Alpa Max Vs Arca Swiss Rm3d

[Jack Flesher]

I feel the need to jump in and clarify a few things since there seems to be some misunderstandings on how the Arca works.

First, the focus helical on the Arca is part of the body and is therefore "generic" to all lenses. To get from infinity to really close, you may need all 4 revolutions with a longer lens.  However with a shorter lens like a 40, you get from 0.5M to infinity with about one turn.  The knob for that "turn" is about 4" in diameter and has 34 main graduations around the ring, with 5 divider hash points marked off between each whole number.  You can easily interpolate to between each hash mark for basically 10 graduations between each whole number, or 340 distinct positions for one full revolution.

That may be a more fine set of gradations than most folks want, but I can tell you that with a 40mm lens over the IQ180 back, you can easily "see" the exact focus plane move pretty significantly in the image between one whole number mark.

Moreover -- and this may be the most salient point -- the Arca has tilts.  When you impart tilt, even very minor adjustments to focus distance can have a magnified effect on the relative position and effect of that now tilted focus plane.

Thus, for a camera with tilts, having very precise, fine focus control is a clear benefit.  With a camera that does not tilt, it is possibly less necessary -- meaning close enough often is.  So while the existing arrangement for the Alpa without tilts may be more than adequate, the extra resolution on the focus knob of the Arca is welcome, especially if you impart tilts.

Finally, the amount of "time" it takes to do one complete revolution on teh Arca is trivial -- we're talking a fraction of a second.

Hope that helps clarify,

[Wayne Fox]

Thanks Jack.  I've never really tried using tilts much for focus, but always assumed once you start tilting things, distance scales sort of go out the window.  Maybe I'm wrong and the trick is being able to know exactly where to measure when figuring out your focus plane.

As far as precision, maybe some clarification on the Alpa as well.  I only have the HPF rings for my 70mm.  When shooting with the 47 or 35 I'm either focus stacking, using a limited version of hyperfocal distance (basically I shoot each lens at f/11 and f/16 and figure out where infinity fails to satisfy me and set it there), or I know there isn't anything close enough to worry about it.  So I haven't put the HPF rings on the wides (yet anyway).

  On the 70mm, the focus ring turns over 3/4's of a revolution and does make the ring much larger although prob more like 3".  The HPF ring divides that into 50 increments, and there are 5 sub increments. It's actually quite easy to set the lens "between" each subincrement about the thickness of the line so the line covers the space between the two. However, the precision at close range is rather impressive, 10 of the 50 cover only 2.65 feet up to 3.31 feet, so 50 minor steps to go 9 inches.  The next 10 get you to 4.4 feet, and overall 40 of the 50 major steps, or 200 of the 250 minor steps covers the range from 2.65 feet 16.6 feet - so thats 200 steps to cover about 167 inches.  Of course, one challenge that makes them less useful than what I understand about the arca is the fact they are non linear, so each substep doesn't represent "x" inches.  This means you have to sort of look at the two major steps, do some math in your head, and figure out what each substep represents in distance. Example, on the closest subset, each step represents 0.12" while at the 40th major step, each substep represents  about 5.5" so even if you factor in using in between each sub-increment for distances between 14.3 to 16.6 feet your control is probably only around 2.5" per step. From there it scales and become pretty inaccurate so if you need precise focus at 192 feet, tough to do on the Alpa with the HPF rings ... so then your having to depend on whatever tools the back has for you.

They certainly are both great systems.  I know alpa prides itself on precision, and won't do sliding backs.  After calibrating mine and realizing .01mm spacing between the back and the body does make a difference I can see their point.  Some method to account for that is critical, sounds like Arca's system works just fine. I certainly hope Phase realizes it the precision problem (certainly Joseph Holmes has documented it well). The DF really should have a focus calibration function much like dSLR's, to compensate for this as well as the difference between focus screen and sensor.

