Life Member
  • Content count

  • Joined

  • Last visited

  • Days Won


ilkka_nissila last won the day on 13 August 2014

ilkka_nissila had the most liked content!

Community Reputation

155 of my posts have been liked


About ilkka_nissila

  • Rank
    Master Member
  • Birthday 01/01/01

Profile Information

  • Gender
    Not Telling
  1. You could just add a WR-R10 + WR-A10 to the 10-point connector and it performs the following functions: - triggers SB-5000 flashes by radio, with full functionality, flash control settings in the menu (I have them in my menu, very quick to access). Fires the flashes very reliably indoors and outdoors, unlike optical triggering, no frequent eye closures like the pop up gives, no long recycle times, you can fire at the pace of the remote flash which in the case of the SB-5000 is very fast - allows one to trigger the camera remotely without a cable (using WR-T10 which comes with the WR-10 kit) - allows one camera to trigger other remote cameras. - you can continue to use your old flash in the hot shoe together with the radio triggered remote(s). The SB-5000 is expensive, admittedly, as a new unit, but it does have advantages: the flash is surprisingly small, compared to SB-910/900, it has a fan so it keeps itself cool, and allows fairly rapid firing at or near full flash energy. It has the radio receiver built in and so since the WR-R10 doesn't need a battery either the radio triggering is carried out entirely without requirement of maintaining extra batteries for the triggering devices, just keep the flash itself and the camera charged, and you're good to go.
  2. Since this would require completely immobile subject over time (or expect artifacts), I find very limited interest in such technologies. I'm used to reconstructing stuff from incomplete data and it really is a crap shoot - it can work, or not. Thus it is essential that the process is not done by an automated in-camera algorithm but with plenty of user control on a real computer.
  3. That could be very complicated as Nikon seems to change the protocol quite often to accommodate different lens types. And these protocol specifications are not public. Furthermore, I am sure that Sony wants to make money by selling their own lenses instead of helping Nikon or Canon make money. I believe to achieve good AF performance on a mirrorless camera, it is essential to provide a motor that can perform small adjustments to focus precisely and reproducibly. Ultrasonic motors perform large focus movements quickly but they seem to stutter about a lot when used with contrast detect AF. The new Sony 70-200/2.8 FE includes two focusing motors: one is the ring ultrasonic motor, and the other apparently a linear motor. My interpretation is that the ring motor is used to fast changes over long distances and the second motor performs precise fine adjustments. To get this kind of performance Nikon lenses would need to be revised with similar motor configuration, or at least a stepper motor in place of the ultrasonic. Nikon have been introducing a few "AF-P" lenses with stepper motors but so far they are consumer entry-level zooms. The live view AF on compatible Nikon DSLRs becomes suddenly very similar in performance to mirrorless cameras: no stutter, no noise, the focus just snaps silently in place. I think Nikon made this change to support better live view and video AF but also it is something that is likely needed to support good AF on any future mirrorless platform (through adapter or not). Canon's STM lenses are similar and they also have made several primes with that technology. Sony is definitely thinking out of the box in their numerous new technical solutions but I don't know if the advancements help users of existing non-stepper motor equipped DSLR autofocus lenses. I suspect all-new lenses are required for best performance.
  4. I think shooting at high frame rate is completely pointless at an event like White House press conference. I think 3-5 shots should be enough to capture the spirit of the event visually. There isn't much happening. The noise problem is created by numerous photographers shooting at high frame rate and accentuated by poor positioning of microphones.
  5. There are no native fast superteles (300/2.8, 400/2.8, 200-400/4, 600/4 etc.) available for the mount (with adapters, full performance is not available) so I find it difficult to believe photographers who specialize in sports could switch (not to mention the cost of switching, if the lenses were available). If Sony had launched at least a couple of those lenses at the same time, it might have made a different impression of who the are targeting with this camera. However, for other applications where silent operation is a significant benefit, and fast long lenses are not required, this camera seems very interesting. I imagine some wedding, music, theatre, dance photographers would love it.
