Nikon V1 - shooting 4K 60fps raw for $200

[1tkman]

One thought is to use your Aperture software to color correct and then convert Raw files to Tiffs, then QuickTime Pro ($30-40 upgrade that might also be found free on an operating system installation disk) could import these images and generate or combine the Tiffs into small sequences/clips. iMovie could then be used to edit the assortment of clips.

It's the slow motion component that's new to me and I would imagine Final Cut has options that would benefit. Don't forget most programs have 30 day trial periods :)

[1tkman]

It does 30 fps for 1 second or 60 fps for half a second.

I am confused on settings, should I set it on 10,30, or 60?

[screwbiedooo]

There are at least four advantages to shooting this way, even if your rendered film will be 1080p:

• 12 Bits color depth (as opposed to 8bit color of standard DSLR video codecs) are a quantum leap in image quality, giving you 48 times more color information per pixel. This results in a much improved dynamic range, radically improved possibilities for grading, changing white balance in post and, if necessary, fixing exposure. In 8bit codecs, on the other hand, dynamic range is limited to 8 f-stops or, when a greater dynamic range gets squeezed into 8bits, color banding.
• zero compression artefacts/codec degrading in the footage.
• using the extra resolution to crop/zoom into the image in post. This also drastically improves the results of software stabilization.
• oversampling and scaling to 1080p gives an ultra-detailed image where every single pixel is sharp. (Sometimes even too sharp...)

Ok thanks for the explanation. I can see the benefits for cropping and image stabilisation. But isn't it just an awful lot more work, for a moderate gain in quality? What I don't understand is why we are not already hitting peak quality with large sensor cameras such as the GH1 for example which is already downres-ing its footage in a ratio of 5:1 to output 1080. Shouldn't those pixels all be amazing already or is there some flaw in firmware processing?

[cantsin]

One thought is to use your Aperture software to color correct and then convert Raw files to Tiffs, then QuickTime Pro ($30-40 upgrade that might also be found free on an operating system installation disk) could import these images and generate or combine the Tiffs into small sequences/clips. iMovie could then be used to edit the assortment of clips.

 

Don't use iMovie. It uses a very weak, old, 8bit/4:2:0 codec for all its projects, Apple Intermediate Codec (AIC). You'll throw away most - most likely: all - of the advantages you get from the Nikon 1/raw workflow if you edit in iMovie (or a similar entry level video editing programs).

[cantsin]

I am confused on settings, should I set it on 10,30, or 60?

 

That depends on the frame rate of your project, and whether you want (a brief moment of) slow motion, normal speed or accelerated speed. (But, without wanting to be impolite to you, if you are new to these things, then you really shouldn't make movies with the Nikon 1 raw "hack", but use the standard video recording function of your camera. What we're discussing here is borderline esoteric geek stuff for quality freaks and experimentalists.)

[cantsin]

But isn't it just an awful lot more work, for a moderate gain in quality? 

 

Yes, it is. It makes absolutely no sense if you are a mainstream (or even a professional) videomaker. It's similar to cooking some esoteric high end food whose preparation takes 10 hours instead of fixing a normal meal in half an hour. It's practically unusable for documentary or narrative filmmaking. But if you are an experimental filmmaker, an animator, or someone who makes very short but strongly visual films (which could also be an advertising clip or a music video), or maybe even someone who still dabbles with analog 16mm filmmaking, hand-developing and esoteric lab processes, then this can be more satisfying than dealing with the limitations of standard video.

 

What I don't understand is why we are not already hitting peak quality with large sensor cameras such as the GH1 for example which is already downres-ing its footage in a ratio of 5:1 to output 1080. Shouldn't those pixels all be amazing already or is there some flaw in firmware processing?

 

As regards to oversampling, the downscaling algorithms in cameras are a big problem. They actually decrease image quality because they need to work in real time on the relatively slow camera hardware. It's not just the in-camera downscaling that is inferior btw., but also the in-camera debayering of the raw sensor data into an RGB signal, and of course the in-camera encoding into 8bit MPEG-4 video (all of which a GH1 needs to do within 1/30 of a second).

 

Think of the camera chip as the worst quality, quickest-and-dirtiest raw converter that exists, particularly in video mode. Conversely, you pay with a lot of processing time on your computer that will be spent when you shoot raw and develop each frame with some raw converter software.

[screwbiedooo]

Aha thanks a lot. In tat case is the tele crop mode on the GH2 the zenith of hd video quality (no rescaling)?

[cantsin]

Aha thanks a lot. In tat case is the tele crop mode on the GH2 the zenith of hd video quality (no rescaling)?

 

No - because you get an in-camera-debayered/color-rendered, in-camera-noise filtered 8bit signal with this mode as well. Compared to shooting raw with the Nikon 1, you don't get the quality gains from oversampling in post.

[1tkman]

That depends on the frame rate of your project, and whether you want (a brief moment of) slow motion, normal speed or accelerated speed. (But, without wanting to be impolite to you, if you are new to these things, then you really shouldn't make movies with the Nikon 1 raw "hack", but use the standard video recording function of your camera. What we're discussing here is borderline esoteric geek stuff for quality freaks and experimentalists.)

