Nebula 4000 Stabilizer – Part I

UPDATE February, 2015.  Am now happy with the Nebula.  Feel it best with the Sony A6000 or BMPCC.  It works with the A7 and GH4 of course, but is easier to balance and run with smaller cameras.  I no longer use any weight “mods” like those seen in the picture below; indeed, I don’t recommend them.

On YouTube, I saw that Dave Dugdale, at, was spending hours-upon-hours trying to figure out how to get a Nebula 4000 camera stabilizer working.  At $700, the device promises major league Steadicam quality in the palm of your hand.  Sample footage from China, where the device is made, was/is certainly tantalizing.  But back at Dave’s studio, the Nebula was shaking his GH4 around like a bad dog with a rag doll.  The software looked only marginally easier to configure than the cockpit of the Space Shuttle.    I thought, ‘better him than me’.  Maybe I’ll get into it after he figures it out; that is, if it doesn’t all end in tears (which was seeming more and more likely).

Nebula with Yaw balance fix

Then, in passing, the photographer/D.P., Red Wylie told me he had a Nebula and, since he didn’t have a GH4 to use anymore (having sold it to me), I could borrow the Nebula and try it out.  He warned me that he couldn’t configure the Yaw anymore and that he might have a ‘dud’.  The only dud was me–I’ve now spent two solid days with his Nebula 4000 and I barely know what end is up anymore–gimbal pun intended!

After watching Dave’s latest video I decided I had to share what meager knowledge I’ve accumulated so far.  Some of it might be wrong, very wrong.  When it comes to the Nebula 4000, at this point in time, even good guesses are probably welcome by fellow sufferers.

The goal of a stabilizer is to keep an object, in this case a camera, in the same position no matter what movement the person holding it makes.  If I lean down, the stabilizer would lean the camera up, to compensate (pitch).  If I twist my hand a bit in the clock-wise direction, the stabilizer would twist the camera an equal amount counter-clockwise (roll).  If I point the camera a bit to the left, the stabilizer would move it to the right (yaw).

For each axis (pitch, roll and yaw) there is a motor that will move the camera.  In a perfect world, the Nebula would know that a 1-volt surge for 1 millisecond might move the camera up an inch if I moved it down.  Unfortunately, temperature, weight, speed of movement, battery power, etc., all influence the net-effect of any voltage applied to the motor.  In short, the Nebula is like a golfer who can’t predict exactly how the ball will roll on any particular grass.

So what does the stabilizer do?  It practices hundreds (?) of times a second, and through a feed-back loop, eventually points the camera in the exact same direction it was pointed when you first turned it on.  Like a golfer, after a few tries it can figure out the power, speed and grip strength it needs to get the ball exactly where it wants.  Unfortunately, you have “configure” the Nebula with how much power it should use when it tries, how much error it will tolerate, and how quickly it will try again.  Every camera (weight) necessitates different settings.  Will you be running or standing still.  Should it keep the camera really still while you’re standing still?  Or keep it reasonably still while you run?  There is NO perfect configuration for the stabilizer (just like there is no one perfect club for a golfer).  The stabilizer can’t predict what movements you’re going to make.  You must prepare it.  Configuring a stabilizer is an art.  I call anything an ‘art’ when I stink at it ;)  For example, when the stabilizer shakes it does so because the power and timing does not match the object it is trying to balance when not moving.

Too much power and the camera will over-shoot the target (sort of like Shaquille O’Neil’s free-throws).   Too little power and you have an air-ball.

Before the Nebula can balance anything with its motors you need to balance it on the device, physically.  This takes us to the first serious design “problem” of the Nebula 4000.  The camera platform is not balanced itself (without the camera).  It expects that the camera weight will be added to each axis, in a certain direction.  That is, each axis begins “top heavy”, so to speak.  The reason I (and Dave) have so much trouble balancing the cameras is one must “solve” two centers of gravity at the same time (times 3 for each axis)–the device’s arms and then the camera’s.

The worst “top heavy” balancing problem is the Yaw.  You can only adjust it by moving the sled forward or back three positions.  To allow more better tuning, I have added counter weights to the Nebula where I can move weights small increments to get a perfect Yaw balance.

Above is a photo.  This post is very rough, I will re-write it later, with more info, and publish better photos.

I also place the camera on a clear, see-through CD case on top of a dowel (or any round object).  I then move the camera back and forth until it is perfectly balanced.  I then mark where on the camera it is.  By knowing the 3 centers of gravity for my camera I know theoretically exactly how it should sit on the Nebula. I find this is very helpful.







