Nikon Df

[Tim Ernst]

I've got a deep, dark question for you DF/D4/D800 users and figured the talent and experience on this forum would shed some light to help. I'm working on a night photography project over the next six months and looking for a second camera specifically for high-res, high-ISO, pinpoint star photos. I've been using a d800e for the same types of images since it came out, and have had good results, but it is right at the edge of being too noisy, so I wanted to see if there was a better option at this point. I use a medium-format system for the rest of my work, so this camera would only be used for high-res, high-ISO, pinpoint star work with the one lens. I'm considering both the Df and Canon 6d. I could always just get a second d800e, but wanted to see if one of the others would work better for this job.

I need a camera that can shoot 15 seconds at ISO 6400 with the least amount of noise (I use 14-24mm Nikkor, usually at f2.8). I know both the Df and Canon 6d beat the d800e with lower noise at higher ISO, but I also need to make uprezzed prints in the 40x60 inch range. So the gist of my question is this - to make big prints (and high-res reproductions in books and calendars), am I better off starting with a native file with lower noise and lower resolution (Df/6d, more rezzing up required), or a native file with more noise but higher resolution (d800e, less rezzing up required)?

I've found reviews of the various cameras at high ISO to be of little use, so wanted to ping you guys to get your thoughts - any are most appreciated. Thanks!

Tim Ernst in Arkansas

 

[glenerrolrd]

Tim

I maybe missing something here ..but if you are shooting at 15sec and ISIO6400….the first step I would take is to increase my exposure and lower my ISO . The difference between 6400 and 1600 in any of these cameras is significant both in reduced dynamic range and desaturated color . I do realize that beyond a certain point long exposures can be problematic .

My feeling based on having a D4 and D800E (and now a Df) is that the D800E will produce a noticeably better large print even up to ISO6400 ….but I would be pushing hard to get the ISO down to at least 3200.

The D4 and Df sensor (and the in camera process) is tuned for reportage and sports where fine tone separation isn t much of a factor .

 

[Tim Ernst]

Thanks Roger - the problem is that an exposure longer than 15 seconds won't give me pinpoint stars, so that is the issue (despite what most other folks report - I need pinpoint stars at 100% view, and at longer exposures than 15 seconds with the 14mm lens the stars begin to trail, so I need 6400 at 15 seconds to keep them at pinpoints). High ISO in the daytime is different than high ISO star fields, where 90% or more of the image is basically dark shadow, and noise is most noticed. I do make star photos with the 800e at iso 3200 and 6400, but I was hoping for a better image than what the 800e can do at night, hence my quest to look into the Df/D4 and 6d sensors. If these sensors can't produce a cleaner file when rezzed up than the d800e, then I'll just get another one of those.

 

[segedi]

I'm not experienced in astrophotography, but wouldn't a tracking mount allow for longer exposures and solve most of the issues? Seems a better option than a lower megapixel camera.

 

[Tim Ernst]

Thanks Segedi, but a tracking mount will blur the landscape. I've got two of them, and use them for other types of star photos, but not when including the landscape (when I use a tracker, I use my medium-format system). I'll be shooting at ISO 6400 for 15 seconds with the 35mm camera, that's pretty much what I am locked into for these images, and am just looking for the best sensor for the job...

 

[glenerrolrd]

Thanks Segedi, but a tracking mount will blur the landscape. I've got two of them, and use them for other types of star photos, but not when including the landscape. I'll be shooting at ISO 6400 for 15 seconds, that's pretty much what I am locked into for these images, and am just looking for the best sensor for the job...

You have a tough requirement ..but I now understand . This is so specific I think you need a test ..maybe an order at Amazon would give you enough time to compare .

 

[AreBee]

Tim,

...the problem is that an exposure longer than 15 seconds won't give me pinpoint stars...

I was going to suggest that you take into account the portion of sky you are shooting in order to squeeze every last bit of shutter speed you can out of an exposure. For example, for a given time period, trails will be longer if FOV covers a portion of sky near the equator than it will near one of the poles. Unfortunately, given your use of the 14-24mm Nikkor, FOV is, um, astronomical?

What about deliberately capturing star trails so that ISO can be reduced, then undoing the former in post using something like this method?

 

[mjr]

Have you tried with noise reduction set to on with the D800E? I do quite a bit of star stuff and if i'm not stacking then there's no problem with the second dark exposure, it actually does quite a good job. I've found with my Zeiss 21 that I can't do longer than 8 seconds without getting star movement, the second 8 second exposure is not a problem for me. The other way is obviously 2 exposures, 1 on the tracking mount for the stars and a second for the foreground at much lower ISO and blend them.