[Rod.Klukas]

For anyone interested that may have missed it, I have published a video on the Arca-Swiss R system on my blog:

http://rodklukas.com/wp/arca-swiss/video-introduction-to-arca-swiss-r-cameras/

The first half is filmed in the studio - it takes you through the parts and accessories of the Rm3Di camera and how they function. The second part is making a photograph in the field.  This will explain the system in some detail. Take a look, it may clear up your questions.

[Chris Eyrich]

I feel the need to jump in and clarify a few things since there seems to be some misunderstandings on how the Arca works.

First, the focus helical on the Arca is part of the body and is therefore "generic" to all lenses. To get from infinity to really close, you may need all 4 revolutions with a longer lens.  However with a shorter lens like a 40, you get from 0.5M to infinity with about one turn.  The knob for that "turn" is about 4" in diameter and has 32 main graduations around the ring, with 5 divider hash points marked off between each whole number.  You can easily interpolate to between each hash mark for basically 10 graduations between each whole number, or 320 distinct positions for one full revolution.

Thanks for pointing this out, Jack, this is a common error in Arca-Alpa discussions: You don't use 5 revolutions with each and every lens.

If you have a lens with a small hub, you stay within one revolution, e.g., with the Schneider 28mm. The supplied distance table for the SK28 runs from 0.5m (on the helicoid: 23.2) to 25m (0.4), so this is less than a full revolution. If you look at the Arca-table for the SK43, you go from 0.5m (RED 26.7) to 25m (1.1), that is, you are close to two revolutions. If you look at the table for the Schneider 90mm, you use use up to five revolutions but if you focus somewhere between 3.4m and infinity, focusing takes place within one revolution. Put differently, the precision of the Arca way of focusing increases with a large focus hub BUT one must keep in mind that you have large movements in the close range and these get smaller if you approach infinity.

One more point: Whether focusing an Arca-lens (without the E-module) or focusing an Alpa-lens with a HPF ring using a Disto, you have to have the respective table at your hands. Take the SK 43mm: on the Alpa HPF you have an explicit mark for 7.29m and another one for 4.89m. The table will tell you that the small marks between correspond to 6.63m, 6.09m, 5.63m, 5.23m respectively. On the Arca table, looking at approximately the same range, the following distances and corresponding values are given: 7.5m (3.7), 7m (4.0), 6.5m (4.3), 6.0m (4.7), 5.5m (5.1), 4.8m (5.9).

It's worth having a look at the Alpa-HPF distance tables to get a more concrete idea of what you can do with HPF rings (which you don't get for all lenses).

(Apologies for my usage of meter instead of feet!)

Chris

[TH_Alpa]

@ Wayne

you aren't wrong, the distance scale goes out when tilting (or swinging). For a tilt of let's say 5 or 6 degrees to get your sharpness plane set almost "flat" (parallel) to the ground for a landscape shot (to get DoF without stopping down too much), the extension of the now DoF is not longer from front to back (far & near points), but one has to know that the DoF in this case extends perpendicularly to the plane of sharpness. In this case (with the plane lying "flat") up and down. When you are focusing in this tilted setup the sharpness goes up (sky) and down (ground). Therefore one can't use the distance scales any longer as such.

@ Chris & Wayne

thanks for explaining so precisely what I try to explain since yesterday and which has been wrongly presented in this thread since its beginning: the Alpa HPF is in fact, and despite some comments which may induce in error, as precise as the Arca helico focus ring, actually more precise when one looks at the tables from both cameras. And in the contrary to the Arca helico ring, it is less complicated and faster to use, IMO. I can only suggest anyone interested to know what are the facts and to try both systems side by side: take your time to make some real-life shots with both systems and come to your conclusions.

