Arca Rm3di

[Frederic]

Agreed on all you said, you've studied the topic a lot more than I have.

I guess my feeling is I doubt LV will deprecate the use of very fine helicals and favor "faster mounts". For instance the RS40 has 360° to go from infinity to 0.9m on the rm3d, it's not *that* insanely precise.
Since I don't shoot landscapes tilt is not something I use a lot. Focus stacking is more useful with high MP backs in architecture, a process where very precise focusing is welcome, whether with LV or not.

Now if that's suitable for landscape work or the OP needs is another matter, indeed.

 

[JGR]

Some really interesting comparisons have been made here, one point regarding tilt and focusing at the hyperfocal distance...

I'm not sure approximations with tilt using printed tables will provide optimal results, particularly with high resolution digital sensors. When I was using the D3x with the PC-E lenses and live view, I noticed a big difference when tilt was slightly out, so much so that I have a handful of images that I have not made available as prints on my website.

Regarding setting the lens focus at the hyperfocal distance, this will set the angle of the focal plane when tilt is applied so why would one set the lens at the hyperfocal distance? Am I right in assuming some of you are doing this as a starting base?

Lets assume we set the camera up to take an image of a sweeping view with no vertical subject matter, the land fairly flat with a mountain in the distance. Assuming the camera is approx 5-6 feet from the ground and set the tilt to 1 degree to bring the foreground into focus. If you have the lens set to hyperfocal distance, aren't you going to be throwing away a certain degree of depth of field due the the angle of the focal plane being to steep? By adjusting the focus nearer to infinity, you would be adjusting the focal plane to better suit the typical scene described.

I think regardless of which of the two cameras (techno & rm3di) is used, tilt will need to be set using the ground glass for precision, which one would be better would come down to the ground glass on each camera, loupe used and one's eyesight. This is were I think live view would be a welcome feature for these tech cams.

That been said, doesn't the IQ back's have this feature now? or is the refresh rate still too slow?

 

[torger]

Regarding setting the lens focus at the hyperfocal distance, this will set the angle of the focal plane when tilt is applied so why would one set the lens at the hyperfocal distance? Am I right in assuming some of you are doing this as a starting base?

Lets assume we set the camera up to take an image of a sweeping view with no vertical subject matter, the land fairly flat with a mountain in the distance. Assuming the camera is approx 5-6 feet from the ground and set the tilt to 1 degree to bring the foreground into focus. If you have the lens set to hyperfocal distance, aren't you going to be throwing away a certain degree of depth of field due the the angle of the focal plane being to steep? By adjusting the focus nearer to infinity, you would be adjusting the focal plane to better suit the typical scene described.

I think regardless of which of the two cameras (techno & rm3di) is used, tilt will need to be set using the ground glass for precision, which one would be better would come down to the ground glass on each camera, loupe used and one's eyesight. This is were I think live view would be a welcome feature for these tech cams.

That been said, doesn't the IQ back's have this feature now? or is the refresh rate still too slow?

I've attached an image showing what happens with the wedge if you focus at hyperfocal distance. It's ideal for scenes with horizontal ground and covering as much height as possible. But as you say, if you don't have anything high objects close-by it's usually better to aim the focal plane straight at the horizon. This is also easy to do with an RM3Di, just set the focus at infinity and you aim (approximately) at the horizon. To have some alternatives in-between one could make a table or use a tilt DoF app, so you could use your Leica Disto to aim and find an angle, and check in the table which focus distance you should use for that angle and tilt combination. The problem with these inbetween angles is that unlike hyperfocal/infinity the focus plane angle will depend also on the amount of tilt used (ie when you change tilt, you need to change focus distance to keep same focus plane angle), so a table would be messy. Maybe a DoF app with tilt support is the way to go.

Depending on the shape of the foreground you can make all sorts of of interesting compromises. The messy/impossible forest scene I've attached I shot with a 90mm at about 4 degree tilt f/22, also tried to focus stack it (which worked) but I liked the tilted one-shot image better despite that resolution suffers a bit from diffraction. In this case I chose to set the focus plane along the log going out in the right corner and aim it a bit high, the water below is slightly out of focus but I think it gives a nice 3D feel to it.

One thing you can do with a the Techno and other view cameras which you can't(?) with RM3Di is that you can actually aim the focus plane downwards even with forward tilt. Not that you need to use it often, but you can . The thing is that in order to do so you need to move the lens *closer* than the infinity setting.

