DPreview First impression review 5D IV

This wasn't supposed to be a topic about 5D Mark IV? All I read about in the last 14-15 pages is about some guy from DPReview trying to explain how the focusing system from 1 Dx Mark II works. And by the way, playing for a couple of days with cameras and then writing a few words and publish some photos taken with those cameras...call it anything you like, but not a proper/real review. I mean, who's buying cameras/lenses based on DPReview reviews? I know I don't. much prefer to read reviews on forums and then go and rent cameras in order to do my own testing, rather than relying on some bad jokes, pompously called by today standards reviews.

That being said, I hardly wait to put my hands on 5D Mark IV. For the moment I'm still happy with my 6D, so I have enough time to play with 5D Mark IV before I make the financial effort to buy it.
 
Upvote 0
StudentOfLight said:
Don Haines said:
dilbert said:
Sharlin said:
...
What I'd like to see are tests where there is less subject separation and more confusing foreground/background elements than in birding, as it seems to be specifically those sorts of situations the DPR claims Canon's tracking has trouble. Also, using the full area mode in such confusing situations.
...

Like when you have to shoot "through" wire fencing?
been there, done that..... The AF system said BIRD!!!! no, FENCE!!!! no, BIRD!!!! no, FENCE!!!! and kept jumping back and forth...... so rather than spend a week re-reading the AF manual, I put it into manual...
the RTFM approach - Resorting To Focusing Manually ;D

Often the focus limit range switch will stop that behaviour, obviously depending on the situation.
 
Upvote 0
ritholtz said:
Canon Rudy is everywhere on net explaining 5d4 duel pixel raw.Looks like he knows how Canon iTR focusing suppose to work. Not sure why DPR ignores his suggestions and material.

I think what comes into play when heels get dug in, is ego. Some folk are less receptive than others, which seems strange considering that we can always learn something new if we're willing to be receptive, but that's human nature. I know from personal experience! ;)
Wonder if Rudy would ever consider directly addressing any of DPR's criticisms?

Jack
 
Upvote 0
AF tracking is has so many variables. How is your subject moving and how fast? What sort of lighting conditions? How much do they change and how drastically while shooting? What lens are you using? Some can lock focus far faster than others.

I was watching something a few days ago where the AF was getting tested with the 85L 1.2, a notoriously SLOW (by comparison) lens to AF lock. Probably not the best lens to try and test a camera's AF potential with.

I tend to take AF opinions with a grain of salt without knowing what the variables were, and trust me I've shot in some absolutely insane focusing conditions with great results from a 1DX and 1DX2 but I was using a 70-200 IS II and a 200-400 F4 L. Those things are lightning fast. Had I been using an older lens or that 85L, my results would not have been as good.
 
Upvote 0
DPR has just looked at the focus tweaking. It was a very preliminary analysis yet the wrote a damning conclusion. I must admit that they do seem to be up themselves. It's the pompous, carping tone that grates.
 
Upvote 0
AlanF said:
DPR has just looked at the focus tweaking. It was a very preliminary analysis yet the wrote a damning conclusion. I must admit that they do seem to be up themselves. It's the pompous, carping tone that grates.
Looks like Canon already documented limited use of this functionality. Still DPR is trying to sound it like Canon failed solution to MFA. At least that is how readers are interpreting based on the way it is presented. Canon should have named it to software Image Nano adjust.

https://www.youtube.com/watch?v=HcQ9MSRRvn4
 
Upvote 0
rishi_sanyal said:
Jack Douglas said:
Why harp at poor Rishi when there is stuff like this out there. This guy is hilarious.

https://www.youtube.com/watch?v=Z_FtiZB9DYQ

Jack

OK, seriously, we've had our differences but drawing a comparison against that guy is just, plain unnecessary. :)

I agree. Still he gets views and I guess that pays his bills. :)

Jack
 
Upvote 0
The effect supposedly gets better with longer focal length, distance to subject and aperture, and they tested with an 35mm/1.4 at only 87.5cm distance, and an 85/1.8 at who-knows-how-up-close since the subjects head pretty much fills the frame. Hardly realistic shooting scenarios for me. I'd assume I'll get more out of it with my 85mm and 135mm at real working distances. I'm not expecting miracles, but testing two focal lengths at rather close distances and talking about the technology like "miss, try again in future" is hardly worthy of an article. It's a disgrace for dpreview to draw conclusions from these two samples and present them as facts.