Sigma dp2 Quattro Review

In early 2009 I bought my first Sigma,  a dp1.  I have been a life-long picture-taker, but that camera re-ignited my passion for photography. I currently use a Sigma dp2s and dp1 Merrill (dp1m).   This review is a bit negative on the Quattro, but please understand that it is written from a frustrated lover of Sigmas, not a stranger.   Ignoring the new body shape, the Quattro, like other Sigma/Foveon cameras, delivers medium format quality at 1/40th the price and a fraction of the size.   I can’t afford medium format  and though I have full-frame cameras, they are bigger, bulkier and more expensive than similar Sigmas.  When I can work patiently, and want a film-like image, all Sigma cameras, including the Quattro, have a unique sensor technology, coupled with superlative optics, that produce a 3-D look.  Here are some photos from my early Sigma cameras.  Sigma dp1 and dp2s photos.

Before I received the Quattro I hoped for two improvements, a more responsive camera and Sigma Photo software that loaded and saved images in a reasonable amount of time.   My first wish has been realized–a Sigma camera that isn’t frustratingly-slow.  Unfortunately,  every time I put this camera in my bag I become irritated.  The camera is about 2 inches wider than my other cameras.

If the Quattro were my only camera I wouldn’t mind.  But what photographer would make the Quattro their only camera? (It’s too limited).  Most serious photographers have multiple cameras.  Therefore, this camera, for photographers carrying multiple cameras, or extra lenses, is the proverbial “poke in the eye.”  The camera is a bag-hog.  It is one of the great benefits of Sigmas Merrills that they gave you Nikon D800 quality in a small size.

What's that sticking out of my bag?

What’s that sticking out of my bag?

I could lay the Quattro down flat, but then it would push the other camera out.   Here is the Quattro compared to a Sony a6000.


In these photos I forgot to remove the Clearviewer on the Dp2Q that doesn’t come with the camera.

Past Sigma cameras have been utilitarian, in a good way.  The bodies are a bit slippery, but an aftermarket grip, or 2-cents worth of gaffers tape, fixes that to the user’s preference.  The batteries were small, but that never bothered me.  I had the choice of going light, or bringing more batteries.  With the Quattro, you’re stuck with a big camera, which now makes choosing it over a full-frame more difficult.

Here is a progression of Sigma Cameras.


Dp2Q, Dp1m and Dp2s. The gaffers tape on the Dp1m is for grip.

Onto the good.  The new Quattro design uses a concept that drives bayer sensors, interpolating single pixel colors to increase resolution at the expense of color accuracy.  The Quattro still samples three colors, vertically, like film, but uses a top, higher resolution layer to increase resolution.  In a sense, the Quattro is like the original DP series in full-color sampling, but with some added high density pixels (4 to 1, hence “Quattro”).  Excuse me if I don’t go into the details, there are better sources for this analysis (like the Sigma forum on DPreview).  Suffice it to say that in good light, the Sigma Merrill cameras are the most color-accurate consumer cameras you can buy.  The Quattro isn’t quite as good as a Merrill, but it delivers greater speed and over-all resolution.

I try to use a Sigma camera whenever I can.  I don’t have any special love/attachment to the Sigma brand.  They simply deliver the best images when their limitations don’t get in the way.  So my first question is how well the Quattro will compare my other APS-C camera; that is, which should I use more?  Lately, I’ve been using the Sony a6000, and have been very happy with it.  Sony has really got their act together with their cameras (standard hot shoes for one!).  They have been pushing new technologies for a few years now.  Their hard work is paying off.

I wanted to see for myself  how much the a6000  suffers from typical bayer problems against the Quatro, or if the Quattro’s somewhat partial “bayer-like” sensor change would craete bayer-like problems .  (The test has nothing to do with Sony.  Any Canon or Nikon camera would show the same issue.)

When I try to show why I think Sigma images are better than bayer images most people don’t see it.  It seems subjective.  My goal with the following test is to show objectively, what is going on, deep within the images, of both the Foveon and Bayer technologies.  Again, the difference between both cameras is not obvious to most people and I’m not arguing that it should be.  For those who print big, or need fine color detail, there is no need to defend these cameras.


Here is a close up of the interpolated pixels from the Sigma Quattro.  As desired, the white fabric between colors is white, as it is to my eye.  The dark spots are actually true-to-life, some of the black threading can be seen below the white-stitching.


Sigma dp2 Quattro

Here is the same rendering by the a6000, bayer sensor camera.


a6000 (bayer)

This is a problem of ALL bayer sensor cameras.  When a green pixel, say, borrows red and blue values from neighboring pixels it can come up with a value that is not neutral.  You can see how the red bleeds a bit into the white, and how the black allows red pixels to slightly saturate to orange on the white.  The farther one stands away from a print or screen, the less visible these differences are.  Indeed, in most cases, one cannot see the difference.

