High ISO test. Nikon Dƒ vs. Nikon D700

My D700 is 5 years old. It was the first D700 in Austin TX off the truck and out the door. I've shot well over half-million images on this one D700 alone (I had the shutter replaced right around the 500,000 mark). It's time give the ol' girl a rest and bring in a new workhorse and I chose the Dƒ as my D700's successor. I've had all of the other small-body Nikon FX cameras, but I didn't think any of them bested the D700 which was why I stuck with it. Like a lot of people I was hoping that a D4 in a D800 body would come out, but that's not going to happen and the Dƒ just happens to fit in with what I'm looking for in a camera for the most part. No, it's not perfect, but it's the closest thing to a D700 that Nikon is going to release, for awhile at least, and I have no interest in lugging around a D4 (nor spending $6K on one).

Anyway, I don't want to get into a detailed review on the Dƒ just yet. I'm still working on getting familiar with it and I want to get a few more shoots under my belt with it. I had a little time on my hands tonight so I decided to do a quick and dirty high ISO shoot out between the new Nikon Dƒ and the old standby for low-light shooting the D700.

I love my D700 and it has stood the test of time. It's been my go-to camera for low-light concert photography as well as sports and just about everything else. Suffice it to say I've always been impressed with the D700's performance when the light gets low and the ISO's are cranked. But I just had to see how the Dƒ stacked up against the D700 when it came to ISO.

I shot each image at f/8 using a Sigma 35mm f/1.4 | A using a tripod and remote release. The Dƒ images are on the left and the D700 on the right. The images were shot using fine JPG with NR and sharpening off. They were imported into Lightroom 5 and the default LR5 color noise reduction and sharpening was applied. I would have preferred to use RAW, but as of yet there's no RAW support.

The images from each respective camera are pretty much equal all the way up to ISO 1600. One thing you will notice is that the Dƒ resolves fine detail better than the D700 due to the slight increase in resolution so it looks a little sharper and the Dƒ also has a much more substantial increase in dynamic range showing more detail in highlight areas where the D700 has little. The Dƒ also has a more accurate White Balance, both were set to Auto.

Alright, so here we go. Take a look at ISO 1600. Pretty much dead even here.

Once we crank it up to ISO 3200 the D700 starts to break up a little but the Dƒ holds it together a little better. 

At ISO 6400 things take a big downturn for the D700. I'll have to say that ISO 3200 was usually my limit for the D700 unless I was in a pinch. But as you see at ISO 6400 the Dƒ is still holding more fine detail and the noise is less "chunky". ISO 6400 on the Dƒ is easily usable. 

Pushing it up to ISO 12800 isn't something I'd ever really thought about doing. But the Dƒ still holds it together very well even at this high ISO. The noise in the shadow areas is extremely well controlled in the Dƒ while the shadow areas are a mess in the D700. In the highlight areas the detail is still very fine in the Dƒ while the D700 is pretty messy. To be fair, ISO 128000 is not in the D700's native ISO so I wouldn't expect it to perform that well. I mean it's not bad compared to a lot of other cameras. 

I didn't bother running the Dƒ up to it's H settings. ISO 12800 is higher than anyone really needs in my opinion and the H settings aren't really optimal so I just stopped the comparison there. 

So what are my conclusions? I think the Nikon Dƒ is nearly two stops better than the D700 in ISO performance. The Dƒ controls shadow noise better and holds quite a lot of fine detail even at ISO 128000. Let's take a look at a couple of direct comparisons. The Dƒ at ISO 12800 and the D700 at ISO 3200.

The Dƒ noise is noticeably chunkier, but the fine detail is still there. Keep in mind that these two images are two stops apart. That means you can make images with the Dƒ in 4X less light than the D700 and still have relatively comparable output. That's quite impressive indeed. 

Just for fun let's throw the D5300 into the mix as well. First here's a look at the D5300 at ISO 12800. Pretty much unusable. Noise is out of control and detail is obliterated. 

Taking a look below you can see at ISO 1600 the D5300 clearly shows finer detail, but the results are pretty similar. The Dƒ is almost THREE stops better than the D5300! Of course this isn't a fair comparison. Th Dƒ costs $2000 more than D5300, it better perform better. The D5300 is a great camera for it's price point and can do HD video whereas the Dƒ can't, so there's that. 

The Dƒ isn't a direct replacement for the D700 for many reasons. Build, battery, lack of vertical grip, inferior AF system etc... But if you need the best IQ in low-light and don't want to spring for the D4 this is the camera. 

