Nikon D4 Firmware Update
- Thread starter last_SHIFT
- Start date
OlegKalyan
Well-known member
squig, how much of a difference you see between the cameras? in persent, roughly?
D4 still attracts me mostly for upbelievable resolution, color depth in CX mode.. D800e possibly is the sharpest, but when you consider reducing noise in high iso, if needed, even at 1600, it all balances out.. (D5200 is a great little cam, no doubt, but for resolution only nothing beats GH2) .. actually have to take it back, tested, now D4 looks like it resolves more in CX mode.. another good news!
D4 still attracts me mostly for upbelievable resolution, color depth in CX mode.. D800e possibly is the sharpest, but when you consider reducing noise in high iso, if needed, even at 1600, it all balances out.. (D5200 is a great little cam, no doubt, but for resolution only nothing beats GH2) .. actually have to take it back, tested, now D4 looks like it resolves more in CX mode.. another good news!
Last edited:
squig
Veteran
Well I'll have a guess at resolved lines given that the 5D MKIII resolves 700 lines and the D800 is measured at 780 lines and my D5200 out-resolves the 5D.
My guess is-
D800e/VAF 780 lines
D5200 750 lines
D4 700 lines (may be less)
For resolution: C100 1075 lines :2vrolijk_08:
OlegKalyan
Well-known member
Thank you, C100, something to think about, if not getting F5, for a little more money 
where is the test measurements for C100 1075 lines, please!? D4 in CX mode seem to resolve much more than FX, maybe close to C100?
where is the test measurements for C100 1075 lines, please!? D4 in CX mode seem to resolve much more than FX, maybe close to C100?
squig
Veteran
Thank you, C100, something to think about, if not getting F5, for a little more money
where is the test measurements for C100 1075 lines, please!? D4 in CX mode seem to resolve much more than FX, maybe close to C100?
C300 is 932 lines measured in this test http://alfonsoparra-development.s3.amazonaws.com/uploads/docx/file/158/eosc300_test_ev.pdf
Not sure if I saw 1075 lines in an independent test or whether that was Canons documentation.
CX mode is probably close but the 2.7x crop with an F mount is completely useless to me.
Cosimo Bullo
Veteran
Squig, this test between the GH2 and D800 has me really wondering about the sharpness of the D800. Just looks madly soft in comparison. The tester is using what I still consider to be the best of the best of the driftwood patches, Sedna.
https://vimeo.com/63263001
https://vimeo.com/63263001
squig
Veteran
That's a very interesting test. I transcoded the file to ProRes 442HQ and graded and post sharpened it, nothing scientific given that it was a 17Mbps compressed file but it's still quite revealing. Firstly the sharpening settings on the 2 cameras weren't evenly matched. The GH2 has edge halos (look at the branches in the last shot) which looks like a Nikon or Canon would look at +2 sharpness (which is too much for my liking and screams video).
I think the D800 with +1 sharpening and the GH2 with -3 sharpening would be very close in appearance. The edge halos get much worse when you try to grade and post sharpen however some of the busy GH2 shots looked really good graded and slightly post sharpened. Secondly you're comparing a 145Mbps codec to a 24Mbps codec, an externally recorded D800 ProRes file would even the playing field. Thirdly they're different lenses, you can't accurately judge sharpness with different lenses.
Overall I'm confident I'd get a more filmic image from the D800 with a VAF filter and a Ninja, before now I thought the D800 only had an edge in dynamic range but looking at that footage I think with some effort it can match the GH2 in resolution in the final output image.
I think the D800 with +1 sharpening and the GH2 with -3 sharpening would be very close in appearance. The edge halos get much worse when you try to grade and post sharpen however some of the busy GH2 shots looked really good graded and slightly post sharpened. Secondly you're comparing a 145Mbps codec to a 24Mbps codec, an externally recorded D800 ProRes file would even the playing field. Thirdly they're different lenses, you can't accurately judge sharpness with different lenses.
Overall I'm confident I'd get a more filmic image from the D800 with a VAF filter and a Ninja, before now I thought the D800 only had an edge in dynamic range but looking at that footage I think with some effort it can match the GH2 in resolution in the final output image.
Last edited:
To be honest I think that the dislike of the 2.7x Crop is overblown. For one thing if you imagine using an 18mm lens you're getting about a 50mm image from it. For many scenes that would be a perfectly usable look. A Tokina 11-16mm would still be usable even tho it won't be wide. Since it's not the ONLY crop factor it's only an option and a good one for many angles of view up to telephoto. Just not a wide angle look, which IMO isn't used as much as people make it sound. I know from my work that wide angles are not the majority of the shots in any one project I do. If wide angle is needed you do have that FX option on the D4 or you could just have a D5200 on hand for your wide shots.
OlegKalyan
Well-known member
PAVP agree with you, when you think of 2.7 crop, there is no such a big difference with X2 crop on GH3 for example, or X2.2 on BMCC.
Full frame can be used as well, for "background out of focus, good latitude, skin tone, closeups, faces" shots. I have Sigma 10-20 3.5 lens, on wide side it makes 27 mm, which is wide enough.
If full frame, lots of stuff out of focus aestetics is the goal, yes, D800e for price/quality is unbeatable. for wide shots with low iso, GH2 is great, to me it looks sharper than C300, C100.
Full frame can be used as well, for "background out of focus, good latitude, skin tone, closeups, faces" shots. I have Sigma 10-20 3.5 lens, on wide side it makes 27 mm, which is wide enough.