Have a look at the Alpa video on how to mount the HPF ring:

http://www.alpa.ch/en/products/lenses/alpa-hpf/alpa-hpf-alpaschneider.html

The Alpa HPF Ring exists for following lenses:

Rodenstock HR Digaron-S 5.6/23 mm = RDS 23
Rodenstock HR Digaron-S 5.6/28 mm = RDS 28
Rodenstock HR Digaron-W 4.0/32 mm = RDW 32
Rodenstock HR Digaron-S 4.0/35 mm = RDS 35
Rodenstock HR Digaron-W 4.0/40 mm = RDW 40
Rodenstock HR Digaron-W 4.0/50 mm = RDW 50
Rodenstock HR Digaron-S 4.0/60 mm = RDS 60
Rodenstock HR Digaron-W 5.6/70 mm = RDW 70
Rodenstock HR Digaron-W 5.6/90 mm = RDW 90
Rodenstock HR Digaron-S 4.0/100 mm = RDS 100
Rodenstock HR Digaron-S 5.6/180 mm = RDS 180

ALPA AAA Apo-ALPAR 4.5/35 mm = AAA 35
ALPA AAA Apo-ALPAR 4.5/45 mm = AAA 45
ALPA AAA Apo-ALPAR 4.5/55 mm = AAA 55

Schneider Apo-Digitar 5.6/35 mm XL = AAS 36
Schneider Super-Angulon 5.6/38 mm XL = SSA 38
Schneider Apo-Digitar 5.6/43 mm XL = AAH 43 / SAD 43
Schneider Apo-Digitar 5.6/47 mm XL = AAH 48 / SAD 47
Schneider Super-Angulon 5.6/47 mm XL = SSA 47
Schneider Super-Angulon 5.6/58 mm XL = SSA 58
Schneider Apo-Digitar 5.6/60 mm XL = AAH 60
Schneider Apo-Digitar 4.0/60 mm N = SAD 60
Schneider Apo-Digitar 5.6/72 mm N = SAD 72
Schneider Apo-Digitar 4.0/80 mm N = SAD 80
Schneider Super-Symmar 4.5/80 mm XL = SSS 80
Schneider Apo-Digitar 4.5/90 mm N = SAD 90
Schneider Apo-Digitar 5.6/100 mm N = SAD 100

Schneider ALPA Apo-Switar 5.6/36 mm = AAS 36
Schneider ALPA Apo-Helvetar 5.6/48 mm = AAH 48
Schneider ALPA Apo-Helvetar 5.6/60 mm = AAH 60

I've never really tried using tilts much for focus, but always assumed once you start tilting things, distance scales sort of go out the window.  Maybe I'm wrong and the trick is being able to know exactly where to measure when figuring out your focus plane.

As far as precision, maybe some clarification on the Alpa as well.  I only have the HPF rings for my 70mm.  When shooting with the 47 or 35 I'm either focus stacking, using a limited version of hyperfocal distance (basically I shoot each lens at f/11 and f/16 and figure out where infinity fails to satisfy me and set it there), or I know there isn't anything close enough to worry about it.  So I haven't put the HPF rings on the wides (yet anyway).

  On the 70mm, the focus ring turns over 3/4's of a revolution and does make the ring much larger although prob more like 3".  The HPF ring divides that into 50 increments, and there are 5 sub increments. It's actually quite easy to set the lens "between" each subincrement about the thickness of the line so the line covers the space between the two. However, the precision at close range is rather impressive, 10 of the 50 cover only 2.65 feet up to 3.31 feet, so 50 minor steps to go 9 inches.  The next 10 get you to 4.4 feet, and overall 40 of the 50 major steps, or 200 of the 250 minor steps covers the range from 2.65 feet 16.6 feet - so thats 200 steps to cover about 167 inches.  Of course, one challenge that makes them less useful than what I understand about the arca is the fact they are non linear, so each substep doesn't represent "x" inches.  This means you have to sort of look at the two major steps, do some math in your head, and figure out what each substep represents in distance. Example, on the closest subset, each step represents 0.12" while at the 40th major step, each substep represents  about 5.5" so even if you factor in using in between each sub-increment for distances between 14.3 to 16.6 feet your control is probably only around 2.5" per step. From there it scales and become pretty inaccurate so if you need precise focus at 192 feet, tough to do on the Alpa with the HPF rings ... so then your having to depend on whatever tools the back has for you.