Just remember, this is pixel peeping. In the real world the differences aren't quite as noticeable. 

Later this week I'll get to a full review on the Dƒ. It's a quirky camera, but this little test has proven to me that the Dƒ is indeed the low-light king in Nikon small-body FX cameras. 

Leica M8 vs. M9 ISO comparison

Before I bought my M9 I spent quite a bit of time researching, trying to find out how much I stood to gain in buying an M9 instead of an M8. I found plenty of talk on the forums about how the M9 was 1-stop better than the M8. I found that statement a lot, but there wasn't much in the way of empirical proof. There were images all over the place but I couldn't find a direct comparison. Trying to compare images of two different cameras at the same ISO settings of different subjects in different lighting situations really wasn't helping. So basically all I had to go on was word of mouth.

I was actually very surprised that nobody had done a side-by-side comparison of these two cameras. As more and more people are moving over to the Leica system because M8's and M9's are becoming relatively affordable I decided to do a direct side-by-side of the M8 and M9 so that people like me, who are on a relative budget can have some good data to look to when trying to decide whether the M9 is worth the price premium over the M8. Right now a used M8 is about $2000, an M9 is about $4000, and the M9-P (exactly the same guts as an M9) is about $5800.

Before I get into the test, there were some anomalies between the M8 and M9 that I noticed that had to be adjusted for. First of all, the M9 underexposed by one stop consistently. I ran the full test 3 times and finally set the M9 to +1EV to get a good exposure. Also, both the M8 and M9 were set to Tungsten WB, but the colors were vastly different. I attribute this to the stronger UV/IR that I had on the M8 lens. I shot in DNG and set the white balance on the white square on the color card in the test scene. The results are similar but not an exact match.

To keep a similar field of view I used a Leica 50mm f/1.4 Summilux lens on the M9 and a Voigtländer 35mm f/1.4 Nokton on the M8 (46.6 equiv.) Both cameras were set to Aperture priority and the lenses were stopped down to f/8. The images are 100% view screenshots from Lightroom 5. I used the LR5 defaults for sharpness and noise reduction of 25% sharpness and 25% color NR. (I also ran a set with no adjustments, which I can post if anyone asks).

The M9 images are on the left and the M8 on the right. Click on the images to view full size.

As you can plainly see the M9 does indeed give about a one stop advantage. At ISO 1250 the M9 is easily usable whereas the M8 is borderline. At ISO 2500 the M9 could work in a pinch, but the M8 in completely unusable. Even after trying a number of different noise reduction techniques I could not get a usable image out of the M8 at ISO 2500. 

That being said, I regularly use the M8 at ISO 1250 when shooting concerts and I've had good results. It seems that in high contrast situations like concerts where you have extremely bright and extremely dark tones that the shadows can be toned down to make the noise a little less noticeable. I haven't had much practice shooting the M9 in these conditions and I'll report back later on that. 

As far as daylight shooting goes the M8 and M9 are pretty much equal up until 640 where the M9 edges out the M8, but not by much. If you're a daylight shooter and not big into cropping get a 28mm f/2 or a 35mm f/1.4 (or both) and keep it down to ISO 640 and you'll be fine. If you shoot at night a lot that extra ISO stop of the M9 may make a difference. 

The M9 does have some advantages for the price difference. It is a full-frame camera and has more pixels to play with if you like to crop, and the menu system has many refinements, such as easier exposure compensation settings, manual entry for non-coded lenses, easier ISO setting in 1/3 steps. Full-frame also allows you to go wider since you have no 1.3X crop. 

So is the M9 worth twice price of the M8? I'd say on the whole, no. If the M9 were only $1000 more than the M8 I'd say get an M9 over the M8, but at double the price I'd have absolutely no problem buying two M8's and running with a 28/35 set up. 

If you know you have your heart set on full-frame, then don't mess around. Just get the M9. But if you can live with a slight crop or you want to get into the Leica system without spending more than a high-end DSLR kit will cost you the M8 is a really good camera. I really like the M8 paired with the Voigtländer 35mm f/1.4 Nokton. This is a great normal starter kit for less than $2500. 

Obviously, this is just my opinion. In the end I knew I wanted a full-frame camera, I had the money, and I got a great deal on a used M9-P, but, I'm also keeping my M8. It makes a good companion to my M9 with the lenses I have.

You can check out my work with the cameras on my flickr site. Obviously I've been working with the M8 longer so there are much more photos than with the M9.
Flickr Leica M8 set
Flickr Leica M9-P set