Canon warns against use of SanDisk CFast cards with EOS-1D X Mark II

It's been said the issue is with images less than 16MB, so I think rather than the card not being fast enough (which would affect larger images too) the issue is either:
(1) Images that are less than the card's erase block size (which I'm guessing is 16MB, although that seems large) aren't written immediately as it wants some more data to arrive so it only does one erase/write cycle rather than writing this data then erasing/writing again (plus table updating, as every write will go to a different part of the flash) when a little more turns up.
(2) The card's write buffer is 16MB and the data is sitting there waiting to be written to the flash.
Then in both cases a firmware bug means on power-off the data doesn't get written successfully. Of course there's also (3) which is a combination of the two.

Jun 30, 2016permalink

Oh and (4) The data is written but the table saying where it is doesn't get updated, as you wouldn't want to write that all that often. (In a mechanical disk Sector 12345 is always at the same point on the disk, with flash it will move about with every write so you need a table saying where it is now.)

Jun 30, 2016permalink

weixing

SanDisk notice for this: http://kb.sandisk.com/app/answers/detail/a_id/19576

Jun 30, 2016permalink

BobT3218

I have a similar problem but in my case it's with an Olympus E-M1 and a couple of Lexar SD HCI 32GB cards. The first image after switch-on is often but not always corrupted. I've had the camera's CPU replaced for other reasons but the problem persists. I format often and always in camera.

Jun 30, 2016permalink

Plus, Sandisk (as well as Samsung and other major brand name cards out there) has been proven in many tests and studies to be reliable, so again I'm sticking with this is a CANON software/hardware problem, not the card. I wouldn't be surprised if it happens with Lexar, Samsungs, and other CF cards (particularly the low-speed cards).

Jun 29, 2016permalink

Since Samsung aren't arguing (see link above) this is most unlikely. Plus no way are Canon going to annoy a bunch of Pros by misleading them, it would cost them a ton of business.

Jun 30, 2016permalink

just how many CFAST 2.0 cards are in use out there in cameras for you to stake that claim?

Jun 30, 2016permalink

Glad my camera doesn't use CF cards. And the one i had that did, I never had a problem. This sounds like a software/hardware glitch, and NOT the card or the brand.

Jun 29, 2016permalink

CFast, quite different to CF except in size (serial interface, not parallel).

Jun 30, 2016permalink

nicolaiecostel

I warned you, in the article that announced the purchase of SanDisk by WD, I said that they will interfere and mess things up, and you contradicted me.

Here is the proof !

Jun 29, 2016permalink

I think you should put away your tinfoil hat. The merger of WD and Sandisk happened on Oct 21, 2015.

Check this page here:
http://www.bhphotovideo.com/c/product/1087556-REG/sandisk_sdcfsp_064g_a46_64gb_extreme_pro_cfast.html

First product reviews show up on November 25, 2014, almost an entire year before the merger. Maybe the WD ninjas dropped into Sandisk headquarters to pre-ruin this product because the Nikon time wizards wanted to mess with the Canon 1Dx2?

Jun 29, 2016permalink

Tristan Jones

Just no. Based on facts. no.

Jun 29, 2016permalink

paul simon king

thou art wise beyond thy ears
:o

Jun 30, 2016*permalink

nicolaiecostel

Those time travellers, man. I tell you ..

Jun 30, 2016permalink

Miron09

I avoid SanDisk. Failed mechanically. Samsung cards much more robust.

Jun 29, 2016permalink

George Zip

Reading some of these comments, you would think a small country was just destroyed.

Jun 29, 2016permalink

Probably will be when they landfill all the dud cards...

(Just joking, surely a firmware fix for the cards plus a camera firmware fix to help with any dud cards left around.)

Jun 30, 2016permalink

Favorable Exponynt

good old reliable sd has never failed on me :p

Jun 29, 2016permalink

ozturert

At least Canon, unlike Nikon, acknowledges these issues as soon as they understand the problem.
Does anyone in Nikon really follow the world news?

Jun 29, 2016permalink

tvstaff

I'm not buying this. My Canon 1DX MKII bricks using the Lexar 128gb CF 1066x. Not CFast... CF!!!

To those who bought their cameras with the 64's deemed as defective.... Then a replacement should be sent to each owner. However, the fact I had issues with a Lexar
CF leads me to believe the "issue" is with firmware vs. the cards. But why should I believe what I experienced first hand ;) Honor? What ever happened to honor?

