D800 is here..

[arthurking83]

Yeah, did you see the D4 performance data?

It seems XQD is the way to go in future(see other thread!)

[Chris G]

Yeh I did check that out, files where screaming across

I'm surprised Nikon didn't add the XQD to the D800 but I guess if they did it would of be counter productive for sales.. lol

[arthurking83]

On another note, check out the speed of the Lexar card at the top of the D800 list compared to the same card on a Canon 5DIII!!
Much faster than on the D800 ... so it seems that Nikon don't seem to have the camera to card interface as optimised as Canon seem to have got.

It'll be interesting to see how well the Canon D1X performs compared to the D4, as the Canon is technically faster(in ultimate FPS rates) yet it only has CF card slots.


Anyhow.. back onto the D800 topics .... The D800 was probably signed off on just before the quake/tsunami issue in Japan, and compounded with the Thai flood issue, the delay was probably close to a year or maybe more.... it was a very late update for the D700 even by Nikon's standard timing.
It seems that the D800 was never going to be fitted with a XQD slot, and as I remember, Nikon pushed hard in late 2011 to have the XQD format ratified by the Compact Flash Association.
They almost certainly did this as they planned from day one to have the D3 successor fitted with XQD, and some of the other members were tardy in approving the format.

So the D4 was barely lucky to have an XQD card slot at all, as in late 2011 the XQD format was finally ratified, and abotu 3 months later Nikon announced the D4!
It's rare for a totally new card format to be 'conceived' and a camera to have one in such quick time.

The XQD format was supposed to be the CFast card format, which is the successor to the CF format, but the CFast format obviously morphed and evolved as time went by, and we ended up with XQD instead.

[ricktas]

On another note, check out the speed of the Lexar card at the top of the D800 list

Ouch, checked the price out too. Over $300.00 for 32gb, or you can get the 128GB version for around $1000.00. . I could buy a D800 or 3 cards..LOL

[rodw]

Ouch, checked the price out too. Over $300.00 for 32gb, or you can get the 128GB version for around $1000.00. . I could buy a D800 or 3 cards..LOL

Now you know why I didn't buy a super fast CF card, I think 60 mb/s will be fine for my needs
SD is much more affordable than CF and I prefer them over CF (as I have bent CF pins on a few readers over the years) and have never had one fail.
Still not sure if the camera itself can write fast enough to keep up with the speed of the cards past 60 or 90 mb/s. Maybe there might be some advantage reading data off the card if you use a USB3 interface.

[ricktas]

Now you know why I didn't buy a super fast CF card, I think 60 mb/s will be fine for my needs

Agree, I tend to take my time setting up my shot etc, then take one frame. Recompose etc, and go again. I don't think speed of the card will be a big issue to me. And the D800 looks not to be designed as sport photographer camera etc anyway, its FPS rate is not high enough.

[arthurking83]

Ouch, checked the price out too. Over $300.00 for 32gb, or you can get the 128GB version for around $1000.00. . I could buy a D800 or 3 cards..LOL

Yeah! LOL!

Prices are a bit extreme, but you have to take in context.
I purchased my fast cheapie card about 2 or three years back for $95 .. when a similar Sandisk was about $200-250 (depending on how much you like to gamble) .. that's 8G for $100 and it gives me approx 60Mb/s .. but I have yet to test it out properly on a faster connection(or timings on my camera). I just know it's fast enough.

Therefore this faster Lexar 32G card is actually cheaper in a total Gb/$ equation. But I still prefer to have my data spread out over more cards than to have them all on the one card.

[rirakuma]

Hey Arthur can you give me more info on the patriot memory you bought? For those getting a D800 its probably worth the money getting a faster card. When I take a picture in live view it has almost 5s of blackout screen, im guessing its because of the slow write time of the crappy sandisk that I have. Its actually quite annoying, the camera is pretty much unusable during that write time (only happens in live view)

[rirakuma]

On another note the live view histogram is a really neat feature!

[Tommo1965]

On another note the live view histogram is a really neat feature!

yes that does look like a useful tool...but i dont shoot in liveview much..a d800 might change that

[Xebadir]

Have decided after that great article posted by Chris (and after deliberation and more reading) I'm going to go with a 32GB '60mb/s' CF and a 32 Extreme Pro SD (95mb/s class 10) as my video available card (price is pretty reasonable for these too). Works out the best for my needs, where I generally don't need a burst rate anyway. Crossing my fingers that 3 D800s will arrive at my supplier next week in this next shipment as I would like to learn the camera before the states trip.

[Chris G]

Yeh that's exactly what I'm doing John as my photography does vary from time to time like the 2 weddings I have coming up one being September and the other in November. Luckily for me they are both for family, but I still wish the D800 would hurry up so I get properly acquainted with in that time before the first one in September as it will be my first FF cam.

