Here are some thoughts current as of this posting. This is a complex subject with no clear answers. Thinking has evolved over time but I think the thinking has matured at this point (Dec 2011)

To summarize, the Gitzo Systematic series allows for interchangeable tops, which include the stock flat plate as well as various center columns and a leveler. RRS currently makes their own compatible replacement plates and levelers. The top is held in place with a single very robust bolt that compresses the circular "mount yoke".

A failure could result from either a loose bolt (definitely) or, more speculatively, some minor deviation from perfect spec that causes loss of full contact (or sufficient friction) between the yoke and mating top plate lip.

Some argue that regardless of the quality of the product or the robustness of that yoke bolt, the mere existence of that (very arguably unnecessary) single point of possible failure is unacceptable. I am in that camp. I am very conservative about this but I also know the value of the gear on top of my tripod.

People used Gitzo Systematics for years with no recurring reports of plate separation. My assessment is that no one thought much about this until the most recent generation, for reasons that will become clear.

Reports of plate separation popped up coincident with the new Safe Lock version of the top plate on what is generally (but not always) the "V1" models, for example, on the GT3541LS verses the GT3540LS. The GT3530LS is also a "V1" model, due to peculiarities in the historical Gitzo model nomenclature and their use of "version" as the last digit of the model number. Any Gitzo Systematic with a Safe Lock plate would be included in this group.

(I am not suggesting that the Safe Lock feature is unsafe- it is possible that very slight changes to the machining of the yoke or the lip of the plate itself was the cause, or possibly a short term QC problem, or many other possible speculations- and it would all be wild speculation. But it is a fact that the multiple reports of separation were coincident with early samples of that model series and it was almost immediate. The presence of the Safe Lock upper plate surface merely identifies the era when the problem reports started. It surely is unrelated to this problem)

It is also possible that the specific model era many of us identify with this problem is a mere coincidence. Remember, nothing for certain is known about this potential problem.

I have an older G1410 Series 4 Systematic. I used it for 4 years, at least, before I probably ever even checked that yoke bolt. It was just out of sight and out of mind.

Historically Gitzo has never recommended carrying a loaded tripod over the shoulder. Some other makers of other related products, such as some heads, also do not recommend that practice - when pressed. It seems to me it is a subject most makers want to avoid but when pressed do not warrant their gear for that use. That makes sense because a heavy lens such as a fast telephoto puts tremendous stress on the components in an "unnatural" way, especially when combined with the impact force inherent with the up and down motions from even gentle walking, much less a rapid stride.

Some users of the 'V1' series Gitzos reported loose or less than fully tight yoke bolts out of the box and that was speculated to be the reason for the relative rash of separation reports. It is likely Gitzo responded to that, paying closer attention at the factory. It is also possible that the tripods involved were toyed with at a store and the yoke bolt loosened. We will never know and cannot know the answer to that.

Some users that either reported plate separation or reported the plate "migrating" upward (but not separating) also reported that the yoke was tight and that was something they were very aware of and were careful to tighten the yoke. One of those users is a mod here and I respect his opinion and honesty in his assessment of that. No one will ever know the Ultimate Truth of that aspect of this issue.

Another Nikonian reported success roughing up the mating surfaces of the yoke and mating plate lip. He suggested maybe the machining was "too good", or something similar.

I have not seen plate separation reports for some time now, but I do not crawl the net looking for them either. But I suspect if it were still occurring with any frequency we would hear about it here and elsewhere because any incident is generally widely reported and recirculated. (NSN) came out with an ~$70 add-on plate that eliminates any chance of separation. I have used one since day 1 on my own GT3541LS and I would personally not use a Gitzo Systematic without one because I do carry a 500/4 over my shoulder. I hike my 500 a mile or more on most outings, although I try to use the carrying stance illustrated in my current avatar, which among other benefits relieves lateral stress on the support. However, when working through vine infested areas and other brush it may be necessary to do some sort of over the shoulder carry, or something that puts the same stress on the mount.

Because so many long lens users are now using the NSN solution or something similar, it is likely the "high risk group" has shrunk considerably and that alone may account for the lack of recent reports.

I do run into long lens shooters with the "V1" Systematic models that are using the stock mount and have not had a problem. They are always very surprised to learn of the potential problem. I always show them my NSN plate and give them a brief explanation of the potential problem.