The Quattro is better at keeping colors separate.  Bayer cameras “smear” color, but it isn’t noticeable to most people.  Physiologically, we’re more sensitive to contrast than color fidelity.  Again, at the resolution we view most images at, these differences are not apparent.

Even with pure colors, bayers have trouble exactly calculating proper pixel color through horizontal interpolation.


a6000 (bayer)

Now the Quattro image

Sigma dp2 Quattro

Sigma dp2 Quattro

What these tests tell me that on the basis of pure IQ, within 8 stops of dynamic range, the Quattro is better than any other bayer camera.  However, the quality comes at a huge cost  in speed and flexibility.  Sigma has made improvements with the speed of the camera, but has not improved the software and has added a radical body design to the mix.

In pixel-deep, pure image-color-IQ, no other camera touches a Sigma, not even the venerable Nikon D800, IMHO.   For black-and-white photography, Sigmas are also exceptional.  Bayer cameras introduce tonal inconsistencies, due to the different color filters on each pixel.  The problem is not academic.  Leica has a b/w-only camera, without bayer filters, which sells for $9,000.  Because Sigma cameras use vertical sampling, they also produce exceptional b/w data.

If Sigma is reading this, here is my wish list.  In general, FOCUS all your camera resources on making the best, small, low-light ISO camera available.  You’ve been doing this for years, but I feel the Quattro is getting off-track.  1.) Rebuild SPP from scratch to make is fast.  It just needs minimal adjustment sliders, white-balance and exposure adjustments (only what can’t be done in Photoshop).  EXPECT everyone to use PS, or something similar.   2.) extend the grip forward on the Merrill (like other cameras) and put in a bigger battery to run a STRONGER processor.  If the Quattro doesn’t have superior IQ to the Merrill, drop it.   Don’t worry about making the battery last, per number of shots, make it powerful to clear up the camera buffer as fast as possible.  And again, EXPECT most photographers to carry more than one camera; the Quattro is too wide.

If the Merrill shot as fast as other cameras at ISO 100, and you could process those files as fast as a Canikons, it would be, at ISO 100, a camera most photographers could not ignore.  If more people bought Sigmas I’d breath a little easier.  I don’t want to shoot in a photographic world without these truly unique, and beautiful images.  Finally, a quick test-shot of my daughter, f2.8, the first one I took with the Quattro.







Sony a6000 and Fujian 35mm 1.7 c-mount TV lens

I would hesitate to admit it, but if I had to take any lens to a desert island, it might be the $30 Fujian 35mm 1.7 c-mount lens made for CCTV video cameras.  This is a lens never made to pass any DxO test.  Until recently, I used it on an EOS-M; however, I needed a loupe or reading glasses because the EOS-M doesn’t have a view-finder.  A week ago,  I bought an a6000 body so I could get back to one of my favorite lenses with an EVF.  With a $20 Nikon to Nex adapter, I have plenty of other lenses to use on the a6000 or a7.

Here are some of the test photos I took with the a6000.  I’ve said it elsewhere, this is the best $30 I’ve ever spend on photographic equipment, period.  Click on any of the images to see a larger version.

This is the corner of Pemberton and Rindge.  I’m tired of looking at this corner, but not with this lens.



Or the park I walk Pepper to every day.


When I want to draw attention to a solitary feeling


Harsh light works with this lens, not against it.


At a flea market


And finally, I love this photo.  Don’t ask me why.


Here’s a portrait I took with the lens on an EOS-M

Fujian 35mm used for Portrait

You can get this lens on Amazon or Ebay with a c-mount to NEX adapter usually thrown in.  I’ve used this lens on MFT also, but the look isn’t the same.  The smaller sensor of the MFT doesn’t show as much blur.



Fixing Chinese Focal Reducer to fit on BMPCC

I bought a focal reducer to use with my Nikon 24mm 2.8D on my Blackmagic Pocket Cinema camera.  Getting an effective 0.7 * 24mm, or 17mm starting at 1.8 (1 stop reduction) intrigued me. Andrew Reid, the genius behind EOSHD, is a big proponent focal reducers.  As a hobbyist, I couldn’t justify the cost of a Metabones adapter.  So I bought a Chinese knock-off on Ebay from C. Kee for $96, shipping included.  It was shipped within 24 hours and arrived exactly one week later.  Off to a good start!

Unfortunately, it would only turn a few millimeters onto the BMPCC, just enough to mount, if I handled it very gingerly.   If I forgot about it for a second, the lens would go crashing to the ground.  Anyway, I had promised Andy Lee, an expert on lenses and the Panasonic g6, who gives great advice on the EOSHD forums, that I would shoot some test footage.  First I shot Focal Reducer on BMPCC then Focal Reducer on Panasonic GF3 (which the adapter fit on perfectly).  He approved and said he was going to order one (he shoots Panasonic so wasn’t worried about fitting it to a BMPCC).

