How good are Rollei 6000 series lenses compared to Digitars?

[bjny]

I read that the Schneider 35mm Apo Digitar XL resolves 100 LINE PAIRS in the center.

Does anyone know what the 55mm PCS Super-Angulon, 90mm + 150mm Apo-Symmars, and

150mm Tele-Xenar for the Rollei 6000-system is capable of resolving?

[erik scanhancer]

You might like to have a look at the Schneider website (http://www.schneider-kreuznach.de/

foto_e/foto.htm). There is quite some tech info on several of these optical designs. The Apo-

Symmar 150 for example:

 

http://www.schneider-kreuznach.de/foto_e/sr_apo_L/pdf/SR_AP_56_150_R47555_2AE.pdf

[bjny]

Thank you Erik. I really don't want to learn how to read MTF charts. I just want to know the

LINE PAIR figures for a head to head comparison.

[q.g._de_bakker]

Billy,<br><br>Knowing resolving power alone is a very poor indicator.<br>That's why they use MTF, and that's why it's good to learn to read MTF graphs. ;-)<br><br>Digitars are used in digital capture, right? To how many line pairs does the resolution limiter, also known as anti-aliasing filter, reduce resolution? 40 lp/mm? Less?<br>The MTF performance of the 35 mm Digistar at 60 lp/mm doesn't look very impressive. But given that the 'softener' is involved, i guess that doesn't matter much. ;-)

[bjny]

Medium format digital backs do not use anti-aliasing / lowpass filter as 35mm DSLR's do.

[david_littleboy__tokyo__ja]

"digital backs do not use anti-aliasing"

 

But they should (and some offer an AA filter as an option). Without an antialiasing filter, you get nasty aliasing artifacts with digital sensors. The apparent gain in sharpness from not using one is largely bogus. Digital images cannot correctly (or even acceptably) resolve patterns with sharp edges above about 2/3 of the Nyquist frequency (only sine wave patterns can be resolved in this region; sharp-edged patterns turn into a mass of Moire). The patterns that are blurred by an AA filter are patterns that would not be resolved in any useful manner. With a correctly antialiased imaging system, the images sharpen up nicely with a bit of USM.

 

This is even more true of Bayer CFA sensors, since aliasing often results in nasty color artifacts. Nowadays, most of the DSLRs have inadequate AA filters, so you get garbage like this with, e.g., the Canon 5D. Sigh.

 

http://www.pbase.com/davidjl/image/52118414/original

 

To get back to the original question, though, smaller format lenses have much higher lp/mm ratings. The lens resolution in lines per image height falls quite slowly as the format decreases, so consumer dcam zoom lenses have MTFs at 100 lp/mm that are similar to Leitz and Zeiss lenses at 33 lp/mm.

 

Even worse, getting all hot and bothered over lp/mm resolution is a waste of time; you simply don't see it in real images. Photographically useful detail at 32 lp/mm is pretty rare in real life. (The sainted Hassy 80/2.8 at f/5.6 has an MTF at 40 lp/mm that's under 30% over 60% of the frame. Sorry, QG, but combined with Velvia's MTF at 40 lp/mm, that's simply not adequate to capture pictorially significant detail. (The 100/3.5's a much better lens.))

[bjny]

Thank you everyone for the replies so far. I'd like to purchase some lenses adequate to use

with the latest 33MP & 39MP digital backs. I'm wondering if the 55mm, 90mm, & 150mm

combo mentioned above is as good as Schneider's Digitar series.

[david_littleboy__tokyo__ja]

The short answer to your question is that it's inconceivable that there is any problem whatsoever with any Rollei 6000 lens with a digital back. From fisheye to the longest telephoto.

 

"Thank you Erik. I really don't want to learn how to read MTF charts."

 

There's really nothing particularly difficult about MTF charts; they're just a graph of how badly the lens messes up (contrast, with 100% meaning no mess up) vs. how close the lines are together (frequency in line pairs per mm) vs. how far from the center you are. The MTF charts directly tell you exactly what you want to know.

 

Also, the math here isn't much more than grade school arithmetic.