Considering all the above I can only come to the conclusion that I cannot come to any firm conclusion, other than that the inherent design of the systematic is not ideal for over the shoulder carries. There is too much money on top to worry about one-in-a-million lightning strikes. I do not believe enough is known about this problem to attribute it to user error or manufacturing or design defects. I do know that that line is very blurry and it is a fact that the stock systematic is subject to that single point of failure, at best dictated by the tightness of one bolt- that is an accident waiting to happen.

If you are looking for trouble, all tripods have one potential single point of failure, which is the 3/8" bolt connecting the head to the mount. In theory that bolt could shear, resulting in the head plus payload falling to the ground. I'm ignoring that in the following because there is no solution to that potential problem, nor have there been many reports of that bolt failing, except maybe with some very inexpensive tripods. I don't ever recall reading about a Gitzo bolt failing, or the Markins titanium bolt supplied with their Mountaineer replacement plate (TB-20/TB-21/Tb-30).

All tripods with center columns have a potential failure point where the 3/8" head stud screws into the center column. Somewhere in there, there is some glue, epoxy, or some other adhesive holding the female threaded insert in place. Although not commonly reported, I worry about that with my monopod, and would with any center column tripod (including my Series 2 Mountaineer if I had the center column installed). Years ago there was a very widely circulated report of at least one Induro tripod failing in this way.

RRS and Feisol, which both make Systematic type tripods with removable plates, took an approach using 3 grub screws to retain the top plate. There is arguably safety in numbers since it would likely require all 3 screws to loosen before the plate separated and long before that happened the user would or should notice the plate rattling. I have handled both of those maker's products and looked at this issue carefully.

The Markins TH-300 solution for Series 3 Systematics replaces the entire mount and costs $100 more than the NSN solution. For that $100 you get the following benefits:

1. It supports the stock hook, which the NSN plate does not. On the other hand, I have a carabiner looped through my NSN plate and I actually prefer that to the stock hook. That because I can attach small items and do an over the shoulder carry without worrying about things falling off the hook. That is a very personal preference and I can see people preferring the hook over a carabiner.

2. The package is lighter and somewhat cleaner although I think the NSN plate is a clean solution- you would have to look closely to see it.

3. Markins claims improved vibration performance. I have no way to evaluate that, not owning the TH-300. Nor have I seen any attempts yet to do those tests. It would be interesting to do controlled tests but that would require some effort, swapping mounts at least twice.

4. Aside from the bolt issue common to all tripod, there is, in my opinion, no conceivable way for the TH-300 to fail. It is too much robust high grade aluminum.

I believe the Markins Mountaineer replacement plates (TB20/21/30) are similarly safe because although it is a two piece design (it screws onto the mount) the threads involved are too robust to worry about failure and there are far too many turns of the thread to worry about the thing spinning off over time. I have owned one for 7 years now and I have studied this carefully.

Markins also makes the TH-200 for Series 2 Mountaineers, which replaces the mount and the Markins TB-20/21 to create one integrated mount solution. Markins does not currently make a Series 5 replacement so in that case the NSN solution is the only commercial "fix" I am aware of.

How fast does the moon move?

Welcome to my new blog .  I thought this was a good subject for my first entry; something I don't want to have to re-type in forum posts very often!

I want to dispel a frequently misunderstood concept here- the moon does not move nearly as fast as most people think!

The true rate of movement of the celestial sphere (the stars) at declination zero across a 12mpx DX sensor is 1.3634 pixels per second per 100mm of focal length. The moon actually moves about 3% slower (approximately 1/29.5 slower than the stars). The moon moves fastest at declination zero and varies +/- about 27 degrees, so it moves even slower at or near the extremes but the effect is so minor that we can ignore it for our purposes.

That means that the moon moves 2.6 pixels per second with a 200mm lens. That's not very fast and it implies you could shoot the moon at as slow as about 1/2s at 200mm and get only one pixel of blur.

With a 1000mm lens, it implies a movement of 13.3 pixels per second. That implies that a 1/15s exposure will yield about 1 pixel of blur, and a 1/30s exposure about 1/2 pixel of blur. As proof I offer this image:

shot at 1000mm with a 500/4 Ai-P and TC-301 working 1000mm, at 1/30s on a D200. And indeed I also shot images at 1/15s to see how far I could take it and decided that 1/30s was the slowest I could shoot without a very minor increase in blur- and it was very minor and would not be noticeable except with a direct comparison. The image was shot with a D200, which works out to about 1.22 pixels/second/100mm FL of motion blur.

