Nikon d5100

18 replies [Last post]
lowlight
Offline
Joined: Dec 6 2010

This may be a silly newb question. The new Nikon d5100 shoots 1920 x 1,080 video, so is it correct to assume that this automatically means 16:9 aspect ratio? I can't see that ratio mentioned in the specs. For those interested in the camera...

http://www.europe-nikon.com/en_GB/product/digital-cameras/slr/consumer/d5100

It has an interesting 'night vision' function where ISO is boosted to 104,000 (equivalent), articulated LCD panel, 3.5mm mic jack, continous autofocus etc. Worth a look.

Richard Payne
Offline
Joined: Sep 15 2000

Yes all standard HD is 16:9.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Correct, all HD is 16:9.

Any camera having 'night vision' is responding to infra-red rather than visible light. IR response can be a problem in a normal camera because it produces a video response to non-visible light. That can pollute some saturated colours, depending on the type of surface radiating it (so a dark blue cloth can appear to be red or even green). It can also make the black level unstable, if the IR pollutes the RGB sensors unequally. All this makes the footage a nightmare to grade properly, and that's why the cheaper cameras aren't allowed for high-end TV and film production.

Normally, there's an optical filter in the camera to absorb IR and stop it getting to the sensor. This is apart from the other optical filter which prevents excessively high image frequencies getting to the sensors, the stuff that causes spatial aliases. If the IR filter can be swung out of the optical path, then you've got an IR camera. In the best of them, this is actually a filter wheel with two filters in it, one to stop IR and pass visible light, the other to pass IR and stop visible light.

These filters can be a bit expensive, and they seem to be just about the first component to be left out when the camera cost is being reduced. Beware.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

lowlight
Offline
Joined: Dec 6 2010

Richard and Alan, many thanks for your confirmation about 16:9. Regarding the 'night vision' function, I have a Sony Handycam with nightshot so aware of the IR filter, but I believe the Nikon D5100 does not rely on removing an IR filter. Although Nikon calls the mode 'night vision', this is perhaps a little misleading. There is no IR illumination provided with the camera and, as I understand it, they simply maximise the potential sensitivity of the sensor to visible light frequencies and boost it to an equivalent of just over 100,000 ISO, similar to the Nikon D3s. It's black and white images only, and I don't know how effective it is until I see some stills/video examples. In complete darkness it would be as blind as any other camera. I like the idea of being able to use every available photon flying around, and it's likely sensors will be developed for mainstream cameras that are even more sensitive than this one. Serious generation 3 night vision scopes costs thousands of pounds so this is an interesting alternative avenue. If I am wrong about this night vision mode I would be happy to be corrected, as if IR is a factor I may be able to improve the night vision capability with IR illuminators.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

You don't need the camera to have an IR source for it to respond to IR. There's a lot of IR out there anyway. Just winding up the internal gain in the camera will raise the noise levels immensely, and is probably not what they're doing.

Solar energy reaching the earth's surface comprises roughly 5% UV, 40% visible, 55% heat as IR and longer wavelength radiation. Unless there's a filter in the camera to separate visible from IR, then all you have is gain controls, which will always produce high levels of noise. Manufacturers often don't talk about the internal workings of cameras at all, so you have to devise tests to see how it works, and that's what I do.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

lowlight
Offline
Joined: Dec 6 2010

Thanks Alan, that's intriguing because if Nikon are utilising IR in some way in this particular 'night vision' mode, it opens up possibilites. I should mention the reason I am interested in this particular field is nocturnal wildlife filming. Incidentally, off thread but noticed your location and have fond memories of drinking at the Well House and White Hart, not far from you, when a student back in the 70s at Epsom School of Art.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Yes, indeed. I'm in the house nearest to The Well. 4.5 minutes across the field, 5 minutes back (uphill).

Incidentally, you can do your own IR experiments with cameras: get a consumer remote control, one that you know actually works. Point it into the camera lens, and press a button. If you see the led flashing in the viewfinder/output/saved picture, then the camera is responding to IR. All the remote controls I've measured emit 'light' at 933nm, which is well into IR and is invisible.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

col lamb
Offline
Joined: Jan 2 2010

Yet another DSLR capable of HD video but probably not to broardcast standard.