Jun 29, 2016permalink

garyknrd

At least they acknowledged it. Good to know.

Jun 29, 2016permalink

belle100

Pathetic!! It's one of the most expensive camera with one of the most expensive cards. Great! I'm lovin' it.

Jun 29, 2016permalink

JaimitoFrog

Haven't experience this after 3000 shots on mine. I wonder how frequent this occurs.

Jun 29, 2016permalink

Is the part number of your card the same as the affected ones? Also do you shoot raw only, as it seems to only affect files under 16MB (or so I understand)?

Jun 30, 2016permalink

Joe Ogiba

Canon should have used UHS-II SDXC slots. The 2000x UHS-II SDXC cards are very reliable and more than fast enough with much lower cost.
http://www.bhphotovideo.com/c/product/1191363-REG/lexar_lsd128crbna2000r_pro_2000x_uhs_2.html

Jun 28, 2016permalink

Minimum write speed guarantee of 30MB/sec. That is really slow compared to CFast (2.0). The Lexar 256GB CFast cards are writing in excess of 400MB/sec.

The 1Dx2 is way better with CFast, just some teething pains with any new(ish) standard.

Jun 28, 2016permalink

more than fast enough?

4K at 60 fps is something like 800mb/sec

Jun 29, 2016permalink

If 800Mb/sec is fast enough then there would be no need for XQD nor CFast. The 1Dx2's 16fps at ~30MB is 480MB/sec and the D5's 12fps at ~43MB is 500MB/sec. Faster cards help maintain the buffer so the cameras can keep shooting at top speed. It's not a common occurrence but I've hit the buffer limit on my 1Dx and 1D4 occasionally.

Jun 29, 2016permalink

Daniel L

Thanks goodness Joe is not on Canon technical advisory board.

Jun 29, 2016permalink

Martinka

SD cards can be easily mechanically damaged -> not suitable for heavy pro use.

Jun 30, 2016permalink

The 1DxmkII 40/60p video data rate is 800Mb/sec which is 100MB per sec and not challenging for fast CF cards, let alone CFast (the last Lexar CF card I bought was rated for 155MB/s writes: http://www.lexar.com/products/lexar-professional-1066x-compactflash-card ).

Jun 30, 2016permalink

the watcher

The funny thing is that Canon currently distributes exactly the same type CFast card as a bonus with purchase of new 1DX Mark II. I've got SDCFSP-064G-G46B. I wonder if I can claim a replacement of the card...

Jun 28, 2016permalink

I imaging a firmware update program for Win/Mac will be along quite soon...

Jun 30, 2016permalink

Marcel

Was SanDisk not for Sale ?

Jun 28, 2016permalink

Imagine

Sandisk was bought by Western Digital mid May.
They've paid $16B, so I don't think the CFast issues - even if known - "hurt" them.

Jun 28, 2016permalink

noflashplease

Uh huh. It's not Canon's problem, but Canon is correcting it with a firmware in early July? Unless SanDisk is writing code in Canon's firmware for them, I'd definitely say it's Canon's problem.

Jun 28, 2016permalink

It looks like the actual "bug" is with the SanDisk CFast cards. I conjecture that they experience a write cache-commit delay of the last block (of apparently 16MB) sent to the card. There is also probably an API call in the CFast standard that can force a cache commit. My guess is that Canon will now force, rather than assume, the commit during the power-down sequence.

Jun 28, 2016permalink

If that is the problem with these cards, then how come other camera brands do not suffer from this issue?

Jun 28, 2016*permalink

They most likely do, but Canon uses CFast, other's XQD. There is no much CFast out yet.

Jun 28, 2016permalink

Bernard Carns

@Fujica
Other brands do suffer.
I believe Sandisk CF cards have had problems with some Blackmagic cameras until there was a firmware update.

Jun 28, 2016*permalink

Kenneth B

@AprilW if Canon is shutting down without telling the card to flush its buffer, don't know why you would do that on flash in he first place, that is an error on Canon's part

Jun 28, 2016permalink

Andersonm

There's probably a very large number of these cards around. It seems like something that could be fixed very easily in firmware (e.g. store a dummy image and delete it before shutting off?), so they might as well do it. Doesn't imply camera error.