This is a half-assed Test at best, made worse by throwing in the microfocus adjustment tidbits as filler, which have absolutely nothing to do with the DPRAW adjustments.

Also
Overall, traditional 'AF (lens) microadjustment' is a much more powerful tool for achieving pinpoint sharpness and ensuring any particular lens is properly calibrated to your body.
is _completely_ missing the point of the tech. I can AFMA all day, if my subject and/or I move during shooting I will still get slightly out of focus pictures, which I then may or may not be able to correct. Noone in their right mind would use DPRAW to adjust focus on all their images because they were to lazy to AFMA. Are they for real?

Again, I'm not saying the functionality is the best thing since sliced bread and will solve all focus problems in post, in fact I don't know if my RAW converter of choice (C1) will ever support it and I don't care. But I can see the flaws in this "test" from miles away.
How come they haphazardly get to play with a 5D Mark IV while I'm eagerly awaiting shipments and could actually put it to some good use?
 
Upvote 0
Jack Douglas said:
Why harp at poor Rishi when there is stuff like this out there. This guy is hilarious.

https://www.youtube.com/watch?v=Z_FtiZB9DYQ

Jack

When I first started I used to be a sheep and would believe all the bs that reviewers like this would spew forth. I've found though that it's best to use the gear yourself and find out what you need in a camera. I don't need a high fps camera for 99% of my shooting so someone claiming that a camera is a turd because it doesn't have cfast doesn't translate for me. I love landscape photos and so the dr issues should be a bigger deal to me but they aren't. People turn the most minuscule issues into deal breakers. Even if a certain technology has never been available in the past people act like great photos can't be taken without said technology.

This will be a great camera. I'll probably buy one when the price goes down a bit.
 
Upvote 0
ritholtz said:
Canon Rudy is everywhere on net explaining 5d4 duel pixel raw.Looks like he knows how Canon iTR focusing suppose to work. Not sure why DPR ignores his suggestions and material.

We're not ignoring Rudy's material. We've enjoyed and learned from his technical documents and expertise many times over the years.

The point I've tried to make in numerous posts at this point is that Rudy's description of iTR does not disagree with ours. Neuro is trying to over-interpret Rudy's words to fit his preconceived notion of how auto selection with iTR works. I'm saying that if you simply do the experiment - focus on a face with some other object in the foreground - it's quite clear that Auto does use the metering sensor for initial point selection using scene/subject analysis.

Nothing Canon's manual, nor Rudy Winston, says contradicts that.

Furthermore, how iTR works is irrelevant to this entire discussion about our motocross piece - whose simple crime (apparently) was to point out, amongst many other points, that Auto area appeared to struggle on the 1D X II in that situation, which we actually spun around positively by framing it as 'but switching to manual point selection for Auto area mode instantly helped'. It was an opportunity to educate our readers re: how to get the best result out of that system. Just as it was an opportunity to educate our readers that, for whatever reason, Auto area selection appeared to be more successful at picking a subject of interest on the D5, in that situation. After all, the piece was a shootout of the D5 vs the 1D X II.

Yet we're blamed and discredited for comparing the two?

We pointed out our findings, and yet here at CanonRumors in the comments about the article, one might conclude our entire piece was only about Auto area AF - a mode we rarely use as anything but a fall-back when all else fails in the heat of the action.