D90 to D800, I'm going to be having so much fun..

[arthurking83]

Hey Arthur can you give me more info on the patriot memory you bought? .....

I just checked MSY's site(the PDF version) and they only have the cheaper 266x cards listed at the moment.

I got mine about two or three years back now, it's been quite heavily used and no problems at all(so far).

If there's an MSY store not far from you, it may be worth going in and asking or seeing for yourself.

here's an image of mine:

D300_DSH_5107.JPG

There is no brand name on the card, only the P symbol with the stars at the top of the card and the rear face is bare metal only with a faint imprint of the serial No.
I remember the blister pack had the name Patriot on it.

I just had a peek on Ebay, and have noted(and 'watched') a couple of interesting cards listed there. One of which is the Duracell 600x which gets a high score on Rob Galbraith.
$50ish for a fast 16Gig card .. landed!
Unfortunately a quick check on CheapChips hasn't revealed anything of value. Good cards and all that, with the high end Lexars listed at 'reasonable' prices .. but I still think it's just too much for not enough gain.

[arthurking83]

Hey Mongo, I was adding my other reply as you added yours.

For (predominantly) shooting birds I wouldn't choose the E model.
The lower specced low pass filter will most likely cause moire effects in the high frequency patterns in the birds feathers.

Where there is high frequency repeating patterns, such as many fabrics, or distant building textures, or birds feathers, you risk getting moire patterns.
You can get software that reduces the effect, but this is not the same as not getting the effect in the file in the first place.
It(the software) reduces the resolution of the fine detail anyhow, so why bother trying to get it in the first place.

There are particularly good reasons to have the E model, and because I primarily concentrate on this genre, Landscaping comes to mind as a reason to get the E model.
It's very rare to get repeating high frequency details in nature like that.

I've been reading that a lot of folks (on other fora) are interested in an E model for Macro, where you'd expect the extra detail level to be an advantage, but again, in many macro situations you get very fine repeating patterns that will cause moire ..... think of a fly's eyes or very fine hair like structures in a flower.

I hate moire, and while this was a long time ago that I had to deal with it, and software may have come a long way since then, this was with the D70s.
A camera notorious for producing moire whenever it could!
Strange little fact about this camera, is that it used the sensor from the D100, which was less susceptible to moire, I suspect in an attempt from Nikon to get better details. But on the D70/D70s, it kind of backfired. The D70 models obviously had weaker AA filters than the D100 had.
But it was well known too that the D100 image could be brought back up to the same level of detail as an image from the D70's could be got!
USM! Appropriate use of USM got images back up to similar standards of detail to the D70 cameras.
It seemed that Nikon relented with that sensor's AA filter, because they also used it in the D50 and then D40 cameras, but they never had moire issues like the D70s did.

It is true tho that in some cases you will get detail in the image from an E model that you may never get from the non E model, as you would expect from a lower strength AA filter on the same sensor, but you risk seeing this moire issue as well.

There is a great example of the issue on the Nikon site:

http://www.nikonusa.com/Learn-And-Ex...slr-tech-moire

While it's great that you can ultimately get more detail from the E, look at the last image to see the moire patterning.
One thing tho, and that is that even tho the fabric in the kimono did produce moire patterns in the image from the D800E and not the D800, if you look at the sample images from the D800E on the Nikon site, the fabric in the kimono doesn't produce moire colour. So the image can be captured(or processed) to eliminate the issue .. just something to be aware of.

[mongo]

Hey Mongo, I was adding my other reply as you added yours.

For (predominantly) shooting birds I wouldn't choose the E model.
The lower specced low pass filter will most likely cause moire effects in the high frequency patterns in the birds feathers.

Where there is high frequency repeating patterns, such as many fabrics, or distant building textures, or birds feathers, you risk getting moire patterns.
You can get software that reduces the effect, but this is not the same as not getting the effect in the file in the first place.
It(the software) reduces the resolution of the fine detail anyhow, so why bother trying to get it in the first place.

There are particularly good reasons to have the E model, and because I primarily concentrate on this genre, Landscaping comes to mind as a reason to get the E model.
It's very rare to get repeating high frequency details in nature like that.

I've been reading that a lot of folks (on other fora) are interested in an E model for Macro, where you'd expect the extra detail level to be an advantage, but again, in many macro situations you get very fine repeating patterns that will cause moire ..... think of a fly's eyes or very fine hair like structures in a flower.

I hate moire, and while this was a long time ago that I had to deal with it, and software may have come a long way since then, this was with the D70s.
A camera notorious for producing moire whenever it could!
Strange little fact about this camera, is that it used the sensor from the D100, which was less susceptible to moire, I suspect in an attempt from Nikon to get better details. But on the D70/D70s, it kind of backfired. The D70 models obviously had weaker AA filters than the D100 had.
But it was well known too that the D100 image could be brought back up to the same level of detail as an image from the D70's could be got!
USM! Appropriate use of USM got images back up to similar standards of detail to the D70 cameras.
It seemed that Nikon relented with that sensor's AA filter, because they also used it in the D50 and then D40 cameras, but they never had moire issues like the D70s did.

