Has a website at www.camnostic.com
Joined on Aug 21, 2013


Total: 33, showing: 1 – 20
« First‹ Previous12Next ›Last »

I could have used the customizing focus limiter range last night shooting nighthawks. That sounds actually useful, if rarely so.

Link | Posted on Aug 26, 2021 at 19:56 UTC as 3rd comment
In reply to:

LJ - Eljot: For me, the studio scene comparison tool does not work properly. I gives the same size for comp and print. That is the case on different computers with different OS and different browsers. I have no idea where that could come from.

To the topic. The A7s beats the A7R at ISO 409200, but only because the A7R tops out to stops earlier. But pushing it gives pretty much the same result.

That is interesting, because more than 10 years ago the amount of read noise per pixel was much higher and lead to different results sometimes.

I find that the studio scene comparison tool is a bit frustrating (similar to the issues with the-digital-picture version of it) where I wind up with different proportions of the scene due - I think - to have the images were captured and stored. It may be that I'm just not grokking how to do it correctly, but it may be in the future that DPReview has a clever programmer script the creation of a parallel database of pictures based on object recognition/sizing.

This is not to criticize the tool. It's a great value to the community.

Link | Posted on Aug 16, 2021 at 14:11 UTC

Finally, Panasonic's DFD (Depth from Defocus) system is vindicated. Perhaps Lumix cameras can now blame "temperature wobble" for their video quality :)

Link | Posted on May 28, 2021 at 16:51 UTC as 49th comment
On article Why are modern 50mm lenses so damned complicated? (923 comments in total)

I especially appreciate the history of the canonical designs. Put much in perspective.
It brings up an interesting trend for today, though: with the complexity of correcting elements, etc., I'm not sure we're going to be able to establish canonical designs going forward. When lenses had 2-5 elements, it was easy to call out categories of designs, but when it's a soup can full of 20 elements, few lenses will resemble others significantly. Ray tracing software, along with coatings and new materials may have allowed designers a level of creativity that will confound Rogers of the future. But thank you, Paleo Roger for this great overview of the present.

Link | Posted on May 9, 2021 at 15:58 UTC as 183rd comment
In reply to:

CanonKen: No issue reusing the 2 year old optics, but lack of control ring seems lazy even if there was some tooling, modification cost and they raised the prices more to compensate.

In fact, the most professional setup will possibly be the EF 600mm f/4 III along with the control ring adapter, giving maximum functionality; or the EF 600mm f/4 II with the control ring adapter if the overriding concern is image quality.

This would change if there was a stabilization benefit, AF benefit, or other nice thing coming from the RF mount contacts. In the absence of that, this is disappointing.

Perhaps what Canon is really saying here is that an actual big white for mirrorless is enough years out that there is room for a cemented-adapter version in the interim.

I was going to upgrade my 600 II, but will hold off pre-order until there's evidence of it being outperformed in something other than weight. (Though my arms would appreciate the weight difference too, they can't convince my wallet to agree.)

Link | Posted on Apr 16, 2021 at 14:47 UTC

I find myself thinking that the "class" that the A1 falls into includes the Canon R5 at least as much as the 1DX3. The R5 is a whole lot more similar to the A1 than the 1DX3, being mirrorless and having a higher resolution sensor. Wildlife and sports photographers are flocking to it as much as the 1 series DSLR.

Because of this fuzziness in class, I wouldn't lead with the message that the A1 has class-leading DR. My first thought in reading that was that it beat the R5, which apparently would be incorrect. I see that Bill Claff's analysis over on photons2photos.net indicates that the R5 edges it...

Maybe Rishi's tests were different, though. My sense from the article is that the R5 was left out, which doesn't make much sense to me if the A9 variants are included. I've owned all of them, aside from the A1, so personally I naturally consider them of a class.

Link | Posted on Mar 15, 2021 at 01:22 UTC as 11th comment | 4 replies

A missed use: using these to keep batteries going overnight with outdoor remote camera setups. Getting a thermostat-controlled, battery-powered device with low enough voltage to last a night is actually really hard. Noted here:

Link | Posted on Feb 10, 2021 at 16:40 UTC as 15th comment

Actually not their first lens. Others are cine-oriented. See here:

Link | Posted on Jan 8, 2021 at 18:45 UTC as 20th comment

Gannon, I know that writers often don't write the headlines, but the "... Stagnant - For Now" bit really caught my attention. It was a bit of suspense. Did DPReview thing Nikon was about to fall off a cliff, or did they think that Nikon had a new camera to announce that would boost sales? Would they release more pro-level lenses?

Upon reading the article, of course, I find that there is hope held out that the change in share might be upward, and that it would be due to the Mark II versions of the Z6 and Z7. That was not what I expected. But thanks for the roller coaster.

Link | Posted on Oct 26, 2020 at 20:46 UTC as 100th comment | 1 reply

One of the reasons that the industry's "Uncle Roger" is so valuable is that he'll surgically answer good questions that pop up in the comments. I worry for him now, coming over to DPR, as the both the quantity of forum dwellers and ratio of pedants is so very high. I hope Roger manages to get some good interactions out despite the firehose of input.