We're not the ones that made a big deal about Auto area AF performance between the cameras - in fact, it's not even mentioned in the conclusion of that article.

Fair criticism is fair criticism - in fact, I took suggestions from some CR members on board in the 1D X II vs. D5 slideshow, whilst also pointing out the unfair criticism, if you remember, here:

http://www.canonrumors.com/forum/index.php?topic=30411.msg611230#msg611230

But unfair and unjust criticism, the kind this particular thread is rife with, needs to be called out.
 
Upvote 0
Loibisch said:
The effect supposedly gets better with longer focal length, distance to subject and aperture, and they tested with an 35mm/1.4 at only 87.5cm distance, and an 85/1.8 at who-knows-how-up-close since the subjects head pretty much fills the frame. Hardly realistic shooting scenarios for me.

This is a half-assed Test at best, made worse by throwing in the microfocus adjustment tidbits as filler, which have absolutely nothing to do with the DPRAW adjustments.

Basically, the impact of image microadjustment varies inversely with DoF, so wider apertures and higher magnification (longer FL and/or closer subject) will increase the amount of adjustment possible. At its best, it's still not going to be a significant change.

When the 5DIV press release was leaked a few hours prior to the official announcement, CR published it as usual – without the footnotes. Someone asked what the *** footnote for DP-RAW stated, and I summarized the actual footnote (which I had read from the source of the leak) as follows:

neuroanatomist said:
***Dual Pixel RAW may not be all it's cracked up to be. Caveat. Another caveat. Maybe one more caveat for good measure.

I suspect some assumed I was being facetious, but I wasn't. For reference, here are the three caveats:

[list type=decimal]
[*]Sufficient adjustment volume and compensation effect may not be achieved, depending on lens in use and shooting conditions.
[*]Adjustment volume and compensation effect vary depending on camera position (landscape or portrait).
[*]Sufficient adjustment volume and compensation effect may not be achieved depending on the shooting conditions.
[/list]

Since DPReview published the press release, I assume they bothered to read it (but maybe I have fallen victim of the classic ASSumption trap). How anyone could read that and come away with the idea that anything greater than minuscule effects in select situations was possible is beyond me, unless their reading comprehension skills were less than even rudimentary. But even if they missed that, Rudy Winston (Canon tech rep / mouthpiece) published an article on Canon DLC five days ago making the limitations abundantly clear, using words like 'tiny', 'subtle' and 'just a tad'. Anyone with the tiniest, most subtle, or even just a tad of journalistic skill and integrity would have found and read that article prior to reviewing the feature.

But instead, DPR posts a rush-job 'test' of the feature – can it be used for major or even minor focus shifts? "No," they conclude...which is exactly what Canon already told us. They suggest that higher levels of sharpening could enhance the microadjust effect, but don't bother mentioning that the microadjustment itself increases image noise,which would be further accentuated by increased sharpening (Rishi, you really should incorporate that bit into the article, since you always take such pains to highlight how noisy Canon's sensors are, it seems you missed a bet to have another go at that old saw, by highlighting how this 'could-have-been-great' feature is not only barely effective, but makes Canon's horrible noise problems even worse if you try to use it).


Loibisch said:
Overall, traditional 'AF (lens) microadjustment' is a much more powerful tool for achieving pinpoint sharpness and ensuring any particular lens is properly calibrated to your body.
is _completely_ missing the point of the tech. I can AFMA all day, if my subject and/or I move during shooting I will still get slightly out of focus pictures, which I then may or may not be able to correct. Noone in their right mind would use DPRAW to adjust focus on all their images because they were to lazy to AFMA. Are they for real?

