Nikon D90 astromod VS. Nikon DF unmodified

It is fact that Nikon’s DF is among the most sensitive camera’s available on the market today. Its FX format CMOS chip offers 16.2 million pixels. The corresponding pixel size of 7.3μm is thus large compared to most other state-of-the-art cameras (with typical sizes of less than 5μm). As a result the Nikon DF has much better low-light, high-ISO performance.

However, as all unmodified cameras also DF’s CMOS detector is covered by an infrared (IR) blocking filter. This is unsatisfactory for astrophotography, in particular when imaging nearby star-forming regions. The reason is that young, massive stars emit hard UV radiation that leads to the ionization of the surrounding hydrogen. Subsequent recombination of free electrons with ions then produce strong emission lines such as the Hα line at approx. 656nm (in the red part of the spectrum). This wavelength unfortunately is already blocked by the IR filter found in almost all digital single-lens reflex (DSLR) cameras.

For that reason, some companies such as DSLR Astro Tec in Germany recently have specialized on modifying DSLRs. Different modifications exist, the one for astrophotography is basically a replacement of the IR-blocking filter with a clear-glass filter. This modification drastically increases the sensitivity of the camera at the wavelength of the Hα emission line. This modification comes at the cost of the camera’s white-balance, which then needs to be set manually. However, for astrophotography this doesn’t play a role anyway.

Since I own an unmodified Nikon DF and a modified Nikon D90, I was wondering how these two cameras would compare to each other when imaging star forming regions such as e.g. M8, the Lagoon nebula. In order to perform the test, I have used my Nikkor AF-S VR 200-400mm 1:4 lens, operated at 400mm f/4 and took images of the nebula using both cameras. In both setups the exposure time was set to 30 seconds at ISO 800. The result is shown below. Both images were taken in raw format and only brightness and contrast were adjusted in the same way. The result makes clear that an astro-modified D90 clearly outperforms even Nikon’s low-light market leader, the Nikon DF.

Nikon D90 astromod vs Nikon DF unmodified

Testing Nikon TC-17E II and TC-20E III with
Nikkor AF-S 70-200 1:2.8 ED VR and
Nikkor 200-400 1:4 ED VR

For more than one year I am now carrying Nikon’s 2x teleconverter TC-20E III in my camera bag. I bought it from a local store in good used condition, with intent to get more reach with my Nikon D300 (which is APS-C sized) and the Nikon AF-S 70-200mm f/2.8 VR lens. Since this lens is very fast and its image quality superb, the 2x teleconverter would still allow for high shutter speeds at f/5.6 on bright summer days when doing wildlife, e.g. bird photography.

So far the theory, but after taking my first shots with the 2x converter attached to Nikon’s 70-200mm f/2.8 VR, I was really disappointed with the results. Images taken at the widest aperture through the TC are of poor quality and very smooth, not sharp at all. Stopping down improves the quality, but still not to a level I would be satisfied with.

Now comes the surprise! Just recently, I got hold of a very nice and sharp Nikkor AF-S 200-400mm f/4 ED VR lens, which came together with the teleconverter TC-17E II, both in very good used condition. When using the 1.7x teleconverter on that lens for the first time, I was really “shocked”, because the image quality was only slightly degraded and very sharp. Next, I attached the 2x teleconverter TC-20E III to the Nikkor AF-S 200-400mm f/4 ED VR as well and was likewise astonished by the image quality, which was still good and reasonably sharp.

Remark: Teleconverters and Autofocus performance

Autofocus is getting much slower with the TCs attached. However, although the D300 is not explicitly mentioned on Nikon’s TC compatibility chart, apparently the camera supports f/8 autofocus and the Nikkor AF-S 200-400mm f/4 ED VR will autofocus when either of the TCs under consideration is attached.

TC Image Quality Comparison using SpyderLensCal

In order to make a fair comparison, I decided to setup a typical lens calibration session with SpyderLensCal (the distance was 5m, so that enough focus path was left on both lenses). That way, I would get a fair image quality comparison of the lenses and the TCs, and would at the same time calibrate all my camera+lens+TC combinations. Both, SpyderLensCal and my camera were mounted on a tripod. Shots were made with my D300 using different AF finetuning settings. Vibration Reduction (VR) was turned off, ISO set to 200 and the largest aperture was chosen using aperture priority mode. The resulting shutter speed was always faster than 1/500s. SpyderLensCal and my D300 were brought onto the same optical axis through leveling SpyderLensCal with its integrated bullseye bubble level and the camera using the hot-shoe to level with a common level meter.