Thanks for pointing this out, Jack, this is a common error in Arca-Alpa discussions: You don't use 5 revolutions with each and every lens.

If you have a lens with a small hub, you stay within one revolution, e.g., with the Schneider 28mm. The supplied distance table for the SK28 runs from 0.5m (on the helicoid: 23.2) to 25m (0.4), so this is less than a full revolution. If you look at the Arca-table for the SK43, you go from 0.5m (RED 26.7) to 25m (1.1), that is, you are close to two revolutions. If you look at the table for the Schneider 90mm, you use use up to five revolutions but if you focus somewhere between 3.4m and infinity, focusing takes place within one revolution. Put differently, the precision of the Arca way of focusing increases with a large focus hub BUT one must keep in mind that you have large movements in the close range and these get smaller if you approach infinity.

One more point: Whether focusing an Arca-lens (without the E-module) or focusing an Alpa-lens with a HPF ring using a Disto, you have to have the respective table at your hands. Take the SK 43mm: on the Alpa HPF you have an explicit mark for 7.29m and another one for 4.89m. The table will tell you that the small marks between correspond to 6.63m, 6.09m, 5.63m, 5.23m respectively. On the Arca table, looking at approximately the same range, the following distances and corresponding values are given: 7.5m (3.7), 7m (4.0), 6.5m (4.3), 6.0m (4.7), 5.5m (5.1), 4.8m (5.9).

It's worth having a look at the Alpa-HPF distance tables to get a more concrete idea of what you can do with HPF rings (which you don't get for all lenses).

Chris

[TH_Alpa]

It looks like others are seeing as well the difference .01mm makes when shimming precisely his back. You won't actually see it before you have 2 shots to compare, side by side, and will then always believe that your not-shimmed back is perfectly sharp where you want it to be. The point here being that this will become more and more important with the sensors' resolution going up: in clear, why buying a IQ 180 when you don't use it at its full capacity, buy a 22 or 33 MPx back and you will get approximatively the same IQ.

Thierry

I know alpa prides itself on precision, and won't do sliding backs.  After calibrating mine and realizing .01mm spacing between the back and the body does make a difference I can see their point.

[john milich]

out of curiosity, does anyone know, for the alpa and the arca, how many of those "ticks" on the dial correspond to, say 0.01mm of actual lens movement?  that number being the alpa shim tolerance.

and what about backlash, or is everyone approaching their target number from the same direction?

[dchew]

I have a rough estimate when shimming that 0.01mm = 1 minor tick mark. I've shimmed about half a dozen times, so not the most experience out there. I haven't noticed any backlash effect, but that would be interesting to check.

I too think Jack is spot-on. Technically the Arca can be more accurate and that might be a welcome feature when applying tilt.  Of course having 50 turns would be more accurate still!  Both of these systems are awesome. Like Wayne, once I realized focus stacking could be used instead of tilt for the majority of my situations, I felt the Alpa was a better fit for me.

If you want/need tilt on WA lenses, then the Arca is the only choice. But if not, then it comes down to how you like to work and personal preferences.

Dave

out of curiosity, does anyone know, for the alpa and the arca, how many of those "ticks" on the dial correspond to, say 0.01mm of actual lens movement?  that number being the alpa shim tolerance.

and what about backlash, or is everyone approaching their target number from the same direction?

[del_pscc]

The Arca R focus helix is approximately 110mm in diameter, has 172 tick marks (numbered every 5 ticks, from 0 to 34), and moves the lens in or out 2.5mm per revolution -- i.e., 2500/360 = 6.94 microns/degree, or 2500/172 = 14.53 microns/tick mark.  Lacking a vernier scale, about the best that could be achieved in terms of repeatability is 1/2 a tick mark, or 7.27 microns.  Arca claims 10 micron repeatability for the Rm3di <http://www.rodklukas.com/resources/Arca-Swiss_NewProducts_2010.pdf>, which coincides with the minimum shim increment from Alpa.