It is true tho that in some cases you will get detail in the image from an E model that you may never get from the non E model, as you would expect from a lower strength AA filter on the same sensor, but you risk seeing this moire issue as well.

There is a great example of the issue on the Nikon site:

http://www.nikonusa.com/Learn-And-Ex...slr-tech-moire

While it's great that you can ultimately get more detail from the E, look at the last image to see the moire patterning.
One thing tho, and that is that even tho the fabric in the kimono did produce moire patterns in the image from the D800E and not the D800, if you look at the sample images from the D800E on the Nikon site, the fabric in the kimono doesn't produce moire colour. So the image can be captured(or processed) to eliminate the issue .. just something to be aware of.

Mongo sends his very big thanks to you Arthur for for the detailed explanation. Mongo was afraid that may be the answer - now he is sure.


There were reasons for the question. The obvious one is that the D800E must have an application and no shortage of buyers and it is a little sharper. Clearly, there are limits to its use- horses for courses it seems


The thing that got Mongo going down the track to that question was also reading that the Fijifilm X-pro 1(for general use) has done away with the low pass filter and gets sharper out of the camera images to start with. The appear to have achieved this by randomizing the red, green and blue pixel arrangement in lots of 6 x 6 pixels. Very clever ! Clearly, Nikon has not done this or has not thought the idea worth while or fiji’s patent is preventing that from happening.


Whilst it is generally true that each successive generation of camera brings real improvements on the last and that the D800 is one of the best things announced by Nikon since Nikon invented sliced bread , Mongo now thinks that the D800 is probably one or two generations short of being ideal. It is possible that its successor may incorporate something like the fujifilm technology (giving sharper out of the camera images so we do not have to work so hard in PP trying to get that lost sharpness back) and also Sony’s mirrorless technology giving less vibration and very quick frame counts.


Just a thought - and again, thanks Arthur.

[mongo]

now that the D800 is really close by the look of it, Mongo has been giving some serious thought as to whether or not to get the D800E model. He knows all that stuff about the low pass filter etc but is the practical difference ? The filter reduces the sharpness but also reduces ill effect too.

This is likely to be Mongo's last camera, so he wants to make the right choice for his type of shooting - long lens wildlife, portraits , macro and landscape (Suppose it does not leave much else ???).

Would particularly love to get Arthur's opinion on this stuff but all information/advice gratefully accepted. will see your replies in the morning - Mongo is going into short overnight hibernation now.

thanks

[arthurking83]

I was curious to know for myself how well this card does, so I did a RG type speed test.

I can't give an accurate assessment of what the Patriot card can ultimately do, but I can explain how the D300 limits in in some way.

I used another of my cards .. a 4G Sandisk Ultra II to compare it with. I also have two other main cards in the form of 8G A-Data cards, that I originally purchased for my Toshiba PDA only .. not for use on the cameras. Those two were purchased cheaply only for their size(at the time) as I needed masses of space for lots of maps.

So the Patriot seems to write too quite quickly, in fact quicker than the D300's buffer can empty! .... this is why I can't give you an accurate assessment of how fast the card ultimately is.
It maxes out the USB2 card reader at 17-20Mb/s in both read and write times .. that's as fast as I can get from the USB2 card reader.
I'm hoping to get a USB3 card reader one day soon too, but it's not a priority.
The Sandisk card I have is only 9-10Mb/s by comparison.

On the D300 with the Patriot card, I can get more than the standard 17 frames before the buffer limit is reached.
If you know Nikon cameras, you'll know that when you half press the release, you see on the top LCD screen (and in some viewfinders) an r figure.
That is, something that looks like r12, or r09, or in the D300's case r17. That's the current buffer limit, and 17 is supposedly the maximum.(note you can set this to a lower figure if you wish too!)

With the Patriot card, my buffer doesn't slow down until I get 20, or sometimes 21, images exposed .. that is, the buffer is clearing out quicker to the card than the buffer can fill up with.
The Sandisk Ultra can't do this at all. It hits the 17 image limit, then there is a definite slow down, a sec or two and you can easily feel for this and stop shooting.
With the Patriot card, you can't. You have to watch the image counter on the review screen, or count frames and then forcibly stop shooting to shoot only 17 frames .. even when you hit the 20th or 21st image, it still shoots quite rapidly in comparison to the Sandisk card.
With the Sandisk card, the camera is simply locked up until the buffer clears a little bit, then you can shoot more again in a burst... and so on.
As would be expected of an older model camera like the D300, the hardware is more limited, in that the buffer clearance times is limited by the camera's hardware, not the card.