Then richg101 on the forum talked about taking it apart and checking the lens distance into the body.  When I did that I realized I could take the whole thing apart.  I studied it some more and the best I could figure is the flange on the adapter was getting stuck on the mount-flange.

Here is what the adapter looks like, looking at the MFT mount side.  Obviously, the adapter didn’t look like this when it arrived.  I had sanded it a bit trying to get it to fit.

Here’s another view.  If you look closely, you can see three flanges that twist under the mount flanges (which have springs underneath).  When the lens turns far enough a small bolt on the camera inserts into the adapter and locks it into place.

Here you can see a flange in profile.


In the BMPCC you can see there is a slight ridge/block that probably holds the spring into place.  I believe the adapter flange hits against this, ever so slightly.


First I take off the Nikon mount of the adapter.  There are 3 little screws.


After unscrewing a small locking screw,  I screw the lens out


Here are the three parts to the adapter.


And now I sand the inside of the adapter flanges down a bit.  I did a little, tested, until it fully locked into the camera.


Some advice.  Wear glasses.  You don’t want grit in your eye.  Eye doctors are expensive.  Also, make sure you wash/blow all grit away from the adapter before mounting it on the camera.  The smallest particle can show up on your sensor.

I may have ended up with a “bad” copy of a BMPCC mount or adapter.   These adapters don’t have names, so it is hard to see who is buying what on Ebay and Amazon, etc.  I’ve read reviews of what looks like the exact same adapter working on someone else’s BMPCC.   The good news is that all is not lost if you experience the same problem I had.  I hope you don’t though!

Now that I have it working, I ordered a Zenit 16mm.  With the adpater, I’d have 0.7 x 16 = 11mm time the 2.8 crop of the BMPCC, or 32mm.  Test footage to follow later!

Sigma X3F TIFF Conversion Issues For Blown-out Skies

The Sigma Photo Pro (SPP) software, which must be used to open X3F files, has trouble exporting to TIFF, clipped or blown-out parts of an image, like a cloudy sky, unless you manually adjust exposure and fill.  If you let SPP export in “auto” mode, there is a high risk that you will not be able to recover highlights, from the TIFF, in Photoshop(PS) or other editing software.

A long-time Sigma user pointed out that Foveon sensor files are more diifficult to work with because “They (Sigma) have two problems. (1) The layers have strongly overlapping color response that is not matched to the color theories used for RGB color management. (2) the third layer is very noisy.”

At this writing, there is no X3F to DNG converter and I doubt, for long-winded technical reasons, there ever will be.

If you must have details in the blown-out areas of the image, you must export for that (adjust exposure in SPP for the clouds), then, in PS, bring up the exposure in the part of the image that had it’s exposure reduced to bring out the sky.

Adobe Camera Raw does a significantly better job with bayer sensor images, good enough that one can often use the “raw” TIFF (many even work with JPGs) to post process with little loss of image information.  However, if one can, only usually loads the image into Photoshop through Adobe Camera Raw (ACR)

The root problem is that you can open RAW files in Photoshop from just about any camera but the Sigma.  This means that you have to convert Sigma X3F RAW files to TIFF files first, if you’re going to use other image editing software.  Or put another way, you have to adjust exposure and fill twice.  First in SPP, then in PS.  For Sigma images, the TIFF becomes the “RAW” file for PS.

Although a Bayer / Photoshop workflow is much easier than working with Sigma images, I still prefer Foveon. As nice as the D600 is, there is a 3-dimensional look to Sigma images that, IMHO, it just can’t replicate.  That said, the dp1m is not the best choice for low-light or fast photography.

Here are some images that I hope illustrate these issues.

First, a Sigma DP1M photo.  This is a JPG copy of a TIFF created by Sigma Photo Pro, at its default setting (that is, no auto or manual adjustment)


Second, a photo from a Nikon d600.  A TIFF was created by Photoshop’s Camera RAW filter, again, no adjustments


Here I reduce the exposure on the dp1m image by 1 in PS.  


Yes, doesn’t seem bad. Now I do the same for the Nikon version (I’m quickly selecting using a wand, so excuse the jaggy nature)


Dropping the exposure doesn’t give me an image I want, for either camera.

Now, here is the Sigma where I have tried to adjust the sky using a curve, to suit my taste.


Doesn’t get me where I want to go.  Now here I do the same thing with the Nikon TIFF


As you can see, I am able to recover the sky pretty well in Adobe’s default TIFF creation.  The bottom line is the Adobe camera RAW created a much better TIFF in default mode than Sigma Photo Pro did with the X3F file.