Jun 28, 2016permalink

FredVasilescu

It's very likely Canon is working with SanDisk to address this problem. So ýes,there might be some code from SanDisk in the new f/w release. 3

Jun 28, 2016permalink

@Kennethe B - It may be that SanDisk, for certain commands, is prematurely reporting "operation complete".

Might be a performance optimization gone wrong... Wouldn't be the first time a performance optimization by a flash manufacturer interacted negatively with another feature. For example, I have it from a very good off-the-record source that the infamous Samsung Superbrick problems which affected quite a lot of 2011/2012-era Samsung eMMC chips (such as those used in all Galaxy S2 models) were the result of a performance optimization interacting in an unexpected fashion with the eMMC/SD "secure erase" function. (In these cases, a secure erase command could, in certain cases, cause permanent unrecoverable wear leveller data corruption, effectively bricking the eMMC itself to the point where not even JTAG recovery houses could fix dead phones)

It may be that a certain command sequence that is 100% OK per the CFast specification triggers a SanDisk wear leveller bug.

Jun 29, 2016*permalink

A little more on the issue I described above - It affected maybe 0.1% or less of users. But scale it across thousands of users, and Bad Things happen. Samsung originally, rather than use the kernel patch I linked in another post, altered their recovery binary to not issue secure erase. So if someone ran "standard" AOSP 4.0 recovery binaries - 0.1% of them would permanently damage the eMMC because the eMMC claimed to be standards-compliant but wasn't.

Samsung went out of their way to suppress any information that the chips might be defective for over a year. (Google discovered the issue during Galaxy Nexus development, forced Samsung to fix the chip, but Samsung kept using buggy chips in their own phones. Samsung went out of their way to punish the Google employee who slipped up and mentioned to someone that the problem bricking hundreds of devices was a known issue.)

Jun 29, 2016*permalink

wear leveling happens behind the scenes.

it's difficult to say if the camera would have knowledge that sandisk is moving blocks around.

then it powers down quickly.

in theory it shouldn't be a problem.

or sandisk is reporting that a write has completed, when it's still in temporary volatile cache.

I'm also curious on where the heck are all these cameras that use CFAST 2.0 like some are blathering on about..

Jun 29, 2016*permalink

@rrccad I really wouldn't be surprised (since I raised it as a possibility) that the Sandisk is prematurely reporting a complete write in certain situations.

A good parallel to this is NFSv2 vs NFSv3 - NFSv2 requires the server to only report that a write is complete when it is 100% committed to disk. However some servers can be put into "dangerous mode" where they report completeness early (a violation of the spec but a MASSIVE performance improvement). NFSv3 allows for non-synced writes (with a way for the client to force a flush), but imagine an NFSv3 server that reported "operation complete" for certain syncs when it wasn't actually 100% done. May be a similar thing in these Sandisk cards.

Jun 29, 2016permalink

Why Canon why?
Second BIG problem with the 1Dx MKII and playing the blame card is really poor.

IF this problem is caused by Sandisk then why oh why Canon do other cameramanufacturers NOT suffer from this problem?

Its really poor to blame others while the problem described more sounds like a buffer problem with the 1Dx MKII.

Jun 28, 2016*permalink

SanDisk cards outside of spec.

Jun 28, 2016permalink

Still the problem is not in other camera brands - did they all knew about this problem beforehand? Or did Canon just not keep to the specifications?

Jun 28, 2016permalink

@Fujica. Please give me some examples of cameras with CFast, long enough on the market.

Jun 28, 2016permalink

Astrotripper

So, which other CFast equipped cameras are not affected by this? And how do you know they are not affected?

Jun 28, 2016permalink

Astro....
Pentax K1, K3 II, K 50 etc ... are NOT affected by this.
Ok?
:o)))

Jun 28, 2016permalink

Jan Itor

Odpisan
Are they film cameras?

Jun 28, 2016permalink

Imagine

If the issue is not reproducible with Lexar or other CFast 2.0 cards, it might be Sandisk's fault. Won't be the first time. Don't get why you are so fast and committed to dismiss Canon's claim on not being their fault.

Jun 28, 2016permalink

They don't have CFast.