Results

The distance between the camera chip and the calibration device was always 5m, but since the focal length changed with each lens+TC combination, I decided to scale down each frame to a focal length of 200mm and then make equal crops around SpyderLensCal’s ruler and save a JPG file. That way, all images can be compared on a pixel-by-pixel basis and more easily displayed here. However, down-scaling and cropping does not have any effect on the results and all images shown below are very good representations of the true RAW images I have taken.

Nikkor 70-200mm f/2.8 ED VR @ 200mm f/2.8

Nikon_70-200_VR_05_AFp10

AF +10

Nikon_70-200_VR_04_AFp05

AF +5

Nikon_70-200_VR_03_AF000

AF 0

Nikon_70-200_VR_02_AFm05

AF -5

Nikon_70-200_VR_01_AFm10

AF -10

This basic setup of camera and lens gives already good results, even without AF finetuning. However, slight frontfocus can be identified and an AF correction of +5 seems to give sharpest results.

Nikkor 70-200mm f/2.8 ED VR + TC-17E II @ 340mm f/4.8

Nikon_70-200_17x_05_AFp10

AF +10

Nikon_70-200_17x_04_AFp05

AF +5

Nikon_70-200_17x_03_AF000

AF 0

Nikon_70-200_17x_02_AFm05

AF -5

Nikon_70-200_17x_01_AFm10

AF -10

With the 1.7x teleconverter attached, the image quality decreases and it seems that the frontfocus issue is getting worse than without TC. Moreover, the overall smoothness makes it hard to find the best solution. However, AF finetuning of +10 gives good results.

Nikkor 70-200mm f/2.8 ED VR + TC-20E III @ 400mm f/5.6

Nikon_70-200_20x_05_AFp10

AF +10

Nikon_70-200_20x_04_AFp05

AF +5

Nikon_70-200_20x_03_AF000

AF 0

Nikon_70-200_20x_02_AFm05

AF -5

Nikon_70-200_20x_01_AFm10

AF -10

With the 2.0x teleconverter attached, the image quality decreases quite drastically and strong frontfocus can be identified. The tendency of how AF finetuning changes the results is clearly seen in the above images. The total focal plane shift is so large that my final best result is found with a AF finetuning value of +20, which is not shown above. However, in the real world I would not consider using this combination since the image quality is very poor.

Nikkor 200-400mm f/4 ED VR @ 400mm f/4

Nikon_200-400_VR_04_AFp10

AF +10

Nikon_200-400_VR_03_AFp05

AF +5

Nikon_200-400_VR_02_AF000

AF 0

Nikon_200-400_VR_01_AFm05

AF -5

This lens is really great and extremely sharp out of the box. However, also here a slight correction for frontfocus, i.e. an AF finetuning value of +3 was found to give best results.

Nikkor 200-400mm f/4 ED VR + TC-17E II @ 680mm f/6.8

Nikon_200-400_17_04_AFp10

AF +10

Nikon_200-400_17_03_AFp05

AF +5

Nikon_200-400_17_02_AF000

AF 0

Nikon_200-400_17_01_AFm05

AF -5

In contrast to the poor performance of the TC-17E II in combination with the Nikkor 70-200mm f/2.8 VR lens, the image quality here is reasonably good, in particular after applying an AF finetuning value of +7.

Nikkor 200-400mm f/4 ED VR + TC-20E III @ 800mm f/8

Nikon_200-400_20_04_AFp10

AF +10

Nikon_200-400_20_03_AFp05

AF +5

Nikon_200-400_20_02_AF000

AF 0

Nikon_200-400_20_01_AFm05

AF -5

In contrast to the extremely bad performance of the TC-20E III in combination with the Nikkor 70-200mm f/2.8 VR lens, the image quality here is still reasonably good, in particular after applying an AF finetuning value of +7.

