Why not more pixels in the M camera?/ 36 MP {merged}

[Martin B]

Advertisement (gone after registration)

I wonder how many people who work in Leica R&D at Wetzlar are following this thread and laughing their asses off...

 

I hope they follow, but I am not sure if there is much to laugh about. They better vest their time implementing new tech for new camera generations. There is a lot to do for them. 

[Martin B]

Nikon will have a 120 MP camera by then, you'll have the same problem.

 

Where is it written in stone that Leica M is always a step behind?! Formerly in film days Leica M was very competitive with cameras from other brands and even a step ahead sometimes. Funny excuses you have....

[Paul J]

I am doing both - shooting film and also enlarging B&W film based negatives to make prints. A bit unfortunate that the resolution difference in this discussion is abused by some to justify lower MP FF sensor usage - that's all what is behind. But I have to agree from my own experience that a 35 mm negative does not have the resolution of a 36 MP FF sensor. Just two weeks ago I took the same photo with Fuji Velvia 100F slide film and with my Sony A7R at ISO 100 from the same spot. Both photos were taken with the same lens (CV 21/1.8 M-lens) and correctly focused, same aperture used. The slide film positive was scanned with a Plustek 8200i scanner at 3200 dpi. I cropped the similar area of the original photos at 200% magnification. 

 

Slide positive at 200%:

 

Digital photo (36 MP FF) at 200%:

There are a few variables that we need to consider first.

 

The crops are both 200% but the digital image is quite a bit closer and the point of view is different. Are these lenses the same?

 

The scan also looks quite soft. The first place to check sharpness of film is the grain - With a dum scan, a 200% crop of a 3200dpi scan would show sharp and well defined grain even on medium format.

 

If the two images were shot from exactly the same position and a you had an HQ drum scan I think you might see less of a difference.

[lct]

Where is it written in stone that Leica M is always a step behind?! Formerly in film days Leica M was very competitive with cameras from other brands and even a step ahead sometimes. Funny excuses you have....

 

Last time Leica was technically competitive was with the M3 if my memory serves me well. Following the latest trend is not exactly Leica's cup of tea .

[Martin B]

There are a few variables that we need to consider first.

 

The crops are both 200% but the digital image is quite a bit closer and the point of view is different. Are these lenses the same?

 

The scan also looks quite soft. The first place to check sharpness of film is the grain - With a dum scan, a 200% crop of a 3200dpi scan would show sharp and well defined grain even on medium format.

 

If the two images were shot from exactly the same position and a you had an HQ drum scan I think you might see less of a difference.

 

As I stated above, the same lens was used. The scan is sharp, there is nothing wrong with the scan - simply the slide is not as sharp as the digital image. A drum scan would not help either here. The crop is not 100% from the exact location in the images - I moved a bit between them  when the photos were taken. But this does not affect the outcome. I would agree with you when comparing high MP sensor sharpness with LARGE format film though. Below the originals:

 

 

[sblitz]

As I stated above, the same lens was used. The scan is sharp, there is nothing wrong with the scan - simply the slide is not as sharp as the digital image. A drum scan would not help either here. The crop is not 100% from the exact location in the images - I moved a bit between them  when the photos were taken. But this does not affect the outcome. I would agree with you when comparing high MP sensor sharpness with LARGE format film though. Below the originals:

 

 

 

True enough. If I was shooting for an architect's or designer's catalog the digital image is the far superior choice. If I wanted a photo to convey the feeling of this walkway, the film version is, to me, so much better. Although, to be fair, the film shot needs a dose of WB to get it to the color palette of the digital

[Martin B]

Advertisement (gone after registration)

True enough. If I was shooting for an architect's or designer's catalog the digital image is the far superior choice. If I wanted a photo to convey the feeling of this walkway, the film version is, to me, so much better. Although, to be fair, the film shot needs a dose of WB to get it to the color palette of the digital

 

You are absolutely correct - purposely I did not want to make the images look similar with lots of PP. The reason why I shot digital and slide film was simply to see exactly the difference in color - I did not expect to see the same resolution, this was not the purpose of my initial test. I just wanted to see how this slide film brings out the colors in comparison to the digital camera which I used. The colors in the slide film are the more accurate ones - the whole walkway was filled with this yellowish tint, it looks more natural. The digital photo is the one which looks a bit unnatural and colder - I could change the WB to a warmer level, too if I wanted.

[Paul J]

As I stated above, the same lens was used. The scan is sharp, there is nothing wrong with the scan - simply the slide is not as sharp as the digital image. A drum scan would not help either here. The crop is not 100% from the exact location in the images - I moved a bit between them  when the photos were taken. But this does not affect the outcome. I would agree with you when comparing high MP sensor sharpness with LARGE format film though. Below the originals:

 

 

A Plustek scanner is not even close to a drum scan. You don't even have sharp grain in your scan.