DPR missing the point? Wow, there's a shocker. ::)
 
Upvote 0
benperrin said:
Jack Douglas said:
Why harp at poor Rishi when there is stuff like this out there. This guy is hilarious.

https://www.youtube.com/watch?v=Z_FtiZB9DYQ

Jack

When I first started I used to be a sheep and would believe all the bs that reviewers like this would spew forth. I've found though that it's best to use the gear yourself and find out what you need in a camera. I don't need a high fps camera for 99% of my shooting so someone claiming that a camera is a turd because it doesn't have cfast doesn't translate for me. I love landscape photos and so the dr issues should be a bigger deal to me but they aren't. People turn the most minuscule issues into deal breakers. Even if a certain technology has never been available in the past people act like great photos can't be taken without said technology.

This will be a great camera. I'll probably buy one when the price goes down a bit.

Re: the 'bs that reviewers like this spew forth' and turning 'the most miniscule issues into deal breakers', allow me to quote the conclusion of the motocross piece being referred to throughout this thread as evidence of our incompetence, bias, overexertion on features, etc.:

"Obviously, both the Canon EOS-1D X Mark II and Nikon D5 are in a league of their own when it comes to professional sports cameras...

In the end, we saw firsthand the advantages of a ridiculous burst rate and practically bottomless buffer while shooting objects that move very, very fast. We saw that the 1D X Mark II shoots at a noticeably higher burst rate than the D5, and that it is able to track subjects noticeably better than other Canon cameras, including the EOS 7D Mark II and original 1D X. However, the Canon still can't match the Nikon's uncanny ability to track objects reliably and accurately as they progress across the frame while also coming toward or moving away from the camera. Regardless, both are impressively capable photographic machines worthy of professional sports photographers."

This is what we actually write (is it unreasonable?). Then we get labeled as being biased/incompetent by people who dislike one aspect of the entire piece where their camera didn't win. Then that's the conversation spread on these forums, which then turns CR readers off from potentially useful info they could've benefitted from. Like unfocused, who wrote: "I am not thrilled with any of the Canon autofocus tracking for sports photography. I'd love to be able to focus on a single player and then have the autofocus continue to track that player as they move through the scene. Maybe there is a way to do that, but I sure can't find it in any of their manuals or online resources." He would've benefited from reading our piece, but was probably told not to, because DPR is a joke, after all, right? Another with the same sort of feeling despite knowing of the manual point selection for Auto area AF might, on the other hand, benefit from knowing that there are cameras that are more reliable in this regard (1D X II over 5D IV, for example, or, yes heaven forbid, D5 over 1D X II).

Yet readers are turned off from this very information because of erroneous claims of bias and incompetence. And that is sad. I know of people who, for example, really did care about DR, yet never knew about the superiority of some cameras in this regard, because was fed erroneous information right here on CR that (1) it doesn't matter, (2) DXO is paid off, and (3) that DPP solves all your DR problems. So he was led to believe there actually weren't any differences, which is patently false. That's the kind of misinformation that was spread for years (probably still is), and it's particularly worrying when it comes from someone that otherwise appears knowledgeable and perfectly capable of logical reasoning. Because that sort of person earns trust, and yet then must be willfully misleading people by disallowing any negative conversations about his favorite brand by balking at, and continually trying to discredit, any source of any negative discourse. I saw the crusades against jrista, with dilbert, etc. - all of whom are, I'm sure, trolls in everyone's book here, right? Is that the truth though? It's a similar story now for AF subject tracking.

What ends up happening is that posts and experiences like Sharlin, who said: "At least with my lowly 80D - and I fully admit I'm a newbie at using complex AF systems - with full 45pt mode the AF just often gets distracted and locks on contrasty background elements when it loses the actual subject for a moment. OTOH with enough subject separation it seems to work pretty well. The Live View Face&Tracking is definitely superior, though, as seems to be the case with the 5D4 as well."

... get buried amidst all the erroneous claims of bias and incompetence. He sees exactly the behavior that we refer to and have been digging into the causes (relying too heavily on phase information vs. metering sensor for subject tracking, and the ramifications on which shooting scenarios this works vs. doesn't work well for). He sees this either because his shooting scenarios highlight these shortcomings, or because he doesn't have blind bias, or some combination thereof. But all he hears here is how he shouldn't listen to us (whether it's us telling readers how to more effectively use their system, or comparing it to other cameras/systems), and that's a missed opportunity for education.