Conclusion

Teleconverters decrease AF speed, in particular in low-light, low-contrast situations. However, when using a Nikon body which allows to autofocus at f/8, AF is still working considerably well even with the Nikkor AF-S VR 200-400mm f/4 lens. The image quality of teleconverters can drastically change when using different lenses. In the case presented here, either of the two teleconverters, TC-17E II and TC-20E III performed very bad on the Nikkor AF-S VR 70-200mm f/2.8 lens, i.e. producing very smooth images. On the other side, when attaching to the Nikkor AF-S VR 200-400mm f/4 lens, the image quality was only slightly decreased (in particular the 1.7x converter performs very well) with images that are reasonably sharp. However, the loss of light is then significant and such combinations presumably only work in environments that provide a sufficient amount of light.
 

 

Nikkor Lens Comparison for Astrophotography
AF 80-200 f2.8 ED vs. AF-S VR 70-200 f2.8 ED

I bought two used Nikon lenses, both very similar in their specifications, which is not surprising as the Nikkor AF 80-200 f2.8 ED is a precursor of the Nikkor AF-S VR 70-200 f2.8 ED (see Ken Rockwell’s history page). With the highest bid, I got the 80-200mm for 280 EUR + 20 EUR shipping from the big bay and for the 70-200mm a private seller in Sweden claimed 8200 SEK, equivalent to 870 EUR. Then I asked myself if it is really worth spending 570 EUR more on the new lens, especially when doing astrophotography, where vibration reduction (VR) and fast autofocus is not needed.

To answer this question, a simple startest was performed. Both lenses were mounted to a Nikon DF (FX format chip: 36mm x 24mm). The camera was then fixed on a tripod without any further star tracking. The scenery was a strongly light-polluted sky. At ISO 1600 an exposure time of 4 seconds was chosen and the camera was pointed towards NE. The startest was performed using an aperture of 2.8 and 4.0. The resulting images are shown further below for reference. The cutouts (center, top right, top left) are 100%, when the images are viewed in their original size.

Conclusion

This Nikkor AF 80-200mm f/2.8 ED I got from Ebay is bad for astrophotography, no it’s terrible! The star images at focal lengths of more than 80mm are frustrating, producing very strong coma effects more or less all over the image and even stopping down to f/4 does not make a big difference. Interestingly at 80mm the quality is OK and the coma disappears in most parts of the image. I am really surprised by this result, because the lens operates really good under daylight conditions when sufficient light is available. This is proven by the testshot below, which is taken at 200mm f/8 and a shutter speed of 1/800 at ISO 800.
On the other side, the Nikkor AF-S 70-200mm f/2.8 VR is superb. It shows only little coma over all focal lengths, with slightly better results when stopped down to f/4.0. Note that the reason for elongated stars is due to the rotation of the Earth during the 4 second exposure and not necessarily coma.

Although really bad for astrophotography, the Nikkor AF 80-200mm f/2.8 ED is really good under daylight conditions; here a shot at 200mm f/8 with a shutter speed of 1/800s and ISO 800. The image shown is a cutout without further processing of the full frame image.

Although really bad for astrophotography, the Nikkor AF 80-200mm f/2.8 ED is really good under daylight conditions; here a shot at 200mm f/8 with a shutter speed of 1/800s and ISO 800. The image shown is a cutout without further processing of the full frame image.

