Nikon D800 hacked with 50Mbit/s high bitrate video option

[Michael1]

But don't forget that canon and nikon have different ISO value, cause of the different dynamic range.

A 5d at ISO 1250 is like a d800 at ISO 640.

Look!  ;)

 

Exactly.  I have never understood these goofy tests comparing ISO from two or more different cameras.  If the person who did this video can't see the difference in luminance between these two cameras at the same ISO, then they need to buy a waveform monitor!  Obviously, it has to be corrected for these arbitrary ISOs.

 

Michael

[Michael1]

I've always liked the cinematic look of the D800 and D600.  I wouldn't say Blackmagic cinematic, but very nice, nevertheless.  Personally, I think Nikon is sitting on a gold mine, if they would ever put their efforts in that direction.  Obviously, the sensor and some of the processing is already there.  Now they just need to finish the job.

 

Michael

[leafcutter]

I have two old Nikon 3100s so I patched one yesterday to 56mb and did some side by side shooting in the garden - just foliage with a bit of wind, flowers, kids running around, various toys and balls being thrown etc.  I didn't approach it rigorously in any way, I just put the cameras down on a chair, hit record, and then made some stuff happen in front of the lens.

 

The file sizes of the 56mbs camera were 2.5 times as big as the standard files so I think I patched it correctly but I have to say I really struggled to see any appreciable visible difference between the two.  For two and a half times the data cost I was (possibly mistakenly) expecting to be able to see a difference as soon as I hit play but in the shots I took it didn't appear to be there.

 

I had to unpatch the camera as I need it for something else this week but I am curious to do some more testing to see what is going on.  I am thinking about setting up an indoor test that I can repeat easily so I can put the two cameras in exactly the same place each time, rather than running them side by side (as this means the footage doesn't exactly overlap).  

 

Based on that one test I'm struggling to see why it would make sense to patch a camera.  Ok maybe there are certain circumstances where there is a big difference (and/or maybe I did something wrong), but if you don't get a big pickup in general shooting conditions then in general it doesn't seem worthwhile.  

[maxotics]

Based on that one test I'm struggling to see why it would make sense to patch a camera.  Ok maybe there are certain circumstances where there is a big difference (and/or maybe I did something wrong), but if you don't get a big pickup in general shooting conditions then in general it doesn't seem worthwhile.  

 

In photography there is mostly one kind of compression, how to use the least amount of data to represent the same STILL image visually.  The Nikon patch does not effect that part of the compression, so you will not see any difference in the image.  When you look at a RAW image on your screen, compared to a JPG that you compressed and is 1/10th the size, can you see much of a difference?  No. Image quality is a function of the quality of the display/print.  

 

In video, there is a second kind/part of compression.  How to use the least amount of data to represent the same MOVING image.  Again, a lot of data can be thrown out when the image moves before we notice it.  So in video compression, another "pass" of compression is applied to the images, taking detail out when the image changes a lot.  So taking a video comparison of a subject that doesn't move would not show a difference in the "still" image compression, which the patch doesn't change.

 

The hack reduces the amount of compression that happens between images.  That is, when the image changes drastically between frames, it throws out less data and you see more detail.  If you look on the web for other comparisons you'll see things like people shooting rain.  With less compression in those sequences the rain drops will look more distinct.  

 

So try a test where there is movement.   However, I don't believe the patch worth it in most cases.  I did it more out of curiosity.

[leafcutter]

 

So try a test where there is movement.   However, I don't believe the patch worth it in most cases.  I did it more out of curiosity.

 

My understanding of what is going on is pretty much as you outlined above.  The test I did had movement - little things like leaves right up to big things like humans but none of it really stood out as 'better'.  I'll report back when I have another go at it.

[Nikkor]

It will look better as soon as you touch the colors ,contrast and/or sharpening , the macro blocks pop right in your face.

Also the shadows look cleaner so you can underexpose and use flat profiles.

[maxotics]

It will look better as soon as you touch the colors ,contrast and/or sharpening , the macro blocks pop right in your face.

Also the shadows look cleaner so you can underexpose and use flat profiles.

 

Ah.  Didn't realize that.  Thanks!

[Nikkor]

Ah.  Didn't realize that.  Thanks!

 

Like this tipical shot with grass where you want to put in some contrast with a curve and the macroblocks rape your eyes.

[nazdar]

Lol, higher bitrates help to keep tiny and quickly moving details to fall apart and makes huge macrobloks smaler. For testing it turn your camera to the runnig stream or water with zillions of small waves and you will see how is your bitrate keeping details. Then when you touch curves and color correction you will see how it keep together. Moving grass, branches with lots leaves in strong wind. Fur, hair, detailed structured surfaces etc