5D MarkIII worth the additional money for astrophotography?

<p>While aware that I've already started a topic about choosing between the MK3 and MK2 when I was already settled on the MK2 prior to the announcement, I'm still at a crossroads in terms of making up my mind.</p>

<p>My current setup is a Canon 40D with the following lenses:</p>

<ul>

<li>Tamron 17-50 f/2.8</li>

<li>Carl Zeiss Jena 135mm f/3.5 (M42)</li>

<li>Takumar 200mm f/4 (M42)</li>

<li>Pentacon 300mm f/4 (M42)</li>

<li>Pentacon 500mm f/5.6 (M42)</li>

<li>MTO 1000mm f/11 (M42)</li>

</ul>

<p>You can see some examples of what I shoot in my photo.net portfolio, although this is quite limited and in many cases outdated. My most recent work can be found on my www.15sunrises.com webpage.</p>

<p>The fact of the matter is that without question, I'm going to be upgrading to full frame, with the Canon 24-105 lens. I've chosen this one as it will finally bridge the gap (~80mm) that I've been missing for quite some time on my 40D. I may also get a wider angle prime (if I decide to experiment with untracked exposures), but would prefer to keep my weight down as my pack is already heavy enough as it is when going on shoots.</p>

<p>Now, in terms of the moon shots, I'm certain that the MKII and MK3 would perform equally as well, there is rarely a time that these shots require an ISO of above 800 on my 40D, therefore I think the difference between the two cameras would be negligible.</p>

<p>My question relates to star shots. Up to this point, I've been stacking tracked and untracked exposures in photoshop, usually with exposures ranging from 3-5 minutes at ISO 640-800. This means that a single shot basically takes around 20 minutes to complete (5 minutes tracked + dark frame subtraction and then 5 minutes untracked + dark frame subtraction). Ideally, I would like to get this time down in order to be more productive, and would like to do this by increasing ISO. I also think that the 24-105 will be higher quality than my Tamron, so I may not have to stop down as much to keep everything sharp. I'm wondering what people think the quality difference will be in terms of ISO1600 and ISO3200 files in RAW between the two cameras. All I've heard so far is information regarding improvements in JPG noise, which seems to be due to new processing in the chip. I will <strong>always</strong> be using a tracking device, but just want to get the times for individual shots down. When printing at A2 sizes, trailing is already readily visible with 30second exposures at 17mm on a crop, and I'm further looking to start printing at A1 or even A0 sizes, where trailing will be evident at even shorter exposures.</p>

<p>Really don't know if my point got across at all here, but I've been wavering between these two since the announcement of the Mark3 and am looking for any advice. Thanks.</p>

<p>My crystal ball is a bit hazy today, so here is the stock advice: wait until it is released, rent it and test it in real life, to see whether it suits your purpose. Simple. Any other answer at this stage is pure speculation.</p>

<p>Just regarding</p>

 

<blockquote>

<p>I also think that the 24-105 will be higher quality than my Tamron, so I may not have to stop down as much to keep everything sharp.</p>

</blockquote>

<p>I find the 24-105 to be extremely sharp; however at 24mm f/4 there is a lot of vignetting, probably enough to make it difficult to use for star shots at that aperture. Also, always turn off the IS for multi-second exposures.</p>

<p>On the 5D mk II -vs- mk III high-ISO RAW performance, it's hard to say until tests have been published. I would guess the mk III will only be a little better, maybe half a stop. Either will be better than your 40D.</p>

<p>If you are getting star trails on a 30 second exposure with a tracking device, I'd guess you need a better tracking device. Instead of a new body, have you looked into a decent telescope and an equatorial mount.</p>

<p>Thanks for the reply Alan, I usually don't trust any lens wide open and would be stopping down the 24-105 regardless especially since I'm planning on printing quite large. Thanks also for your speculation regarding the RAW performance, I thought personally that the difference would be between 0.5 and 1 stop as well, which doesn't quite make it worth the money for me. 2 stops on the other hand (like the JPG performance) would sell me though! The issue for me is that waiting means missed opportunities this month.</p>

<p>Hi Bob, I'm not getting trails with the tracking device, I just wanted to emphasize the fact that I will be continuing to use a tracking device regardless of the high ISO performance. Many people seem to say that trailing is still OK at 30seconds with a ~24mm lens, but I find this level of trailing to already be unacceptable.</p>

<p>I can offer an opinion but is is not tested with long exposures. I came from a 40D and moved to a 5D2. I think you could safely bump up to iso 1250 with a 5D2 and get similar results which might reduce your exposure times by about half. Also, perhaps certain lenses might also cut down the times as well. For example - maybe the 135L might be able to be shot wide open (f2 or perhaps f2.2) and also cut down those times a bit?</p>

<p>Perhaps you could rent a 5D2 and 135L and evaluate to see where things stand while the 5D3's work their way to availability and the reviews start coming in. Its a bummer that the recent 5D3 dpreview iso sample images were not from RAW's.</p>

<p>A question - do you really have to perform 1:1 dark frames or can you just do one every Nth exposure and still get acceptable results.</p>

