Jump to content

The Leica Q2 and my attempt at a grand unifying theory of sensor size and AOV


Recommended Posts

[Note to moderators: this was deliberately posted in the CC forum, and is not intended for the Leica forum.]

 

This post does indeed discuss the Q2, but yet it's ultimately about photography in general - but for now, "hold that thought". Yesterday, Leica announced the Q2. The original Q was launched in 2015, almost four years ago. Let me be blunt - at the time I really didn't get it.

 

I mean, I did sort of understand it, but not completely. I saw very clearly that its lens and sensor combination, a 28/1.7 mated to a 24Mpx CMOS, was oddly superior to a lot of other cameras with the same sized sensor, even if those cameras had higher photosite densities. I remember pixel peeping the charts on DPReview and feeling somewhat enlightened.

 

Here, Leica pulled off a product which seemed to be ahead of everyone else, until you had to remind yourself that Sony was there first. But, Leica did it better. Apple did not invent the portable MP3 player, but it did make the best one. The iPod was announced in 2001, three years after the Diamond Rio PMP300 (and the Rio wasn't even the first such product).

 

But Apple did such a great job that its predecessors effectively ceased to exist. In the case of the Sony RX1 vs the Q, the gap is by no means that dramatic. The RX1 still has advocated today, both amateur and professional. It's by no means a forgettable camera. But the Q was a surprise hit for Leica, and they had no idea how popular it would be.

 

The Q, with 24Mpx, is superior to the Sony RX1R II with its 42Mpx (although that was probably because the Sony included a variable low-pass filter). It matches the Nikon D800 and D800E, those having 36Mpx sensors (sidenote: if you really must choose between those two superceded cameras, skip the D800E).

 

It is almost a match for the D850, with its very high resolution, being 45Mpx; and the Canon EOS R, with 50Mpx. Though in these cases, 50Mpx is 'merely' a 50% increase in resolution over 24Mpx. Also, the lens in the Q is exceptional - because it has to be.

 

Which brings me to my first point - although this will entail a pissing contest, so please be warned. If you took any DSLR with a 36mm sensor, regardless of brand (there are four that I can think of), and tried to replicate the functionality of the original Q, you'd go backwards instantly.

 

Not that you ought to replicate the Q, but if you did... think about it. You'd have arguably lesser performance: frame rate, resolution, weight, size, complexity. The only thing you'd lose with the Q, due to contrast AF, is AF performance. This might be a big deal for some. Although when it comes to manual focusing an AF lens, the Q's dual control system is, on paper, the best in the business.

 

Look at this size comparison of a DSLR with a 28/2.8 compared to the original Q - ay caramba:

 

Link: Compact Camera Meter

 

But why would you even think about replicating the Q with another camera? Simple: because of the superiority of digital over film in some areas (I said "some", don't @ me) you can perform moderately tight crops to a digital file and still be left with a nice image. This means that you can mount one lens - a prime lens, mind you - and have it perform the duties of three or even four. This only holds true to a point. Ironically, film handles extreme crops slightly better. Compare a frame of Super 8mm film to a frame of 35mm film, then perform an equivalent crop from a digital file, and see if I have a point or not.

 

My contention is that you can take a medium or large sensor of high resolution, cut it into four, and still have a file capable of enlargement. Let me give a practical example. Take a 28mm lens that out-resolves both a G9 and an A7. Take a shot with the G9, then, without changing perspective, take a shot with the A7. Now, crop the A7's file to match the AOV of the G9's file. The G9's file will have more resolution - close to double. But the cropped A7 file, at 6Mpx, can be enlarged to a moderate size.

 

The original Q, the Typ 116, had a very nice 24Mpx file. Its output was 7.5Mpx files when cropped to 45° (50mm crop mode). Now, the Q2 has 47Mpx to play with. The 45° AOV is now captured with 14Mpx. According to many photographers, including me, this is more than enough for any application - but that's after cropping, not before, which is the point I'm making. The 31° AOV (75mm crop mode) is captured with a respectable 6.5Mpx.