That said, on his smaller blog, he did so much good for the industry. Perhaps the bigger platform will do even more. Rooting for him!

Link | Posted on Oct 19, 2020 at 15:31 UTC as 51st comment
In reply to:

Horshack: Whether or not the firmware update will translate into faster buffer clearing times (and commensurately, deeper buffers) will depend on where the bottleneck is in those cameras - Expeed, XQD/CFE, and/or internal memory bus. It'll also depend on whether there are any issues in Nikon's specific implementation of CFE, either firmware-related or in the PCIe lane configuration they chose.

I did a deep-dive of XQD vs CFE performance on the Z-series and found the bottleneck to be in the single Expeed chip on those cameras. I posted my findings here:


One more point: for those who think the CFexpress card concept is a "hole," as one put it, I can say that I'm very happy to have them in the R5. Yesterday I was shooting my kids throwing a ball for my dog - one of those times you need the 20 FPS to get the right crazy brittany expression. We were goofing around and managed to take 1600 images in about 20 minutes, and I never once waited due to a full buffer. The offload was amazing. This was with pictures that averaged 56 MB from a 45 megapixel sensor.

Of course, that sounds like a nightmare to some, but there are some applications where it's useful, and us wildlife (I guess the brittany qualifies) photographers will use it if we got it.

I was shooting one camera with a Prograde Cobalt 325 GB and the other with the Sony Tough 512 GB. A fast SD card would have worked too, but I'd have missed a few of the shots. The 4K HQ video I took wouldn't have worked on the SD cards, but I could have used normal 4K, and it'd have been fine.

Link | Posted on Oct 12, 2020 at 15:22 UTC
In reply to:

Horshack: Whether or not the firmware update will translate into faster buffer clearing times (and commensurately, deeper buffers) will depend on where the bottleneck is in those cameras - Expeed, XQD/CFE, and/or internal memory bus. It'll also depend on whether there are any issues in Nikon's specific implementation of CFE, either firmware-related or in the PCIe lane configuration they chose.

I did a deep-dive of XQD vs CFE performance on the Z-series and found the bottleneck to be in the single Expeed chip on those cameras. I posted my findings here:


The review over on camnostic.com of the major CFexpress cards...
... shows that Horshack could be right. It was just updated on Sunday. It'll be hard to know before they release the firmware upgrade. Camnostic will be renting all those bodies to run the same cards through and give comparisons.

They just came out with a review of CFexpress card readers last night too, which is where they got most of the data to update the cards review...

Link | Posted on Oct 12, 2020 at 15:14 UTC
On article Opinion: Camera names are getting ridiculous (706 comments in total)

Couple points to add, Richard...
1) Worse than commonly-used letters are letters used with every model that aren't useful in distinguishing them. For instance, why include "EOS" in front of every Canon camera? There is also a D in all of their names; yet they no longer make non-digital cameras.

2) Lens names are the real problem. The third party manufacturers are the worst. Sigma reformed this a bit with their Art series, striving for simplicity.

3) The "mark" used in the names sounds very British, and I have a suspicion that DPReview had something to do with its popularization. Sort of has Barney's fingerprints on it. Even though it likely predates him, it would be fun to blame him.

Link | Posted on Oct 11, 2020 at 15:28 UTC as 267th comment | 3 replies
In reply to:

DWilkinson: I really wish Sony overhaul its color science completely. Great cameras and specs but I had to sell my a7iii as the color was the worst of any system I ever had. Almost unfixable in many instances.

I once thought that the Sony colors were pretty bad when relying on auto white balance. And, in fact, they used to be worse than Canon's AWB. After owning two A7r4s and the couple versions of the A9, I've come to a more subtle conclusion: the AWB feature takes a couple seconds to get it right. I find that when firing a burst, the first few shots can be quite off, but after a few exposures, it generally is as good as the Canon.

And, yes, AWB isn't some amateur feature, as it's pretty much required for some forms of pro shooting. I find that sports/action/wildlife requires it, as panning can radically change mixed lighting sources.

Sometime around the release of the 70D, Canon really nailed this. Prior to that, AWB was a source of problems sometimes. I think Sony started to nail this with the release of the A9 (although I didn't own the A7r III, so perhaps it was slightly earlier). But Sony's still isn't as accurately available for the first few frames.

Link | Posted on Jun 29, 2020 at 20:33 UTC

The newer model is better. Presumably some cameras will actually be persistently able to connect to the web and upload, rather than requiring a fiddly manual connection process. If this is so (or becomes so), then it's a great way to not worry about loading pictures. They go to cloud automatically, and then they download to computer automatically. Lightroom, meanwhile, monitors and ingests from the folder automatically. Well conceived.

To do this right, however, they need the following:
- A camera connection process that is done automatically, without even needing to initiate it.
- A status indicator noting that an image up on the cloud has indeed been loaded down to a computer.
- The ability to set images NOT to delete unless they have this confirmation of download.
- Yes, some sort of assurance that the service won't suddenly be withdrawn 18 months from now.

Do the above, and this would actually be a reason to rely more on my Canon body than my Sony body.

