A number of claims are circulating regarding the high ISO performance of the D4 compared with the D3 and D3S, notably that the D4 offers around a one stop advantage over the D3S despite the former's higher pixel pitch.
My own tests suggest in general terms the D4 actually offers around the same noise levels across the image as a whole as the original D3. This is by no means a poor performance, the D3 continues to be an excellent high ISO camera. So what is the basis for claiming the D4 offers a one stop advantage? Remember the D4 offers around 40% more resolution than the D3. Once the D4 images are downsampled to match the resolution of the D3, the D4 emerges as the better performer, possibly rivalling the D3S, but not exceeding it. Looking closely at the character of the noise, the D4 has more chroma noise and Capture NX2 reports the camera is always performing some noise reduction at high ISOs on NEFs, even with 'off' selected in camera. Significantly though, the D4 does have far less noise in the blue channel than the D3, meaning high ISO shots under tungsten light sources will have lower noise.
At the other end of the scale though, noise levels at lower ISOs are virtually non-existent. Overall, it looks as though Nikon has delivered its promise to produce a more rounded camera with a better balance of high ISO performance versus resolution.
Tom Parkes is a London based corporate and editorial photographer
Friday, 6 July 2012
Wednesday, 4 July 2012
D4 Firmware Update 1.02 - It's Tough Being Green
Released this morning, the D4 firmware update 1.02 attempts to address the monitor problems detailed in the previous post:
"Gamut for Adobe RGB images displayed in the camera's monitor has now been changed. This enables more vivid display of images."
I've updated my D4 and can confirm the monitor display has improved somewhat, particularly in relation to the rendering of skin tones. The overall green tint is still visible but appears reduced.
My D4 was supplied with firmware v1.01 installed and it's worth noting that the installation process has changed somewhat from earlier cameras. Full details are available on Nikon's support site. The update took a fair while to instal itself, with no visible progress for several minutes after the progress bar had completed its journey across the screen. Just leave it alone and it will complete.
Two things are particularly interesting, first that Nikon appears to be rapidly addressing snags in the D4 and second, for the first time ever, Nikon actually emailed me to inform me of the availability of the update. If you haven't registered your camera on-line with Nikon then its worth opening an account. As an added benefit you will have instant access to the serial numbers of your equipment should anything be lost or stolen.
"Gamut for Adobe RGB images displayed in the camera's monitor has now been changed. This enables more vivid display of images."
I've updated my D4 and can confirm the monitor display has improved somewhat, particularly in relation to the rendering of skin tones. The overall green tint is still visible but appears reduced.
My D4 was supplied with firmware v1.01 installed and it's worth noting that the installation process has changed somewhat from earlier cameras. Full details are available on Nikon's support site. The update took a fair while to instal itself, with no visible progress for several minutes after the progress bar had completed its journey across the screen. Just leave it alone and it will complete.
Two things are particularly interesting, first that Nikon appears to be rapidly addressing snags in the D4 and second, for the first time ever, Nikon actually emailed me to inform me of the availability of the update. If you haven't registered your camera on-line with Nikon then its worth opening an account. As an added benefit you will have instant access to the serial numbers of your equipment should anything be lost or stolen.
Saturday, 30 June 2012
D4 Monitor
I'm now working with the D4 alongside my D3 and D3S bodies and one of the first questions I had concerned the colour accuracy of the monitor. I'd read various comments from users who found the display to have a green tint. Nikon says the new display is more accurate than that used in previous cameras but the short answer is, yes it does appear slightly green in some light. The truth is, every iteration of the professional cameras has a different monitor tint. The D1 was very blue, the D2X less so, the D2XS was noticeably yellow in comparison and the D3 tended towards magenta, with the result that moving from the D3 to the D4 the green tint appears greater than it actually is. There is a temptation for camera manufacturers to adjust the display to show the image to best effect and so the slight green tint of the D4's display offsets the very slightly magenta-tending Nikon skin tones perfectly. Does the green tint matter in the real world? No, not in my view. At times you're aware of it, at others not. It very much depends on the ambient lighting and it has to be remembered that the magenta tint of the D3 monitor would often disguise a green tint from fluorescent lamps.
New on the D4 is an ambient light sensor for the monitor, or auto brightness. My advice is to turn that off unless you're using the camera in live view mode on a tripod. The reason being that the reading used to set the brightness level is taken once, using a sensor on the rear of the camera. If anything blocks the sensor, such as your face as you lower the camera, the display will be too dark as a result. The camera also tends to set the display at too low a brightness indoors. It works fairly well in sunlight where it's otherwise very difficult to see at all, but the nature of the system introduces another variable to take into account when assessing the image. I've settled on a brightness level of -1 for indoors and either just accept the image will look darker that it actually is when viewed in sunlight or I'll use a hoodloupe for a better view. If I was working outside for a long period I would consider adjusting the display brightness, or even using auto, especially if clients were viewing the display with me, but otherwise it's one of those setting I don't like to change. It's all to easy to forget you made a change when you move back inside or the sun goes down.