  6. Thanks for your comments, Ann, this is a very helpful thread (as are all your threads). I find that auto area AF quickly grabs onto off-center subjects as well; I don't notice a strong preference to lock on a central subject. According to Nikon "The camera automatically detects the subject and selects the focus point. Priority is given to the faces of any portrait subjects detected." I find auto area AF to work well for figure skating pairs and ice dancing, as the subjects are people so there are faces to detect, and I typically want the focus to be placed on whichever subject has a more open face towards the camera in a given situation. The couples spin around themselves and around each other in a rapid succession and I have found that I'm woefully unable to manually move the focus point around fast enough to choose the open face, so I read Nikon's NPS AF tips for the D5 and they recommended Auto Area AF for this. It has worked out well although it can't 100% guarantee the focus is on an open face, it does seem to give a very high focus keeper rate and almost never focuses on the background, so if the face isn't in focus, some part of the couple is. If there are multiple subjects to choose from, auto area AF, in my experience, tends to cycle through them, switching from one subject to another every few seconds. I guess this is very annoying if you are looking to control what is happening but for me it is sufficient that it searches for faces and focuses faster than I can in this situation. I have tried to use 3D tracking with and without face detection for ice dancers in both WIDE and NORM area, with very little success. As the couple spins around fast, it only takes a couple of seconds in some cases before the main subject is lost to the tracking system. A few times I've managed to get the system to track up to 5 seconds or a bit longer than that but this feature doesn't seem well suited to ice dancing. I think the spinning and frequent switch of the position of the partners is too much for it. However, even though it loses track of the selected subject, in most cases it still continues to focus on some part of the couple, so I guess one could not call it a complete failure. It just can't hold onto the selected part of the pair for any length of time. Since the light level is quite low I am not using any TC. I just crop if I have to; unfortunately this is more often than I'd like. I completely understand this; in my case its use relies on a fairly large subject (full body image of a person, with room around the subject) and long enough distance that the subject is mostly within the depth of field and the face is often forward from the rest of the body. I use group area AF as a first choice when I am not able to hold the single focus point steadily on the subject due to the rapidly changing trajectory of the subject (e.g. singles figure skating). I have had much better success using group area AF than dynamic area AF in this scenario but I now understand the distinction between how the two modes are supposed to work and when to use one and not the other. As you say, when the point to be focused on is not the closest point of the subject to the camera, there is risk that group area AF finds the closest point and not the intended point and focuses on it. In this scenario, dynamic area AF works better. When there are a group of skaters, if I want to focus on the middle of three skaters, for example, instead of the one closest to the camera, then dynamic does this well, with the caveat that the photographer has to hold the main point on the subject as squarely as possible. I used this to shoot synchronized skating and found that dynamic 25 point mode did in fact help me focus on the intended skater whereas group would have given less control and often selected a closer subject or arm of the subject. However, for singles I found the movement of the skaters so fast, and following changing trajectories, that I wasn't able to keep the subject in focus using either single point or 25 point dynamic, and group AF rescued the situation excellently. I guess you could call it my favorite for difficult to follow subjects that are following a nonlinear trajectory which I am not familiar with. Yes, this is what I've noticed as well, the dynamic area AF really requires great skill from the photographer in the new cameras. With the D810, it would take quite a while before the camera focused on a background subject compared to D5 where the shift is very quick if there is something to focus on, and the photographer lets the main focus point slip into the background. However, with some practice, this can be used to advantage, as one is able to precisely pinpoint on the subject and yet get some assistance as long as the selected focus point isn't squarely on a detailed area, and there is less waiting time when switching between subjects. I have not yet experimented much with the delay and subject motion setting; I am keeping delay at 2 and subject motion at normal for the moment. I think face recognition is useful if the subjects are people, however, I don't know if it works for animals; probably not so well. The manual is not exactly clear about how face detection is used in different AF area modes; I'm believe single point and dynamic area AF do not employ face detection, but auto Area AF does, and 3D tracking does if face detection is selected in the custom settings. Group area AF is said to use face detection in AF-S mode but nothing is stated about AF-C.
  7. Auto-area AF does not allow the user to specify the subject but the camera decides what is the subject and focuses on it. Furthermore it frequently switches to a different subject when there are multiple potential subjects (e.g., people) in the frame. 3D tracking uses the user's input to identify the main subject and sticks to it as the subject moves in the frame. Auto area AF is continuously evaluating what is the most likely subject and often cycles through the options that the camera has identified. I believe it uses the RGB matrix as does the 3D tracking, and both are capable of prioritising faces as well, if the user sets this option. Group area AF also seems to use the RGB matrix (as it is only available on cameras with high resolution RGB matrix sensor) but the choice is restricted in area to the group. It can use face priority in AF-S mode. For me the dynamic area AF is problematic and for now I have restricted the selection to single point AF, Group Area AF, 3D tracking and Auto Area AF as dynamic is so eager to focus on background features if the point slips even slightly or the subject is too small.