Peter Karbe talk about lenses.

[Photoworks]

Advertisement (gone after registration)

Leica has publish a great video how their lenses are made.

 

[Arrow]

Very interesting! Thank you for sharing!

 

[knisely]

Thank you for the link. This is a fantastic presentation, there is so much information. It was a great way to spend two hours during lockdown!

 

Edited July 16, 2020 by knisely

[scroy]

Really nice talk.  There was a lot to go through. It was quite interesting to be walked through his thought process.

I really enjoyed it.

[CharlesL]

He seemed to be in the mood of someone looking back over his career.

When asked about Chinese lenses with wide apertures, he not only stressed the tasks of precise manufacture and lots of quality control. He also alluded, with a tone of moral disapproval, of the low wages of Chinese workers.

Could you explain one thing? He repeated many times, modern lenses resolve so well that you should not stop down, unless you need the depth of field. Sometimes he almost seemed to toss out the proviso about depth of field. Anyway, what now-obsolete shooting situation did he have in mind?

 

 

[lx1713]

I think I will need more than 2 passes thru this interview. It's a very helpful interview. Thanks!

[jplomley]

Advertisement (gone after registration)

Wish there had been a comment concerning where the hell the rest of the primes are and what the future road map looks like.  The wait is getting beyond ridiculous when you look at how rapidly Hasselblad was able to deliver glass for their X1D II

[MarkinVan]

3 hours ago, jplomley said:

Wish there had been a comment concerning where the hell the rest of the primes are and what the future road map looks like.  The wait is getting beyond ridiculous when you look at how rapidly Hasselblad was able to deliver glass for their X1D II

The Hasselblad has 10 lens and the SL has 8 plus the TL lens and adapters for the M lens plus many other lens.

The waiting is the hardest part.

 

 

[Photoworks]

6 hours ago, jplomley said:

Wish there had been a comment concerning where the hell the rest of the primes are and what the future road map looks like.  The wait is getting beyond ridiculous when you look at how rapidly Hasselblad was able to deliver glass for their X1D II

did you say Hasselblad ? the company that was late delivering camera by 1 year and  firmware always in beta!  I am glad I didn't get that camera.

[Aperture 2]

Fantastic presentation, a lot of answers to questions I had regarding lenses design and capabilities vs sensors !

[Steve Ash]

I really enjoyed the presentation which provides competent answers to frequently raised questions. 

My take away from this great presentation: 
1) There is still improvement potential for the already awesome M-lenses. 
2) Older M-Lenses have a resolving power of 24Mp. Thus a 24Mp sensor is in balance with old lenses.

3) Newer M lenses are capable of more and need higher resolution sensors to get the system in balance. Or in other words, before investing in better lenses from Leica’s pov, it is wise to improve on the weakest part, i.e the sensor. 

4) Due to this I now would be keen to know what the actual resolving capacity of each of my lenses is.
5) As long as both lenses and sensors have room for improvement the "balance approach" presented by Karbe makes very much sense and is currently followed by Leica as seen with the M10-R.
6) Karbe left one scenario out of discussion: What happens if one of both is at its technological end? In my pov then the balance approach is at its ends and a higher resolving power of the other will still improve the image quality. 

Another consequence would be that for a given set of lenses higher resolution sensors will improve the image quality or the capability to enlarge the image without artifacts.

[Jeff S]

2 hours ago, Steve Ash said:

I really enjoyed the presentation which provides competent answers to frequently raised questions. 

My take away from this great presentation: 
1) There is still improvement potential for the already awesome M-lenses. 
2) Older M-Lenses have a resolving power of 24Mp. Thus a 24Mp sensor is in balance with old lenses.

3) Newer M lenses are capable of more and need higher resolution sensors to get the system in balance. Or in other words, before investing in better lenses from Leica’s pov, it is wise to improve on the weakest part, i.e the sensor. 

4) Due to this I now would be keen to know what the actual resolving capacity of each of my lenses is.
5) As long as both lenses and sensors have room for improvement the "balance approach" presented by Karbe makes very much sense and is currently followed by Leica as seen with the M10-R.
6) Karbe left one scenario out of discussion: What happens if one of both is at its technological end? In my pov then the balance approach is at its ends and a higher resolving power of the other will still improve the image quality. 

Another consequence would be that for a given set of lenses higher resolution sensors will improve the image quality or the capability to enlarge the image without artifacts.

You might want to read Roger’s Appendix: “Why Perceptual Megapixels are Stupid” in his following article....

https://www.lensrentals.com/blog/2019/10/more-ultra-high-resolution-mtf-experiments/
 

Jeff

[Steve Ash]

vor 7 Minuten schrieb Jeff S:

You might want to read Roger’s Appendix: “Why Perceptual Megapixels are Stupid” in his following article....

https://www.lensrentals.com/blog/2019/10/more-ultra-high-resolution-mtf-experiments/
 

Jeff

Thanks Jeff,

I take reference from the link

quote

Roger’s rule: If you have either a crappy lens or crappy camera, improve the crappy part first; you get more bang for your $. I just saw a thread for someone wanting to upgrade to the newest 60-megapixel camera, and all of his lenses were average zooms. I got nauseous.

unquote

I see this as confirmation of what Karbe is saying. It is in also not in contradiction what I am saying in case one of both is at its technological end. 

