RF accuracy chart

[LCT]

For those interested, here's a little chart of the R-D1's rangefinder accuracy for lenses at full aperture, based upon the formula b' = (e * f^2) / (k * z) where b' is the effective base length of the rangefinder, e the visual acuity (0.0003 at approx. 1 arcmin), f the focal length, k the aperture and z the circle of confusion (0.020mm for APS-C).
The effective base length (EBL) of the RD-1's rangefinder is 38.2 mm without magnifier (38.2 mm mechanical base length * 1.00 VF magnification) and 49.66mm with the 1.3x Megaperls magnifier (38.2 mm mechanical base length * 1.30 VF magnification).

 

[Roberto]

For those interested, here's a little chart of the R-D1's rangefinder accuracy for lenses at full aperture, based upon the formula b' = (e * f^2) / (k * z) where b' is the effective base length of the rangefinder, e the visual acuity (0.0003 at approx. 1 arcmin), f the focal length, k the aperture and z the circle of confusion (0.020mm for APS-C).
The effective base length (EBL) of the RD-1's rangefinder is 38.2 mm without magnifier (38.2 mm mechanical base length * 1.00 VF magnification) and 49.66mm with the 1.3x Megaperls magnifier (38.2 mm mechanical base length * 1.30 VF magnification).

Great!! Any chance you can do this for the Bessa R too? )please( 😛

 

[LCT]

Great!! Any chance you can do this for the Bessa R too? )please( 😛

No problem if you give me:
- the mechanical base length of the rangefinder
- the magnification value of the viewfinder

 

[Roberto]

Thanks!
The Effective base lenght (from Cameraquest) is 25.16mm.
It has a magnification of .68 and the phisycal base lenght should be the same as R-D1 (in fact 38.2 *.65 = 25.976)

Rob.

 

[LCT]

According to Cameraquest, the VF magnification of the Bessa R is .68x and the mechanical (or physical) baselength of its rangefinder is 37mm.
This gives an EBL (effective baselength) of 37*.68 = 25.16mm and leads roughly to the same results as those of the R-D1 due to compensation between a 1.5 times smaller EBL (25.16mm vs 39.2mm) and a 1.5 times larger circle of confusion (0.030mm vs 0.020mm).

 

[Roberto]

Thanks a lot! Very appreciated 🙂

 

[ferider]

Hey, this is great !

Very much confirms my experience !

Note that your Bessa R + magnifier numbers correspond to the Bessa R3
without magnifier 🙂

Roland.

 

[haagen_dazs]

this is a very cute chart
good stuff 🙂

 

[ferider]

LCT, Just went in and took some other cameras as well, using your formula.
Hope I did no mistake:

IMAGE REMOVED. UPDATED BELOW ...

Cheers,

Roland.

 

[Roberto]

LCT, Just went in and took some other cameras as well, using your formula.
Hope I did no mistake:


Cheers,

Roland.

Cool! 😀 What about M6/M7?
Rob.

 

[Rafael]

I believe that Roland calculated the results for the MP on the basis of it having a .72x finder. This means that the results for the M2, M7 (.72x) and the M6 (.72x) would be the same as those for the MP. (Correct me if I'm wrong Roland 😉 ).

 

[Rafael]

By the way, thanks for doing these calculations you guys.

 

[BillBingham2]

Roland,

Can you add in the 35/1.2 also?

Interesting about the 90/4 on the CL being red. Are you just pickier than Leica/Minolta?

Send me a copy of your excel sheet if you would please. Very cool.

Thanks.

B2 (;->

 

[ampguy]

Great stuff!

I have a question about the coc, I've seen .02 used for various FF M bodies and in the calculations above, only the RD1 is using .02, everything else, even the M8 is > .02. Where is the coc value coming from?

And if I can focus in some of the red areas, does that mean I have an invisible built-in magnifier, or am I just lucky?? 😀

 

[RichC]

If anyone's interested, the RD-1 chart, in a simplified and more numerical version, is on my R-D1 website: http://www.richcutler.co.uk/r-d1/r-d1_05.htm

Ampguy, see here for choice of CoC: http://www.richcutler.co.uk/r-d1/r-d1_02.htm. And, yes, you're just lucky!

In practice, I find that if I'm printing on paper larger than A4 (= 10 inches to you Americans!), the 0.02 mm CoC is too large...