What if I try to do this with the X3F generated TIFF?


No matter what I try, I can’t get the clouds to look healthy.

However, I can get a good image from the dp1m image, by adjusting the X3F’s exposure in SPP to favor the sky then bringing up the houses with curves in PS (reverse selection to sky).  In the end, I can get to where I want to go, but I can’t sleep-walk through RAW processing with Foveon images like I can do with bayers.  Probably for the best because I should pay attention anyway!


Here is what the TIFF looked like


I believe I may be able create better “default” TIFFs by generating two sets from SPP, a normal, and one with the exposure set down 2 stops.  I can then let Photoshop merge the images and get the best of both worlds.  For another day…

In the meantime, here is Ted’s workflow, as explained on a DPReview forum:

I use SPP to produce a neutral image adjusting only the exposure comp. slider to get the histogram as wide as possible, checking each color in turn. Some might leave a little space each end, others might not. Very important, IMHO, to use ProPhoto working space for the review image. Some back off SPP’s sharpening as it does sharpen at zero setting, particularly visible at sharp edges. I also back off the saturation slider a smidgeon. The aim here is produce a neutral image, not one that pleases the eye! Then save as a 16-bit TIFF, still in ProPhoto color space.

So, for me (SD9, SD10, SD14, SD1), all SPP 3.5 sliders at zero, except:

Exposure comp: as required

Saturation: -0.3

Sharpness: -0.7

Any proper editor has much better tools for tone curves, color adjustments, sharpening, re-sampling and conversion to color spaces other than ProPhoto.

If you don’t already, you might want to try editing in PS, but staying in 16-bit ProPhoto, until the time comes to ‘save as’ (for me) sRGB JPEG. A 16-bit neutral TIFF in ProPhoto color space is the best for flexibility, IMHO.

Photographing Old Homes – Part I

Old homes are beautiful to look at, yet they seldom photograph well. Our brains perform a marvelous trick when looking at homes.  For example, they see through all those ugly telephone, power and cable lines.  In the following photo, most of the major culprits of unfortunate-house photography are in view. First, a blown out sky. It was bluer, more nuanced. Trees have covered of up much of the house.  Modern street signs that don’t match old home architecture.  Lastly automobiles. Not only do they block the view of homes, and neighborhoods, almost all of them are ugly.

Except in the early morning, and late evening, houses do not light well. If you expose for the outer exterior, the porch and shadows go dark. If you expose for the shadows, the rest of the house over-exposes. And again, the sky, is always too bright in normal shooting conditions.

The following photos are what I’ve taken in trying to figure out how best to photograph houses. I don’t love any of them. They are here only to illustrate my learning process so far. Since it is the winter, I’ve taken these photos in the middle of the day, when it is warmest. When I go for the photos I want, in the Spring, I will take them in the morning or evening when the light is best.

In the following photo I have abused the image adjustments quite a bit to get balanced lighting. Some like the effect, I find it artificial and, though pleasing, phony.

In the following photo I have Photoshop’d out the telephone wires and darkened the sky. The lighting is good. However, the tree, though not overly distracting on its own, casts shadows on the house that greatly compromise the image.

As you can see, snow will almost always over-expose.  I could do HDR, but I’d like to avoid it if at all possible.

I pass this home almost every day, walking to, or from, Davis Square. I know the owners spend a lot of time on it and its garden. Unfortunately, old homes have a lot of detail and the detail over-whelms the photograph.  So far, the most important thing I’ve learned it the biggest obstacle to pleasing house photography is overwhelming detail.

Many photographs are taken of a house at an angle. They look pleasing because they create perspective lines; lines that converge into a distant point. My problem with them is they do not give the real feel of the house, as seen in person.

The more photographs I take, the more I believe that good photographs are always simplifications of something we like in a view. If you can’t simplify the image, it doesn’t generate the same emotion as what you experienced while viewing the scene. The following photo doesn’t do anything for me, though if you were to stand in front of this tree and house in real-life, you would find it quite wonderful.

There is probably a way to take this scene and make it into a good photograph. My point is that the more stuff going on in the scene, the more difficult that is.

The following is a house that is fairly simple. I’ve taken countless photographs of it and have yet to find one I really like.  The problem with this house, I believe, is the light next hits the front of it during the winter.  I’ve already started consulting the direction of the sun throughout the year to dictate which houses are best to shoot in each season.

Here is a close-up, which I feel is close, but not there.

Shoot from directly in front of the doorway, while adjusting the perspective of the image in post, is what seems to work best for me. If you shoot directly in front of the house, but not directly at the door, the house is less inviting.

However, for a house that is balanced, squarish, the image looks okay

The bushes, the snow, and the house to the right of the following, overwhelm the fine detail of this house.


