in these circumstances is a strong indicator of a resonance issue.
It is very hard to damp complex mechanical systems at all frequencies as parameters are interelated and often have opposite effects. The art of silencing a car in all circumstances, for example, is a specialised business with entire companies only dedicated to vibration analysis and dampening.
Specifically on your questions:
question 1: because the impulse produced between 1/80s and 1/125s has the right frequency to induce the resonance
question 2: most probably because the duration of the flash is shorter than 1/4th of the resonance frequency. It is short enough to not record the full amplitude of the vibration at resonance
question 3: because this body plus lens configuration has different mass and vibration modes and because the VR of the 70-200 has different natural frequencies. The global system resonates then at different frequencies, which are either not produced with this lens/body configuration or are produced at a shutter frequency outside of the range you use
Many thanks Bruno. That is a very clear explanation. So based on the uploaded photos do you think there is something wrong with my copy of the lens?
-
Not a calibration issue.
Some lenses are worse than others, we had a loooong thread about it in May,,,
Bruno is right on the money I believe.
-
1 Why are the handheld photos (with VR on) below 1/60s consistently sharp and between 1/80s and 1/125s showing double contours?
The vibration from shutter and mirror has the largest amplitude right after the shutter opens. Over time the vibration amplitude decays. At longer exposures, the largest amplitude vibration affects on the beginning of the exposure so most of the light that is recorded renders the image sharply.
Mirror vibration and shutter vibration probably have different frequency ranges.
So, if you use a solid tripod, EFCS and M-UP you should get sharp pics at all speeds in most conditions (wind may still cause vibration). If you need to hand hold, you should preferentially use a fast shutter speed which takes care of subject movement as well as all kinds of camera shake.
2 Using flash @ 1/100s with VR On handheld does not show double contours. Why is that?
The flash duration is about 1/1000s or it can be even shorter. Thus the vibration doesn't have time to make an impact on the image sharpness.
3 It seems to me that the issue of double contours is caused by mirror shock. Is that right? and if so why don't I see that on my Nikon 70-200 f/4?
The 300mm has greater magnification due to its focal length so any vibration is magnified as well. The PF is also a very lightweight lens for its focal length which can accentuate the problem. Less mass, easier to vibrate.
I typically use the 300 PF for photographing moving subjects, mainly people, at concerts, fashion shows, and sports. This is hand-held work. I also occasionally use it for landscape and close-ups (most of the time on tripod, but occasionally hand held with VR).
My most typical shutter speeds for the moving subjects are from 1/500s to 1/1250s, with VR set to SPORT mode. This means the lens won't fight sudden changes in subject trajectory as it would in NORMAL mode, and it is easier to keep the AF sensor squarely on the subject's face during the action than with VR OFF, and the viewfinder is more stable so I can get more precise compositions. I do not really notice a decay in sharpness at these speeds due to having VR in SPORT mode. The benefits outweight any drawbacks in my case. If the subject is moving slowly or I am shooting indoors in low light, I may drop the shutter speed to 1/250s or so; this has worked out well but sharpness is typically better at faster speeds. If I have to shoot at 1/160s or slower, I don't have much in terms of expectations. I fire a number of frames in that case and hope to get an acceptable one. This is just as much to subject movement as it is to imperfect stabilization of the lens. A heavier lens such as my VR 200/2 II gives some additional leverage in shooting in low light but even with that, slow shutter speeds often yield images which are affected by subject movement if I go to speeds of 1/100s or so.
I rarely use VR NORMAL on the 300/4 PF. It is intended for relatively static subjects and I prefer to use a tripod for most shots of such subjects, to get guaranteed sharpness. For tripod use I keep VR OFF.
-
I'm not able to judge if there is nothing wrong with this lens but I can say that this behaviour is most probably not wrong, although not welcome.
All lenses have their strong and weak points. When they were mechanically simple, we had to deal with sharpness at a specific aperture or flare or ghosting, etc ... . Lens designers were already making compromises between the many possible flaws. With mechanically complex systems, like lens + VR, they must deal with new categories of flaws, and just like before, they cannot solve everything in one lens.
-
I'm not able to judge if there is nothing wrong with this lens but I can say that this behaviour is most probably not wrong, although not welcome.