The scale that  Arca uses appears to be based on the circumference of the focusing mount: 110 x pi = 345.6, which, when truncated to an even integer, yields 172 ticks and a final diameter of 109.50mm.

Dave.

As you might have guessed, I've received my Rm3di and an SK 43XL, but am waiting for a viewfinder and a way to mount a back on the camera.  So  I'm reduced for the moment to admiring the engineering. 

[dchew]

Well, looks like my previous post was way off.  A quick look at the Alpa distance tables show 0.03333mm per degree for Rodenstock lenses and 0.02778mm for Schneiders.  Each minor tick is 1 degree. 

Dave

[buckshot]

If you want/need tilt on WA lenses, then the Arca is the only choice...

Or the Cambo Wide DS/RS (with their T/S lenses), or the Linhof Techno (with every lens). Both offer simultaneous Tilt and Swing (unlike the latest Arca), and the movements on the Linhof are yaw free.

[David Klepacki]

Or the Cambo Wide DS/RS (with their T/S lenses), or the Linhof Techno (with every lens). Both offer simultaneous Tilt and Swing (unlike the latest Arca), and the movements on the Linhof are yaw free.

How does the focusing precision of the Cambo and the Linhof compare to the Arca?  Also, do you find the yaw-free behavior of the Linhof to give you more accurate focusing capability when using tilt and swing movements?

[Clyde RF]

I have no doubt about the precision of the Arca rm3di focusing mechanism, but it would be very helpful for me to understand more about the most effective use of the camera. While the correct application of the helical, evf, drum, tables, (and maybe e module), along with possible calibration of the back, will quickly and accurately cover any base lacking the inclusion of tilt use; and some formulas are available to work well  providing an effective generic form of tilt focusing using said devices and another set of tables, I still need to know more. I am coming from 4 by 5 ground glass focusing using the method posted by Q.T.Luong on the LFphoto.info site under "How to Focus the View Camera" (procedure 2). This method is very specific to each image, and takes into account the periodic importance of finding the exact degree of tilt best suited for each situation, or whether tilt  provides any advantage. Even more significantly, it also covers the situation in which the near and/or far focusing points are above or below the tilted focusing plane. So it would seem that when using the rm3di without employing focus stacking, there would still be times when I would need to go visual, either by tethering (which I would prefer not to do with landscape) or using some approach to viewing the screen on the digital back. As the ground glass is of little use focusing digital backs, I'm wondering why they are included at all (except for roll film use). The ground glass is sometimes mentioned as being useful for composition, but these references must be meant to apply to film use or in the absence of the evf, which I would think would be a better device for accomplishing that purpose. These are my ponderings in which I'm sure I'm missing a good bit, so any response from those in the Arca camp will be greatly appreciated.     

[David Klepacki]

I currently use the RM3D.  I find the Arca viewfinder to be indispensable in the field.  It can be used off-camera to investigate different shooting angles, locations and focal lengths (including shifting) without having to setup the tripod.  I find it to be extremely accurate in its framing.

However, composing on the 6x9 ground glass is very useful, even for digital use.  Maybe it can be considered a luxury and not a necessity, but most photographers who try it do enjoy the larger image with which to make their final composition.   The large 6x9 view with grid is especially more convenient when planning stitched shots.  And in general, the viewing experience with the Arca reflex binocular is infinitely more enjoyable than trying to compose an upside-down image using a loupe as with most other technical cameras.  Although, I do not exactly like the extra bulk of this binocular.  As usual, there is no "free lunch" and this is the price you must pay for such capability.  Supposedly, Arca will be introducing a more compact reflex monocular soon.