If full frame, lots of stuff out of focus aestetics is the goal, yes, D800e for price/quality is unbeatable. for wide shots with low iso, GH2 is great, to me it looks sharper than C300, C100.
YES. For me the D800, D5200, D7100 and D4 are all good cameras offering a different look, so it's not so much an issue of what's best, but more of a tool for a job. You can get the look you want in different situations from come cameras and other cams can't give you that look. Since the GH2/3 gives a deeper depth of field, you can really get GREAT wide shots with lots of detail. Other times in a scene with a beautiful woman you might want a softer look with a really Shallow DOF. A Nikon might be a better camera for that look. I do believe tho that the Nikons are above the detail threshold that I personally have. Any camera I use has to at least reach a certain detail level which is north of the 5Dmk3.
OlegKalyan
Well-known member
Did some more testing, pretty confident now, D4 in FX mode looks very similar to D5200 in terms of resolution, with definite advantage in lesser noise , it's much cleaner.
D800e with oplf, by Mosaic, had an edge by about probably 75-100 lines of resolution, the noise amount seems pretty equal to D5200.
D800e with oplf, by Mosaic, had an edge by about probably 75-100 lines of resolution, the noise amount seems pretty equal to D5200.
last_SHIFT
Active member
Hi Oleg,
Thanks for all your tests and feedback!
I was wondering on the D4 at what ISO do you start seeing significant noise? I love my D800 - but noise can creep up pretty fast.
Here are some high detail wide angle frame grabs from the GH1 vs D4 after the latest firmware upgrade. The D4 is with the Nikon 24-70mm f2.8 at 24mm. The GH1 is with the event hack and has the Olympus 12mm f2.0. I shot in aperture priority with equivalent apertures (D4 at f8 and the GH1 at f4; the D4 at f11.0 and the GH1 at f5.6). Focus point is the knob o the branch in the center. Unfortunately this was hand-held and somewhat windy so some of the softness is motion blur. But to my eye, the GH1 footage has more resolution at equivalent apertures.
D4 vs GH1
D4 vs GH1
The wind died down so I went out again with my tripod to shoot a high detail scene. For the D4, I used the 50mm f1.8G at its sharpest aperture (f5.6). For the GH1, I used the Leica 25mm f1.4 at its sharpest aperture (f2.8). Fortunately, the 50mm at f5.6 on a full frame sensor is optically equivalent to the 25mm at f2.8 on a micro 4/3rds sensor. So under those settings, the footage has the exact same depth of field. Here are the frame grabs:
The GH1 has significantly more resolution than the D4. To make sure this wasn't a fluke, I repeated the test at the next set of equivalent apertures (the D4 50mm f8.0 and the GH1 25mm f4.0):
And one final test at the next set of equivalent apertures (the D4 50mm f11.0 and the GH1 25mm f5.6):
Judge for yourself, but I think the D4 footage has significantly less resolution than the modestly hacked GH1.
D4 vs GH1
The wind died down so I went out again with my tripod to shoot a high detail scene. For the D4, I used the 50mm f1.8G at its sharpest aperture (f5.6). For the GH1, I used the Leica 25mm f1.4 at its sharpest aperture (f2.8). Fortunately, the 50mm at f5.6 on a full frame sensor is optically equivalent to the 25mm at f2.8 on a micro 4/3rds sensor. So under those settings, the footage has the exact same depth of field. Here are the frame grabs:
The GH1 has significantly more resolution than the D4. To make sure this wasn't a fluke, I repeated the test at the next set of equivalent apertures (the D4 50mm f8.0 and the GH1 25mm f4.0):
And one final test at the next set of equivalent apertures (the D4 50mm f11.0 and the GH1 25mm f5.6):
Judge for yourself, but I think the D4 footage has significantly less resolution than the modestly hacked GH1.
Here are two stills of the same scene captured by the same cameras at the same time and then cropped to 16:9 and downsized to 1920x1080 pixels by the same editing software. From this you can see the D4 (50mm f5.6) and the GH1 (25mm f2.8) are both able to capture stills with a full 1920x1080 resolution.
Danielvilliers
Veteran
I think until we test some uncompressed hdmi out it will be difficult to make final judgement. In the first photo we can clearly see that the depth of field is different on the two shots. But one thing looks a bit weird for me and I think it could be a codec issue, if you look at the branches in the front the resolution are quite close but the finer details are completely blurred in the branches further back. Someone, one of the codec expert, said somewhere that the Nikon codec did not have something (Forgot what it is), I a thinking if it not blurring/mudding finer details to keep it in the 24 mbit region, like the mud in the gh1. As we have seen between gh1 and gh2, not every 24 mbit codec are created the same.
The image pairs have the same depth of field. For example, a 50mm lens at f5.6 on the D4 is optically equivalent to a 25mm lens at f2.8 on the GH1. You can see that better by comparing the stills. The loss of background detail in the D4 footage is, I suspect, due to insufficient bit rate in the codec. The loss of foreground detail in the D4 looks more like line skipping to me.
It looks like dvxuser.com downsized the frames I uploaded, here are the full 1920x1080 video frames:
http://www.flickr.com/photos/noirist/sets/72157633183796903/
It looks like dvxuser.com downsized the frames I uploaded, here are the full 1920x1080 video frames:
http://www.flickr.com/photos/noirist/sets/72157633183796903/
Last edited: