Nikon D800s All Discussions

PB_PM

The mirror is the real issue. Example, the Canon 1DX can shoot faster than 11FPS, but only if the mirror is locked up and auto focus is disabled.

sevencrossing

when increasing fps with the D4 Nikon have a bigger body to play with
a battery pack on a D800 increases the size but can only hold batteries

PB_PM

As far as I can tell the size of the body has nothing to do with shooting speed. The D700 could pull off 8FPS with AA batteries or the EN-EL4a in the grip. In the case of the D700 and D800 body only, the voltage of the standard batteries, EN-EL3e and EN-EL15, simply don't have the oomph to push enough power to the camera (Nikon specifically told reviewers that was the case with the D700).

sevencrossing

Products are build for a price point, they know that if they up the price there will be fewer buyers
+1
It does not matter whether you want a fast car, a fast computer, a fast lens, a fast camera or a fast women; Its going to cost you

The D400 thread proves their is an interest in a Hot hatch DSLR but there will always a limit to how much even keenest amateur photographers, will pay for a few more fps
The well heeled are already catered for with the D4s and the Bugatti Veyron

paulr

TT +1 I believe it took 4 years to develop the Nikon 1 camera so that did not happen overnight either

TaoTeJared

@snakebunk - if you work in that realm you are working with and around people who do understand it. I have always worked in realms where I have been part of that bridge that crosses that gap. 99% of people really don't understand it at all and don't even understand the difference between a electrical/hardware engineer and a software engineer/developer. They just all work on computers right?

Bottlenecks are quite simple (in order) - Shutter/mirror mechanism combo, Power (shown by the increase of FPS with grip), Processor, buffer, write speed to the cards.

Shutter/mirror mechanism combo: Must have the physical ability to move at those speeds. This is not a small feet. The D4/s certainly has a different mirror mechanism to get it out of the way. The Digi cams, phones, etc. that have high FPS do not use mechanical shutters but turn the sensor on and off. At this time the image quality suffers quite a bit (in noise) as it is trying to get power at the same time it is recording the image. It appears across all companies, above 6 fps, camera's shutters and mirrors can handle it. Above that, you start to see "digital capture" or what I described above, leaving the shutter open and zapping the sensor to get shots. You also loose true continuous AF and get into more of the "predictive" tracking of systems which is just another way of saying they are guessing what will be in focus rather than the pro bodies actually taking the AF reading.

Power: Pretty self explanatory and is shown by the increase of FPS with grip on all systems and for the Pro bodies having different batteries.

Processor: They can only process so many MB per second. D4 was 10fps but 11fps with AF locked. 16mb x 11 = 176mb/s. D800 is 6fps (with grip) x 36 = 216mb. My experience with the D800 is that the file sizes are rarely above 30mb. Using that you get 180mb/s or just about the same as the max of the D4. So with everything, the processor can probably only handle 170-190 mb of image data per second. My guess it is the good ole computer multiple value of 192.

Sidebar: The D4s only adds 1fps, but has a whole new AF system to track subjects - that is the real big news. From hearing pros talk about the new Canon Pro bodies that were better than Nikon, this is what needed improvement. This is probably where the Expeede 4 comes into real play - not FPS.

Buffer & Writing to cards: I believe it covered for the most part. The buffer would be the simplest but that just holds the images and doesn't impact the initial burst. Eventually it will fill up and the bottle neck becomes the writing to cards. My guess is that it is sitting with a 256mb chip for the D4 & D800. This is where XQD (168mb/s) cards come into play (as will the new CF cards) where they can clear 2/3rds of the buffer every second. That would also give you 6 seconds before the buffer is full for a D4 and 3 seconds for a D800, which lines up with the specs and what I have seen.

There is the 10,000 foot view of bottle necks. There probably are more complex issues that we just don't have the ability to see.
-----------------------------------
I shoot my D800 with the grip all the time. Realistically, I don't get 6fps. 4-5 at the max. The D4 could do 11fps with the AF locked and the D4s 11fps with continuous AF. An increase of 0. Sony's A7/r can only do 1-2fps. Canon 5D Mark III 22MP can only do 6 FPS with a 1/3rd less pixels. The dream that a D800s with files twice the size mind you, will magically double it's FPS is just unrealistic.

