l_d_allan

Smart-phones?
It is with very mixed feelings that I acknowledge how good they've gotten in the past several years. As the saying goes, "the best phone is the one you have with you."

Agree. The article barely mentions (or even ignores) how much modern IBIS helps reduce the need for fast lenses. Coupled with very good high ISO performance, f/4 is often satisfactory with marginal light.

My non-stabilized Canon EF lenses perform MUCH better with inexpensive adapter to my Sony a7R2 with IBIS ... even ignoring the MUCH better Exmor sensor with ultra-res and ultra-DR and dual-base-ISO magic.

Interesting. However, my less-than-fully-informed impression is that coatings have gotten MUCH better over the past decade or so.

I'm pleasantly surprised that DPR has significant editorial independence. Whew (and thanks for the clarifying reply).

The article barely mentions a disadvantage of the fast primes ... lack of balance on a small full-frame body. I rented a Sigma 105mm f/1.4 for my Sony a7R2. It has a huge, heavy front element. I felt the need for a monopod.

That lens would have been more at home on a Canon full-frame body like a 5d4. Or a Nikon D850 with modern Sony sensor with ultra-res and DR.

And I'm underwhelmed by the hyper shallow depth of field look. With a 3/4 portrait, one eye will be in focus, and the other eye and nose are out of focus. No thanks.

With two people, f/4 or f/5.6 works better with 70mm to 150mm of focal length.

I'd guess the Amazon mother-ship won't be happy about this article. Anti-G.A.S.?

This Sony full-frame defector from Canon is still waiting for an affordable 100mm f2. Their excellent 85mm f1.8 isn't enough different in focal length from my excellent but relatively overpriced 55mm f1.8. My Sony 28mm f/2 is extremely affordable.

To me, the best prime values are the Canon EF's for DSLR, adapted for my a7R2. I'd like to have eye-detect, however. Ka-ching!

For older Canon DSLR's like the 6d and 5d3 with DIGIC 5 and older, MagicLantern implements exposure tools based on RAW, including histograms, blinkies, ETTR, Zebras, and also ....
Waveforms

I'm unclear on the status of cameras with DIGIC 6 and newer.

I am a Yuge fan of ML, but its non-availability for the 5d4 was part of my reason for defecting from Canon to Sony a7R2.

Ah ... thanks for the feedback. It didn't occur to me about the timing. Duh.

The relative sizes of the Earth / Moon look about correct with the Earth several times larger.

Nice image.
However, I wonder if the Earth + Moon has a lot of "blooming" from (unavoidable?) overexposure.

The moon appears to be about one Earth diameter (or less?) from the Earth.

However, with the Earth's diameter of about 8,000 miles and the Moon about 239,000 miles from the Earth, I'd expect both the Earth and the Moon to be Much Smaller.

I also wonder how large the Sun would appear.

Inquiring minds want to know?

Unfortunately, ML development seems to have been stagnant for several years now. The last "port" was apparently for the 5d3. Nothing for the 5d4, 5ds/r, 6d2, etc.
Bummer.

Review?

Was I the only one who had trouble finding one of the .... if not THE ... most important spec .... price?
(about $800 at B&H and Amazon ... maybe I didn't look hard enough, but I never did come across the MSRP in the review)

This may be over-simplistic from an less-than-informed entry-level astro-photographer, but a fuzzy logic yes-no decision tree might be:

* using star tracker ... star eat'ing may matter
* otherwise ... probably doesn't matter

I wonder: might the a7Riii's capability of ultra-resolution using sensor-shift make a difference wrt star eat'ing?

Jim K ... to the rescue ... yet again?

Sort of agree, but at least to me, it is a "who cares" to have 100...00 stars vs 100...01 stars vs 100...02 stars captured by the sensor of my 42 Mpix a7R2 when doing star trail photos. Same with Milky Way.

Actually, with star trails, you can probably have too many stars tracing out arcs. I'll tweak ACR sliders to tune how many arcs are involved, with some or many being suppressed.

I suppose if you were really doing serious stuff like exo-planet detection, it would definitely matter. However, my impression is that use case requires specialized gear, as well as that much more "attention to detail".

But I agree that the option to eat or not-to-eat being an in-camera menu choice would make sense, partnered with post processing enabled or disabled.

I can understand the default being the current behavior.

My impression is that for what might be considered "entry level" astro-photography of star-trails and wide angle captures of the Milky Way (including panos), the star eater issue is a minor to negligible matter.

It may come into play with deep space captures with longer focal lengths and a star tracker.

I own a Sony a7Rii rather than a Nikon, but it has a similar "Aptina conversion gain changing trick".

My practice is to that 99% of my captures use either ISO 100 (first choice if enough light) or failing that, ISO 640 and work with ISO-less capability from 640 thru 12,800+.

ISO 640 on the a7Rii is where the "trick" happens ... that would be ISO 400 on the Nikon D850.

I especially tend to very rarely use Auto-ISO. The exception would be when I might show people who are the subjects the image in the LCD ... they could be understandably concerned if the image is perceived to be significantly under-exposed (and greenish from UNI-WB).

* Read the manual for the equipment.
* Re-read the manual some months later to gain insights into subtle features you may have missed on first reading.
* Get photography related books thru your local library.
* Join forums related to general and specifics and ask plenty of questions.

And I thought "talking head" videos were boring ...

Perhaps it makes sense to work with Roger C. at LensRentals to out-source quantitative measurements? IIRC, he is already working with another web-site that provides tests of multiple lenses from OLAF.