[buckshot]

How does the focusing precision of the Cambo and the Linhof compare to the Arca?  Also, do you find the yaw-free behavior of the Linhof to give you more accurate focusing capability when using tilt and swing movements?

The focussing action of the Cambo is comparable to the Alpa - i.e it utilises Schneider's standard helicals. It should be possible to use the Alpa HPF rings on the Cambo, although the bars will need to be removed and some machining of the mounts may be required. I still maintain that a simple sticker over the current distance scale, with markings from (say) 1 - 100 is all that's needed (plus a lookup table per lens a la Arca).

No bellows camera is as easy to focus as pancake camera with a helical, but with the release of the IQ backs this is rather a moot point, since focus mask/live view may be all that's needed to make focussing a breeze. That said, there are many photographers using the Techno who don't have a problem focussing - it may well be this whole issue is one of those things that gets blown up out of all proportion in the world of photography forums. Hell, if I'd just dropped $40k+ on an IQ180 I wouldn't like anyone telling me I don't have the most accurate focussing system in the world in front of it 

For wide angle lenses the yaw free movement isn't so important - but for longer lenses it's nice not to have to chase the image around the gg.

[JdeV]

No bellows camera is as easy to focus as pancake camera with a helical, but with the release of the IQ backs this is rather a moot point, since focus mask/live view may be all that's needed to make focussing a breeze.

I haven't tried the new IQ backs but the sensor is a CCD and live view is likely to suck (though I would love to be proved wrong). In my experience Focus Mask is of marginal use (mostly not much cop and totally useless in difficult-to-focus low-contrast situations where you need it most).

This issue as a whole is absolutely not unnecessary fussing. If you shoot with a wide-angle lens you can't accurately hit correct focus without a great deal of care and all the technological assistance you can get. As is frequently pointed out, 35mm DSLRs are currently very good and the fairly small resolution gain from view cameras is easily wiped out by inaccurate focus.

[DanLindberg]

I suppose that IF the live view will work with the new IQ backs in a good way, then everything else is of no importance. Make the composition through the viewfinder and then finetune both composition and precise focus via live view. However,  should it not work as many seems to suspect, then I think that a "smallish" macbook air 11" could do the trick even in the field. If you use a lot of shift and tilt/swing, then this could be an alternative and always be spot on. I am hoping that the live view on the back will be workable though, that would be fantastic!

[Wayne Fox]

I suppose that IF the live view will work with the new IQ backs in a good way, then everything else is of no importance. Make the composition through the viewfinder and then finetune both composition and precise focus via live view. However,  should it not work as many seems to suspect, then I think that a "smallish" macbook air 11" could do the trick even in the field. If you use a lot of shift and tilt/swing, then this could be an alternative and always be spot on. I am hoping that the live view on the back will be workable though, that would be fantastic!

I always thought live view was a limitation of the backs processing power.  If live view is decent tethered seems getting decent functionality with the new backs could be possible?

[design_freak]

I always thought live view was a limitation of the backs processing power.  If live view is decent tethered seems getting decent functionality with the new backs could be possible?

Otherwise there would be no progress... you remember what was the quality of 6 years ago, and what is it now? I think almost anything is possible. Just can not have everything in a very short time.

[nazdravanul]

I might just have to go with Alpa simply because of the awful Arca-Swiss communication. Arca-shop.de are a real nightmare to deal with : 3 weeks and still no response to my email with technical questions, despite a follow-up phone call and another email. Alpa has been really good so far in dealing with my questions, and that is quite reassuring.
I wish I could find, in Europe, a better Arca-Swiss representative than arca-shop.de so please, if anyone knows one, do let me know. For my needs the Arca Rm3d seems like the better product, but I refuse to deal with the ridiculous response times. When the time comes, later in the year, Alpa will probably get my business, simply based on good communication, rather than product preference (not that the Alpa Max or STC are bad products , on the contrary, they appear to be outstanding, yet ... ).