 

6.5Mpx is somewhat less than the ideal of, say 10-16Mpx, but it's more than the Digilux 2 which has 5Mpx, and more than the Nikon D2H (arguably the best sports camera of its day) which has 4Mpx. Remember, too, that the D2X had a crop mode which gave files of a little under 7Mpx. And these 6.5Mpx files will arguably surpass the quality of almost any 12Mpx phone camera. Although if you saw the files of the 2018 iPhones, you might not want to bet money on that until you've done a direct comparison.

 

All of this is great, but it only holds true up to a given AOV, then it rapidly collapses. I've been thinking for several years about relating AOV to sensor size. For example, I'd shoot birds and sports with Micro 4/3 and not full frame, but landscapes with medium format and not APS-C. The announcement of the Q2 was a reminder that this kind of thinking should probably be discussed a bit more.

 

I've attached a simple chart which attempts to match AOV to sensor size, based on multiple variables such as weight, performance, image quality and what not. Just note that this is a mere attempt at formalizing this sort-of-new way of thinking about focal length and sensor size. It's merely a hypothesis at this point, and I'm sure that someone out there will be able to refine the numbers.

 

AOV_vs_focal_length_ideal.thumb.png.73282cffa4915d5290c63c8d2ef6680a.png

 

References

DPReview on the Q Typ 116: Leica Q In-depth Review

DPReview on the RX1R II: Ultra-compact: Sony Cyber-shot RX1R II review

Ars Technica on the Diamond Rio: The Diamond Rio PMP300: Can this classic 18-year-old MP3 player still cut it?

An early review of the Q2 by Jonathan Slack: Leica Q2

A review from 2012 of the Sony RX1 by Steve Huff: The Sony RX1 Digital Camera Review Part 1: An intro to a full frame compact POWERHOUSE – Steve Huff Photo

Link to comment
Share on other sites

When I was working abroad on engineering projects, I used a little Olympus P&S with a fixed, 28 mm lens to document progress, and was perfectly satisfied with the results (my Leica stayed home).If someone told me I could enjoy the benefits of interchangeable lenses by cropping the results, my reaction would be less than charitable. In official engineering jargon, "Don't xxx on my leg and call it rain."
  • Like 1
Link to comment
Share on other sites

If someone told me I could enjoy the benefits of interchangeable lenses by cropping the results, my reaction would be less than charitable.

Everything is a crop of everything larger. 6x4.5 is a crop of 6x6. 6x6 is a crop of 6x9. 6x9 is a crop of 4x5". 4x5" is a crop of 8x10". Etc. You can't crop forever, but it works very well.

Link to comment
Share on other sites

Everything is a crop of everything larger. 6x4.5 is a crop of 6x6. 6x6 is a crop of 6x9. 6x9 is a crop of 4x5". 4x5" is a crop of 8x10". Etc. You can't crop forever, but it works very well.

That is a gross over-simplification. A classic 4x5 camera has a resolution of about 40 l/mm, whereas the resolution of a high-resolution digital FF camera is about 200. Cropping a 47 MP camera from 28 mm to 75 mm brings the resolution down to about 50 l/mm. In the 'day," most newspapers printed 4x5 images at 1 column size or less at 60-80 dpi. Once the medium of choice changed to 35 mm, there was no practical difference. However the difference is easily observed in a postcard sized print cropped in the same manner.

 

Are you trying to tell me it was rain?

Link to comment
Share on other sites

  • 4 weeks later...
I see what you’re saying, but still, there’s something deeply unsatisfying about the idea of paying $5000 for a camera and only using 1/4 of the sensor.

 

If you did that all of the time, I would agree. (And I have still not put a DX lens on my D700.)

 

But the other way is that the $5000 camera allows cropping when needed, still with enough resolution.

 

With the cheaper camera, you need to fill the frame on every shot, to get enough resolution.

-- glen

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...