 

You moved a fair bit between shots, you are at least a few meters closer. Yes that does have an effect on comparative resolution and it does also put testing parameters like comparative focus into question.

[Martin B]

A Plustek scanner is not even close to a drum scan. You don't even have sharp grain in your scan.

 

You moved a fair bit between shots, you are at least a few meters closer. Yes that does have an effect on comparative resolution and it does also put testing parameters like comparative focus into question.

 

C'on...the slide positive does not have lots of grain like in negative photos to focus on! Yes, this is not a scientifically accurate test where all the parameters are accurately the same, but even the slight movement of my position does not make a difference in the resolution which I showed above. Saying that this slight difference would put the focus at infinity(!) in question is simply not correct. Please show me your comparison 35 mm format photos to prove me wrong here.  

[Chaemono]

I wonder how many people who work in Leica R&D at Wetzlar are following this thread and laughing their asses off...

”Tell me, Mr. Anderson, what good is a phone call when you are unable to speak?” Or to paraphrase Agent Smith, what good are 42 MPx if your lenses are unable to resolve them?

 

And now my signature line, and upfront the G Masterly resolution that everyone at Leica is laughing about.

 

 

https://www.smugmug.com/gallery/n-x8Mwmw/

 

https://www.smugmug.com/gallery/n-Jfdr66/

 

https://www.smugmug.com/gallery/n-9FSKSS/

 

https://www.smugmug.com/gallery/n-9J3jzS/

 

http://www.towerjazz.com/prs/2017/1030.html

Edited March 3, 2018 by Chaemono

[Paul J]

C'on...the slide positive does not have lots of grain like in negative photos to focus on! Yes, this is not a scientifically accurate test where all the parameters are accurately the same, but even the slight movement of my position does not make a difference in the resolution which I showed above. Saying that this slight difference would put the focus at infinity(!) in question is simply not correct. Please show me your comparison 35 mm format photos to prove me wrong here.

A drum scan will show sharp and defined grain even in a large format transparency scan. This is the first way to tell a good scan from an average one.

 

Even a £15K hasselblad flextight scanner won't give as much detail as a proper drum scan.

Edited March 3, 2018 by Paul J

[Martin B]

A drum scan will show sharp and defined grain even in a large format transparency scan. This is the first way to tell a good scan from an average one.

 

Even a £15K hasselblad flextight scanner won't give as much detail as a proper drum scan.

 

I don't question this - but even if you get the positive a bit better resolved you won't get a native resolution close to a 36 MP FF sensor. It's just not there because the resolution is not provided in the film-based slide. If you go for 4x5" large format film size, yes, you get there - maybe even with some medium format (not sure, never used medium format film). But I would estimate that the resolution of 35 mm film is below 20 MP (and resolution is not the reason why I still use 35 mm film). But I am in line with you that for digital, I prefer higher MP resolution in the FF sensor size than 24 MP. 

[pgh]

C'on...the slide positive does not have lots of grain like in negative photos to focus on! Yes, this is not a scientifically accurate test where all the parameters are accurately the same, but even the slight movement of my position does not make a difference in the resolution which I showed above. Saying that this slight difference would put the focus at infinity(!) in question is simply not correct. Please show me your comparison 35 mm format photos to prove me wrong here.  

This is still close enough to real world application.

I've drum scanned (and flatbed scanned) plenty of perfectly sharp, low ISO, negatives - at 11000 dpi. So, massive files. Practical resolution is different though. 

A 24 mp digital file makes a better print. Enlargement looks a bit different on the film, and the film might look better once a certain resolution is passed and the image is so big that the grain/noise becomes apparent, but at most sizes between say, 11x14 and 30x40 - if we're talking about resolution and how the print holds up - I will take the 24 mp digital file any day. It's not even a question - and I don't think 35mm film has been a viably competitive medium resolution wise (ephemeral qualities are a different thing) since the D700 and the 5D - because besides resolution the whole pain in the ass process of drum scanning, inversion (if negative) etc is rarely worth the trouble. I'm surprised this argument is even coming up. There are reasons to shoot 35mm, but resolution....I haven't heard that really argued for like 7 years now.

 

In short - your demo here is close enough to my scanning experience (that I did for years) that I'd say it's a good practical representation of the difference between the mediums. 

Edited March 3, 2018 by pgh

[Tailwagger]

Assuming there's an appropriate supplier, its hard to imagine the next SL not moving to 36mp or more.  If it ever was, it really can't remain competitive as a pro camera otherwise., S or no S. 