I see that 30 is the magic number, the longer shooting time at 10 left me puzzled. V2 is on the way and I am curious to see what differences my glass can possibly produce with these cameras.

Has any one produced anything yet with resolve? I do consider this software a superior method for color correcting.

[YouFirstEye]

Nikon CX sensor (used in Nikon 1 v1) size comparison.

 

http://***URL removed***/reviews/nikonv1j1

[screwbiedooo]

No - because you get an in-camera-debayered/color-rendered, in-camera-noise filtered 8bit signal with this mode as well. Compared to shooting raw with the Nikon 1, you don't get the quality gains from oversampling in post.

 

Ok. Anyway, it's all pretty much by the by, cos as far as I can see the 1080 quality coming from the hacked panasonics is pretty well good enough for 99% of us. I am just fascinated by this discussion as I am considering getting hold of a V1 for the 60fps in order to capture fast action stills into water. I see it as a fairly limited camera, except for this one great function (and the AF).

[1tkman]

Nikon CX sensor (used in Nikon 1 v1) size comparison.

 

http://***URL removed***/reviews/nikonv1j1

Thanks for posting this link. How do you interpret this information regarding the sensor size?

[YouFirstEye]

Nikon decided to use Nikon CX sensors for the Nikon 1 series. The sensor is lager than those in camcorders or super 16mm film. But It is small than micro 4/3, APS-C and full-frame.

http://library.creativecow.net/articles/solorio_marco/hdslrs-for-video/DSLR-Sensor-Sizes.jpg

That is a giving constraint of Nikon 1 v1. For smaller sensor such as Nikon CX, two immediate concerns are low-light performance and wider depth of field (DOF). Nikon 1 v1 is 10.1 MP camera which is not too high for the sensor size having a reasonable low-light performance (Nikon 1 v2 is 14.2 MP camera.) These concerns are addressed by Nikon with faster lens such as Nikon 1 18.5mm f/1.8 len and Nikon 1 32mm f/1.2. However, I would like to see a telephoto len at f/2.8 or faster. To achieve narrow DOF, the len needs to be close to the subject at its largest aperture.

[1tkman]

Nikon decided to use Nikon CX sensors for the Nikon 1 series. The sensor is lager than those in camcorders or super 16mm film. But It is small than micro 4/3, APS-C and full-frame.

http://library.creativecow.net/articles/solorio_marco/hdslrs-for-video/DSLR-Sensor-Sizes.jpg

That is a giving constraint of Nikon 1 v1. For smaller sensor such as Nikon CX, two immediate concerns are low-light performance and wider depth of field (DOF). Nikon 1 v1 is 10.1 MP camera which is not too high for the sensor size having a reasonable low-light performance (Nikon 1 v2 is 14.2 MP camera.) These concerns are addressed by Nikon with faster lens such as Nikon 1 18.5mm f/1.8 len and Nikon 1 32mm f/1.2. However, I would like to see a telephoto len at f/2.8 or faster. To achieve narrow DOF, the len needs to be close to the subject at its largest aperture.

Would the larger number of pixels in the v2 be a concern with regard to noise?

[YouFirstEye]

_{I don't have a V2 to compare with V1. Someone here mention that V2 has more noise than V1. This may be true but not sure by how much. To overcome this, use fast len and lights. Go for lower ISO (i.e., no more than 800) if possible. Also perform some noise reduction in post.}

[kenuck]

Would the larger number of pixels in the v2 be a concern with regard to noise?

 

That was my concern too when I got my V2 but according to DXOmark testing the low-light performance is paradoxically slightly better than the V1, although overall scores for the V2 were lower than the V1 (based on poorer scores for color depth and dynamic range).  Nikon must have worked some magic to get improved low light scores - it's a pity they didn't apply those improvements to the 10mp sensor. See V2 vs V1 comparison:

http://www.dxomark.com/index.php/Publications/DxOMark-Reviews/Nikon-1-V2-review-Evolution-stasis-or-regression/Comparisons

 

As YouFirstEye mentions, I always try to keep it under ISO 800

[Andrew Reid]

DXOMark is weird.

 

I see a significant decrease in low light performance with the V2 due to the 14MP res over 10MP. Really is nosier than the V1 at high ISOs but the extra stuff compensates for it.

[1tkman]

DXOMark is weird.

 

I see a significant decrease in low light performance with the V2 due to the 14MP res over 10MP. Really is nosier than the V1 at high ISOs but the extra stuff compensates for it.

 

I appreciate your observations Mr. Reid!

Have you had any luck with filters? I would imagine you capture some real artwork with large glass and customized settings.

[Alexander Wetzel]

https://vimeo.com/6553361 if you don't count you don't care

[YouFirstEye]

Trying out my new Nikon 1 18mm f/1.8 prime len.

 

http://vimeo.com/63494373

 

https://vimeo.com/63538616