The IQ liveview is like the ground glass -- some say it's perfectly useful and they use it all the time, and some say it's next to useless. I have not used it myself so I don't know what I would think. The problems are the usual, slow refresh rate, requires ND filters in bright light, and too noisy in poor light. (The best liveview is made by Canon I think, considerably smoother than Nikon's.)

 

[f8orbust]

Once you start tilting the lens, focus effectively becomes the means for changing the angle of the plane of focus about the hinge line, thus I've never really subscribed to talking about 'hyperfocal distance' when tiliting the lens - unless what is meant is the focus distance that maximises the depth of field 'wedge'. Typically, the most simple implementation is to focus at infinity, tilt the lens so the hinge line is at your feet and then the plane of focus will be parallel to the ground. However, this means you're losing out on the dof behind the plane of focus. It's better to have the plane of focus angled up in order that the far limit of the dof behind the plane of focus is just in the ground (I guess whatever this distance is could be called the 'hyperfocal' distance). For instance, using a 35mm lens on a DB with a CoC of 0.003mm, with the lens 120cm off the ground and shooting @f11, focussing at about 3m and adding a forward tilt of 1.7 degrees will put the hinge line at your feet and the far limit of the dof just into the ground, with the near limit at about 40 degrees to the horizontal. The actual plane of focus will be at about 22 degrees to the horizontal. It all sounds a bit hellish, but in practice focus at about 3m and tilt the lens 1.5 degrees and you should be good. Focus too far and the plane of focus will be shallower (further edge of the dof deeper in the ground); focus too close and the plane of focus will be steeper (further edge of the dof above the ground). So, if in doubt, focus a little bit too far away.

Tilts/swings done on the fly (i.e. with no chart of presets to hand) really need to be observed, so you need to shoot tethered or use a ground glass to see exactly what's going on, especially in systems where you're going to possibly need to refocus as well, since you're effecting a base tilt (e.g. long lenses on pancake cameras). Given the size of todays average ground glass, this takes a good loupe. The new Linhof and Studio / Silvestri one, 12x with a square base (for getting into the corners) is perfect. If you use an IQ back then you also have live view as an option, or you could use the shoot-check-adjust-shoot-check-adjust routine. I find a ground glass infinitely more satisfying than the shoot-check-adjust routine.

In terms of setting fractions of a degree, I found the Techno to be easier than the RM3d/i, simply because the gearing is finer. The current set of Alpa T/S adapters allow the finest setting of T/S, but unfortunately are limited to only certain lenses.

Disclaimer: Throughout all this, I have assumed the sensor plane is vertical and the ground plane is perpendicular to it.

 

[f8orbust]

Because each lens only uses a 'range' of the throw of the helical on the RM3D/i, it is actually possible to bring a lens closer than it's infinity position.

The RM3D/i has clearly been designed as a camera on which it is possible to repeatedly set certain distances with greater accuracy than any other tech camera. In this it excels. However, what you gain on the swings you lose on the roundabouts, and because the helical pitch is so fine, you lose the ability to 'pop' areas of the image in and out of focus quickly - which is what you need to do to initially find focus on the groundglass. I found trying to move the focus ring quickly enough to achieve this impossible. It was designed to be used with a table of distances, and this is how it operates the best. Just don't leave them at home.

This also affects tilts done on the fly, as once I've set a tilt I like to move the angle of the plane of focus fairly quickly (initially) so my eye can pick it up and see where it is and what it's doing.

If I was an RM3D/i landscape shooter, I would probably have a chart of tilts prepared, for certain focus distances and lens heights. These would do for, maybe, 75% of the situations encountered. For 100% - and as a landscape shooter specifically - I still think you need to be able to use the groundglass and snap focus quickly. The groundglass that was on the RM3D/i I used was 'ok' but nowhere near the newer Linhof or Alpa ones - it would be one of the first things I would replace.

 

[torger]

Focus stacking is more useful with high MP backs in architecture, a process where very precise focusing is welcome, whether with LV or not.

If one wants to use focus stacking the helical focusing is indeed nice.

I've done focus stacking with the Techno, and it works but is kind of hackish. What you do is that you tape a ruler to the base, find the starting and end-point of your scene, note where on the ruler it is, and then shot at f/16 and move 0.5mm between each shot. It works well, but with close foreground it can be many shots. For f/11 one have to reduce movement to about 0.3mm between shots and for that I'd like to have some special ruler with 1/3 mm markings.