 

The M is another story.  Put me in the somewhat indifferent category, assuming adding pixels doesn't compromise noise and DR.  Certainly no more than 36. That said, I'd be far happier to retain  24MP or a slight bump, say 28, in exchange for wider DR, better highlight handling, lower base ISO, improved high ISO performance.  Overall, I'd prefer increasing flexibility as opposed to pixel count. 

[Paul J]

I don't question this - but even if you get the positive a bit better resolved you won't get a native resolution close to a 36 MP FF sensor. It's just not there because the resolution is not provided in the film-based slide. If you go for 4x5" large format film size, yes, you get there - maybe even with some medium format (not sure, never used medium format film). But I would estimate that the resolution of 35 mm film is below 20 MP (and resolution is not the reason why I still use 35 mm film). But I am in line with you that for digital, I prefer higher MP resolution in the FF sensor size than 24 MP.

But your estimations are based on this kind of scan.

 

There is quite a lot more resolution there than in your scan (depending on wether your image is sharp to begin with). I think you would be surprised at how much a drum scan will reveal. Using a £300 bayer CCD scanner is really like using a cheap lens to shoot it in the first place. The lens is cheap, the film is rarely even flat and/or uniform, the bayer CCD can't replicate film resolution very well in fine detail, especially in colour. There's all sorts of aberrations and artefacts and the scan ends up mushy.

 

Sharpness and resolution are different things. That is the most common misconception about film. The grain softens the linear sharpness but it doesn't necessarily remove resolution. A high quality drum scan of a 35mm transparency is more likely to have around 12MP of sharpness, 18-20MP of luminosity resolution and 24-36MP of colour resolution.

 

I'm not saying the film image will be as necessarily sharp as the digital image, and I think the 36MP will still have more resolution that the film image, but what I am saying is your scan is really mushy and the gap between the two will likely be much less with a drum scan.

 

No argument from me re more resolution in digital sensors.

Edited March 3, 2018 by Paul J

[Neko]

Sometimes I print big so I will be more than happy with a 36-42 mp Leica M. I love M system for its compactness and amazing lenses. I like to go to the mountains walking or by bicycle for landscape shooting. A light tripod, my monochrome and the 50APO and I get medium format quality. With a 42mp M I would get the same quality in color.

[pico]

It is too much effort, but it would be interesting to see if those wishing for more pixels come

right back when their wish is satisfied, whining for the next step upwards.

 

That's the way they are.

 

Born complainers. Nothing is ever good enough. Must suck to be them.

Edited March 3, 2018 by pico

[dau]

I am doing both - shooting film and also enlarging B&W film based negatives to make prints. A bit unfortunate that the resolution difference in this discussion is abused by some to justify lower MP FF sensor usage - that's all what is behind. But I have to agree from my own experience that a 35 mm negative does not have the resolution of a 36 MP FF sensor. Just two weeks ago I took the same photo with Fuji Velvia 100F slide film and with my Sony A7R at ISO 100 from the same spot. Both photos were taken with the same lens (CV 21/1.8 M-lens) and correctly focused, same aperture used. The slide film positive was scanned with a Plustek 8200i scanner at 3200 dpi. I cropped the similar area of the original photos at 200% magnification. 

....

The negative will be 24mm by 36mm in size, which is about 0.945 by 1.417 inches squared. Scanning this negative at a resolution of 3200 pixels per inch will yield an image of 3024 by 4535 pixels.

 

Your are comparing a scanned image of negative with not quite  14 million pixels to an image made by a sensor which has 36 million pixels. What is the comparison supposed to show?

Edited March 3, 2018 by DAU

[mmradman]

Nowadays access to high MP full frame camera is much easier than acces to a drum scanner, for many convenience wins even if it is only 24MP.

 

We had this MP discussion on SL thread and well known LUF member and nature photographer illustrated limitation of 24MP when shooting his subjects of choice - birds.

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

There are quite legitimate reasons to wish for more pixels, and they have nothing to do with print size or cropping.  The main reason to me is moiré and aliasing suppression. If one uses high-quality lenses on high-frequency subjects, be it foliage or branches or a bird with finely structured feathers, or many other subjects, this can be a real problem.

 

No need to be insulting.

 

I'm sure that Leica is aware of the tradeoffs. For the moment they seem to have decided that 24 MP is the optimal compromise for the highest possible image quality in most cases. I  would say they are in a better position to judge than we are.

However, the CL shows that they now have access to the technology to reduce pixel size whilst maintaining image quality.  That would mean that a 36 MP FF sensor may be coming.