New on the D4 is an ambient light sensor for the monitor, or auto brightness. My advice is to turn that off unless you're using the camera in live view mode on a tripod. The reason being that the reading used to set the brightness level is taken once, using a sensor on the rear of the camera. If anything blocks the sensor, such as your face as you lower the camera, the display will be too dark as a result. The camera also tends to set the display at too low a brightness indoors. It works fairly well in sunlight where it's otherwise very difficult to see at all, but the nature of the system introduces another variable to take into account when assessing the image. I've settled on a brightness level of -1 for indoors and either just accept the image will look darker that it actually is when viewed in sunlight or I'll use a hoodloupe for a better view. If I was working outside for a long period I would consider adjusting the display brightness, or even using auto, especially if clients were viewing the display with me, but otherwise it's one of those setting I don't like to change. It's all to easy to forget you made a change when you move back inside or the sun goes down.
Friday, 1 June 2012
Selective Batch Download Added
Just added to the main site, clients can now select a group of images to download in a batch. Instructions for using this new feature have been added here:
Image Viewing and Download
Image Viewing and Download
Wednesday, 16 May 2012
Image Viewing and Download
This post provides guidance for visitors to the main site on viewing and downloading images
- Click the link to enter the gallery
- Enter the password if prompted (passwords are always lower case and with no spaces)
- Please ignore any options to sign in or open an account, these are for a different client group. The password is all you will need
- Moving your mouse around the gallery will trigger larger roll-overs for a better view and enable the full caption to be read
- Click on an image to open it in the full window
- Click 'Download' and choose a size for the image. This will immediately initiate the download to your default location for downloaded files
To download all of the images:
- Click 'Batch Download'
- Select Images to Download - Either 'Select All' then 'Continue' or choose a group of images to download together. The window will scroll, supports shift-selection, dragging a box around a group of images or CTRL + Click on a PC or CMD + Click on a Mac to select a discontiguous set of images
- Choose a size for the downloaded images from the drop down menu and click 'Proceed to final step'
- Click the link which appears. In the case of a large gallery, several links will appear. Click ALL of them
- The download will be to your default location for downloaded files and will be in the form of a zip file which will decompress when complete.
Nothing happening? Some browsers, notably Safari, will give very little indication that a file is downloading. Rest assured the download has initiated. The file will be quite large and this is a good time for a screen break.
Still not convinced? Try this small gallery of three images:
Click the link and enter the password: test
- In the single image view, you can use the left and right arrows on your keyboard to advance to the next image
- You can return to the gallery view by clicking the title of the gallery
- Selecting 'original file' where it is available as the download size will generally be faster as no additional processing of the image will need to take place before it is transmitted
Password not accepted? If you copied and pasted the password from an email, please retype it. Some email applications will add invisible data which will be picked up along with the visible text
Are you logged in? The password system covered in this post does not require you to be logged in or open one of our free accounts. If you have an existing account, please log out from the system via the 'My Account' tab in the top right of the screen. This will avoid any conflicts
Download speed - the system will send images as fast as you can receive them. An individual image should download almost instantly. Users in corporate environments may experience lower speeds due to local bandwidth and security restrictions but a good average speed for batch downloads is five to ten images per minute. Clients with high speed connections not being shared with other people should see download speeds of better than 20 images per minute
Can't find the downloaded file? - Search your computer for the file name of one of the images
'Right clicking' is not used anywhere in the system
Browser related issues. The system works with the current versions of all popular browsers but if you do have a problem you can't resolve it's always worth trying another browser available to you on your system
Still having problems? Please use the contact form or give us a call and we'll help
You will have received an invitation to view a gallery of images in the form of a link and password
- Click the link to enter the gallery
- Enter the password if prompted (passwords are always lower case and with no spaces)
- Please ignore any options to sign in or open an account, these are for a different client group. The password is all you will need
To download individual images:
- Moving your mouse around the gallery will trigger larger roll-overs for a better view and enable the full caption to be read
- Click on an image to open it in the full window
- Click 'Download' and choose a size for the image. This will immediately initiate the download to your default location for downloaded files
To download all of the images:
- Click 'Batch Download'
- Select Images to Download - Either 'Select All' then 'Continue' or choose a group of images to download together. The window will scroll, supports shift-selection, dragging a box around a group of images or CTRL + Click on a PC or CMD + Click on a Mac to select a discontiguous set of images
- Choose a size for the downloaded images from the drop down menu and click 'Proceed to final step'
- Click the link which appears. In the case of a large gallery, several links will appear. Click ALL of them
- The download will be to your default location for downloaded files and will be in the form of a zip file which will decompress when complete.