As far as I've understood most that do focus stacking with an RM3Di rather than covering the whole range they pick out some important foreground object and some important background object and snap a picture for those, and let stuff inbetween be a bit out of focus. The problem is that if you want to cover the whole range you may quickly end up having to shot 10+ pictures for one scene. In a scene where you have a tree up close, then a stretch of water, and then some background on a distance this two-shot focus stack will work very well. In an unbroken line of landscape it may not work so well. This type of stacking can be done on a Techno too but is not very practical, as you'd have to slide in the glass and refocus to take the next shot, which is a bit slow and there's a high risk of disturbing the camera.

Stacked scenes rarely is rarely free from artifacts (due to focus-breathing, defocus shadowing caused by close objects, movements by wind etc) either, and to me I rather make a slightly diffraction-suffering image (smaller f-stop) and/or a tilt "compromise" than risk having multi-shot artifacts. From an artistic standpoint I also like the concept of capturing the scene in one exposure.

But again, it depends on shooting style, some prefer stacking ahead of tilt also in landscape.

 

[stngoldberg]

Because each lens only uses a 'range' of the throw of the helical on the RM3D/i, it is actually possible to bring a lens closer than it's infinity position.

The RM3D/i has clearly been designed as a camera on which it is possible to repeatedly set certain distances with greater accuracy than any other tech camera. In this it excels. However, what you gain on the swings you lose on the roundabouts, and because the helical pitch is so fine, you lose the ability to 'pop' areas of the image in and out of focus quickly - which is what you need to do to initially find focus on the groundglass. I found trying to move the focus ring quickly enough to achieve this impossible. It was designed to be used with a table of distances, and this is how it operates the best. Just don't leave them at home.

This also affects tilts done on the fly, as once I've set a tilt I like to move the angle of the plane of focus fairly quickly (initially) so my eye can pick it up and see where it is and what it's doing.

If I was an RM3D/i landscape shooter, I would probably have a chart of tilts prepared, for certain focus distances and lens heights. These would do for, maybe, 75% of the situations encountered. For 100% - and as a landscape shooter specifically - I still think you need to be able to use the groundglass and snap focus quickly. The groundglass that was on the RM3D/i I used was 'ok' but nowhere near the newer Linhof or Alpa ones - it would be one of the first things I would replace.

Since I am more than 55 years away from my last physics course, I tend to take a less technical approach to get the best focus.
1. I laser measure the midpoint distance of the area I want to maximize focus and coordinate my helical ring to the number on the chart supplied by Arca
2. Tilt my lens 1 degree per every 32 mm of focal length of my lens
3. Set my aperture to f8
4. Take an image and check focus on LCD in the critical areas of the image (when I am indoors I prefer to tether to my computer using Phocus) Those of you to have a phase one IQ DB have it ready made with the focus mask-I envy you
5. Usually I find focus fine at this point, but occasionally I find that f11 is necessary and I once had to use f16
For a 73 year old grandfather of six this avoids the hinge line and all the other jargon which I don't choose to master at this point in my life
Stanley

 

[torger]

Since I am more than 55 years away from my last physics course, I tend to take a less technical approach to get the best focus.
1. I laser measure the midpoint distance of the area I want to maximize focus and coordinate my helical ring to the number on the chart supplied by Arca
2. Tilt my lens 1 degree per every 32 mm of focal length of my lens
3. Set my aperture to f8
4. Take an image and check focus on LCD in the critical areas of the image (when I am indoors I prefer to tether to my computer using Phocus) Those of you to have a phase one IQ DB have it ready made with the focus mask-I envy you
5. Usually I find focus fine at this point, but occasionally I find that f11 is necessary and I once had to use f16

I understand the tilt part, it brings hinge line approximately to the ground for eye-level tripod height, but the way to set focus distance seems to me that it would cause a quite random result. After tilting the focus distance + tilt angle in combination decides the angle of the focus plane which may not have anything to do with maximizing sharpness at the original focus distance.

However, if you often shoot wide angle that first chosen focus distance will probably often be past the hyperfocal distance and then the focus plane angle quite slowly lowers towards straight out at infinity, so it will work quite okay. But, if focus distance happens to be shorter than the hyperfocal distance it is likely that the focus plane shoots off into the sky.