Here’s another house, where dirty snow, which makes winter shooting difficult, kills the image.

One way to reduce the number of distractions is to shoot with a lens that vignettes. The following is shot with a Fujian 35mm 1.7 CCTV c-mount lens on a Canon EOS-M.

I love these types of photographs. But the blur also makes one lazy in finding the best sharp image for the whole house.

Current Workflow

  1. First, spend as much time as you can, in front of the house, looking through the viewfinder. Move around to find the best position and angle—through the camera—that gives you the feel you want. I wish I could visualize what the camera will do, but I never can. I need to experience the house as the camera sees it.
  2. Try to get an image where the objects in front of the house, like telephone lines, signs, etc., can easily be removed (through Photoshop say) in post. A telephone line over a window, for example, is a lot more difficult to remove than when it is over siding. Of course, don’t take a shot just to get rid of distracting power lines, the distractions are just something you don’t want to overlook.
  3. In post, use perspective tools to straighten the house. As a camera purist, I don’t like this, but I’ve concluded that you can’t get the emotional feel of a house through camera/lens alone. Auto adjustment sometimes works, but generally you have to play with the horizontal, vertical and distortion adjustments until you get an image that you respond well to. Our eyes are very sensitive to crookedness. A photo needs a plane, either vertical or horizontal, that our brains can rest on.
  4. Use dodging tools to create brighten the windows and porches; that is, brighten and darken areas of the house to move the eye to where you want it to rest.
  5. Usually, select the sky and make it darker

January 2014

I could never get Magic Lantern to run on the EOS-M without some difficulty.  So I bought a Blackmagic Pocket Cinema Camera (BMPCC).  What I didn’t expect was how easy it is to shoot with the camera.  Davinci Resolve, software primarily designed to color grade footage, can now create time-line edited videos.

There are many blogs that point out shortcomings of the camera.  The people at Blackmagic must cringe and scream every time they read one.  I’ve been shooting film/video since super-8.  The last time I witnessed such a jump in technology is when Sony came out with Video-8.  That was in the 1980s I believe.

What I’ve come to realize, over the past six months, in working with RAW video, is that most people, for one reason or another, don’t want to engage in the technology.  It’s still a small world.  Also, video is, at heart, a narrative medium.   A well-written, prepared video is more important than the recording medium.

Anyway, I just wanted to post something to say, I’m still as deeply distracted in the tech as I ever was.

On the camera front.  Nikon replaced the shutter assembly of my D600.  I keep wanting to sell the camera because it’s big and I like to travel light.   However, every time I pick up the camera it whispers in my ear, ‘don’t blame me when you need to shoot low light or want a super-rich portrait and your other cameras fumble around’.

I now have a 24mm, 85mm and 24-85 (amazing how much distortion is in that lens; I just have to get over it and fix in post).  Miss the Tamron 24-70/2.8.  Was that lens sweet!

My Sigma DP1 is acting weird.  I’m close to pulling the trigger on a DP1M.

I absolutely love the EOS-M with the c-mount adapter and $30 Fujian 35mm 1.7 lens.  Yes,  I now reach for the cheapest camera I have for most of my shots!  Though I still use the Sigma DP2S a fair amount.

I’ve noticed that you can’t get discount Sigmas on Amazon from the U.S. anymore.  Canon doesn’t sell the 11-22 STM lens in the States.   Andrew Reid of EOSHD pointed out that Americans like big stupid cameras (or their iPhones).  So sad and true.

I still want a take with me everywhere camera.  Been considering the X100s, but it’s a little bit too cute for me.  Maybe the GX7?  Most likely I’d end up with the Sigma DP1M.  No other camera brings tears to my eyes, both joy and sorrow, like those cameras!

Magic Lantern and Blackmagic RAW Video

Why RAW? The short answer is video that has a film look, natural light and color.

RAW video can provide a look and feel that consumer video can’t. It accomplishes this by saving the original color detail and brightness information. Consumer video, what I’ll call H.264, the name of the most popular compression CODEC in use today, is designed to deliver a pleasing moving image—sharp and with rich colors—at a data rate that does NOT exceed most consumer electronic devices’ maximum bandwidth (usually under 4 megabytes per second, or around 30 megaBITs per second).

Here is a frame taken from H.264 video shot with an EOS-M.  Study these images closely.  They are straight from the camera, on a light-table, with controlled exposure.

Here is a frame from Magic Lantern RAW shot with the same camera:

Notice how the “RAW” colors are evenly bright, unlike the H.265 where the more primary the color, the brighter it is (because compression adds contrast).  You will notice more noise in the RAW image, but with it, more detail and dynamic range.

