Rishi Sanyal

Rishi Sanyal

DPReview Administrator
Lives in United States Seattle, WA, United States
Works as a Deputy / Technical Editor
Has a website at www.rishi.photography
Joined on Feb 25, 2014
About me:

Although I'm a scientist by training, having recently completed my Ph.D in biophysics, photography has always been a huge passion of mine. It's been an incredible opportunity to meld these two interests together here at DPReview!

Comments

Total: 2169, showing: 1 – 20
« First‹ Previous12345Next ›Last »
On Sony: An eye on focus article (648 comments in total)
In reply to:

Skipper494: If you'll pardon the pun, I fail to see the point, as most advanced photographers are used to point focus and exposure, with recomposition, so just a few points in the center would suffice most of us.

And the ironic thing is that it's perhaps pros that benefit *most* from this sort of tech - b/c it allows them to focus on more important things, like the composition, the lighting, the moment. So they can make even more compelling photos, have an even higher hit rate, which itself leads to a better chance of awesome photos.

The same arguments were made back when autofocus was introduced... and what happened to those arguments? They quickly became irrelevant, while autofocus allowed for photos never before thought possible.

Direct link | Posted on Jul 6, 2015 at 05:01 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

DigitalWalnut: Any plans for a low-light AF stress test between the big players? I'm curious to see how Nikon's latest system does against the massive spread of cross-type points on the 7DII.

We've already done something of the sort in our Nikon D750 review [here](http://www.dpreview.com/reviews/nikon-d750/7), where we tested the -3EV claim of all AF points.

The off-center points of the D750 did continue to focus in lower light than the 7D2.

Cross-type sensitivity & low-light sensitivity are somewhat orthogonal (though not entirely), & you can definitely have one without the other. The reason cross-type sensitivity is somewhat relevant in low light, I think, is b/c in low light situations, subjects tend to have lower contrast, and so 'looking' along two axes for detail to evaluate phase differences is always better than looking along one. So that in case one such detail has more contrast than another orthogonal detail, your system will use it instead, and so therefore may be quicker.

Important to keep in mind though that non-cross-type sensitivity doesn't mean your subject has to align exactly along the sensitive axis but, instead, have some projection along that axis.

Direct link | Posted on Jul 6, 2015 at 04:57 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

Skipper494: If you'll pardon the pun, I fail to see the point, as most advanced photographers are used to point focus and exposure, with recomposition, so just a few points in the center would suffice most of us.

Focus and recompose doesn't work for shallow primes below a certain focal length, nor does it work for moving subjects. So if you want to shoot moving subjects, especially with fast primes, this matters.

In fact, it *really* matters.

Wedding, newborn, kids, candid portraiture, etc. - all these types of photography stand to benefit immensely from this sort of tech (as well as what already exists in Nikon full-frames).

Direct link | Posted on Jul 6, 2015 at 00:46 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

wombat661: "Imagine the potential for fast and accurate focus with shallow depth-of-field F1.4 primes."
I see you did the test at f4. But you talk about f1.4. You realize f1.4 has a razor thin depth of field. So... can this Sony actually focus track at f1.4. For comparison, even my 3 generations ago old Nikon D7000 can track an electric go-kart coming at me with a 85mm f1.4 lens and get tack sharp focus in moderately low light. In gymnastic I can track and catch a high speed spin on the bars in moderate light with the DSLR. You realize that is FAST!
Every 3 months yet another mirrorless comes along and say they can match DSLR. Each time turns out to be a bust. So can it really, really I mean really really do it this time???? Don't come back in yet another 3 months and say really, this XYZ camera can really match DSLR!

It's not about speed of the subject, it's about the change in subject distance relative to DOF. And that was actually pretty huge for our 50mm F1.4 example, so it's actually far more strenuous than you think.

Read my extensive comments on this in a comment above, [here](http://www.dpreview.com/articles/1652088044/sony-an-eye-on-focus?comment=0574751025).

Direct link | Posted on Jul 5, 2015 at 20:43 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

naththo: Unrelated to this topic but my brother bought Powershot SX710 which is Canon and autofocus is awesome, very quick. I noticed my dad's old Cybershot RX1 done very poorly to compare. These were done in low light. Sad to say Sony isn't still as good as Canon usually. Canon sensor not as good as Sony but AF is totally the opposite!

You're comparing focus at an effective 25mm f/18 equivalent on the Canon vs the Sony's 35mm f/2?

You realize the Canon barely has to even focus at f/18 equivalent, right?

Also, the RX1 (CDAF only) is literally one of the worst cameras for focus made in recent history, and is entirely unrepresentative of pretty much every other Sony camera out there.

Using it as representative of 'Sony AF' is entirely disingenuous, at best. The RX100 is CDAF only and it's miles ahead of the RX1; throw in PDAF on top of that, and the a7R II is far, far ahead of the what the RX1 can do with respect to AF, which should be pretty clear from these videos.