FWIW, some times for clearing out the buffer to the card:
Patriot cleared 21 images in about 13-14sec(call it 13.5sec) at 13.2Mb per image, that roughly equals about 20Mb/s write times to the card. I think this low value is more to do with the D300's hardware, than the Patriot card, as I said, by about image number 9 or 10 in the buffer(of 20-21 images), the green light starts to blink as the images are written to the card. That is, the card is waiting for data from the data chip in the camera.
The other quick test I did, was to shoot 6frames on my D300 to see how quickly they wrote to the Patriot card, and the time to green light out was about 4sec or so from pressing the shutter release.
Sandisk UltraII stopped shooting at exactly the 17 image mark every time. The green (card write) light stayed lit for the duration of the 17 images being displayed on the review screen, and then some more.
In total the time(s) from ceasing shooting, to the green light going out was 27sec or so .. 26.75s, 26.94s and so on.
This card is OK .. but only just. I keep it now because I simply have it. I purchased it thinking ... get a Sandisk, everyone swears by them and so on, but I also got burned by them with another card I also purchased. It got corrupted, I lost a few images but in formatting it again after the corruption, I lost the card .. it only red zero bytes capacity. Some internal chip or something went 'pop'.
The only CF card I've ever had trouble with funnily, and I have a swag of cheapie no-name branded cards from various sources(MSY, ebay, etc).

Apologies for going wayyy OT, but I hope you find your fast CF card .. in some cases there'd be nothing more infuriating than having to wait for a card to clear it's throat just when you really don't want it too(eg. wedding shooting or other fast-ish paced situation .. even if it's only 4 frames in a sec!

[Chris G]

It maxes out the USB2 card reader at 17-20Mb/s in both read and write times .. that's as fast as I can get from the USB2 card reader.
I'm hoping to get a USB3 card reader one day soon too, but it's not a priority.

Andrew I'm pretty sure you are aware of this but just for reference for others, in order for a USB3.0 to "fully work" you need also a USB3.0 port on your computer otherwise it will only transfer that of a USB2.0
Regardless of card type / speed, unless you are upto date with computer hardware within the last year or less your more than likely not to have USB 3.0 fitted on your main board / computer.
Of course this may effect a very large margin of people who choose to rent or are behind in current computer tech by a generation or 2, this will also be the same with Mac computers / lappyies aswell.

As stated, USB3.0 is backwards compatibility with USB2.0 from the Nikon D800 and anything in that matter but I have been reading around that mac users think that they wouldn't be able to transfer from the D800's USB3.0 to the Mac's USB2.0 because they don't have the USB3.0 port..

This statement is: UNTRUE , The only thing that it affects if the transfer rate of that of a USB3.0 compared to USB2.0

[rodw]

Good to see a discussion on chips and a couple of people following my lead but here is a couple of observations.

It is unfair to compare the Sandisk Ultra at 30 mb/s with the Patriot at 90 mb/s (if their 600x is to be believed). But it does show that a 3 fold increase in card speed does not yield much in terms of buffer performance indicating that other restrictions are in play within the camera (eg, 3x increase in card speed only yielded 3 more images in the buffer stakes).

Cheap as chips have a $30 Sandisk USB3 card reader which I bought with my chips. It seems to be fine for me.

If you were going to benchmark card performance you would need to do it on a USB3 reader using a HDD benchmark program. I did this a long time ago with USB2 and FireWire external HDD. Google should unearth something. I'm too lazy to dig back and find out what software I used back then. The results were interesting.

I've got USB3 on my PC and an external USB3 HDD and the performance is blistering so it will be interesting to see the difference in performance between my CF And SD Cards. If I do this, I'd start looking at using the benchmark program while the camera was connected by USB3 and then once performance was understood, look at some speed comparisons of buffer performance against this benchmark data. There is no doubt in my mind that the USB3 interface will NOT be the limiting factor.

[arthurking83]

Hey Chris, that was me(Arthur.. AK83) that said that (quote) about the USB2 slow issue.

I got USB3 covered and have had for a long while now.
All good, I get between 60-90Mb/s from my USB3 enclosed external hdd. So once I get this USB3 reader, the USB interface won't be the bottleneck.

I'm just going to hazard a guess with what I believe this Patriot card to be capable of, and I reckon it'll be a good 40Mb/s once it's allowed to work properly.

LOL on the Mac users USB3/2 concerns.

USB3 is fully backward compatible with USB2 that I know of .. in fact when you have issues with your USB3 connection, the work around is to revert back to USB2 to get the device working again.

Strangely enough I've had this issue and can say from experience .. if your USB3 device suddenly stops connecting via the USB3 connection, there's a very high probability that connecting it to a USB2 port will get it going again!
(had this problem with a problematic USB3 HDD docking station .. that still isn't resolved!)