Jun 28, 2016permalink

@Odpisan,

They are not even CFast spec cameras, and besides, Pentax cameras don't count anyhow. :-)

Jun 28, 2016permalink

SickVodka

@odpisan: LOL troll? you do know they dont even use CF cards right? so CFast is even further away..
You mean besides ARRI and some Sony as far i remember?
you do know that CFAST is NOT New.. ?

Jun 29, 2016permalink

@SickVodka & others

Hahaahaaahaahaaahaahaaahaahaaa
.. can you imagine that cameras which DO NOT USE some kind of cards, have problems with souch cards?

I can't!

But YOU can!

Do you have some problems with your imagination - the same problems with compatibility of facts as Canon has?
:o)))))*
___________________
* what means this:
:o))))))
explanation: the sign :o))))) in koments menas that somebody made a joke. If somebody replies serious on joke, it is a problem. It shows that the person has no sense of humor, which is in high correlation with high intelligence. Sorry my friends - that is a fact.

Jun 29, 2016*permalink

I don't think I've ever heard of :o))) meaning a joke emoticon.

I even tried to search google for this emoticon and I came up empty.

Nothing shows up here either:
http://www.netlingo.com/smileys.php

I think you've outsmarted everyone here, odpisan.

Jun 29, 2016permalink

Colin46

@Odpison,

If you read the posts properly, you would realise that they are talking about Cfast cards

None of thoes Pentax cameras you mentioned or actually any other dslr on the market are currently using Cfast cards apart from the new 1dx Mark 2

Ok?

Jun 29, 2016*permalink

@Colin

I know .. as you see it was a joke .. meaning: if Pentax cameras do not use those cards, they CAN NOT have problems with them - can they?

U should read all my coments.
:o)))))

Jun 29, 2016permalink

>U should read all my coments.

No. Because there is no subtance in it.

Jun 29, 2016permalink

@tonywong

obviously you lost your sence of humor somewhere ... look under the table, maybe it is there ...
;o)))
____ others _________
I am very disapointed that nobody here understand joking - as you are all in deep depression ... it helps if you look through window this beautiful suny day. Go out & make some beautiful photos as I just did. You will feel only joy & happiness. Then you will write some jokes of your own.
Just do it!
:o)))))

Jun 29, 2016permalink

I have a great sense of humor, but you are not funny. The problem from you is that not you decide what's funny, but the one you tell your "jokes". Now please have another funny quote no one wants to read.

Jun 29, 2016permalink

@vscd
Ok .. sorry. Everybody has his own sence of humor & mine is not wanted here as I see on this depressing topic where you all are dissapointed because of Canon's approach to customers.

That was the last time I was joking here on DPreview. It is better to simulate depression & to cry with others about Canon.
:o(::::...

_________________
;o)))))

Jun 29, 2016*permalink

Ramjager

Mmm 2nd big problem?
Pleas Fujica as the owner of 2 1DxMk2's exactly what was the first problem?
I assume you have discovered something that the rest of us owners should know about.
So what is it?
Oh I have now shot 35000 frames with two bodies and yet to see your second issue but have had ten corrupt photos...

Jun 30, 2016permalink

SickVodka

@odpisan:
yeaaa i have a sense of humor.. but a joke that is less funny than a di** slammed into a cactus, does not make me laugh.. (thank you captain obvious!)

please do not get depressed about your jokes.. there are a lot other joke cameras (like the Lytro or maybe soon an other Samsung camera) you can try again ;)
Just shows that pointing out obvious things aren't as funny as it used to be..

11 months agopermalink

YA9Productions

a problem with sigma lenses and now this??
and in the most expensive camera from canon.

Jun 28, 2016permalink

As I see also by other products - not only by photocameras - the big money which is payed for some product does not guarantee high quality any moore. Nowadays you can find high quality products for reasonable price or even better ... for unbeliaveble low price. Products of unknown producers - newcommers on the market .

Jun 28, 2016permalink

FredVasilescu

And sigma is related to canon, how.

Jun 28, 2016permalink

photenth

How can you blame Canon for Sigma badly reverse engineering the lens mount? Sigma just didn't properly implement the mount protocol.

Jun 28, 2016permalink

@photenth - Yeah. It's normal and happens ALL the time that when new bodies are released, they don't always play nice with existing third-party lenses because the bodies use new features or command the lens in ways not previously expected.