A more repeatable method would be to always focus at infinity when tilting, or hyperfocal. "Hyperfocal" is a bit controversial though as it depends on the circle of confusion, and a suitable size of that in the digital world is something many of us like to discuss . (The distant horizon is put at the edge of the DoF with a hyperfocal tilt-wedge and that may not be what you want if you actually see distant horizon)

 

[RodK]

I guess you've seen this intro video? Very informative on how the system is used: An introduction to Arca Swiss R cameras by Rod Klukas on Vimeo

I also like this Luminous Landscape review video Arca Swiss Rm3d Review on Vimeo (note that the reviewers are both dedicated ALPA users)

I have not used the RM3Di, but while waiting for responses from real users I can present some aspects "good to think about" concerning what the camera is capable of.

* tilt is built-in to the body which is a great cost saver compared to Cambo and Alpa.
* tilt is only +/- 5 degrees. For longer lenses/close work (shooting flowers etc) this can be a limitation, but for typical landscape grand scenes its not
* tilt cannot be made diagonally, only forward/backward or left/right, typically not limiting for grand landscape scenes though, as it is almost always about small tilts forward.
* compared to the Techno or other view camera the lens mount cost is high, if you want many lenses the lens cost gets higher.
* if you like to carry around long lenses it can get a bit bulky
* if you want to make pano stitches I prefer a sliding back with click-stops (Techno) rather than gear-turning on the back (RM3Di) as it can be a bit slow
* while the primary operation of the camera is with a viewfinder and thus approximate framing, some do use ground glass for framing, and the rotaslide sliding back can be used if you want to (I've got conflicting info if it's compatible with infinity focusing or not though, I think it is but needs double-check).

I'm a Techno user myself and of course hope that you would choose that instead ;-). Concerning pancake cameras RM3Di would have been the number one alternative though, due to the integrated tilt and high precision focusing, and possibility to extend the system to include an MF-two view camera for greater movement flexibility if would need it.

Movement comparison:
RM3Di: Back Vertical Rise/Fall 30/20mm, Back Lateral Shift +/- 15mm. Tilt +/-5 degrees. Camera can be mounted on the side or upside down to change vertical shift and tilt direction.

Techno: back vertical shift +/-20mm, front rise +20mm, Back lateral shift fixed click-stop on sliding back only -17/0/+17mm, Front lateral shift +/-10mm. Tilt +/-10 degress, Swing +/-10 degrees.

The Techno advantages are best shown if you use many focal lengths including longer ones and often use tilt, and if you would use the ground glass regardless of camera. If you work mostly/only with wides, dislike the ground glass then RM3Di is best. If you shoot at f/8 or f/5.6 often you will love the focusing precision of the RM3Di, if you shoot at f/11 my experience is that the ground glass focusing precision is adequate, but I need a 20x loupe to be satisfied (most only use 6x - 10x)

I am sure you know what you are doing, but a 20x loupe creates more focus problems than it solves as it magnifies the glass pattern so much it can interfere with the image you are trying to focus.
With standard to longs it may be OK, but with wides I believe it becomes a liability. But if you feel it works for you, then stick with it.
That is why most stay below 7x for a magnifier.
Rod

 

[RodK]

Once you start tilting the lens, focus effectively becomes the means for changing the angle of the plane of focus about the hinge line, thus I've never really subscribed to talking about 'hyperfocal distance' when tiliting the lens - unless what is meant is the focus distance that maximises the depth of field 'wedge'. Typically, the most simple implementation is to focus at infinity, tilt the lens so the hinge line is at your feet and then the plane of focus will be parallel to the ground. However, this means you're losing out on the dof behind the plane of focus. It's better to have the plane of focus angled up in order that the far limit of the dof behind the plane of focus is just in the ground (I guess whatever this distance is could be called the 'hyperfocal' distance). For instance, using a 35mm lens on a DB with a CoC of 0.003mm, with the lens 120cm off the ground and shooting @f11, focussing at about 3m and adding a forward tilt of 1.7 degrees will put the hinge line at your feet and the far limit of the dof just into the ground, with the near limit at about 40 degrees to the horizontal. The actual plane of focus will be at about 22 degrees to the horizontal. It all sounds a bit hellish, but in practice focus at about 3m and tilt the lens 1.5 degrees and you should be good. Focus too far and the plane of focus will be shallower (further edge of the dof deeper in the ground); focus too close and the plane of focus will be steeper (further edge of the dof above the ground). So, if in doubt, focus a little bit too far away.