Direct link | Posted on Jul 5, 2015 at 20:38 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

LBJ2: *** In case you missed it: 'magnify' in playback now magnifies the AF point used at capture, rather than simply magnifying the center of the frame. No more scrolling over to the point of interest in playback to check focus; you now have one-click (albeit somewhat laggy) magnification of what was focused on. This extends to all three cameras: a7R II, RX100 IV, and RX10 II.

Thank you for pointing this out. I will find this feature very helpful

Really? That's great if they did! Thanks for the heads up Daniel. We'll have to look into this.

Direct link | Posted on Jul 5, 2015 at 07:04 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

Rocco N: Ok so what does this really mean in real world every day shooting? I can't remember missing focus taking pics of people with either my Panny point and shoot or any of my Nikon DSLRs. It's certainly new but in what percentage of circumstances will all these focus points really matter.

Good point @2eyesee.

Problem is, I have seen shots with my RX100 III where the eyes weren't in perfect focus. So I would imagine that eye-AF will help. After all, we're talking about F4.9 full-frame equivalent at the wide end (which is mostly where I shoot), so it's not like perfect focus on the plane of the eyes is *never* a problem at those apertures on full-frame.

But you're right - much, *much* bigger deal on the a7R II than the RX100 IV.

Direct link | Posted on Jul 5, 2015 at 06:29 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

wombat661: "Imagine the potential for fast and accurate focus with shallow depth-of-field F1.4 primes."
I see you did the test at f4. But you talk about f1.4. You realize f1.4 has a razor thin depth of field. So... can this Sony actually focus track at f1.4. For comparison, even my 3 generations ago old Nikon D7000 can track an electric go-kart coming at me with a 85mm f1.4 lens and get tack sharp focus in moderately low light. In gymnastic I can track and catch a high speed spin on the bars in moderate light with the DSLR. You realize that is FAST!
Every 3 months yet another mirrorless comes along and say they can match DSLR. Each time turns out to be a bust. So can it really, really I mean really really do it this time???? Don't come back in yet another 3 months and say really, this XYZ camera can really match DSLR!

No, the entire first video was shot at 50mm, f/1.4.

You don't have to take my word for it, but if you do wish to, please note that I shoot with 24, 35, and 85mm F1.4 lenses primarily, with a D810 and 5D Mark III, so those are my reference points. Always microadjust all my lenses, as well.

And with that said, I was duly impressed with the a7R II with the Canon mount Sigma lens, and even more impressed with Sony native lenses. Because of the accuracy of on-sensor PDAF just based on theory alone, & the obvious pinpoint accuracy of the a7R II's subject tracking, outside of 'Lock-on AF', anyway.

Also, if you watch the 50/1.4 video closely at 1080p, you'll see the a7R II actually keeping up, moving the focus element as necessary to keep the subject in focus, despite quick, fast movement that constantly threw the subject, Richard, well outside of the razor thin DOF of 50/1.4.

This is *not* some marketing piece saying mirrorless has matched DSLR. We start from the position of skepticism.

Direct link | Posted on Jul 5, 2015 at 06:26 UTC
In reply to:

Dan Bracaglia: For those of you concerned about the very noticeable chromatic aberration in several of these images, I've brought a few back into ACR and attempted to remove the CA. Here are two examples of corrected images:
[Image 1](http://www.dpreview.com/galleries/reviewsamples/photos/3239947/05iso125dan-ca?inalbum=sony-rx100-iv-real-world-samples)
[Image 2](http://www.dpreview.com/galleries/reviewsamples/photos/3239948/01iso125dan-ca?inalbum=sony-rx100-iv-real-world-samples).
I've also left the original edits for the sake of comparison.

Good man, Dan. CA becomes especially evident with lots of post-processing, including highlight reduction and shadow pushing. Easy to remove in ACR though, with not too much of an impact on sharpness unless the CA was really, really bad (many pixels wide) to begin with.

Direct link | Posted on Jul 5, 2015 at 02:54 UTC
On Readers' Showcase: Raiatea Arcuri article (53 comments in total)

Stunning, gorgeous, beautiful scenery. Kudos to you, sir! If you're ever in Seattle, please swing by our offices and teach us all a thing or two. :)

Direct link | Posted on Jul 5, 2015 at 02:52 UTC as 2nd comment
On Readers' Showcase: Raiatea Arcuri article (53 comments in total)
In reply to:

bwabl: Why is it that many modern digital images of landscapes/nature as displayed under this section of dpreview are so color-saturated and overworked to a degree of unnatural kitch? You sometimes feel as if you look up-close into a bowl of multicolored pudding...

Perhaps it's time to calibrate your monitor, or stop expecting the drab JPEG defaults of most cameras that don't have the dynamic range to even deal with most of the scenes pictured here.