 

 

 

[Jeff S]

12 minutes ago, Steve Ash said:

Thanks Jeff,

I take reference from the link

quote

Roger’s rule: If you have either a crappy lens or crappy camera, improve the crappy part first; you get more bang for your $. I just saw a thread for someone wanting to upgrade to the newest 60-megapixel camera, and all of his lenses were average zooms. I got nauseous.

unquote

I see this as confirmation of what Karbe is saying. It is in also not in contradiction what I am saying in case one of both is at its technological end. 

 

 

 

I was referring to the earlier part, speaking to the revolving power of lenses vs sensors, which you cited in your takeaway points. 

Jeff

[Steve Ash]

vor 7 Minuten schrieb Jeff S:

I was referring to the earlier part, speaking to the revolving power of lenses vs sensors, which you cited in your takeaway points. 

Jeff

Ok, lets limit the discussion to Leica lenses.

I understand that all lenses are good enough if stopped down. Karbe is referring to wide open only as this is where the lowest resolving power is. This makes perfectly sense to me. Specifically he showed the mtf for the summilux 35 for f1.4 which is in balance according to his definition with a sensor of 24MP. Sadly he did not talk about the performance at f2.

If you keep the lens a higher resolution will improve the image quality wide open. 

The logical consequence from my pov could be a development of an APO Summicron 35 as sensor development is going on. 

I know the whole discussion is somewhat meaningless as we are spoiled anyway.

 

 

[Jeff S]

21 minutes ago, Steve Ash said:

Ok, lets limit the discussion to Leica lenses.

I understand that all lenses are good enough if stopped down. Karbe is referring to wide open only as this is where the lowest resolving power is. This makes perfectly sense to me. Specifically he showed the mtf for the summilux 35 for f1.4 which is in balance according to his definition with a sensor of 24MP. Sadly he did not talk about the performance at f2.

If you keep the lens a higher resolution will improve the image quality wide open. 

The logical consequence from my pov could be a development of an APO Summicron 35 as sensor development is going on. 

I know the whole discussion is somewhat meaningless as we are spoiled anyway.

 

 

I was referring to the earlier part of the Appendix, not the body of the post. 
 

Jeff

[scott kirkpatrick]

The reason this discussion is going in circles is because of fear of actually discussing the very simple formulae involved.  Roger Cicala's article makes the point that if you have a chain of contributions to an image, the MTF (the percentage of contrast that survives starting with a crisp 100% image of a series of black and white lines) is the product of the MTFs of each contributor in the chain.  So if you have a crappy lens on a good sensor you might be able to improve the lens by 50% but the sensor only by 10%, so of course you upgrade the lens.  But digital sensors are pretty good until you ask them to resolve line pairs that are closer together than twice the pixel spacing (that's the Nyquist frequency).  So we usually want to know if a lens is still giving us information (good contrast) at frequencies approaching the Nyquist frequency of cameras that we have or hope to have someday.

The notion of taking the product of MTF figures is a good way to think about combining lenses, but the MTF of a pixel array is pretty much 100% until the Nyquist frequency is reached, then drops fast as artifacts occur. The best quick check for this with a camera that we have is to look for Moire effects on distant regular latticed objects, such as louvres or textiles.  But Roger C's articles generally plot MTF percentages as a function of distance from the center of the image, and include several spatial frequencies, all much less than typical Nyquist frequencies.

So let's look at one key figure in Peter Karbe's talk, which plots MTF as a function of spatial frequency, and asks where that percentage is good enough to present a strong image at frequencies up to the maximum which the pixel array can resolve.  This is also the readout that Imatest (a chart-based resolution analyzer) will give you.  He compares the latest and greatest SL Summicron with a fine lens from more than 10 years ago:

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Screen Shot 2020-07-16 at 3.45.05 PM (2) by scott kirkpatrick, on Flickr

and you can see that the older lens has reached an MTF of 20 % by the time you get to the Nyquist frequency of a 24 MPX full frame array.  10-20% at the limiting frequency of the camera was what we used to consider good performance. But not Peter.  He wants to see 50% MTF as a working limit, and his latest lens in fact gives us 60% MTF at the 160 line pairs per mm that will charactize a 100 MPx array!  Now do you see why he loves this lens so much? 

Edited July 19, 2020 by scott kirkpatrick

[DadDadDaddyo]

A wonderful presentation. I'm grateful to have been able to see it.

[Arrow]

Concerning the 35 Cron SL´s resolving power: will there be a 100 MPX full frame sensor that will make sense? We are currently at 47 MPX.

Edited July 21, 2020 by Arrow

[Advertisement]

Hallo,

Look here: Leica L-Mount

[caissa]

The trend to higher resolutions is unbroken. So wait for SL3 or SL4 (4 to 8 years). It will probably come, but if it is really useful for me or you, will be seen.
Maybe Panasonic will be there even a bit earlier ...   (S2R or S3R)
In the meantime use the high-res (multishot) option with 187 MP.  

Edited August 1, 2020 by caissa