 

For example, your 39MP back has 39,000,000 pixels in a 48x48mm square (I think, anyway; it might be rectangular). Lets assume that for now. That means that it has square_root(39,000,000), or 6250 pixels in either direction. That's (6250)/48 = 130 pixels per mm, but a _line pair_ takes roughly three pixels to represent without jaggies, so that's 44 lp/mm. None of the Rollei lenses will have any trouble with that.

 

Now it might be more fun to put your US$30,000 back on a camera with movements than the Rollei, at which point a the LF lenses you happen to have at hand might be problematic, since many LF lenses don't do very well at test chart sorts of things (lower magnification means you don't need great resolution). At that point, you should think about the "Digitar" lenses.

[david_littleboy__tokyo__ja]

Looking at that, the "grade school arithmetic" comment could be taken the wrong way: I was trying to be encouraging, not sarcastic/rude. Sorry.

[geoffrey goldberg]

Lots to comment on here, so let's start:

 

- I too could use some help in reading MTF charts. What do the different lines mean, for

example? I've never been able to figure this out. I assume X axis is distance from center,

but is it in mm? Or do I have this all wrong?

 

- the use of AA filter is not necessarily so simple. I (and a bunch of others) have been

excited over the Leica DMR without AA filter, and its 10 mp makes amazing results.

They have a software moire filter, but very few of us who have the camera have ever felt

the need for it. There is occasional moire, but you have to shoot for it in particular.

Otherwise, there is no issue with it. Really. There is a very active thread (some 400 pages

worth!) on it on the FM

forum. Worth reading.

 

- as to quality of lenses, the Schneider 90 is reputed to be awesomely sharp. I have

enjoyed mine, and never found a problem with any degree of resolution. But then again,

the Schneider lenses in general have all been exceptional. I swear by the 60mm Curtagon

and keep it under my pillow at night.

 

- Given the senor sizes, I would suspect any of the Rollei lenses are fine. Problems exist (if

thee are any) at the edges of the frames, where the light angle and the sensor are not

happy together...but with these smaller sensors(not a true 6x6 or 645), one of benefits is

that this potential issue at the edges is naturally "cropped out". Of course, we get the

cursed magnification factor in trade-off.

 

Geoff

[david_littleboy__tokyo__ja]

"I assume X axis is distance from center, but is it in mm?"

 

Yes. The lines correspond to how well the lens renders patterns parallel to lines in the film plane that pass through the center of the image and patterns perpendicular to such lines. For each of those two, you get one for 10 lp/mm patterns and one for 40 (sometimes 30) lp/mm patterns.

 

"use of AA filter is not necessarily so simple"

 

Moire is horrible. And with Bayer sensors, it confuses the heck out of the color calculations, making the crazy patterns you'd see if you looked at the link I provided above. You never know whether or not you've got it, and if you chimp every image during a shoot (which I can't imaging being an option for a pro) and find it, there's no guarantee the retaken shot won't be ruined by Moire again. I simply can't imaging it being acceptable in a professional context. (And a lot of MF back users are finding this out the hard way.)

 

"Problems exist (if thee are any) at the edges of the frames, where the light angle and the sensor are not happy together..."

 

All SLR lenses must have a long enough back focal distance for the mirror to clear. As a result, the angles at the edges are neither particularly acute nor particularly problematic. And the last time I looked, the Rollei 6000 was an SLR. This whole idea is FUD from a certain failing camera manufacturer. I'm surprised it's taken on: it should be obvious to anyone who's ever used a 6x6 SLR that the lens to film distance is the same for an 80mm lens as it is for a 40mm lens.

[desmond_kidman]

Wow, that moire is horrible. One thing that surprises me is that I have not seen anything like this with my Canon 1ds2. Is there that much difference between 1ds2 and 5d?

 

The supporters of MTF over line pair resolution are correct. Ultimate line pair resolution means very, very little in terms of perceived sharpness. However, the MTFs correlate very well with perceived sharpness, as they are a measure of contrast, and edge contrast (acutance) is what the eye/brain perceive as sharpness.

 

An excellent book which explains many common optical terms and MTFs is the Canon Lenswork book. For this very reason I owned one before I ever considered owning a Canon, so good was the section on optical principles. You can get it from Canon for about $20.