Tilts/swings done on the fly (i.e. with no chart of presets to hand) really need to be observed, so you need to shoot tethered or use a ground glass to see exactly what's going on, especially in systems where you're going to possibly need to refocus as well, since you're effecting a base tilt (e.g. long lenses on pancake cameras). Given the size of todays average ground glass, this takes a good loupe. The new Linhof and Studio / Silvestri one, 12x with a square base (for getting into the corners) is perfect. If you use an IQ back then you also have live view as an option, or you could use the shoot-check-adjust-shoot-check-adjust routine. I find a ground glass infinitely more satisfying than the shoot-check-adjust routine.

In terms of setting fractions of a degree, I found the Techno to be easier than the RM3d/i, simply because the gearing is finer. The current set of Alpa T/S adapters allow the finest setting of T/S, but unfortunately are limited to only certain lenses.

Disclaimer: Throughout all this, I have assumed the sensor plane is vertical and the ground plane is perpendicular to it.

The gear ratio of the techno is not finer than the RM3di. This is explained above...

 

[Jae_Moon]

I've had my Rm3di for as long as Jack has, since August of 2011. I had F-Metric Compact 6x9 before that and used both camera with P45 DB.

The focusing is the main challenge with both View and Technical camera whether one is using Tilt or Swing or not. Tilt and/or Swing just makes the challenge bigger, at least by a magnitude of 10. None of currently available methods of focusing; GG, LV, tethered shoot and using a laser range finder, are a perfect solution.

I am only going to talk about 'focusing' here, not 'composing' or 'Tilt.'

For me at least, GG is not an option. I tried too many different types of GG and loupes with my F-Metric and gave up on the idea of using GG with MFDB. Personally, I only have a limited experiences with the LV with MFDB, and I hope that we may have the 'perfect' solution soon. A tethered shoot provides the best possible focus but by means of trials and errors.

The reason why I picked Rm3di is based on the possibility of achieving the perfect focus every time utilizing its very precise 'helical' focus ring and precise measurement of object distances. It is my assumption that Arca-Swiss originally designed its Rm3di with the same design criteria. Arca-Swiss is finally making available 'E-module', which I am only familiar with by reading about it, to measure 'precise' distance.

Both ALPA and Arca-Swiss are providing means of 'precise' focus; HPF rings from ALPA and a built-in Helical focus from Arca-Swiss, with different twists. There are two significant differences between two methods; rotational pitch and indexing method. Many are focusing on either advantage or dis-advantage of the larger rotational pitch of helical focus ring of Arca-Swiss over the HPF ring of ALPA.

I will use the example of using SK35XL on both mounting methods.

As can be seen, Arca-Swiss helical ring needs to rotate approximately 4 times as much as HPF ring. You make your own conclusion if it is good for your workflow or not.

Another difference is the method of marking distance scale. HPF ring is marked with the actual object distance for a particular lens while A-S Helical ring is just simple index (0 to 34.4 for each revolution, up to 5 revolutions if necessary).

Each has both pros and cons; HPF ring eliminate one step of 'look up table' that A-S uses; measure the distance with a laser range finder then dial in the distance while you have to look up a table which converts the measured distance to a focus setting for Arca-Swiss. On the other hand, HPF ring only shows 9 discrete distance marking between 2.5 m and infinity while the look up table from Arca-Swiss shows 34 discrete focus settings for the same distance range. HPF ring has tick mark for each angular degree but has the actual distance marking at every 5 degrees (except first 5 degrees from infinity). For example, it show the distance setting of 9.64 m at 5 degree and 4.86 m setting at 10 degree, therefore, you have to make a quick interpolation if your distance measurement is 7.5 m to find the correct focus setting. Done? The answer is 6.5 degree. ALPA makes it available a print out of HPF ring setting for each degree in PDF format for each lens, and they are 7 pages per lens.

The real advantage of the simple indexing used by Arca-Swiss is when one takes advantage of an iPhone and some mathematic. It is possible to develop a 'lens equation' which converts the object distance into the required angular rotation for an individual lens, and run it on an iPhone. That's what I had in mind when I chose Rm3di and I've been using it from the beginning. Measure the distance with a laser finder, then tap in the number into an iPhone and it will show the exact focus setting; i.e., White 7.3

The development of 'lens equation' requires the combination of 'optical lens equation' and 'curve fitting' by numerical analysis. A simple 'optical lens equation' does not work in entire object distance range since the modern lenses for MF are very complex with multiple lens elements. It is necessary to measure the actual focus settings for different distances or just use the data provided by Arca-Swiss. I did my own measurement.