This might sound strange but I have a Sony RX1 which is unbelievable at night, I get cleaner files from that than the D800 but then it's restricted to the 35mm lens.

Have you got any images to show what you are doing?

Cheers

Mat

 

[Tim Ernst]

Thanks Mat - yes, I do all of that, I'm just looking to see if I can get a better image out of the camera with a cleaner file to begin with and not have to do a lot of work in post other than what I already do. And just trying to figure out if the cleaner sensors would get the job done - sometimes it just needs to be a straight shot with no post work. I need 14mm, but glad to hear someone else believes that old "30 second" BS is only for web-size images, ha, ha!

Arebee - I'm not near the equator nor either pole. But it is true that you can shoot longer before trailing when pointing north than when pointing south (maybe that is what you meant). But I plan to be shooting in all directions, so I'm wanting to cover all bases.

Glenerrolrd - yes, I'm going to do the testing myself with all three cameras, but I thought I would be lazy and see if anyone here at the answer already! I'm also not very good with testing, and sometimes I think we see what we want to see - although since the cost is not an issue (800e & Df are about the same price, and the Canon with adapter is getting close to the same price), I don't really favor any of them at the moment, so I was trying to get someone with practical experience to sway me in one direction or the other.

I appreciate all of your comments!

 

[D&A]

I'm certainly not the one to have the last word (or even the 2nd or 3rd next to last word) nor profess to be be anything but an armchair amateur astro photographer, but I've dabbled in it on and off during the years with a wide variety of cameras. If you going to print that large (40 x 60') with an image that contains fine detail of either part of the landscape and/or pinpoint star points, I think you're going to have a somewhat difficult time up rezing a 16MP file and retain pinpoint detail without it becoming somewhat smeared by algorithms used in up-rezing to such a degree....as good and as clean as the D4/Df "out of camera" files are.

Excessive noise in a file of course has it's own issues especially if post processing noise reduction schemes are utilized but I believe the D800 (e) might be the better tool for this endeavor. Of course nothing substitutes for actually trying out all of the cameras your mentioned in a real world situation and that too would be my recommendation.

Dave (D&A)

 

[Shashin]

Thanks Segedi, but a tracking mount will blur the landscape. I've got two of them, and use them for other types of star photos, but not when including the landscape (when I use a tracker, I use my medium-format system). I'll be shooting at ISO 6400 for 15 seconds with the 35mm camera, that's pretty much what I am locked into for these images, and am just looking for the best sensor for the job...

Tim, a tracker like the Vixen Polarie and iOptron Skytracker have half sidereal rate for landscape work which allows you to double your exposure time. The idea is that it blurs the sky and landscape at the same time.

 

[AreBee]

Tim,

...it is true that you can shoot longer before trailing when pointing north than when pointing south (maybe that is what you meant).

It was.

Did you see the link I provided in my previous post, or is it just that it is of no use to you?

 

[Tim Ernst]

Will - my plan is to not have any blur in either the landscape or the stars, so I don't like using the 1/2 speed option (I have the newest skytracker which has that speed) - any blur shows in large print sizes. Tracking at normal speed and then blending with the foreground is better when I can do that, although a lot of times I would not have a view to the north to align the tracker (doesn't matter as much with the 14mm lens though, so I might give that a try). If that 1/2 speed deal really did work, then I would want to make the exposure time even longer and stop the lens down to f4 - it is a never-ending quest!

Rob - yes, I looked at the link and it does not sound like something that I want to mess with, and probably would not work for my stuff - but I have never seen that done before and it was an interesting read.

Dave - I know going from 16mp up that large is quite a stretch, but I was hoping someone with a D4 or Df had played around with doing something like that instead of dealing with the noise of the larger sensor and would proclaim that it would work. I have made some huge enlargements with just CS6 that held up really beyond what I could have imagined - the software is getting so good. But of course you have to have an exceptional file to begin with - which is what I was hoping the cleaner low-res file might help with, instead of starting with a file that had no noise but only needed smaller rez.

I have the 800e & 6d (waiting for an adapter so I can do a comparison), just wondering if I should also try the Df and compare all three - I would rather stick with Nikon if I could. Sounds like it would probably be a stretch though, but I'm a dreamer...