Like Sigma HSM lenses and Pentax live-view bodies for 1-2 years - Severe hunting. Most likely Pentax didn't use focus motor speed control for normal PDAF, but had the capability in SDM lenses. Then they released a liveview body which needed motor speed control and BAM - Sigma lenses would hunt constantly without ever locking. (CDAF needs the lens to slow down enough for the body to be able to keep up. Same reason Sigma's SA-mount lenses needed firmware updates to work with the MC-11 but their EF-mount didn't - the SA-mount MC-11 firmware release was VERY likely a rollup of their EF-mount "cine body compatibility" releases.)

Jun 29, 2016*permalink

Ramjager

And who would spend 6000USD on a camera and skimp on the best glass in the world?

Jun 30, 2016permalink

Dragonrider

The problem is Western Digital really put a micro hard drive in those SanDisk CFast cards and it takes a while to spool down :-).

Jun 28, 2016permalink

lol .. I remember those microdrives!

Jun 29, 2016permalink

photomedium

They are fixing it, it's all good. Anyways, sandisk cards are just fine.
No need for Canon to play the blame game.

Jun 28, 2016permalink

works fine with lexar,etc ..

not with sandisk..

and it's a blame game?

Jun 28, 2016permalink

Some Lexar CFast cards are having the same problem.

Jun 28, 2016permalink

photomedium

rrcad,
"The statement makes very clear that the problem is caused by SanDisk's cards, not by the camera." Blame game. In reality is like April is saying, other cards may have the same issue. Things can get funny with high capacity cards and the writer should be smart enough to deal with it, canon is making it so now.

Jun 28, 2016permalink

@ AprilW,

I have no problems with my 1Dx2 and Lexar 3400x 256GB cards.

Jun 28, 2016permalink

@Canonisti

It is very strange that Ken Rockwell doesn't write about this problem - isn't it?
>;o)

Jun 28, 2016permalink

No of course not - Its the one special feature that he loves.
Free Write errors on each session make it stand apart from all other cameras.

Jun 28, 2016permalink

:o)))))

Jun 28, 2016permalink

George Zip

Too busy supporting his growing family

Jun 29, 2016permalink

sh10453

You read Ken Rockwell's stuff? I'm disappointed.
DPR should ban you from this site :)

Jun 29, 2016permalink

@sh10435
No, no, nooo ... I don't read that garbage. Only exceptionally by souch opportunities I control the temes without realy reading them - like about Pentax K1 or this one.

Jun 29, 2016permalink

The last time I read his stuff, which was A LONG time ago, his "reviews" seemed more like his opinion (which some reviews in general are based on opinions--but his seemed to lack enough evidence to support his side)--rather than a review. Plus, I just didn't care much for what he had to say, especially claiming that the many kit lenses are junk. Not quite true--some are, some aren't--it just depends on the lens and the body it's sold with--sometimes it's the lens, or sometimes the body just sucks. Anyway, I questioned some of his logic in whether to buy/not buy stuff. Even if he did publish a page on this issue, most probably wouldn't read it anyway.

Jun 29, 2016*permalink

sh10453

I went to his web site once, and only once, about 10 years ago.
I saw what I saw, and never ever went back.
As far as I'm concerned, I don't give a "..." to what he writes.
Of course I'm entitled to my opinion, just like he is :)

Jun 29, 2016permalink

David Stephen

Who uses SanDisk? Lexar4Life! lol

Jun 28, 2016permalink

Some Lexar CFast cards are having the same problem.

Jun 28, 2016permalink

Evidence for this please? Namely the tail end of an image being corrupted with a Lexar card, not some other issue.

Jun 30, 2016permalink

The tail end of an image being corrupted almost sounds like a software issue, as in the system doesn't verify that the end of the image has been successfully written. I mean, you can probably corrupt any file on any body by switching off the camera before the writing is done. I am still sticking with this is more a Canon problem than a CF card problem. I wouldn't be surprised if they push out a firmware update in the next few weeks with a fix for this. It has got to be something not related to the CF card. Either a software glitch or possibly an issue with the internal circuitry of the camera.

Jun 30, 2016*permalink

Actually I went away and had a re-read, both here and the Fred Miranda forum and saw images that were okay but then went wrong when they read them from the card a second time...
http://www.dpreview.com/forums/post/57918957
Note HKFEVER has both has uncorrupted and corrupted copies of the same file.

Jun 30, 2016*permalink

(comment removed)

Jun 30, 2016*permalink

Sergiusbr

Here in Brazil file corruption is worse. O.o

Jun 28, 2016permalink

Are you saying the Zika virus is the root of the corruption problem in country and camera?

Jun 28, 2016permalink

I think he is referring to the way his government currently handles things and that corruption is one big problem in Brazil.

Jun 28, 2016permalink

Edgar_in_Indy

True story

Jun 29, 2016permalink

sh10453

That's because of the PetroBras corruption.
So the firmware will not fix the issue.

Jun 29, 2016permalink

Fred Miranda has a long thread about the problem
http://www.fredmiranda.com/forum/topic/1434550

Jun 28, 2016permalink

jmey26

"The statement makes very clear that the problem is caused by SanDisk's cards, not by the camera. Canon says it may issue a firmware update in early July to address the problem."

Ok so if its not Canon's problem how come they seem to have a fix coming?

Jun 28, 2016permalink

RPJG

You really can't imagine that a firmware update might overcome a problem with the cards?

Jun 28, 2016permalink

https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/drivers/mmc?id=3550ccdb9d8d350e526b809bf3dd92b550a74fe1

It's called a "workaround".

When a piece of hardware does not meet spec, and you determine exactly HOW it is non-spec-compliant, you can work around the problem by avoiding the trigger condition. See, for example:

Is it Linux's fault that some Samsung MMCs suffer unrecoverable damage as a result of a command that is part of the MMC spec version that Samsung claims the part is compliant to? No - it's Samsung's fault. But once it was discovered that the Samsung parts were noncompliant, the operating system was changed to not use the feature that Samsung screwed up.

Not just Samsung - https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/mmc/card/block.c?id=refs/tags/v4.6.3#n2484

Jun 28, 2016*permalink

Revenant

Canon certainly have a problem, even if they didn't cause it. And if it's possible to fix the problem with a firmware update, it's in their interest to do so.

Jun 28, 2016permalink

new boyz

Because Canon "The Fixer" fixes other's problem? Next, they will issue firmware upgrade for Sony and Fuji. lol! No fix for incompatible Sigma lenses though. Strictly Canon only lenses please... haha.

Jun 29, 2016permalink

because it's not something they can fix. they'd have to break their own protocols to make something of sigma's work.

sigma also falsely reports itself to the camera exasperating the issue.

Jun 29, 2016permalink

Looks like Canon is having trouble moving into this century with memory media also, should have gone with Sony's XQD cards.

Jun 28, 2016permalink

photenth

CFast cards are perfectly fine, this is clearly a manufacturing fault and is already fixed at Sanddisks end. Canon probably will push a firmware upgrade that detects these specific cards and does a workaround to make even those compatible.

Jun 28, 2016permalink

the Sony QXD cards that sony doesn't use in their cameras?

Jun 28, 2016permalink

They were designed for Sony video cameras

Jun 28, 2016permalink

Still Sony is not using them in their still cameras - even though those Sony still cameras do 4K video.

Jun 28, 2016*permalink

Sony also uses usb2 in their still cameras, I guess they figure its good enough since they lead everyone in processing.

Jun 28, 2016permalink

I don't understand why manufacturers still include CF card slots (well, I do, but to switch over to SD cards isn't that expensive, unless you'd be replacing dozens of cards or something like that). Granted, the CF cards MAY be able to write faster than some SD cards, I find that SD cards are more reliable, easier to store (in a case or something), and work in almost any computer (CF cards require an adapter in most computers).

Jun 29, 2016permalink

AshMills

Seems a bit of shame they didn't test the main brands before release. Personally I've not had any problems with SanDisk with 100,000's of photos professionally.

Jun 28, 2016permalink

You don't understand the nature of the problem.

Jun 28, 2016permalink

AshMills

Go on then, enlighten me as to what is not in this article. Camera has issue with some memory cards, (including the type bundled with the camera.)

Jun 29, 2016permalink

I suspect some card firmware change to gain a bit of extra performance did them in.

Jun 30, 2016permalink

DualSystemGuy

At least they are warning users about it.

Reports of soft focus and sensor oil across various forums are more concerning - no advisory yet but probably too early tell if it's widespread.

Jun 28, 2016permalink

Ryanide

And these are the cards that were bundled with the camera!!!

Jun 28, 2016permalink