Why can’t you get both, contrast or deep colors? It comes down to convenience and size. RAW video requires the latest memory technology, around 80 megabytes a second for 1080p, or in Blackmagic, which performs some compression, about half that. RAW takes up a lot of storage space. Everything about it takes more time and effort.

Some background.

The total pixels in a frame of 1,920 pixels wide, and 1,080 pixels high, is 2,073,600, or about 2 million pixels. In one second, we watch 30 of those frames, so that 2 million times 30, or roughly 60 million pixels per second. For a minute we’d need 60 million times 60 seconds, or 3,600,000,000 pixels per minute, or 3.6 billion. Yes, when you’re watching your HD TV your eye is viewing 3.6 billion pixels every minute.

What makes up a pixel? A color of course. Colors are often described in their red, green and blue components. That is, every color can be separated into a red, green and blue value, often abbreviated RGB.  When most digital cameras take an image, whether a Panasonic GH4 or Canon, each color is assigned a brightness value from 0 to 16,383 (14 bits). So you need three sets of numbers, red (0 to 16,383), green (0 to 16,383) and blue (0 to 16,383) to numerically describe ANY color. Some simple math tells us that we need a value that might reach 16,383 times 16,383 times 16,383 or 4.3 trillion  As expected, a single 1080p RAW frame from a Canon camera is about 4 megabytes.  

In the above images, the H.264 frame ended up as a 114k JPEG (leaving no extra image data in the H.264 stream).  The RAW frame, a 256 JPEG, originated from a 2.4 megabytes RAW file, which means you can choose less contrast, or more detail and noise.

Even 8-bits (1 byte) per color “channel” is enough to create 24-bit (8+8+8), or 16 million colors. The human eye can see about 12 million colors at best (so we don’t need those 4.3 trillion “RAW” colors).  That allows an H.264 to throw out over 96% of the original pixel data.

A consumer video camera can quickly figure out what we can see, and not see, in an image, so this isn’t difficult.  It takes the “brightest” data and saves it–AND THROW OUT THE REST.  However, the overall brightest image is not always the image we want!  Sometimes we want a dim image with a lot of detail in the shadows.

Let’s go back to the optimum image we’d like to see, 3.6 billion pixels per minute times 24bits (3 bytes). That would be 10.8 gigabytes per minute. As you know, you’re not streaming 10 gigabytes of video to your TV every minute. Video compression does a marvel job of cutting that down to a manageable size

HD 720p @ H.264 high profile 2500 kbps (20 MB/minute)
HD 1080p @ H.264 high profile 5000 kbps (35 MB/minute)

It is the limitations of our computing devices that we can’t have what we really want–10 gigabytes of video data every minute.  If for the sake of argument we had unlimited storage and speed we’d all save and view images without compression.  That’s when they have the greatest fidelity.

Consumer video cameras record video using a “distribution” CODEC, not a photographic storage method. This means they’re making an immediate decision of what part of image to save, and what to throw away. The top image is what they end up with. The bottom image is what the sensor recorded BEFORE being put through H.264.

The benefit of RAW video to me, is that I can decide how to compress the image after it has been taken.  I can make the decision of what the image should look like.  I can get a photographic look.

The new 4K cameras coming out will offer more resolution (4 times more), but resolution will not give me more color depth.  That isn’t to say 4K is phony; only that it doesn’t fix the color-depth problem inherent in consumer-level compressed video streams.  4K RAW is a different matter, of course.

Wasting Time On DSLR Video

But A Used $100 Panasonic GF3 and $30 Lens Win My Heart

For the past couple of weeks I have been looking for a small video camera to complement my Sigma DP1 and DP2S, which take superlative photos, but unusable video.  Knowing I’d want short focal-length, shallow depth of field, I ruled out a camcorder. Having an all-around utility camera to complement the Sigmas, and the ability to use separate lenses, would also be helpful.

Having read a lot about Micro Four Thirds (MFT) cameras, especially the Panasonic GH2, I bought Panasonic’s lowest-end camera, a GF3, with 14mm pancake lens, for $245 on Craigslist (CL).  I then subscribed to Vitaliy’s PersonalView.  For $10 (you can do $5) and downloaded and installed a hack (first Cake 2.3, then Driftwood) for the GF3 which can double, even triple the bit-rate.

I was very impressed with the GF3, but noticed on a YouTube video, Kim Letkeman’ s Video Hacks Compare Four Panasonic Bodies — VLOG4  that the G5 provided as-good, if not better video, without a hack.  It also has a viewfinder and articulating screen.  Since Panasonic has a new G6, the G5 were being dumped on Amazon.  I picked one up with the kit lens for $350.

[Dear GH2 users.  I understand that none of those cameras is a match for a properly hacked GH2.  However, I’m going for small and cheap here].