 

[LCT]

35mm and 40mm lenses are too wide to pose any problem as long as they won't have a f/0.5 aperture.
0.03mm is the usual CoC value for the so-called 'full frame' format (24*36mm). It is the value that is still used by lens makers and DoF calculators nowadays:
http://www.dofmaster.com/dofjs.html
Of course changing that value would change the results but i still retain it AFAIC for the simple reason that it works well for me.
Same for smaller formats which have the same fraction of the FF CoC value as that of the so-called 'crop factor'.
For instance, APS-C cameras like the R-D1 have a 1.5x crop factor, hence a CoC value of 0.03:1.5 = 0.02mm on the same basis.
Now being out of the accuracy range does not mean that all pictures will be out of focus. It is just a matter of hit rate.
For instance, the Noctilux is OK but is not easy to focus on the R-D1. On the other hand, 90/4 lenses are out (by very few) on the Leica CL but are still usable at full aperture with a lower success rate than at f/5.6 that's all. At least that is what i seem to recall to be honest as the last time i've used a CL was 20 years ago more or less.

Ferider, i have not checked your whole chart above but it looks great at first glance indeed.
Edit: Rich, thanks for sharing your charts. Would you have the same in meters by chance? Also could you explain where your 0.018mm value for the R-D1's CoC comes from? Perhaps it is explained in your great site, if so forgive me please.

 

[ferider]

This is fun. Here is an update with more cameras.

IMAGE REMOVED. UPDATED BELOW ...

Bill, LCT, send me your email adress (ferider at yahoo dot com) and I'll email you the spreadsheet.

Best,

Roland.

 

[MikeL]

This is fun.

Bill, LCT, send me your email adress (ferider at yahoo dot com) and I'll email you the spreadsheet.

Best,

Roland.

Fun + Spreadsheets. Roland, by chance are you an engineer? 😉

 

[jvr]

35mm and 40mm lenses are too wide to pose any problem as long as they won't have a f/0.5 aperture.
0.03mm is the usual CoC value for the so-called 'full frame' format (24*36mm). It is the value that is still used by lens makers and DoF calculators nowadays:
http://www.dofmaster.com/dofjs.html
Of course changing that value would change the results but i still retain it AFAIC for the simple reason that it works well for me.
Same for smaller formats which have the same fraction of the FF CoC value as that of the so-called 'crop factor'.
For instance, APS-C cameras like the R-D1 have a 1.5x crop factor, hence a CoC value of 0.03:1.5 = 0.02mm on the same basis.
Now being out of the accuracy range does not mean that all pictures will be out of focus. It is just a matter of hit rate.
For instance, the Noctilux is OK but is not easy to focus on the R-D1. On the other hand, 90/4 lenses are out (by very few) on the Leica CL but are still usable at full aperture with a lower success rate as at f/5.6 that's all. At least that is what i seem to recall to be honest as the last time i've used a CL was 20 years ago more or less.

Ferider, i have not checked your whole chart above but it looks great at first glance indeed.

Nice work!

BTW, I still feel that digital (because of almost zero depth of the sensor) is even less forgiving than film, which has some thickness and when printed (or scanned) show more in focus, because the enlarger (scanner) lens has also some DoF...

When using the RD1s on hyperfocal (yes, although I have been using the M8 more and more, I sometimes still use the Epson...), at least one aperture difference is needed (if not two!) to get really allround sharp pictures on largish enlargements...

The good news is that 800 ISO is no problem and even 1600 ISO (in B&W) is very feasible, when 400 ISO was the norm in film (for me).

 

[ampguy]

Thanks Rich and LCT, I understand the math you are doing, was just curious if you had any primary sources for the coc info. vs the web/tertiary sources you have cited.

I do have pre M8 Leica info. for use of .02x coc, will cite it when I'm at my book.

Also, from practical experience, I have a higher hit rate with the RD1s (new) at 75/1.4 and 2.0 which are *RED*, than with CLA'd M6 .72 with Noct 1, and CLA'd CL with 40/f2 which should be *OK* Maybe the Canadian 75's used .03 for coc??

Would be interesting to see if this table corresponds to other users experiences, I'd guess not, or who in their right mind would use a Bessa or CL??!!

Also, maybe mention distance of subject?? Did you know that all of these setups focused on infinity at a far away subject should be *GREEN* 😀