Here is a review link of the still's aspect of the camera

http://www.dpreview.com/previews/nikond5100/

As with most DSLR's the limitations will probably be the same as Canon 550D, 7D etc models.

If you are contemplating getting one for its HD capabilities please do some research, there are a lot of HD DSLR posts on this forum.

I am perfectly happy with the HD video from my 7D but and it is a big but, DSLR's are definately not point and shoot with respect to HD video and buying the camera is not the only expense......add a mike (if indeed you have a socket to plug one in on the 5100) and a method of actually seeing the LCD screen in daylight (Zacuto ZFinder at c£200).

That said I love my 7D

Col Lamb Lancashire UK ASUS P6X58D-E MOBO, 3.3GHz hex core i7 CPU, 12GB RAM, nVidia GTX580 GPU, W7 64bit, 500Gb boot, 1Tb RAID (Mirror) Store, 500Gb RAID (stripped), Edius 6.05, CS 5.5

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

For what it's worth, nobody is asking for formal tests on DSLRs these days. That probably means that nobody thinks it's worth doing because they'll all fail. Not that I ever pass/fail cameras, I just measure and report.

But, I tested a Nikon D3 a while back, and it made lots of lovely high-level spatial aliasing. Horizontally, it was about the same as the Canon 5D, and vertically it was about the same as the 5D horizontally. I.e. it's symmetrical, which the Canon's aren't. Still rubbish at video though.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

lowlight
Offline
Joined: Dec 6 2010

Thanks Alan and col. Ideally I would like a camera that can work with the highest possible IR nm rating, 900nm is about the minimum. I will be keeping a close eye on D5100 reviews so thanks for that dpeview preview link col.

Yes, you are absolutely right about the importance of thorough research before selecting a DSLR for HD video, and I’ve certainly spent countless hours on forums and manufacturers websites doing just that. My head is full of aliasing, moiré, rolling shutter, frame rates, bitrates, video AF capability, manual control options, audio jacks, lenses etc, and it’s all fascinating stuff.

I recognise camcorders will nearly always out-perform a DSLR for video in the real-world practicality and image stakes, particularly for documentary style shooting where you can’t plan each shot. The temptation of HD DSLRs is you can, in the right circumstances, get a (superficially at least) fantastic movie image along with proper stills capability, all in a small package at reasonable cost. I also rate the importance of good audio very highly because you can ruin good video with a poor soundtrack so 3.55mm jack essential. If I was going 100% for ‘broadcastable’ video I would choose a camcorder, and haven’t ruled that out - the Sony HXR-NX70 is looking interesting for me right now (admittedly not necessarily a broadcast spec camcera). However, there is a pleasing new D5100 video on youtube that shows what the camera can do. Usefully, the editor has included subtitles detailing technical information about lenses and camera settings used on each shot. All done at 1/50 as far as I can see. Nicely done too.

d5100 film Winter's End by Ernie Kohl

http://www.youtube.com/watch?v=kh7uueMmLEA

col lamb
Offline
Joined: Jan 2 2010

Despite Alan's testing results I am more than happy with the optical results on my 7D, any good cameraman can make allowances for the limitations of their kit in their shooting.

Test is out yourself, take an SD or compact flash card to your local dealer and produce some video on it/them, take the card home and see the results for yourself on your PC or MAC.

I had seen video from a friends Canon 5D mark 2 projected onto an 8' x 6' screen, with massess of movement, fog, sharp close ups, great detail and colour rendering, fantastic focus control, whats more with no aliasing and no moire effects visible. I then tested out a 7D and bought one.

To each their own, I will always have a dedicated camcorder (Panny 151 and Sony TG3) as well as the 7D DSLR so I have the best of all worlds, plus Edius 6.2 which takes any video I throw at it with ease.

I did a video of some badgers whilst on hols last year, it was pretty dark and increasing the ISO rating and using my f2.8 L lens I got some great video on the 7D, my Sony TG3 was dire and produced just noise. Got plenty of great images too without even letting go of the camera.