The image was shot at 47 degrees altitude, not ideal but about as good as it gets for that phase of the moon, at declination 26d 47m. It was shot 3 minutes after the start of nautical twilight, about 56 minutes before sunrise.

That image was part of a series of 11 out of 12 days straight (days 16-27 of that lunation) I shot starting Aug 29, 2007. To get 11 of 12 days of clear skies was a miracle that may never be repeated . If I had known I would have such a run I could have gotten the 12th day (lunar day #18 - the 2nd day of the series), which was partly cloudy but with patience could have been shot. But whodathunk?

I can tell you that the above was not a fluke. Over the years I've spent a bit of time verifying the edge of this envelope and adopted 1/30s for 1000mm as my own limit for ideal results. Some of the later dimmer crescents in that series were shot as slow as 1/6s at 500mm with very good results, the seeing being a far bigger issue at the necessarily lower altitudes.

All 11 images (plus the replacement for lunation day 18) are posted to my Phases of the Moon gallery:

with exposure EXIF, which some may find interesting. Some earlier images were shot with a coolpix 990 or 995 afocally coupled to a Questar telescope (with motorized equatorial mount) and should be ignored for our purposes. I personally expose to a certain histogram I like.

All this is irrelevant unless you have a very sturdy tripod and head, in my case an alloy Series 4 Gitzo G1410 and a Markins M20 ballhead. If you get blur it is important to understand why, and it is likely your support or technique, not the turtle's pace of the moon

If your fuzzy moon is not due to it's movement, the next thing to check is focus.  For lunar imaging LiveView is the best thing since sliced bread. I wish I had had a D300 (with LV) when I did that series (all with an MF lens working 500-1000mm and up to F/8 wide open).  If your camera supports LivevView but you thought that was just another gimmick, try it (in tripod mode) the next time you shoot the moon!

Some blame "seeing" for fuzzy lunar images.  In my own very humble opinion, I doubt that "seeing" (atmospheric issues) is a big deal when the moon is over 40 degrees altitude or so- at 200mm. Unless you have very bad seeing and that is usually accompanied by high winds. I say this based on over 100 evenings of lunar imaging I've done at far higher focal lengths. It matters at 1000mm, likely not much at 200. I rarely see good seeing. Shooting it just over the horizon is a different matter - shoot the moon when it is high, at least 40 degrees elevation whenever possible.

If focus is good but your moon is still fuzzy and you have a good tripod, rethink your technique.  If your camera has full mirror up (D300, D200, D2 series, D3) then use it, and count 5 seconds between mirror up and firing the shutter.  If your camera only has "exposure delay mode" then try that, but be aware that the built in 0.4s to 1.0s delay may not be enough, depending on the length and weight of the lens and your support.  If you don't have mirror up, try Long Lens Technique and shoot a lot of images.  Using a remote without delay, hands off, at over 200mm will almost always result in some mirror slap blur.  With my longer lenses (500/4 and 300/2.8) I can often get a better image with LLT than with the remote without any mirror delay.

There are seasons for each phase of the moon when it is highest in the sky. Right now October) the 3rd quarter is best but has to be shot at dawn as was the image below (tonight's moon and that image are both 24 day old moons). The first quarter is worst now, and best in the spring around March. The full moon is best in winter. Crescents are more complicated because of their low altitude and don't quite conform to the basic "seasonal schedule" but are best shot generally in the spring for the new moon and after the summer equinox for the old moon. If you live in Australia (anywhere in the southern hemisphere) reverse all that .

If you are shooting the starry background you can use the same formula (1.3 pixels/second/100mm) as the fastest rate the stars move.  As you approach the celestial poles (North star in the Northern Hemisphere) the rate slows but that is a subject in itself.  Just use this formula unless you have a set of tables and understand celestial coordinates.  These exposures are necessarily long.  You will find that for small enlargements and web page use that you can get away with several pixels of blur before the trailing becomes blatantly obvious.

Here is a sample of the moon shot at 1000mm at 1/30s f/11 with a Gitzo G14120 Series 4 alloy tripod, Markins M20 ball head, D200 and 500/4 Ai-P + TC-301 2x converter.  Just to confirm my computations :-)

 <img src="" width="500" height="500" />