Direct link | Posted on Jul 5, 2015 at 02:50 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

Rocco N: Ok so what does this really mean in real world every day shooting? I can't remember missing focus taking pics of people with either my Panny point and shoot or any of my Nikon DSLRs. It's certainly new but in what percentage of circumstances will all these focus points really matter.

"I can't remember missing focus taking pics of people with either my Panny point and shoot or any of my Nikon DSLRs."

Lucky you :) Try shooting full-frame shots of people at f/2.8, f/2, or f/1.4. Now try that with moving subjects...

Direct link | Posted on Jul 5, 2015 at 00:31 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

njlarsen: I question the choice of lenses for this test. Short lenses are always going to be easier to track than long ones. When you get the actual samples for a real review, please make appropriate tests using a long lens (400 mm at least for a full frame). Don't do the easy part of telling us how it works with a standard lens -- all systems are perfect with that type of test.

Niels

All that said, even an a6000 can initially acquire AF at 300mm equivalent quite quickly, so on-sensor PDAF is getting there. Furthermore, I bet this can be solved computationally - by dynamically varying the effective phase-detect sensor baseline on the fly depending on amount of defocus.

Maybe not, though, as dedicated PDAF modules benefit from looking through 'virtual apertures' that yield more 'in focus' data for the phase-detect sensors to work with (they're looking through something like f/16-f/22 virtual aperture, if I remember correctly, based on Doug Kerr and Marianne Oelund's excellent work). Think about those split prisms - the images you were aligning were somewhat in focus.

I think on-sensor phase-detect sensors 'see' much blurrier data, and I'm not sure how they're going to work around that, or how they do currently.

This is my understanding anyway - hopefully I'm not way off/dead wrong. :)

Direct link | Posted on Jul 5, 2015 at 00:30 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

njlarsen: I question the choice of lenses for this test. Short lenses are always going to be easier to track than long ones. When you get the actual samples for a real review, please make appropriate tests using a long lens (400 mm at least for a full frame). Don't do the easy part of telling us how it works with a standard lens -- all systems are perfect with that type of test.

Niels

Yes, exactly what @abortabort and @fatdeeman said.

Lots of common misconceptions that really need to be re-evaluated.

The *biggest* issue w/ long focal lengths is hunting - initial AF acquisition is challenging due to the fact that things can be *extremely defocused*, and that's where cameras with longer baseline PDAF sensors can perform better as they can make measurements from extreme defocus. But otherwise, shorter subject distances with fast wider primes are likely to be far more taxing - especially when you consider the amount the focus elements have to move to keep tracking the subject. As @abortabort said, there's much more movement necessary w/ faster wide primes at shorter subject distances.

Direct link | Posted on Jul 5, 2015 at 00:25 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

Timur Born: The examples with third party lenses don't exactly leave me with awe, I do appreciate the effort, though. There are too many shots with the nose in focus and focus jumps around too much (ears, nose, eyes + cheeks).

If people are lured to rely on this instead of their own focusing skills (mine are woeful) then it might lead to more disappointment than what the high entry price suggests (body, adapter, lenses that you might or might not own yet).

Exactly what fatdeeman said. It's refreshing to see an audience member here that gets it.

Your observations are exactly what we've found- that the closest thing to this is the Rebel T6s in... 'surprise surprise'... live view, but even then it's markedly slower w/ much more hunting, likely b/c it has far too few PDAF points (or the algorithm, or some combo thereof).

No way does a 7D II or even 1D X subject track with this level of speed, accuracy, and confidence, except when you have just one object isolated in depth, since here the phase-detect data alone can be used quite effectively to subject track. But that doesn't work well for 90% of other scenarios, whereas the pattern recognition/image analysis the Sony is using does. It's a bit strange, in fact, as the 7D2, 5DS, and 1D X *should* be using more pattern recognition w/ the hi-res metering sensor, but it appears they don't for some reason.

And, as you said, the Nikons do this very well, using their metering sensors effectively.

Direct link | Posted on Jul 5, 2015 at 00:23 UTC
On Canon EOS 5DS / SR First Impressions Review preview (3381 comments in total)
In reply to:

beavertown: Dynamic range compare to entry level APS-C D5500 is awful.

Also, 'half' the read noise sounds bad, but take a look at our ISO-invariance widget to take a look at how much that actually affects tones even after the 5-6 EV push. Since we're only talking about a couple electrons here and there, the biggest difference is down in the deepest shadows where SNR is so low due to shot noise anyway that it's largely irrelevant for most users. 'It' being the noise cost from shooting ISO 200 then boosting 5 EV vs. shooting natively at ISO 6400 in-camera.

In other words, 1 or extra electrons of read noise aren't going to make a dramatic to most of your tones, save for the tones you probably shouldn't be using anyway.