Startest at f/2.8

Startest of Nikkor AF 80-200mm f2.8 ED

Testing Nikkor AF 80-200mm f2.8 ED at 80mm f2.8

Startest of  Nikkor AF-S VR 70-200mm f2.8 ED

Testing Nikkor AF-S VR 70-200mm f2.8 ED at 70mm f2.8

Startest of Nikkor AF 80-200mm f2.8 ED

Testing Nikkor AF 80-200mm f2.8 ED at 135mm f2.8

Startest of  Nikkor AF-S VR 70-200mm f2.8 ED

Testing Nikkor AF-S VR 70-200mm f2.8 ED at 135mm f2.8

Startest of Nikkor AF 80-200mm f2.8 ED

Testing Nikkor AF 80-200mm f2.8 ED at 200mm f2.8

Startest of  Nikkor AF-S VR 70-200mm f2.8 ED

Testing Nikkor AF-S VR 70-200mm f2.8 ED at 200mm f2.8

Startest at f/4.0

Startest of Nikkor AF 80-200mm f2.8 ED

Testing Nikkor AF 80-200mm f2.8 ED at 80mm f4.0

Startest of  Nikkor AF-S VR 70-200mm f2.8 ED

Testing Nikkor AF-S VR 70-200mm f2.8 ED at 70mm f4.0

Startest of Nikkor AF 80-200mm f2.8 ED

Testing Nikkor AF 80-200mm f2.8 ED at 135mm f4.0

Startest of  Nikkor AF-S VR 70-200mm f2.8 ED

Testing Nikkor AF-S VR 70-200mm f2.8 ED at 135mm f4.0

Startest of Nikkor AF 80-200mm f2.8 ED

Testing Nikkor AF 80-200mm f2.8 ED at 200mm f4.0

Startest of  Nikkor AF-S VR 70-200mm f2.8 ED

Testing Nikkor AF-S VR 70-200mm f2.8 ED at 200mm f4.0

Measuring Linearity, Readout Noise and Gain of a ATIK383L+ CCD camera

Today I did some tests with my 2-year old ATIK383L+ CCD camera. I am using the camera regularly and you can find some results at my astrophotography page. The camera uses the well-known KAF-8300 17.6mmx13.52m CCD chip with a resolution of 3362×2504 pixels and a pixel size of 5.40µmx5.40µm. The manufacturer promotes it as a camera with very low read-noise of 7electrons, great linearity and with ideal Gaussian shaped bias frames.

So let’s see if that is true for my model.

Linearity

This is the easiest of the performed tests. Thereby the camera was mounted on the focuser of my telescope and the aperture was illuminated with my Flatfield panel. Then several images were taken with increasing exposure times as seen in the table below. Subsequently, the so gained flatfields were corrected for bias and darkcurrent contributions. The resulting mean counts in each individual image were tabulated and plotted against the exposure times showing the linearity of the CCD.

Exp
Time [s]
ADU Poisson
Error [%]
Fit 1
Error [%]
Fit2
Error [%]
Fit3
Error [%]
Fit4
Error [%]
1 650 3,92 135,0    
2 1222 2,86 61,7    
3 1598 2,50 51,8    
4 2108 2,18 36,4    
5 2611 1,96 27,4    
6 3106 1,79 21,5    
8 4111 1,56 13,7 15,8  
10 5098 1,40 9,3 11,3  
14 7073 1,19 4,2 6,1  
18 9045 1,05 1,4 3,2 7,3
22 11009 0,95 0,4 1,4 4,6
28 13911 0,85 1,8 0,0 2,3 4,1
36 17771 0,75 2,9 1,2 0,5 1,8
44 21635 0,68 3,6 1,9 0,7 0,2
52 25353 0,63 3,5 1,8 1,0 0,2
60 29083 0,59 3,5 1,9 1,3 0,6
70 33737 0,54 3,5 1,8 1,5 1,0
80 38357 0,51 3,4 1,7 1,5 1,1
90 42887 0,48 3,1 1,5 1,4 1,1
95 45068 0,47 2,8 1,1 1,1 0,9
100 47214 0,46 2,5 0,8 0,8 0,6
105 49310 0,45 2,1 0,4 0,4 0,2
110 51431 0,44 1,7 0,0 0,2 0,0
115 53484 0,43 1,3 0,4 0,2 0,4
120 55549 0,42 0,9 0,8 0,6 0,7
125 57491 0,42 0,4 1,4 1,1 1,2
130 59456 0,41 0,1 1,9 1,6
150 61486 0,40 11,4      
<error> [%]   14,70 2,68 1,56 0,95
ATIK383L+ linearity test

ATIK383L+ linearity measurement. Linear regression was performed iteratively narrowing down the datarange. Best photometric performance was found in the range above 15000 ADU and below 55000 ADU.

A Linear regression was performed iteratively narrowing down the count range (ADU) in order to find typical errors introduced due to non-linearity of the CCD. A fit to all data points (red) gives a poor result, as expected, with a typical error of roughly 15 percent (see last row in table above). When limiting the range to values above 1000 ADU and below 60000 ADU a correlation coefficient R of 0.9996 (orange data; note that the value in the plot is R2) is found, which is still worse than the number given by the manufacturer (0.9998). Thus, I definitely cannot recommend doing photometry within the full range of 1000 to 64000 ADU as suggested on the ATIK website. Nevertheless, the linearity seems to be good within a range of 10000 to 60000 ADU (yellow). In that range typical errors are below 2 percent only. The best performance is found between 15000 ADU and 55000 ADU, which is the range I would consider as the photometric one.