There are things like trap focus, custom saved menu banks, tweaks to VR and auto ISO controol, and 100's of other refinements that would be much more beneficial than FPS. The insane focus on FPS is just baffling to me.

This focus on FPS tells me two things:
1) People don't want to spend $6,000 for a body (no surprise there and you can add me in that camp.)
2) People are missing shots and think that more frames are the answer. That is true only for extreme situations - like for those shooting fast sports or fast wildlife. For everyone else, Faster FPS will just get you more missed shots. It is about anticipation and knowing your subject, not the camera.

10 years ago the D2xs could only shoot 5fps! That's right, the top of the line sports camera shot the same FPS as the D800 with the grip on. Sports, Wildlife, pro shooters were getting the shots then. If shots are missed, it is the shooter, not the camera. Oh and if you haven't heard, it is a hell of a lot cheaper to improve the shooter.

PB_PM

I shoot my D800 with the grip all the time. Realistically, I don't get 6fps. 4-5 at the max.

Unless you are shooting in DX mode the D800 is limited to 4FPS, regardless of using the grip or not.

TaoTeJared

Various Nikon items says it can - I'm going off of that. I have seen 4.5, 5.5, & 6 fps on various things from them. I have seen the most of 6fps with the grip on things from them and not on DX mode. Like I said, I don't get anything more that 4-5 but I don't really have the means to "truly" test it.

PB_PM

Nikon disagrees with you... http://www.nikon.com/news/2012/0207_dslr_01.htm

The MB-D12 supports the Rechargeable Li-ion Batteries EN-EL15 and EN-EL18, common AA batteries, and the AC Adapter EH-5a/b (with Power Connector EP-5B). When the MB-D12 is mounted on the D800, high-speed continuous shooting at approximately 6 fps* using the DX-format image area is possible. The MB-D12 also offers the added convenience of controls, including shutter-release button, AF-ON button and multi-selector, for vertical shooting. What's more, the same seals as those used for the D800 with its magnesium body offer superior resistance to dust and water.

*When powered by the AC adapter, or AA batteries or Rechargeable Li-ion Battery EN-EL18 with the MB-D12. Measured according to CIPA guidelines.

I can find no information from Nikon stating that faster than 4FPS possible (FX), regardless of the power source.

Edit: A Youtube video also seems to agree with Nikon's official figures. https://www.youtube.com/watch?v=nX9QifNT7sY

Ade

Processor: They can only process so many MB per second. D4 was 10fps but 11fps with AF locked. 16mb x 11 = 176mb/s. D800 is 6fps (with grip) x 36 = 216mb. My experience with the D800 is that the file sizes are rarely above 30mb. Using that you get 180mb/s or just about the same as the max of the D4. So with everything, the processor can probably only handle 170-190 mb of image data per second. My guess it is the good ole computer multiple value of 192.

You are confusing megabytes, megabits, and megapixels. Therefore the math doesn't add up.

TaoTeJared

That is not the discussion Ade. You are confusing the actual message into something it is not. Create a new thread if you want to have a different discussion rather than hijacking this one.

ads

"That is not the discussion Ade. You are confusing the actual message into something it is not."

@Ade was just pointing out your maths was a fair way off on the processing speed @TTJ.

ie D4 - 11fps * 34.3mb = 337.3mb/sec,
D800 4fps * 74.4mb = 297.6mb/sec (or 6fps * 32.5mb in DX mode with the D4 battery).

Also the processing speed wont be a computer multiple - but the size of the buffer chips will be.

But you're right in that the Expeed 4 in the D4s seems to be of more value for AF than for FPS.

TaoTeJared

I don't have a single raw D800 file above 38mb max and that is for very complex scenes. Most are around 28mb for 14bit compressed. I have never had one go anywhere near 74mb or anywhere close. I've never heard of a D4 having 34mb files either. Uncompressed maybe? I always (and just about everyone I have ever met) shoot 14bit loss-less. If you shoot uncompressed, you will not get 6fps for sure.

ads

"I have never had one go anywhere near 74mb or anywhere close. "

Figures are from the Nikon website for uncompressed - you were talking about the *maximum* throughput of the processor @TTJ so they are the relevant figures. You'd used 16 and 36, the megapixel ratings rather than the file sizes in mb.

On the D800 the level of compression doesn't affect the fps, just the number of shots you can fit in the buffer and the time to clear it.