Still, I'd prefer that extra cost to not be there, so I look forward to seeing even more progress in this area, possibly by increasing ADC bit-depth.

Also, it's less about being ISO invariant and more about just having less downstream read noise. In fact, ISO variance *could* be a good thing if it's not due to downstream read noise...!

Direct link | Posted on Jul 5, 2015 at 00:09 UTC
On Canon EOS 5DS / SR First Impressions Review preview (3381 comments in total)
In reply to:

beavertown: Dynamic range compare to entry level APS-C D5500 is awful.

It's quite ISO invariant *relative* to most cameras... about as ISO invariant as it gets, with the D7200 perhaps being the leader here.

It's all relative, right? As sure the D750 may have half the read noise at ISO 6400 as ISO 100 (since you approach upstream read noise levels at higher ISOs, whereas low ISOs include components from both upstream and downstream read noise), but the 6D has 1/13 the read noise at ISO 6400 compared to 100.

Furthermore, I'm willing to bet the largest source of the extra noise at low ISO for the Sony/Nikon cameras is quantization error from the ADCs being too low bit-depth. A totally different issue compared to the off-chip ADC architectures of Canons.

I'd love to see a 16-bit ADC on the D810 to see if pixel-level DR goes above 13.9, which is suspiciously close to the ADC bit-depth... indicating, to me anyway, that the ADC is limiting (unlike on Canons, where it's the architecture itself that's limiting).

Direct link | Posted on Jul 4, 2015 at 23:39 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

CBRWizzer: Great specs, no lenses. Next.

But you still get Wide, Center, and Flexible Spot - all with continuous AF - with 3rd party lenses. Which potentially offers accurate AF-C with third party glass, low-light and focus-from-extreme-defocus issues not wisthstanding. But still, that ends up opening benefits of focus that you might not even get on native bodies with these lenses:

1. Accurate AF without microadjustment
2. Wide coverage of AF point selection
3. Excellent subject tracking as long as you have faith your camera will always automatically pick the right subject (yes, we want Sony to work on changing this implementation).

Point being, even with the lack of such native FE lenses (which are a growing market, btw, hello Zeiss Batis...), there should still be quite a bit to be excited over.

Direct link | Posted on Jul 4, 2015 at 10:17 UTC
On Canon EOS 5DS / SR First Impressions Review preview (3381 comments in total)
In reply to:

beavertown: Dynamic range compare to entry level APS-C D5500 is awful.

"I'm just not so sure about the ISO invariance stacking up so well unless you might have surprise highlights (if they aren't a surprise you should perhaps be shooting at lower ISO in the first place)."

Right, but that's just the thing - if those 'surprise highlights aren't a surprise', then, assuming you wish to keep them, you need to underexpose, be it via actually underexposing your focal plane exposure, or by lowering your ISO.

Either method will run the risk of getting other tones (midtones?) you care about closer to your camera's noise floor, and that noise floor might then show when you try to bring those midtones, that you pushed down to preserve the highlights, back.

And that's simply where a camera with a lower noise floor helps. You run into these issues less frequently. And meanwhile, even 1, 2, or 3 stop pushes are helped, as they still show much less noise than similar pushes from a camera with a higher noise floor (or low base ISO DR).

Sure you know all this though :)

Direct link | Posted on Jul 4, 2015 at 10:08 UTC
On Sony: An eye on focus article (648 comments in total)
In reply to:

njlarsen: I question the choice of lenses for this test. Short lenses are always going to be easier to track than long ones. When you get the actual samples for a real review, please make appropriate tests using a long lens (400 mm at least for a full frame). Don't do the easy part of telling us how it works with a standard lens -- all systems are perfect with that type of test.

Niels

In fact, we tend to see systems shooting 200-450mm equivalent wide open tracking Richard biking toward us or sports scenarios - when we're able to test them - performing quite well.

More challenging conditions are like the one the D7200 faced [here](http://www.dpreview.com/reviews/nikon-d7200/11) at 450mm equiv. f/4. There in the last shot, Richard was ~15m from the camera, so the DOF was ~0.37m. Richard was probably falling out of that DOF pretty easily, and the camera struggled a bit here and there, but still probably nailed a good 80%.

Where even that camera's excellent AF struggled was w/ 24mm equiv shots at f/1.8 in our studio w/ the mannequin. The DOF is so shallow, and our movement so drastic *relative* to this DOF, that the AF system really has to work hard, as it's so easy for the subject to fall far outside of the DOF it was just in.

I think this is why see a better-to-worse trend in AF tests going from telephoto to wide/fast primes w/ most cameras.

Does this make sense?

Direct link | Posted on Jul 4, 2015 at 10:00 UTC
Total: 2169, showing: 1 – 20
« First‹ Previous12345Next ›Last »