sonyalpharumors

To me, that looks very similar to the Canon 500mm f/4 element/ group layout. Perhaps it is the rumoured Sony 500mm f/4 G lens that has been lurking around for a while?

Smart guy? Nah, just an optical designer – who got sucked into looking at this patent last night.

Here is my summary of the claim: A zoom lens with 5 moving lens groups (+ – + – +) can be made smaller and with a wider zoom ratio than a comparable lens with 4 moving and 1 fixed groups, and still deliver high image quality. Five embodiments are demonstrated.

This does not mean a patent on ANY 5-lens group lens – A cursory search in one of my lens dBs found a couple 5-group patents (USA 4963006 160990, Kiyohika, Japan 58_33531 830720) There are, of course, many patenets for compact zoom lenses.

In the lens prescriptions and diagrams, angles are in degrees and dimensions are unit-less: Focal length at each lens’ widest angle is normalized to “1.00″

As an example, here are the specs and my analysis of the first embodiment:

• Half Image Size: 0.812
–> 0.9 x 1.35 (2:3) or
–> 0.8 x 1.42 (16:9)
• Half Field: 43.28° to 1.357°
–> 31.9x (33.7x with distortion)

Performance specifics: (I get as much information from the diagrams as from the text (grin.))
• Wide End:
– High Barrel Distortion (-13%)
– Astigmatism is dominant
– >50% vignetting in the outer field (mostly in the 2nd lens group), 10% at 3/4-field.

• Narrow End:
– Some (2%) Pincushion Distortion
– Coma is dominant
– About 40% vignetting at outer field (mostly in the 3rd lens group), 30% at 3/4 field, 15% at 1/2-field.

Incidentally, the lens is nearly telecentric except at the extreme zoom positions. (Full-field chief ray is incident on the image plane at +1.8° at wide and -3.4° at zoom)

SIZE: If the wide-angle focal length is set 1 unit to = 1 mm, then the front element has an aperture of about 10mm and the assembly has a total length of 23.4 mm – 27 mm. A HD-sensor would only have 0.74 mircon pixels, so let’s look at spot size:

Looking at diffraction performance (using a visible spectrum spread), Spot size (70% energy, an arbitrary but useful performance spec) is given below in microns as a function of zoom position:

efl — Avg Spot Size (70% energy)
1.0 mm — 3.1 microns
2.3 mm — 3.3 microns
7.4 mm — 3.8 microns
15.0 mm — 4.5 microns
33.7 mm — 6.0 microns
Average: 4.1 microns

With an average spot size of ~4.1 microns, a HD detector would sample each spot with 5.5 pixels (a bit overkill (unless you’re doing specialized astrophysics (but I digress…))) Making a detector with 2 pixels/spot would be 388 x 690 pixels.

Now, let’s test the “smallness” claim: Scaling the lens to “half unit” size (5mm aperture and total length 12-14mm) does produce spots that are just slightly more diffraction limited:

efl — Avg Spot Size (70% energy)
0.5 mm — 2.8 microns
1.2 mm — 3.1 microns
3.7 mm — 3.7 microns
7.5 mm — 4.1 microns
16.9 mm — 4.9 microns
Average: 3.7 microns (94% Diffraction-limited)

At this point lenses are getting VERY tiny and I would wonder about manufacturing tolerances being able to produce this performance. Also realize the image plane size has been cut in half (1/4th the area) and HD pixels would be 0.37 microns, which wouldmean 10 pixels/spot. NOT an HD camera. Using 2 pixels/spot would be good for 215 x 383 pixels.

What if the system is scaled to 2x units? Aperture is 20mm and max length is 54mm. Average spot size is now 5.8 microns (No longer diffraction limited) and HD pixels are 3.9 pixels per spot. This is half the point where it could be used for a HD video camera. — at 2 pixels/spot the detector would be 550 x 976 pixels.

At 3x, (30mm aperture, 80mm length) average spot size is 6.9 microns, 2.2 micron pixels, or 3.1 pixels/spot. Still not optimal for an HD sensor. 2 pixels/spot: –> 692 x 1231 pixels

Just using a quick inverse sensitivity tolerancing, the system is just about equally stable with the half-sized model, but the small components are going to be much harder to manufacture and align.

So, what is this useful for? Perhaps a lower-resolution zoom micro camera. Perhaps nothing other than a patent claim for SONY’s clever optical designers, from whom I’ve just learned a bit. (Thanks, guys!)

In conclusion, this is a lens that can indeed be made small with a 30+ zoom range. Is it “better”, or “unique”, and is the patent defendable? — That’s for some smart guy to figure out.

– Bert Pasquale
Member, Maryland Professional Photographer Association, http://MarylandPPA.com