Which of course reinforces your point that FPS depends on more than just the processor...

TaoTeJared

There comes a point when experience overrules marketing material. Obviously you don't own either or you would know the actual file sizes in use. File sizes from the compression does make a huge difference in the FPS. Jpeg is the fastest, 12bit, is slower, 14bit a bit slower yet, and if it is uncompressed, it is slower yet. Shooting Tiff files, basically kiss any sort of FPS goodbye.

If you really want to see a difference on any Nikon body - start playing with turning on and off the Active D lighting, lens corrections, and auto ISO. You will start seeing the difference quickly.

PB_PM

Agreed, average file sized, for even 14bit lossless compressed is around 41MB.

Ade

File sizes from the compression does make a huge difference in the FPS. Jpeg is the fastest, 12bit, is slower, 14bit a bit slower yet, and if it is uncompressed, it is slower yet. Shooting Tiff files, basically kiss any sort of FPS goodbye.

Sorry but that is absolutely incorrect.

I ran a simple comparison test on the D800E and there is zero FPS difference between 12-bit vs 14-bit, RAW vs TIFF. The data is presented below. We can see from the two black vertical alignment lines that in each trial the camera captured 16 frames in the same exact time, regardless of the compression settings.


Compression settings vs. FPS, Nikon D800E (full size)

Anyone may reproduce the test using a simple audio recorder such as Audacity.

SquamishPhoto

He's just confusing FPS with a slowed buffer.

snakebunk

Expeed 4 is rated for 12fps with 24mp sensor, so it should be able to do 8fps with 36mp. If they go to 40mp, you get 7fps, and if 50mp, you get 5.75.

I think Ironheart said it correctly already on top of page 2. Personally I like more fps rather than more mp. I have never experienced the number of pixels in the d800 to be a limiting factor. Sometimes a higher shooting speed is handy as it gives me more images to choose from.

ads

"There comes a point when experience overrules marketing material. Obviously you don't own either or you would know the actual file sizes in use. "

@TTJ you were discussing the various factors that can affect FPS, including processor speed, and you said the processor can limit FPS and quoted how much the processor can process - I was just pointing out that in fact the limit of the processor was higher as a D800 can process 4 uncompressed photos a second - ie that's the theoretical maximum processing power the camera has at its disposal, which is the point you were making

"File sizes from the compression does make a huge difference in the FPS. Jpeg is the fastest, 12bit, is slower, 14bit a bit slower yet, and if it is uncompressed, it is slower yet. Shooting Tiff files, basically kiss any sort of FPS goodbye.

start playing with turning on and off the Active D lighting, lens corrections, and auto ISO. You will start seeing the difference quickly."

You mean buffer size - file format/compression and options like d-lighting absolutely make a difference to how many shots fit in the buffer, how quickly it fills up and how long the buffer takes to clear.

But it doesn't affect the FPS of the body - if I pick up a D800 and press the shutter in full FX mode it will take a max of 4 frames per second regardless of compression until the buffer fills up.

Let's get back on topic. What I am interested in for the D800s is whether the FPS for DX increases too. 6 FPS in FX with the D4s AF improvements would be awesome, if it went to 8fps in DX mode that would be even better...

sevencrossing

ads Let's get back on topic.

Good idea

Given that the price difference between the second value of my D800 vs the price of a new D800s
is likely to be similar to the cost of a new D7100
I think I am more likely to buy a D7100 as a second camera

PB_PM

Or you could just wait for the D800s to come out, and buy a lightly used sample (less than 10k) from the "must have the latest and greatest" crowd, who will order the night the new camera is announced. You could easily get a D800 or D800E for $1900.

Photobug

ads Let's get back on topic.

Good idea

Given that the price difference between the second value of my D800 vs the price of a new D800s
is likely to be similar to the cost of a new D7100
I think I am more likely to buy a D7100 as a second camera

Good idea buying the D7100 for a second DSLR body. You will not regret it.

nek4life

The most recent post mentioned Photokina in September as a possible announcement date. Going by the typical Nikon release cycle when would the camera actually ship? Would it be released soon after or in the middle of January like the D4s? I know this is all speculation, but I'm just curious. Any thoughts?

PB_PM

An early September announcement would normally mean it would start to appear in late September or more realistically mid-late October. Availability will likely be limited for 4-6 months though.