Good luck

Col Lamb Lancashire UK ASUS P6X58D-E MOBO, 3.3GHz hex core i7 CPU, 12GB RAM, nVidia GTX580 GPU, W7 64bit, 500Gb boot, 1Tb RAID (Mirror) Store, 500Gb RAID (stripped), Edius 6.05, CS 5.5

lowlight
Offline
Joined: Dec 6 2010

Thanks col, that's encouraging- horses for courses really. I will just wait until the NAB show is over as there may be some more cameras announced.

jgould
Offline
Joined: Mar 28 2005

You could always buy my Sony A55 advertised on here

lowlight
Offline
Joined: Dec 6 2010
jgould wrote:
You could always buy my Sony A55 advertised on here

Thanks, just looked in classifieds, very nice and presumably v. good value with those lenses. However, I need that 100,000+ ISO speed offered on the D5100, or an alternative camera that can match it. I've been waiting for an announcement on the Sony A77 (assuming that's what it will be called) as it could have high ISOs too, but might not hear until July.

Chris.
Offline
Joined: Nov 5 2000

Alan, curious to see you talking about shooting video and testing a Nikon D3. I didn't think the D3 has a video recording function?

On the subject of the D5100 - afaik all of Canon and Nikon's DSLRs have IR filters. These can be removed by companies like ACS for people who want to get into infra-red photography otherwise they filter out as much IR as they can.

The 'night vision' that is done by cheap consumer camcorders relies on the camera being sensitive to infra-red.

The 'night vision' that Nikon are talking about has nothing especially to do with infra-red - just as you can take a photo at night by using a long shutter speed you can 'see' at night if you use high ISO and a short shutter speed.

Noise is usually a problem at high ISO but the latest generation of DSLRs has it well under control. My Canon 5D-II is better at 1600 ISO than my 1D-II is at 400.

My Canon 5D II does video on a night brilliantly. The first time I tried it out at night I was amazed that I could clearly see the clouds in the sky. I only did it on auto mode as hadn't had the camera long. As a test went to shoot some Xmas lights (which weren't on!) but was amazed at how the camera coped driving along the dark streets (test clip here)

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Pretty sure it was a D3, but it was about a year ago. Shot 720p.

All silicon cameras are sensitive to infra-red, it;s a property of silicon. Some more so, some less, depending on the doping. But all sensors respond to IR unless there's a proper filter in the way.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

MAGLINK
Offline
Joined: Mar 8 2007
Alan Roberts wrote:
Pretty sure it was a D3, but it was about a year ago. Shot 720p.

All silicon cameras are sensitive to infra-red, it;s a property of silicon. Some more so, some less, depending on the doping. But all sensors respond to IR unless there's a proper filter in the way.

It would be a D300S as that is the updated D300 that does video (very badly) the D3 does not do video to my knowledge.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Fair enough. It belonged to a professional DoP who used it as a second camera on HD shoots. he was rather disappointed when I showed him what it did.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

lowlight
Offline
Joined: Dec 6 2010

Hope the night vision thing isn't getting too boring, but imaging resource have reviewed the camera and they mention the mode. The reviewer had to have some light from a lamp (although very little, it was in another room) for the camera to capture an image. Still, I would check if IR illuminators could improve things.

http://www.imaging-resource.com/PRODS/D5100/D5100A.HTM

'Night vision is an obvious choice for the photographer dad wanting to get photos of his children when they're at their most wonderful, which is of course when they're sleeping. It's surprising that they didn't emulate the classic green glow of a night vision scope, but it also shows good judgment, because in the end, we all appreciate a black and white photograph. I used it in low light earlier in the week, but gave it more of a test in real "night vision" light. No joy. So I turned on the bathroom light and it was dim enough that I could barely see well enough to frame the image in the viewfinder. The Nikon D5100 didn't autofocus in this mode, so I had to focus on something in better light using another mode at about the same distance, then reframe my shot. It's pretty darn grainy, but it looks decent.'