Let me tell you, I love me some Adam Wilt. His no-nonsense honest approach to testing is second-to-none, and this article on sensor low-light performance is a great example. Recommend reading this one if you’ve been feeling that your beloved EX1 or HPX has lost it’s lustre with all the recent DSLR video developments. Adam’s comparison test shows that performance is actually pretty close in these modern camera systems at high ISO and gain levels.

I hear it all the time…complete novices and non-techies look at the image from a Canon DSLR and say something to the effect of “I don’t know what it is, but it doesn’t look like video to me. It looks filmic.” There’s a lot of reasons why they might arrive at that conclusion…a shallow depth of field tends to be associated with production value and “films,” and the $2500 Canon 5D MKII offers about as shallow a focus field as you could hope for, with it’s massive sensor size (as a side note, this shallow DOF is also an annoyance, depending on the application). Canon DSLRs also render skintones and color in a very natural and pleasing way. Or maybe it’s the 24p cadence of the video capture that appears filmic. Like I said, there are many factors that contribute, and a few that don’t.

But one more area that the Canon DSLRs shine is in dynamic range. Video traditionally has had a very limited dynamic range, with the best of video cameras rendering only 6-8 stops of usable image information spanning the darkest blacks and the brightest whites. Film, on the other hand, is generally somewhere north of 13 stops, depending on who you talk to and what stocks you are using. So when people like Art Adams do tests on the Canon 5D MKII and show it has 10 stops of dynamic range, there’s one more subtle reason why people may think the image is filmic…it’s approaching the range of detail in film. That’s a significant accomplishment. When the Red One camera came out a couple years back, I recall that they were aiming for 12-13 stops of range, and most folks I believe tested it at around 11-ish stops. A significant development at the time (which has been improved of late with the MX sensor upgrade program).

And there’s reason to believe that DSLRs will improve over time…as I anecdotally recall, didn’t Zacuto recently test the Canon 7D at over 10.5 stops of dynamic range? Or was it 11? I can’t recall ofhand, but do remember if being marginally better than what the 5D offered. So there you have it…pixel pushers like Art Adams (and I say that with the utmost respect and appreciation) are telling us WHY we unconsciously appreciate the images coming from a Canon DSLR. And as a facts guy, it makes me happy that I can associate a factual number with the image that I already knew and appreciated from an aesthetic standpoint. Thanks Art!

Here’s a simple side-by-side comparison of the same test scenes that show how Red and Alexa handle it both pre and post grade. It is not explicitly stated, but I would assume that the Red One was equipped with a MX sensor upgrade. Caveat emptor. Watch below.

Arri Alexa vs Red One from D-Vision on Vimeo.

Unrelated to this test, Bill Bennett has also posted some framegrabs from a recent Alexa shoot that show good highlight handling in a technically challenging situation (sun reflections from silver car). See here.

I’ve been enjoying the Able Cine video series “At the Bench” with Mitch Gross, and the following video overview of Ianiro’s IANILED54 LED light/system is no exception. This is a seriously cool light kit, with a neat 3-in-1 hat trick. It looks extremely versatile. Check it out below.

With the release of Canon’s XF300 and XF305 camcorders, no doubt forum threads and geek discussions will raise the topic of color sampling. These new Canon cameras are equipped to record in the venerable MPEG-2 codec at a bit rate of 50 Mbps and 4:2:2 color sampling (to compact flash media, no less). In contrast, cameras in this price range generally utilize a much lower bitrate and 4:2:0 (HDV & XDCAM EX) or 4:1:1 (DV) color sampling. Canon DSLRs also chroma subsample at 4:2:0.

So while the XF-series of cameras are in general an unsurprising solid evolutionary step for Canon camcorders, the addition of this high bitrate codec at 4:2:2 is a rather significant development! But what is color sampling? If you are a bit confused by all the colons and numbers, this blog post from Adobe should help dispel some of the myths and inform you for that next “who’s camera is better” geeky throwdown discussion. Enjoy!

Jim Jannard has posted some test charts shot by RED, comparing ARRI Alexa with the Red Epic MX. I’ve spent several hours looking over these charts and here are my opinions on what information they reveal about these two amazing camera systems.

Jesse Miller of Midtown Video will once again be hosting a live show tonight from Miami, at 6:30pm Eastern Time (GMT -5). I have it on good authority that Canon has provided an XF305 camera, and during these live shows they generally take questions from the chat room. So if you’ve got a burning question about this hot new 50 Mbps 4:2:2 HD video camera, I suggest you tune in!

Here’s a quick video overview of how Active Shutter 3D tech works. The good stuff starts about 4:30 into the clip. Watch below…

There’s a really solid thread at DVXuser that delves into dynamic range and quality comparisons between the Canon 5D MKII and the recently-hacked Panasonic GH1. Previously the GH1 was an interesting camera hobbled by a frustratingly low codec bitrate, limiting it’s usefulness in the video world. But now that there’s a respectable codec bitrate option in the hacked GH1 firmware, people are starting to take another look at the camera. In the forum thread linked above, shooter Hunter Richards set up a good comparison test that pretty clearly shows sharpness and dynamic range comparisons between the GH1 and the popular 5DM2.

As a further comparison, the 5D Mark II is currently a $2300 camera body and the GH1 lists for $1,190 (B&H prices, buy links above).