 

[D&A]

Tim,

I too have successfully make very substantially large sized prints from detailed subjects with 12-16mp cameras and it can be done with today's excellent post processing software. Astrophotography type images though are something quite different since many of the elements in the image are discrete pin point details and have to be reproduced that way after rezing up. I honestly believe you should rent either a Df or a D4 and test against the D800 for your specific application and compare the two directly. Its the only way to be certain wheich camera gets closer to meeting your objectives.

Dave (D&A)

 

[GrahamWelland]

Tim,

You might want to take a look at some of the testing that David Kingham has done comparing the D4 and D800 series (plus the rest) at his blog:
David Kingham Photography | Nikon Night Photography Shootout

Interestingly you'll see that he is also pretty keen on the 6D for night photography and he did a further comparison with that and the Nikons and ended up selecting the canon.
http://www.davidkinghamphotography.com/blog/2013/11/canon-6d-vs-nikon-d600-high-iso-night-photography

I haven't tried the Df for astro yet as it's cloudy season here in the NW although I'd expect it to perform similarly to the D4 vs D800/600.

 

[Tim Ernst]

Hey Graham, I've been getting info from David's site and enjoy his work. Although quite honestly, I find it difficult to interpret some of this findings due to his test images being out of focus sometimes (kind of basic for a test of this nature). I have a 6d in the bag due to his recommendation, but so far have not been too pleased with it. Which is why I wanted to see what the Df would do, since it is the D4 sensor and he was so high on it.

I've ordered a Df for delivery tomorrow (love holiday shopping deals!), so will hopefully get to take a first-hand look at all three sensors with a crispy-clear moonless night expected here.

THANKS everyone for your thoughts!

 

[Tim Ernst]

THANKS again to folks who commented on my question above in regards to uprezzing nighttime Df files. Just FYI if anyone might be interested, I spent much of last night on a mountaintop shooting pinpoint star photos with the d800e, Df, and Canon 6d, using the same 14-24 lens for all three. I didn't care too much for the user interface of the Df, especially the power on/off and ISO switches, and the fact the ISO is not shown in the viewfinder (I'm hoping that I just missed how to do this - surely there is a way to show the ISO in the viewfinder, right?) I was using all three cameras in the dark which made everything tougher, but I find that way is a good test of how well one person mates up with the controls of a particular camera, and the Df and I just didn't get along too well.

Will - I did try using the ioptron at 1/2 speed and found that even at 30 seconds the foreground had blurred some (naked trees about 100 feet away - great silhouettes against the starry sky - and easy to see if the detail is blurred). But I bet this could work for some scenes to help lower the ISO.

I shot all three cameras at ISO 6400, f2.8, and 15 seconds with high ISO dark-frame on; didn't do any RAW processing (beyond defaults for each) other than change white balance on all to the same point; then I uprezzed all files in CS6 to the same pixel width for a 40 x 60 print via bicubsmooth. I did not do any noise reduction.

I was a little surprised by the results of this limited test - while I could not see the differences from very far away, it was obvious up close to me that the Df produced the best final file, with less noise and more details. The 800e was a close second, and the 6d was the worst on both noise and detail (I kind of liked shooting long exposures with the Canon since it allows immediate use after the end of an exposure while it does its dark-slide thing in the background - wish Nikons would do this!). So it appears to me with this specific type of shooting situation that it is possible to get a better final high-res file with a lower-res, cleaner file to begin with. I may have to use the Df as my primary star camera with the 800e as backup…

Happy trails!

Tim Ernst in Arkansas

 

[GrahamWelland]

Tim

How did you find the live view focus with star points with the Df vs the Canon? I know that my D800 & D600 both had horrible LV noise compared to the relative clarity of the Canons.

I'm sure you'll quickly get used to the different UI of the Df compared to the D800 but it certainly is different.

 

[Tim Ernst]

Hey Graham, I've never been able to get live view to work for stars on the Nikons either, and since I never use live view for any of my normal shooting anyway, I didn't even bother to mess with it on the Canon. I generally shoot a couple of test frames until I can see pinpoint stars when zoomed in on the LCD quite a bit, then tape the lens focus down. On my 14-24 infinity is a ways off from the infinity mark. So it is correct that there is no ISO visible in the viewfinder of the Df? That is going to drive me nuts, and it is not a very long drive, ha, ha! The rest of the stuff I'll learn with use, although the 800e seems a lot more intuitive...

 

[GrahamWelland]

Tim

You should see the ISO display in the VF at the bottom right - at least while you turn or adjust ISO or other exposure values. Obviously you can also display it on the rear LCD too with the info mode selected.

Btw, if you check option d3 and you should have the ISO value displayed instead of the exposure remaining count.