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: 4973, showing: 1 – 20
« First‹ Previous12345Next ›Last »
In reply to:

Absolutic: Rishi please define what you mean by "central region" on a9. I saw that language in metabones field notes but not clear what it means in relation to a9

Good question! Hard to quantitate. Let me see if I can try build a visualization. Preliminarily I'd say the Flexible Spot M mode and all of its immediate surroundings seem OK with most (even telephoto) lenses. Lot more testing to be done though.

Link | Posted on Jun 28, 2017 at 02:13 UTC
In reply to:

lwestfall: How about video AF?

@cyberstudio - that's far too clunky for real-world use. It hunts and hunts (albeit rather visually dampened) and then finally settles. The noise itself means you need pro audio gear: it's not very user-friendly and not worthy of mention in my opinion.

Curious to hear opinions of actual photographers: are you suddenly using adapted canon lenses for video AF? if so, how? It doesn't work reliably enough for my neworn/toddler videography (compared to native lenses), but more than happy to proved otherwise.

Link | Posted on Jun 28, 2017 at 02:11 UTC
In reply to:

Mike Sandman: This is a significant update for A9 owners, and perhaps others. You may want to move this from "Latest Articles" over to the main news board on the home page.

I agree. We'll do so when we have a video demonstration (though it'll be rather amateurish, fair warning, because of time/resource constraints).

Link | Posted on Jun 28, 2017 at 01:16 UTC
In reply to:

beavertown: Breaking news: A9 Banding Issues.
https://www.youtube.com/watch?v=zv1JCHfXDdY

This will get another 2000+ comments for Dpreview if they use it.

(Would not respond to that notoriously belligerent guy, he is wasting his time.)

Fro claims this was under daylight 'and I'm not sure if the artificial lights were on yet or not'.

Given that we've never even experienced banding other some of the harshest artificial light (flickering at 120 Hz, with low duty cycles due to dimming), we think there's literally something wrong with Fro's unit.

I wouldn't extrapolate his findings to apply to all cameras. He himself doesn't - he says 'this is a problem I've experienced with my unit and I can't speak to all units'.

Link | Posted on Jun 27, 2017 at 19:43 UTC
In reply to:

webber15: So...it's just 5fps then (I refuse to recognize an electronic shutter for stills).

Why? The electronic shutter has a shutter rate of 1/160s, barely a stop behind the mechanical shutter (1/300s). That means issues in only the rarest of circumstances. And, no, we haven't had a single instance of banding under artificial lighting. Only a tiny bit of rolling shutter in very fast pans.

I don't know why anyone would 'refuse to recognize an electronic shutter for stills'.

The mechanical shutter is on its way out, just FYI.

Link | Posted on Jun 27, 2017 at 19:40 UTC
In reply to:

lwestfall: How about video AF?

No.

Link | Posted on Jun 27, 2017 at 19:38 UTC
In reply to:

Brian_Smith: Today's Metabones Smart Adapter Firmware Update v57 adds Sony A9 Support. Any tests previous to this update are irrelevant. http://bit.ly/2tNcnDP

Thanks Brian! Testing this now. Seems to support 10 fps with Canon lenses. You're still limited as to the area of coverage depending on what focal length you're shooting (shorter focal lengths offer PDAF over a wider area).

Link | Posted on Jun 26, 2017 at 21:18 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

john Clinch: It does feel like we are debating how many angels we can balance on the tip of a pin.

If we are now at the point where to realise the full resolution of the sensor you need to correct lens, the correct aperture and to focus with in a set of iterations with an external monitor.

Surely we can just say that sensor resolution is adequate. Your only ever see it through luck

PS quite surprised that contrast detect auto focus can't focus the lens accurately. Presumably that will show up as a negative in the review?

You'll be able to achieve it, you probably just won't care. Like I say in the note at the end: "In fact, the difference in focus between our 1st and 2nd set of shots, in the real world, likely would've meant the difference between focus on one eyelash vs. another whilst the entire eye was within the depth-of-field."

Remember in the real world you shoot 3D objects, not 2D charts. Also, then lens will come into play far more than which 24MP sensor you're using... (though sometimes that'll have a role as well: e.g. is there an AA filter? are the microlenses gapless? etc.)

Link | Posted on Jun 26, 2017 at 21:17 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

pixtorial: Sorry, but just more evidence that the technical reviews from sites like DPR should be weighed heavily with skepticism. There are so many variables in the shots from the test scenes, and it would take substantial scientific rigor to create true apples-to-apples comparisons, much more than has been applied. The closest we have to more objective testing are Roger's posts over at LensRentals and a few trusted forum members who have the optical engineering know-how to provide meaningful data.

As many here have suggested, the only true way to evaluate a camera and lens combination is to get out and shoot, process, evaluate, and share.

And I disagree with the conclusion of the author at the end of the article. In the real world, we do evaluate our digital photos at high magnification, because it ultimately does impact the ability to confidently throw away pixels in cropping, latitude in post-processing, and the acuity of the image across multiple publishing channels.

But I of course agree - there's much room for improvement. One thing I'd love to do is start measuring AF speeds of different body/lens combos - an AF latency test of sorts (e.g. from 10x focal length to infinity, and reverse). See if there are differences from body to body (even within a brand), and what you sacrifice when shooting with the F1.4 vs a F1.8 version of a particular focal length. To name a few things.

Link | Posted on Jun 26, 2017 at 21:15 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

pixtorial: Sorry, but just more evidence that the technical reviews from sites like DPR should be weighed heavily with skepticism. There are so many variables in the shots from the test scenes, and it would take substantial scientific rigor to create true apples-to-apples comparisons, much more than has been applied. The closest we have to more objective testing are Roger's posts over at LensRentals and a few trusted forum members who have the optical engineering know-how to provide meaningful data.

As many here have suggested, the only true way to evaluate a camera and lens combination is to get out and shoot, process, evaluate, and share.

And I disagree with the conclusion of the author at the end of the article. In the real world, we do evaluate our digital photos at high magnification, because it ultimately does impact the ability to confidently throw away pixels in cropping, latitude in post-processing, and the acuity of the image across multiple publishing channels.

As the Editor's note says: "In fact, the difference in focus between our 1st and 2nd set of shots, in the real world, likely would've meant the difference between focus on one eyelash vs. another whilst the entire eye was within the depth-of-field."

I think we need an FAQ on what our studio scene is and isn't useful for.

"The much maligned Pentax AF, for instance, proved to be top of the PDAF class, for instance, w.r.t. to AF accuracy and speed" --> Any test that shows that needs to be updated, b/c real-world experience proves otherwise. This is precisely why we don't have truly repeatable test yet - because there are so many variables in AF. Clearly, whoever performed that test that showed Pentax at the top isn't accounting for some serious variables, because in the real world that doesn't bear out (that's not to say Pentax AF isn't adequate for some/many).

That's why right now we use an amalgam of real world sports / action shoots, along with bike / close-focus tests.

Link | Posted on Jun 26, 2017 at 21:14 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

Mariano Pacifico: (26) DPR Staff comments ... Sony apologists I believe ...

Yup that worked. No spaces.

Link | Posted on Jun 26, 2017 at 21:09 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

Mariano Pacifico: (26) DPR Staff comments ... Sony apologists I believe ...

I *think* just putting * around both beginning and end of a word. Sorry in advance to those who think I do it too much :)

Link | Posted on Jun 26, 2017 at 21:09 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

FencerPTS: While it's great and all to get a theoretically best possible test scene result, this change in methodology inevitably opens up caveats. Previously the test scene was the best achievable while manually focusing the lens oon a native 85mm. With the inability of the native 85 to focus precisely there is now a divergence between the test for the A9 and every other camera and lens. While it is clear the intent is the precision of the focus for theoretical best focus, the handicap remains. Shame on Sony foe the release of so imprecise a product yup necessitate this distinction.

The obvious corollary of this handicap is that the test results are not repeatable on the A9 as they are with any other camera and native 85 lens combination. Hence, this retest is both more and less correct than the previous in some fashion.

Remember, it can also be the maximum allowable magnification in live view... which is lower for the a9 than some (many? I haven't checked) other cameras.

Link | Posted on Jun 26, 2017 at 20:40 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

Random Photographer: Forgive my ignorance, but how do we know other cameras aren't suffering from the same flawed test? The A9 was expected to be sharp, thus the rebuttal by the community and the subsequent retest. But what about other Cameras? I for one think that DPR should retest all bodies that are still in production, to get a level and reliable ground for comparison.

... For example, you can't decouple the lens from the body in our test, so it's not really a test of 'how sharp is this sensor'. Which generally isn't that interesting of a question to begin with. We try to get a sharp image so that from a sharp Raw, we can assess JPEG detail, detail retention, noise reduction, color, etc.

One thing that's clear from these comments is that we should probably re-write our Studio FAQ, which is rather outdated.

Link | Posted on Jun 26, 2017 at 20:39 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

Random Photographer: Forgive my ignorance, but how do we know other cameras aren't suffering from the same flawed test? The A9 was expected to be sharp, thus the rebuttal by the community and the subsequent retest. But what about other Cameras? I for one think that DPR should retest all bodies that are still in production, to get a level and reliable ground for comparison.

Random Photographer - every test has tolerances. We don't tout our studio scene as the ultimate test of sharpness, specifically because we would need computerized analysis for that.

Does this article mean all other tests are invalid? No, absolutely not. As I've said, internally we often shoot and reshoot until results line up - in our widget - with what's expected (by comparing to previous bodies of similar resolution/lens). It just so happens that *this time* our initial results did fall a little outside what we should've accepted as 'within the tolerance of our test' (thought it wasn't necessarily obvious in the center).

And remember what my editor's note says: "In fact, the difference in focus between our 1st and 2nd set of shots, in the real world, likely would've meant the difference between focus on one eyelash vs. another whilst the entire eye was within the depth-of-field."

Please keep perspective, & keep in mind the realities of what are studio scene is & isn't useful for.