All lenses have their strong and weak points. When they were mechanically simple, we had to deal with sharpness at a specific aperture or flare or ghosting, etc ... . Lens designers were already making compromises between the many possible flaws. With mechanically complex systems, like lens + VR, they must deal with new categories of flaws, and just like before, they cannot solve everything in one lens.
Thanks Bruno. I don't think this will be a big issue as I would likely use the lens at much faster shutter speeds than the problematic ones and/or mounted on a tripod. It is always good to know that my copy does not stand out as a "lemon".
Regards,
Paulo
-
The vibration from shutter and mirror has the largest amplitude right after the shutter opens. Over time the vibration amplitude decays. At longer exposures, the largest amplitude vibration affects on the beginning of the exposure so most of the light that is recorded renders the image sharply.
Mirror vibration and shutter vibration probably have different frequency ranges.
So, if you use a solid tripod, EFCS and M-UP you should get sharp pics at all speeds in most conditions (wind may still cause vibration). If you need to hand hold, you should preferentially use a fast shutter speed which takes care of subject movement as well as all kinds of camera shake.
The flash duration is about 1/1000s or it can be even shorter. Thus the vibration doesn't have time to make an impact on the image sharpness.
The 300mm has greater magnification due to its focal length so any vibration is magnified as well. The PF is also a very lightweight lens for its focal length which can accentuate the problem. Less mass, easier to vibrate.
I typically use the 300 PF for photographing moving subjects, mainly people, at concerts, fashion shows, and sports. This is hand-held work. I also occasionally use it for landscape and close-ups (most of the time on tripod, but occasionally hand held with VR).
My most typical shutter speeds for the moving subjects are from 1/500s to 1/1250s, with VR set to SPORT mode. This means the lens won't fight sudden changes in subject trajectory as it would in NORMAL mode, and it is easier to keep the AF sensor squarely on the subject's face during the action than with VR OFF, and the viewfinder is more stable so I can get more precise compositions. I do not really notice a decay in sharpness at these speeds due to having VR in SPORT mode. The benefits outweight any drawbacks in my case. If the subject is moving slowly or I am shooting indoors in low light, I may drop the shutter speed to 1/250s or so; this has worked out well but sharpness is typically better at faster speeds. If I have to shoot at 1/160s or slower, I don't have much in terms of expectations. I fire a number of frames in that case and hope to get an acceptable one. This is just as much to subject movement as it is to imperfect stabilization of the lens. A heavier lens such as my VR 200/2 II gives some additional leverage in shooting in low light but even with that, slow shutter speeds often yield images which are affected by subject movement if I go to speeds of 1/100s or so.
I rarely use VR NORMAL on the 300/4 PF. It is intended for relatively static subjects and I prefer to use a tripod for most shots of such subjects, to get guaranteed sharpness. For tripod use I keep VR OFF.
Ilkka,
Many thanks for your thourough answers to my questions and for sharing the way you use the lens. Your experience with the lens will undoubtely help me getting used to use the lens hand held.
Your explanation also explains why I get better results with hand held VR Normal On with M-UP.
So your copy is also prone to double contours when handheld with VR Normal at around 1/100s?
Regards,
Paulo
-
So your copy is also prone to double contours when handheld with VR Normal at around 1/100s?
Regards,
Paulo
(repeated) side info: occasional double contours in the background are not uncommon with VR. The 16-85 DX tends to that at about +- 125/s.
-
So your copy is also prone to double contours when handheld with VR Normal at around 1/100s?
As far as I know this kind of behavior depends on the specific combination of body sample and lens sample, and it cannot really be said that one lens that works without double lines on one camera would not produce them with another. This is the difficulty with this issue: you can't be sure.
I have not noticed any especial tendency for problems with my copy of the lens and my D810; when I got the lens I did run some tests and noticed that some acceptable results could be had at slow speeds like 1/50s but sharpness tended to improve as I increased shutter speed. I almost always keep the MB-D12 on the D810 when I'm shooting hand held with a tele, this helps stabilize the lightweight, high magnification setup. I usually don't go below 1/200s with my lens so I have very little practical data on those intermediate speeds; at 1/250s to 1/200s it seems to work fine but I prefer 1/500s to 1/1250s. I have used it at 1/125s to 1/160s a few times indoors; I would say that I got some passable results but wouldn't call them really sharp. I have once noticed that when I was shooting a concert my shutter speed was 1/320s-ish, and I tried both VR ON and OFF, and the shots with VR OFF were a bit sharper. However, usually I find the opposite to be the case at these speeds. I don't know - I was pointing the lens upwards; perhaps this affected the performance of the VR system?