<p>Hi Brad. I attempt to do landscapes with stars, in this case, I've found that the way to get the most realistic results is to just use one exposure for the sky and one for the foreground. Once I start stacking exposures I find that the stars start becoming almost too predominant as well as the nebulae, etc. which aren't visible to the naked eye while I'm there. Some people find these effects appealing, but I try to get the shots to appear as close to how they look as possible. That being said, in this case, since I'm moving around, I find that the 1:1 dark frame subtraction (in camera) is necessary, as humidity/temperature can be changing, as well as exposure times. While it is a bit more tedious than having and archive of dark frames I could use, I do find it to be the most accurate way...</p>

<p>You need to wait and get some information before making your decision. At the moment we have no data on how RAW files from the two cameras differ. I think it's safe to assume that the 5d MkIII will be lower in noise because of the new sensor design along with the gapless lens arrangement. That means more of the photons entering the lens will hit each pixel and generate more signal electrons. There may also be better firmware noise reduction for JPEGs, but that's a separate issue. Canon did comment on some "on-chip" noise reduction circuitry which I'd assume would affect the RAW files but I have no information (a) if it does and (b) how it does it or what the effect on the data is.</p>

<p>How big the difference will be is hard to say, but I wouldn't be surprised at around 1 stop for the RAW files. However that's just a pure guess. I suspect that the hard data will be out within about a month. I've heard dates of around March 25th as the launch date. I'm sure DxO will have data available by the end of the month.</p>

<p>If you do get a 5D2, you'll have an extra $1200 or thereabouts to put toward a 24/1.4</p>

<blockquote>

<p>If you do get a 5D2, you'll have an extra $1200 or thereabouts to put toward a 24/1.4</p>

</blockquote>

<p><br /><br>

...or maybe towards a Samyang 14mm f2.8.</p>

<p>I just found a set of 5D3 Raw ISO examples here. I think each thumbnail links to a JPEG but there is a link for each RAW (CR2) as well. </p>

<p> <a href="http://www.imaging-resource.com/PRODS/canon-5d-mkiii/canon-5d-mkiiiTHMB.HTM">http://www.imaging-resource.com/PRODS/canon-5d-mkiii/canon-5d-mkiiiTHMB.HTM</a></p>

<p>I occasionally do astro shots with my 5D2. Looking at your current kit I would say you'd be far better considering an appropriate lens rather than shelling out for the 5D3. The best thing I ever did was to buy the 28mm f1.8. It's reasonably priced, it's wide so you can use longer times without trailing the stars and it's got a wide aperture so you can use a lower ISO. I shoot at f2.2 and get beautiful shots with this setup. The ideal lens would of course be the 24mm f1.4L but that's serious money. I'd much rather have a 5D2 + 24mm f1.4L than a 5D3 alone.</p>

<p>Take a look at Ben Canales work. He's one of the best at the moment. He uses a 5D2 and often shoots at f2.8. He's also done some great work with his old 30D so it's not all down to spending huge sums of money.</p>

<blockquote>

<p>...or maybe towards a Samyang 14mm f2.8.</p>

</blockquote>

<p>Good idea, that should work well. A friend of mine takes great astro shots with a Canon 15/2.8 fish-eye.</p>

<blockquote>

<p>...or maybe towards a Samyang 14mm f2.8.</p>

</blockquote>

<p>Great idea. For $1200, you could get three of them!</p>

Here are a few samples from a Canon

Japan site I believe.

 

http://web.canon.jp/imaging/eosd/samples/eos5dmk3/

<p><a href="http://www.northlight-images.co.uk/cameras/Canon_5d3.html">Here</a> is a good site with lots of info and some High ISO image samples</p>

<p>Dave, prime lenses, 1.8 or faster often work out best for capturing the dimmest of sky objects: The Milky Way, nebula etc.<br>

In my opinion, a very fast prime would be a worthwhile investment for astrophotography. Consider a used 5D classic and with the money saved buy a Canon 24MM 1.4L.</p>

<p>Fast primes have more light gathering power than ANY zoom EVER will!</p>

<p>I shot most of this <a href="../photodb/folder?folder_id=856464">astrophotography</a> using a 50MM 1.8 wide open @ ISO 1200 on a 5D classic.</p>

<p>Thanks for the response, great photos. What are the general exposure times for the photos? Also, since you're shooting (I'm assuming) near wide open, what do the prints look like at A1, and also is there any noticeable trailing at A1 print sizes?</p>

<p>I did not make any prints from those shots. Most are 30 seconds. I suspect at 30 seconds some trailing is evident. Some of those were made with an old Canon 10D.</p>

<p>A shot made with the 10D and a Sigma 20MM 1.8 are here:</p>

<p>http://www.saugus.net/Photos/meteor_photography_tips_night.shtml</p>

<p>I sold the Sigma. I was not happy with corner sharpness.</p>

<p>Another thing that happens with star photos is flare, as is shown in the upper left corner of the meteor photo. Stars, as points of light, tend to flare out with some lenses, especially in the corners.</p>

<p>BTW, there is a formula which explains time versus focal length for trails and their visibility in a photo.</p>

<p>Found this by searching for "how to prevent star trails"</p>

<p><a href="http://www.clubsnap.com/forums/night-photography/219390-guide-maximum-exposure-time-avoid-star-trails.html">http://www.clubsnap.com/forums/night-photography/219390-guide-maximum-exposure-time-avoid-star-trails.html</a></p>