Link | Posted on Feb 13, 2020 at 22:14 UTC as 27th comment | 1 reply
On article Canon RF 85mm F1.2L USM review (768 comments in total)
In reply to:

Critical Thinker: GM killer?

I hadn't really thought of it this way before, but my behavior of late is to sell my Sony lens when a new Sigma native Sony equivalent comes out. Just did that with the 24-70 f/2.8 and the 35 f/1.2, and am anticipating doing it with the 70-200 f/2.8 (supposedly any day now) and the 85mm f/1.2 (only vague rumors to date). I always wind up coming out a few hundred bucks ahead selling the GM used and buying the Sigma new, and getting a noticeably better lens.

If Canon's RF halo lenses are as good or better, then few will miss the GM brand over on the Sony side. Some of us would, though, miss the Sigma Art line, as Sigma hasn't figured out the RF mount yet.

Link | Posted on Jan 21, 2020 at 18:49 UTC
On article Canon RF 85mm F1.2L USM review (768 comments in total)
In reply to:

fPrime: Compared to the EF version, I feel there’s something gained but also something lost in the images produced by the R version. It looks like it may draw less romantically than its predecessors IMHO.

The "rendering" argument has been used variously with surpassed lenses over the years. The problem with it is that it means different things to different people. Some notice color temperature differences. Some (many) prefer the way bokeh presents itself. Some don't like a very steep sharpness gradient, which - with modern lenses - can effectively thin the commonly calculated depth of the plane of focus. Many just simply are accustomed to a combination of features and flaws with an old lens.

I think it's worth defining precisely what it is within that constellation of rendering issues (and/or others) that pleases or displeases.

I say this not to criticize the use of the concept, but rather ease my frustration that I never really know what people are meaning, as they're not consistent between themselves. Even if it were expressed in a fashion that let me know how the look made you feel, it would become useful. I'm genuinely interested in these assessments.

Link | Posted on Jan 21, 2020 at 18:35 UTC
On article Canon RF 85mm F1.2L USM review (768 comments in total)
In reply to:

t.c. marino: long time nikon user here,have to give canon very high praise for coming out the gate with some IMPRESSIVE ultra sharp exceptional fast 1.2,1.4 & 2.0 glass.Sony is at least 6 years plus ahead of canon/nikon..and i dont think they have any ultra fast 1.2 glass..do they?Canon in less than 2 years has set the standard for ultra fast 1.2 glass..what about nikon? eventually they will get there..they are wayyy too slow..nikon has not released a fast 1.4 Z lens yet...never mind the 1.2's,anxiously waiting for canon to drop the bomb with a more capable body than the eos R.I will keep shooting my Nikon dslrs..but seriously thinking going with canon for
mirroless..for the lenses..nikon are you listening?

Sigma's 35mm f/1.2 (available on Sony e-mount and the L-mount Alliance system) is in the same league, it appears, as this 85 1.2.

I don't think Sony will be the provider of the crazy-class of lenses for the e-mount. I say that as the owner of several of their GM lenses.

Would buy a Canon RF body in a heartbeat if they had one that shot 7 frames per second or so with continuous autofocus and generally upmarket features such as those equal or better than their 2016 5D4 release. Their stable of orphaned pro lenses makes this potential prospect attractive, especially this 85mm.

Link | Posted on Jan 21, 2020 at 18:21 UTC

Installed update ***and went to prefs to use the "custom" option to make the Develop module stuff accelerate as well***. This is on a recent, tricked out 1Mac 5k with a Radeon Pro 580.

After doing so, I found the most obvious improvement was using local adjustments with the adjustment area coloration turned on. That used to be uselessly slow. Now it's as fast as I can move my mouse. With normal photographs (<=50mp), there is no noticeable slow-down.

With a 350 megapixel panorama, it was balky, but actually useable. Previously, that photograph had to be edited in Photoshop lest the system just stop.

In all, an impressive speed boost in one of the places where it matters.

Link | Posted on Aug 15, 2019 at 02:48 UTC as 16th comment
In reply to:

STEPH RHODES: I wonder if this another D800/A7R situation where they share a similar sensor.
There is a rumored Z8 with 60+MP high resolution sensor coming too.
Or is it an A7RII/D850 situation where Sony releases one sensor and Nikon improves it a little and releases it later.

Sony needed something. Between the major FF manufacturers, they had the lowest resolution "hi res" sensor. And that S1R "hi res" mode is ridiculously amazing.

I do think the hi res mode from Panasonic is industry-shifting. That it hasn't shifted the industry is just a product of the fact that it came out on a new platform that doesn't yet have lenses. When the tech spreads to the other systems, it'll be a big deal.

Specifically, it'll allow all-rounder cameras once again, with landscapers getting their super high res fix, yet having a sensor that can do higher FPS. Provided the manufacturer can do 10fps+ with 40-60 megapixels, which is true now of Sony and hopefully will be soon of the others.

Link | Posted on Jul 16, 2019 at 13:28 UTC
Total: 33, showing: 1 – 20
« First‹ Previous12Next ›Last »