(Via @ninoleitner)

RED’s Jim Jannard made a statement on the Reduser forum recently that doesn’t sit right with me. I quote:

I fully expect the DSLR mfgs to get it right at some point. Make a non-line-skipping 4K camera. At that point, the difference will be RAW 5K and 6K vs. whatever they make. Until then, a line-skipping 1080P camera is just not in the running for a pro camera. Can you make OK images with a line-skipping 1080P camera? Sure. Should you be embarrassed? Yes. We are not in that business.

I saw the Canon commercial about shooting “motion stills”. They should be ashamed. Just try to take a still from their motion 1080P capture…

Jim

I appreciate the crux of what he is getting at…I appreciate his take-no-prisoners, uncompromising attitude. And yes, the current crop of imperfect DSLRs are a bridge technology, we all know it. But disparaging pros who choose to work around these issues for the present reflects rather poorly on Mr. Jannard. So my response to Jim is this:

I’ll be embarrassed for shooting on DSLRs when you ship a camera that can compete in roughly the same price and featureset. Until then, it’s all marketing hyperbole.

-Matt

You see, the problem for my production business is that I’m having a hard time making money with my non-existent Scarlet camera system.

(via Photography Bay, also a few other comments on this from a good EOSHD post)

UPDATE: Jim replied later in the thread to clarify some of his comments. See that original post here (quoted below):

Heh heh… I’m surprised that so many people are reading posts from Reduser! :-) Wow.

Clarification #1. I think that Canon should be embarrassed about pretending that line-skipping is good enough when they have the capability to do much more. I’m sure they will in the near future. Until then, they are delivering much less than they are capable of… which is one of the reasons I started this company. No one in this industry seems to move unless they are forced to.

Clarification #2. No shooting professional has to justify what equipment they use… only their results.

If people want to correctly point out our flaws as a company, they will only be joining my previous admissions. We have no idea what we are doing. We are always late.

On the other hand, I would bet that no company in any industry cares more about their customers than RED does. We have taught the industry about free upgrades, trade ins and customer service at 3am. We have set a new standard for direct communication with our customers, a standard that other companies are trying to figure out how to emulate.

I don’t have any problem with someone pointing out that I have no idea what I am doing… I have said that all along. But it is curious that the industry is falling in line to “be like RED”. All I can say is … “don’t do it. You are following the wrong guy.” :-)

If these other websites are going to post my comments, I certainly hope they will post this one without edit.
BTW… we have taken a lesson from Apple. We will no longer discuss what we are doing until we are done.
Jim

I’ll refrain from adding my own thoughts to his latest comments. There is some good commentary from FreshDV readers below.

Want to see some examples of how Canon 5D MKII footage keys using the various built in picture profiles? Curious to see how the super-flat Marvel custom preset keys in comparison? Watch below.

Green Screen Compositing Test - Canon 5D MKII from Mook Genius on Vimeo.

Patrick over at Stillmotion has posted a quick test shoot they did with one of the first Steadicam Zephyr units. The Zephyr is not yet available for purchase, and in this test they had a limited amount of time to work with the unit. If you’ve been curious about this unit, take a look at the video below. Over at Vimeo Patrick has written quite a bit about the test and their setup, so definitely check that out.

On a related topic, here’s FreshDV’s NAB 2010 video segment with Garrett Brown on the Steadicam Tango (the Tango is a new 6lb accessory that mounts on a Steadicam sled to enable floor to ceiling freedom of movement):
freshdv_nab10_Steadicam&JVC

Here’s an interesting Canon 5D MKII and Canon 7D ISO video noise comparison test. It’s split-screen sequential ride on both cameras through ISO settings from 100-12,800, looking at the same high-contrast, high detail scene (wide exposure range, moving leaves, fine details). For this test, the user has stated that Auto Lighting Optimizer and Highlight Tone Priority modes were turned off.

“Every 5 seconds in this video the ISO will shift from 100, 200, 400, 640, 1000, 1600, 3200, 4000, 5000, 6400, 12800. The camera’s were both set on Manual and after increasing the ISO I then adjusted the aperture to bring the exposure meter back to the middle. I kept the shutter speed at 1/60th since I was recording at 30fps. On the very last ISO setting (12,800) I had to bump the 5D’s shutter speed up to 1/125th to get the exposure centered since I couldn’t bring the aperture beyond f/22. With regard to noise, they both seem to be equally matched until about ISO 4000, where it looks like the 5DMKII pulls further ahead as the ISO increases beyond that point.”

I downloaded the original AVC Quicktime from Vimeo, and to my eye, the 5D appears visibly sharper and exhibits a touch less noise at every setting above ISO 200. Which is not at all surprising, given the difference in sensor size between the two cameras.

Split-Screen ISO Comparison: EOS 5DMK II vs. EOS 7D from Ron Risman on Vimeo.

For a sightly less-compressed comparison, log in to Vimeo and download the 266MB quicktime source file for this video.

Red Digital Cinema has released some footage shot on the new Mysterium-X sensor (housed in a Red One body). It’s a short clip of a lobster from a 4.5K source, downrezzed to a 2K ProRes LT mov, reportedly shot on an (as yet) unknown lens at 800 ISO, T8, 1/48th, with daylight-balanced lighting, and untouched in post. The hive mind is pleased.

As an aside, I wish I could use the phrase “snappy the lobster” in more blog posts. :-)

Frequent Reduser contributor Sanjin Jukic has posted a quick and dirty Red One vs Canon 7D comparison video that shows how each handles a chart that brings out moiré artifacts. If you aren’t sure what moiré is, or are still figuring out how to work around those issues when shooting HD with DSLRs, take a look at this video.

R1vs7D 02 Moire-test from Sanjin Jukic on Vimeo.