Can this method be used with HPF ring? It can be done if the distance markings are replaced with a simple angular degree indexing of 0 - 270, by pasting a simple adhesive tape over distance marking. It is not easy to find 37.5 degree from the scale marked with 9.64, 4.86,3.27, etc.

For me, Rm3di provides the best solution to make perfectly accurate focus under the presently available technologies; Leica Disto D5 and a home brewed program on an iPhone.

I will discuss the subject of Tilt or Swing on a separate post.

 

[torger]

Very interesting description of rm3di use, thanks! I have come to realise that focusing is a very personal issue, each have their own ways. I actually find it easier to succeed when tilt/swing is involved. It may be hard to succeed to cover all objects with the wedge, but actually placing it is often easier than focusing flat.

We often think of tilt in the standard upright grand scene situation, but if you have a shooting style when you use longer focal lengths to make small cutouts of semiabstract parts of the ground or similar you often have the whole camera tilted and need to tilt the focus plane too, these scenarios cannot be solved with a table, but gg must be used. Same for (semi) closeup work. The Techno with its gg is a bit more difficult to use but is more open-ended concerning shooting style.

 

[jagsiva]

The real advantage of the simple indexing used by Arca-Swiss is when one takes advantage of an iPhone and some mathematic. It is possible to develop a 'lens equation' which converts the object distance into the required angular rotation for an individual lens, and run it on an iPhone. That's what I had in mind when I chose Rm3di and I've been using it from the beginning. Measure the distance with a laser finder, then tap in the number into an iPhone and it will show the exact focus setting; i.e., White 7.3

The development of 'lens equation' requires the combination of 'optical lens equation' and 'curve fitting' by numerical analysis. A simple 'optical lens equation' does not work in entire object distance range since the modern lenses for MF are very complex with multiple lens elements. It is necessary to measure the actual focus settings for different distances or just use the data provided by Arca-Swiss. I did my own measurement.

+1

With the eModule, the number of data points and the specific lens behaviour built into the numbers takes the precision to a new level. I am still learning to use mine, but focussing on close objects is just so much easier, and I can do with a much higher level of confidence.

 

[f8orbust]

The gear ratio of the techno is not finer than the RM3di. This is explained above

For the tilt mechanism? Really? From memory, it certainly didn't feel that it was, and given the limited movement of the thumb wheel on the RM3D/i I would be surprised. I felt it was possible to set tilt more precisely with the Techno. A/S should replace the white dot with a small horizontal line or an arrow, and add some divisions between degree markers IMO. Also, tilting on the RM3D/i whilst closely observing the ground glass wasn't exactly easy either (since the control is on the bottom, front, left of the camera) - on the Techno the control is more 'naturally' positioned and easier to operate, since you can grip the whole knob easily - even with gloves on (Full disclosure: I am right-handed).

That is why most stay below 7x for a magnifier.

Try the new 12x loupe from Silvestri - it's a gem for nailing focus, and with its square base you can get right to the corner of the ground glass. The issue of magnifying grain is generally overstated - you can use quite a high powered loupe and still observe changes in the plane of focus.

 

[archivue]

While i enjoy using the rotaslide for composition with my Rm3d, i'm really pleased not to have to use a loupe all day long... the index' system rocks !

 

[torger]

I am sure you know what you are doing, but a 20x loupe creates more focus problems than it solves as it magnifies the glass pattern so much it can interfere with the image you are trying to focus.
With standard to longs it may be OK, but with wides I believe it becomes a liability. But if you feel it works for you, then stick with it.
That is why most stay below 7x for a magnifier.
Rod

I do know what I'm doing . You can read about my focusing tests at:
Review: Linhof Techno in the appendix section. I don't just *feel* that it works, I've made a controlled repeatable test and shown that it works. I haven't tried Arca-Swiss's ground glass though, some say it's not as good as Linhof's

Quite simple: 7x loupe = relatively high risk to miss focus, 20x loupe = very low risk to miss focus. There are people that are more skilled than me though and can do with the 7x nicely, or have a more relaxed view than me on focusing precision. However, a user that feels that focusing is not confident on a view camera really should try using a high magnification loupe. Saying that it won't help or even makes focusing worse without actually testing it is not fair.