Link | Posted on Jun 26, 2017 at 20:38 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

FilmORbitz: I understand the need to manually-focus all lenses/all cameras for the studio shot pixel peepers. But it seems that a big chunk of information is being thrown out by this procedure, namely, the camera/lens combo's autofocusing variability, which you are now revealing to be a non-trivial issue. I for one would like to know whether Camera A is more or less variable in focus acquisition than Camera B, and by how much. Speaks to both the mechanical goodness of the moving parts, as well as the magic behind the curtain. That consumers hope this variability is vanishingly small on a $4500 camera goes without saying, but currently we have no way of knowing.

That's what our AF tests are for. It's important, in testing, to decouple variables.

Link | Posted on Jun 26, 2017 at 20:27 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

Random Photographer: Forgive my ignorance, but how do we know other cameras aren't suffering from the same flawed test? The A9 was expected to be sharp, thus the rebuttal by the community and the subsequent retest. But what about other Cameras? I for one think that DPR should retest all bodies that are still in production, to get a level and reliable ground for comparison.

Instead, we uncovered (working with Bill and Jim) the source of Diglloyd's initial claims of banding in the red channel, attributable to the PDAF masked rows, also uncovering it can cause striping in flare-lit shots. You chose to ignore that finding of ours we reported on on the IQ page of the a9 review.

You could always nitpick 'why didn't they cover this?' - you'll always be able to find something. That's an impossible game to win...

Also - no our studio scene is not flawed in terms of exposure. The belief that 'ISO 100 must be ISO 100' is flawed; in the real world, it's about the performance of a camera when given X amount of light. That's why we control our studio scene exposures.

Some tests, depending on how they're done, do require normalization to measured ISO; we get around this by giving every camera the same amount of light at the stated manufacturer ISO. So if one camera does better than another at the same ISO, that'll bear out in the real world.

Link | Posted on Jun 26, 2017 at 20:17 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

Random Photographer: Forgive my ignorance, but how do we know other cameras aren't suffering from the same flawed test? The A9 was expected to be sharp, thus the rebuttal by the community and the subsequent retest. But what about other Cameras? I for one think that DPR should retest all bodies that are still in production, to get a level and reliable ground for comparison.

As for the rest of your concerns re: standardized testing: I'm completely w/ you. We would love to have all those tests standardized as well. We're trying, but in the meantime, we're trying to provide you w/ meaningful differences based on our comprehensive knowledge/testing of all these cameras, all the time. We provide demonstrations to back up our claims. Yes, they're not entirely standardized tests yet, but that's only because no one (to my knowledge) has developed entirely trustable, standardized controlled AF subject tracking tests (even manufacturers have admitted to us to struggle with this).

The 'ad hoc' tests are demonstrations of what we've experienced having shot thousands of shots with the cameras, under multiple scenarios. It's the best we can do in the absence of current standardized tests.

As for Star Eater issue - we've talked about it on our site by referencing excellent work done by others. W/ limited resources, we've chosen not to repeat already well-done studies.

Link | Posted on Jun 26, 2017 at 20:13 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

Random Photographer: Forgive my ignorance, but how do we know other cameras aren't suffering from the same flawed test? The A9 was expected to be sharp, thus the rebuttal by the community and the subsequent retest. But what about other Cameras? I for one think that DPR should retest all bodies that are still in production, to get a level and reliable ground for comparison.

"you could fine tune a range of five shots and pick the sharpest easily."

There are at least 100 shots we have to shoot (we bracket exposure to pick the most neutral JPEG) for our studio scene and dynamic range tests. To get the level of precision required to discern the first set of shots published for the a9 vs. the second would require far more precise macro rail fine tuning than 5 turns of the macro rail control - more like at least 10, if not more.

That's at least 1000 shots. In other words, we'd have to shoot all the studio tests/exposures (100 shots) per macro rail turn. Then pick the sharpest.

This is what we *wish* to do, but don't have the bandwidth to do currently. Which is why we're looking into automation.

Furthermore, some camera/lens combinations don't necessarily require this - e.g. if they have enough magnification, it's usually possible to get focus with our accepted tolerances. Are they the sharpest they could ever be? Only computerized analysis could guarantee.

Link | Posted on Jun 26, 2017 at 20:07 UTC
On article Now we know: Sony a9 is sharper than we thought (398 comments in total)
In reply to:

EDWARD ARTISTE: On thing though- for the test scene, how about NOT shooting canon bodies with the dinosaur EF 85 1.8. If its hard to tell true sharpness, this particular lens has great focus but a mushy purple haze that pulls the image quality down. Its almost useless to compare the results to other cams, because that lens is just not producing the best pic possible at that mm range. Why not use the 85 1.2?

Glad this was able to be worked out. Calibrating lenses per body is the epitome of painfully, wasted time (in comparison to shooting :) )

How is tech so advanced, yet nikon only now (and not canon) have a auto calibration tool on the cameras. In this regard, cams are stuck in the virtual stones ages.

Because most tests (including our own) show the 85L to be softer than the 85/1.8, with significantly more chromatic aberration.

Link | Posted on Jun 26, 2017 at 20:00 UTC
Total: 4973, showing: 1 – 20
« First‹ Previous12345Next ›Last »