Nothing happening? Some browsers, notably Safari, will give very little indication that a file is downloading. Rest assured the download has initiated. The file will be quite large and this is a good time for a screen break.
Still not convinced? Try this small gallery of three images:
Click the link and enter the password: test
Power user tips:
- In the single image view, you can use the left and right arrows on your keyboard to advance to the next image
- You can return to the gallery view by clicking the title of the gallery
- Selecting 'original file' where it is available as the download size will generally be faster as no additional processing of the image will need to take place before it is transmitted
Having Problems?
Users will find the system to be highly reliable and simple to use, but occasionally problems can occur. The notes below will help to answer the most common questions:Password not accepted? If you copied and pasted the password from an email, please retype it. Some email applications will add invisible data which will be picked up along with the visible text
Are you logged in? The password system covered in this post does not require you to be logged in or open one of our free accounts. If you have an existing account, please log out from the system via the 'My Account' tab in the top right of the screen. This will avoid any conflicts
Download speed - the system will send images as fast as you can receive them. An individual image should download almost instantly. Users in corporate environments may experience lower speeds due to local bandwidth and security restrictions but a good average speed for batch downloads is five to ten images per minute. Clients with high speed connections not being shared with other people should see download speeds of better than 20 images per minute
Can't find the downloaded file? - Search your computer for the file name of one of the images
'Right clicking' is not used anywhere in the system
Browser related issues. The system works with the current versions of all popular browsers but if you do have a problem you can't resolve it's always worth trying another browser available to you on your system
Still having problems? Please use the contact form or give us a call and we'll help
Friday, 4 May 2012
We're Jammin'
"Nikon has received reports of a small number of D4/D800 users who are
experiencing 'locking up' issues, where the camera becomes unresponsive
until the battery is removed and reinserted. Nikon is investigating a
permanent solution to this issue and will update users accordingly. To immediately resolve this issue, users can turn off both the
'Highlights' and 'RGB Histogram' display options in the 'Playback
Display Options' sub-menu found in the 'Playback' menu" Source: Nikon USA
Flashback to the late 80s and the FM2's MD12. A photocall (remember those?) never passed without at least one photographer exclaiming loudly before wrenching the motordrive off the base of the camera to push down the protruding pin which had jammed in the 'up' position and rendered the camera inoperable.
A "small number" can be interpreted as all D4s and D800s shipped to date and given that the majority of those units will be in the hands of professional photographers who received their cameras first through the NPS priority scheme, it's a serious fault. Professional photographers are most likely to be working with both the highlights and RGB histogram displays turned on and pulling the battery to free the camera will of course mean the loss of any images in the buffer.
This speaks to a lack of adequate beta testing and in particular, testing by professionals in the field prior to release where this would have been quickly unearthed.
Flashback to the late 80s and the FM2's MD12. A photocall (remember those?) never passed without at least one photographer exclaiming loudly before wrenching the motordrive off the base of the camera to push down the protruding pin which had jammed in the 'up' position and rendered the camera inoperable.
A "small number" can be interpreted as all D4s and D800s shipped to date and given that the majority of those units will be in the hands of professional photographers who received their cameras first through the NPS priority scheme, it's a serious fault. Professional photographers are most likely to be working with both the highlights and RGB histogram displays turned on and pulling the battery to free the camera will of course mean the loss of any images in the buffer.
This speaks to a lack of adequate beta testing and in particular, testing by professionals in the field prior to release where this would have been quickly unearthed.
Monday, 30 April 2012
17 Inch Macbook Pro Speed Test
Since 2007 I've used Apple's flagship 17 inch Macbook Pro models as my primary computers both on location and back in the office. The move from a desktop plus laptop combination was made possible by Apple's switch to Intel processors and in particular the Core 2 Duo which made working with the D3s 14 bit NEF files a realistic proposition for a laptop. Shortly after, I moved to a cloud-based image storage and file delivery system with the result that wherever I could connect to the internet I was effectively in my office with everything at my fingertips and back at base I could connect my superb Eizo display for critical colour work and retouching. The Eizo is capable of displaying the full range of colour in the Adobe 1988 colour space commonly used in professional imaging and the publishing world.
Five years is the right time to replace a laptop and gain best value in terms of speed improvement versus cost though and I've now added the latest 2.4GHz Intel Core i7 version alongside my existing Core 2 Duo model.