However, compared to the older AF-S 300/4D, I get a much higher rate of focus keepers with the new PF lens, and with the older lens things were sharp only at 1/800s or faster when hand holding it (due to the lack of VR), and even 1/640s seemed slightly blurry. The 300/4 PF does give a significantly better results in my typical use, and part of the excellent in focus rate I get with the 300 PF is due to the benefit given by the stabilizing effect of VR, I believe. A part of it is the faster focus motor, and I also think the nature of the PF's in to out of focus transition is such that it gives a particularly clear phase signal, leading to one of the best in focus rates with any lens I have used, and this is particularly noticeable when shooting into backlit subjects. I can easily live with quirky VR at slow speeds. I am aware of it, and I accept that this is a limitation of the lens, and simply do my best to work around those limitations. If you find that every shot at 1/100s to 1/160s is blurry with a distinctive double contour, you may want to try another copy of the lens, as this is not what my lens does on my cameras. It is simply a little less sharp and trustworthy at 1/125 or 1/160s with VR on than it is at 1/250s or 1/500s, or a little less consistent.
For me the 300 PF has been a wonderful addition to my kit. In the summertime I practically always have it in my bag, and I use it a lot. I would like to have the f/2.8 as well, for the higher contrast, and better image quality in dim and soft light, but cannot stand slightly flexible tripod mounts so I am waiting for the next iteration of the series to get that sorted out. (Probably the FL version will be so expensive that I can forget about it once it is launched. ;-)) I don't believe I could ever give up the PF though, it is so much fun to use and makes the bag lighter.
-
Moving from the D7100 to the D500 the VR issue in the SS range of 1/80-1/200s was reduced a lot. So it is indeed camera body dependent.
As most here I usually use it as SS > 1/320s without any issue.
The old 300/4D AFS got me back to Nikon 8 years ago and now the 300/4E VR keeps me firmly in the Nikon camp, my favorite and most used lens !
-
I've had my 300mm PF only for a short while and have also noticed some loss of VR effectivity in the range of +/- 1/125s.
I'm very grateful for all the good advice given in this thread. As soon as it gets warmer I intend to shoot airshows with prop driven planes.
Now for a nice prop disc blur 1/125 to 1/250 should be used. I hope for an acceptable keeper rate. I'll be using it on the D750.
I still have to learn a lot with longer lenses as this is my very first tele ever. Maybe a battery grip could improve the resonance issues.
Does anyone here use the 300mm PF on the D750?
-
I've had my 300mm PF only for a short while and have also noticed some loss of VR effectivity in the range of +/- 1/125s.
I'm very grateful for all the good advice given in this thread. As soon as it gets warmer I intend to shoot airshows with prop driven planes.
Now for a nice prop disc blur 1/125 to 1/250 should be used. I hope for an acceptable keeper rate. I'll be using it on the D750.
I still have to learn a lot with longer lenses as this is my very first tele ever. Maybe a battery grip could improve the resonance issues.
Does anyone here use the 300mm PF on the D750?
Hi Daniel,
I only have the lens since last week and so far only tested it on my D810. I also have a D750 but haven't mounted the 300mm PF on it. I will share my experiences with the D750 when I have some "evidence".
-
Many thanks to all for your insights and recommendations regarding Nikon's VR technology and using the 300mm PF in the field.
I will go on holidays tomorrow and will have the opportunity of using the lens in the real world. So far I think the 300mm is a great addition to my lens collection and I'm looking forward to use it more and learn how to use it properly.
Regards,
Paulo
-
Has anyone tried the 300mm PF handheld on the D7200?
-
There is a longish thread that also mentions VR issues with this lens:
http://nikongear.net/revival/index.php/topic,2077.45.html (http://nikongear.net/revival/index.php/topic,2077.45.html)
-
Many thanks for the link Simone.