Thus, highest photometric precision is possible for countrates above 15000 ADU and below 55000 ADU.

Gain Measurement

There are several methods available to measure the gain. I have chosen the method described by Michael Richmond. So the following steps were performed:
1) a pair of L-flats was taken with varying exposure times (2,4,8,16,32,64 sec)
2) a set of 3 dark frames with the same exposure time was taken
3) individual darks were combined using the average value (masterdarks)
4) masterdarks were subtracted from the appropriate flats
5) sum of each pair of flats was calculated and devided by 2: sum*.fits
6) difference of each pair of flats was calculated: diff*.fits
7) gain is the inverse of the slope of the plot: mean vs. variance, where the variance is: RMS2/2.

IMAGE Mean   IMAGE RMS Variance
sum2.fits 1524 diff2.fits 89.0 3957.8
sum4.fits 2747 diff4.fits 114.1 6509.4
sum8.fits 5516 diff8.fits 159.5 12720.1
sum16.fits 10975 diff16.fits 218.5 23871.1
sum32.fits 21841 diff32.fits 306.0 46818.0
sum64.fits 43342 diff64.fits 426.4 90908.5
ATIK383L+ gain measurement

ATIK383L+ gain measurement. A linear regression (green curve) was conducted on the measurements of the mean response in ADU vs. the variance. The gain is then the inverse of the slope.

The gain is then the inverse of the slope when doing a linear regression of the mean vs. variance plot.
A gain of 0.48 e/ADU was found for the ATIK383L+.

Readout-Noise

In order to measure the readout-noise of the CCD camera, bias frames taken throughout several hours of observations were used. The readnoise was measured following the procedure described here:
1) 9 bias frames were taken and combined using the median (masterbias)
2) the masterbias was subtracted from each individual bias: Rdnoise*.fits
3) readnoise is the standard deviation in these images (see table)
4) finally the masterbias was analysed using the histogram and a Fast Fourier Transform (FFT)

IMAGE NPIX MEAN STDDEV MIN MAX
Rdnoise1.fits 8529394 0,6629 19,56 -124,5 1885
Rdnoise2.fits 8529394 3,382 19,84 -123,5 1046
Rdnoise3.fits 8529394 -2,212 19,49 -262 644
Rdnoise4.fits 8529394 -1,562 19,57 -129 5063
RdnoiseB1.fits 8529394 -0,4971 20,17 -118 9360
RdnoiseB2.fits 8529394 -0,7299 19,89 -122 518
RdnoiseB3.fits 8529394 0,8872 19,95 -163 426
RdnoiseB4.fits 8529394 -0,467 19,89 -114 523
RdnoiseB5.fits 8529394 1,106 19,95 -116 709

I noticed that the bias-level can vary slightly (order of 20 ADU) during several hours of observations. Therefore, it is the best to keep track on the bias level and take some bias once an hour. For this test, I had to create two individual masterbias frames in order to extract the correct readout-noise. The result is not affected by that.

The average value of the readout-noise for my ATIK383L+ is: 19.8 ADU. Using the gain calculated before, a readnoise of 10 electrons was measured.

Thus, the level of the readout-noise is indeed very low with only 10 e. However, it is not as low as promoted by the manufacturer.

In a next step, the quality and homogeneity of the masterbias was examined using the histogram and a FFT (see below).

Histogram of the masterbias

Histogram of the masterbias showing a Gaussian-like and thus random noise dominated distribution

FFT of masterbias

Masterbias image (left) and FFT of the masterbias (right). A column of an increased bias level can be easily identified in the image. Additionally, the FFT indicates the existence of a non-random large-scale noise pattern.

An examination of the historgram reveals that the bias is of course not an ideal Gaussian as promoted by the manufacturer. However, it is Gaussian-like and thus noise dominated. Furthermore, a column of increased bias level can be identified in the image and after performing a Fourier Transform, it can be seen that some non-random pattern exists. This is indicated by the “cross” in the middle of the FFT image. However, the relative value of the pattern is very low.

I would conclude that the camera’s bias indeed is noise-dominated and very smooth, certainly allowing for faint details to be detected.

Finally the performed measurements show that the ATIK383L+ is a CCD camera that can be used for scientific photometric applications when limiting the linearity range as suggested. Due to its smooth bias it is furthermore an excellent tool for high quality astrophotography.