Next to colour accuracy, nothing matters more in a photographer's workflow than speed and so I was interested to discover just how much faster the new computer would be. I devised a series of tests based around typical real-world procedures to compare the two Macs.
The contenders are a 2007 model Intel Core 2 Duo Macbook Pro 2.33 GHz, 3GB RAM (the maximum) OSX 10.6.8 and the new 2.4GHz i7 with 8GB RAM (again the maximum) and OSX 10.7.3.
PhotoMechanic 4.6.8:
Time to ingest 20 Nikon D3S 14 bit NEFs (288GB of data) using an ExpressCard 34 compactflash adapter, write an IPTC caption and rename the files:
Core 2 Duo - 18 sec
i7 - 12 sec
Copy the files to a new folder:
Core 2 Duo - 8 Sec
i7 - 2 Sec
Nikon Capture NX2 2.3.1:
Process 20 Nikon D3S 14 bit NEFs - convert to jpeg and save:
Core 2 Duo - 1 min 40 sec
i7 - 50 sec
Process 20 NEFs - apply noise reduction (better quality), USM and save as losslessly compressed NEF:
Core 2 Duo - 8 min 49 sec
i7 - 2 min 21 sec
It's clear that PhotoMechanic is gaining a useful increase in speed but NX2 is the real beneficiary of being free to run in 64 bit mode, with a tremendous boost in productivity. It's also interesting that the new Macbook is noisier than its predecessor, with the fan running gently at all times, causing a detectable exhaust flow from the left side of the case and the last batch test really causing it to pour on extra coal and immediately run the fan at full speed whereas the old model would have to be running at full capacity for an extended period before audibly fanning.
On the subject of colour accuracy, the display on the i7 as shipped was excessively blue, too contrasty and delivered over saturated colours. It looked wonderful, but it wasn't accurate. After profiling with my i1 Display Pro it was near identical to the earlier laptop which was always within a whisker of the profile it shipped with following calibration. The resolution of the i7 is 1920 x 1200 which is a nice resolution for a 24 inch display but leaves text too small to be read comfortably on the built in 17 inch display without risking eye strain. Fortunately, many applications, including browsers, now support zooming and CMD + scales the text up to a comfortable size.
Five years is the right time to replace a laptop and gain best value in terms of speed improvement versus cost though and I've now added the latest 2.4GHz Intel Core i7 version alongside my existing Core 2 Duo model.
Next to colour accuracy, nothing matters more in a photographer's workflow than speed and so I was interested to discover just how much faster the new computer would be. I devised a series of tests based around typical real-world procedures to compare the two Macs.
The contenders are a 2007 model Intel Core 2 Duo Macbook Pro 2.33 GHz, 3GB RAM (the maximum) OSX 10.6.8 and the new 2.4GHz i7 with 8GB RAM (again the maximum) and OSX 10.7.3.
PhotoMechanic 4.6.8:
Time to ingest 20 Nikon D3S 14 bit NEFs (288GB of data) using an ExpressCard 34 compactflash adapter, write an IPTC caption and rename the files:
Core 2 Duo - 18 sec
i7 - 12 sec
Copy the files to a new folder:
Core 2 Duo - 8 Sec
i7 - 2 Sec
Nikon Capture NX2 2.3.1:
Process 20 Nikon D3S 14 bit NEFs - convert to jpeg and save:
Core 2 Duo - 1 min 40 sec
i7 - 50 sec
Process 20 NEFs - apply noise reduction (better quality), USM and save as losslessly compressed NEF:
Core 2 Duo - 8 min 49 sec
i7 - 2 min 21 sec
It's clear that PhotoMechanic is gaining a useful increase in speed but NX2 is the real beneficiary of being free to run in 64 bit mode, with a tremendous boost in productivity. It's also interesting that the new Macbook is noisier than its predecessor, with the fan running gently at all times, causing a detectable exhaust flow from the left side of the case and the last batch test really causing it to pour on extra coal and immediately run the fan at full speed whereas the old model would have to be running at full capacity for an extended period before audibly fanning.
On the subject of colour accuracy, the display on the i7 as shipped was excessively blue, too contrasty and delivered over saturated colours. It looked wonderful, but it wasn't accurate. After profiling with my i1 Display Pro it was near identical to the earlier laptop which was always within a whisker of the profile it shipped with following calibration. The resolution of the i7 is 1920 x 1200 which is a nice resolution for a 24 inch display but leaves text too small to be read comfortably on the built in 17 inch display without risking eye strain. Fortunately, many applications, including browsers, now support zooming and CMD + scales the text up to a comfortable size.
Subscribe to:
Posts (Atom)