Most stay with a 7x loupe because of this grain myth that noone cared to put to the test. I can understand back in the 8x10" and 4x5" days that people did not want to use more than 7x, you really did not need more and then it's nicer to have a wide field of view and grain-free image. I'm surprised that the same large format film recommendation is repeated for digital where the format is much smaller. Of course you need more magnification.

Actually, if optics made it possible an ideal magnification would probably be around 40x, but it is difficult to make such a loupe, you would then need to design it as a pen microscope, and the tiny field of view could be a problem.

And yes you do see grain, but the effect is similar to the difference of watching a picture on a tiny screen like on an iphone, and then watch the same image on a larger screen, with the same resolution, you still get a better understanding of the image.

The goal is not to see a pretty picture, the goal is to see when the focus peaks (=contrast peaks), and it's considerably easier to do that when you *without effort* see a larger image with grain and all, also with wide angles. Wide angles does not make it different, oh well one thing is different, the high magnification loupes cannot be tilted so I do have a 10x tiltable loupe to for looking into corners of wides.

If you do not believe me I can only recommend to repeat the test described in my Linhof Techno review, where I clearly focused with higher precision when using the 20x loupe rather than the Silvestri 10x (which is more like a 6x though).

The attached image shows a realistic simulation of what you see through the 20x loupe, based on a true macro photography of the Linhof Techno standard ground glass. Yes the image is a bit grainy, but contrast differences when turning the focusing knob is much easier to see than with a smaller magnification. (The image is showing a detail of a Swedish bill which is quite low contrast print in itself.)

 

[torger]

By the way, here's an example of a shooting style that fits a view camera very well, represented by the excellent British landscape photographer David Ward:

Into The Light | Gallery

Actually he uses 4x5" Technikardan as he loves Velvia 50 (which was to be discontinued, but is now back, woohoo!), but digital it could be a Techno. Why I mean this style is well-suited to view camera is that you have many images with no visible horizon, camera pointing down and flat(ish) subjects that require tilt. Shooting that style with a RM3Di would be a waste, as you would have to use the ground glass anyway.

I've attached one of my own images too. It was shot with a 120mm lens, 10mm fall, the camera pointed downwards, and ~2 degrees of tilt. How would you shoot that with an RM3Di? Using the ground glass I suppose.

Not saying that everyone should point their camera towards the ground and shoot pebbles, leaves, rocks and minimalist patterns, but as you do get to pay a lot for the helical focusing mounts you should think about what you want too shoot and how often you end up using the ground glass anyway.

 

[photomgraphy]

once i found a very interesting article on luminous landscape about focus a tilt lens for 35mm cameras: Focusing Tilt Shift Lenses

the author - david summerhayes - provided a formular to calculate the angle of tilt needed depending on the hight of the optical axis above the decided plane of focus. (sensor plane perpendicular to it).

just use the arcsine of your lens' focal length divided by the distance from the object plane to the axis of the lens.

measure the hight of the middle of your lens above ground and calculate (or read a selfmade table based on the formular) your tilt angle to get the ground in focus.
if you are using a standard setup with the sensor plane leveled, not pointed downwards no problem.
but theoretically it should be no problem to use this method when pointing the whole camera downwards too - just measure the distance to the ground not vertical but in the same angle in which the sensor plane is tilted to the ground. should be the same affect as if you were a little higher with vertical sensor (meaning slightly less tilt angle), or am i totally wrong?

i tried it once and it worked very well.
in his description he first gets the right angle and then focuses anywhere in the scene where he can find good contrast.
that's a little advantage to groundglass users ;-)

 

[f8orbust]

An easy formula to calculate the tit required to put the hinge line at your feet:

t = 90 - arctan h/f

where t = tilt; h = height of the lens and f = focal length at infinity. Assumes this distance (f) is the same as the focal length i.e. not applicable for some lens designs. Also assumes camera back isn't tilted.

e.g. a 120mm (=12cm) lens which is 120cm off the ground. Then h = 120, f = 12:

t = 90 - arctan 120/12 = 90 - arctan 10 = 5.7 degrees

...or you can just use the ground glass

 

[photomgraphy]

For me, Rm3di provides the best solution to make perfectly accurate focus under the presently available technologies; Leica Disto D5 and a home brewed program on an iPhone.

would you mind to provide your little home brewed program?
unfortunately i'm not able to program one ...