And if you were to just step back and ask: 'who is more likely to be biased: the user of only 1 system, or the multiple users of multiple systems who are required to fundamentally understand principles of AF and then use all camera systems'? As I've said before - we're certainly biased, but toward photography-accelerating technologies, not brands. It's our job to dissect which technologies are better at what use-cases/scenarios, so you can judge how suitable a particular camera/lens/system is to you for your needs. At some point we have to offer our opinions, sure, and we do when it comes to how good/bad performance along any one axis applies to certain use-cases. But we're transparent about why we think so, so you can decide whether to listen to or ignore our findings.

But what perplexes me is the anger with which data/opinion we present is vehemently attacked and ascribed to all sorts of malicious intent and incompetence in an effort to discredit. When further 'discussion' uncovers we actually knew what we were talking about all along, the other side can't publicly admit that, because then they'd look stupid, so they double down. And that's the vicious cycle (I, on the other hand, have very publicly admitted when there was an oversight: http://www.canonrumors.com/forum/index.php?topic=30411.msg611230#msg611230. And, of course when there wasn't).

On Pentax Forums I'm an anti-Pentax Canon fanboy, after the Pentax K-1 review. On Sony forums after the a7 II review, I was labeled a 'hack' scientist for daring to suggest the a7/II's low light Raw noise performance fell to the level of the best APS-C sensors, despite all our controlled lighting/SNR tests proving it. And after the Nikon VR analysis, I was labeled an anti-Nikon Canon fanboy for daring to suggest there were vibration issues with Nikon because, if there were, how could knowledge of it not be more widespread (nevermind a lot of people felt safe 'coming out of the woods' and admitting they'd seen the same issues after we published the article - a reminder that tones of discussions in forums can be stifling to knowledge).

What neuro attributes to anti-Canon bias when I joined is actually us just revamping and modernizing our testing. We didn't actually analyze Raw files before (but when we did, it brought to light just as many issues with Sony files as it did Canon). AF analysis was often restricted to one paragraph; now, we study multiple aspects of it across multiple systems every day - because it is probably the most complex thing about a camera you can test (so I fully appreciate the concerns re: 'how can you test AF objectively?' -- we think about this every day). And not all our tests show Canon to be bad - that would be ludicrous. The 1D X II scored an 89% Gold, Canons nail Z-axis refocusing for steadily approaching subjects almost all the time in our tests, Canon's DPAF *does* nail subject tracking, we have the highest regard and only positive things to say about all of Canon's newest lenses, as well as Canon colors, like the following quote from my 5DS review:

"JPEGs exhibit the pleasing Canon colors we've come to expect, particularly when it comes to skintones. I spend countless hours fiddling with Nikon and Sony Raw colors to get the skintones I desire; compare that to the mere minutes I spent on the model shots on this page."

The words of an anti-Canon Nikon/Sony fanboy? Or just the words of someone who tries very hard to be realistic about positives and shortcomings of every system, just trying to share our observations and test results?

What's particularly funny re: the meat of the discussion here in this thread is that: for all the claims of our emphasis on Auto area selection (which as I understand none of you care about, and neither do we except as a fall-back method), the reality is there are only a couple sentences about the actual auto selection not being very successful (or as successful as the D5) in the motocross shootout scenario, with the real information being that Canon users should probably switch to manual point selection in Auto area.

It wasn't even talked about in the conclusion of that piece. So I'd ask: who is it that's making a big deal of Auto area AF? Us? Or the critics?

For your reference, the piece being referred to: https://www.dpreview.com/news/6990762465/motor-drive-and-motocross-with-the-nikon-d5-and-canon-1d-x-ii
 
Upvote 0