I also bought a $20 adapter for my old Nikkor 50mm 1.4, siting in a drawer.  Wanting something less telephoto, I found a Minolta-mount 28mm/2.8 lens for $30 on CL, bought the adapter, and now had a couple of MFT cameras with a 14mm/2.5, 14-42mm, 28mm/2.8 and a 50mm/1.4.

Santo, who I met buying the 28mm lens from CL, uses an un-hacked GH2, and put up some very impressive videos on Vimeo.  Here’s one of a model train and the NMRA Model Train Show.  I watch his videos and think, ‘Max, you waste all your time comparing cameras instead of just using them!’

Then I saw an Olympus E-P1 for $150 on CL, that had a c-mount lens and Olympus 50mm/1.8 with adapter.  I bought it mostly for the c-mount lens (not knowing it was only $30), figuring I’d give the body with a lens to one of my children.  It turned out that the Olympus lens didn’t flare as much as the Nikkor, all the way open, and the E-P1 is a nice piece of metal hardware.  My daughter even said she wanted it!  Nice going Olympus.

I discovered that the C-Mount TV lens (Fujian with effective 35mm 1.7 for $30) delivers beautiful video, which is fun and easy to shoot.  I never would have tried that lens if it weren’t for the inspiring GH2 Shooter’s manual from Andrew Reid’s EOSHD blog and mini-publishing company.  I highly recommend his blog and publications.

There is only one limitation that bugged me.  On the GF3 (or, I believe, any Panasonic consumer MFT below the G6, you can control aperture, shutter speed, but not ISO).  In other words, if you expose for a brightly lit porch and pan to a darker area the camera will boost the ISO to better expose the picture.  In “iA” mode, you can slow the camera’s desire to change ISO if you pan from one level of light to another. Except for that exposure-lock limitation, everything else about the GF3 is superb (as a really inexpensive camera that does video).

With the GF3 I can use almost any lens, and it even has focus-peaking before I shoot.  For a couple of hundred dollars more, the G5, provides a view-finder, hot-shoe and articulating screen. (The current G6 adds true exposure lock and a mic input.  Currently, it is a hot camera so I’m waiting for prices to drop.)

The biggest weakness of MFT cameras is their color depth.  However, they are as equally sharp as any DSLR (if not sharper after hacks are used).  Also, you can get shallow depth of field with a fast lens (the $30 c-mount lens is f1.7 !).


DP1 28mm, DP2S, 42MM, Panasonic GF3 with Effective 35mm TV lens, and the D600 for a sense of size.

Can’t Leave Well-Enough Alone

Instead of going deeply into my new GF3, G5 and lenses,  I broadened my tests and began comparing them to my D-600.  Big mistake!  Here is some video I took during my honeymoon period with the D-600.

A DSLR is capable of better quality video, but in real world use, I ended up with problem-filled footage that, net-net, is worse than that from the MFT cameras (like Panasonic). Almost every time I shoot video with the D600 I score it ‘10 for color saturation, 2 for focus and 2 for basic exposure, white-balance, etc.’  When I shoot with the GF3 I score it 8,8 and 8.  Again, that’s now a $100 used body!

I wondered, would Santo’s work have been improved if he used DSLR video?  I think the answer is no.

The more I tried to get good footage out of the DSLR, the more disappointed I became. As many bloggers have said before me, DSLR video quality is a mixed bag.  The cameras work against you, not with you.

The following is a quick test, the Nikon D600 vs the G5 indoors. The Nikon has the Tamron 24-70mm/2.8 and the G5 the 14mm/2.8 pancake.  I don’t caption either video with which camera is being used.  If you’re a video nut like me, you’ll know which is which.  If you can’t figure out which is better, then believe me, stay away from DSLR video ;)

Here is a test of the two cameras outdoors.


Yes, coming from camcorders, shooting slow-moving subjects, with shallow depth of field, during a nice bright day, and using LiveView under the shade of an apple tree, brought near tears of DSLR joy to my eyes.  In every other situation I cursed a whole range of annoyances and shortcomings: moire, rolling shutter, artifacts, false colors and equipment not really designed to shoot video.

Perhaps I could work around all that.  However, it would never cure the fact that I’m used to getting perfect RAW images in photography.  At the end of the day, DSLR video is a string of middling-quality JPEG images (video).  And the rolling-shutter effect is very irritating.

Perhaps only RAW video, from the 5D Mark III, hacked Canon 50Ds or Blackmagic will give me the real color look I want.  As much as I am amazed at the video that comes out of the D600 on a good day, it is still no match for the video that others are getting from raw-based video.  DSLR video, especially of faces, is blotchy and pasty.

For now, I’m sticking with Panasonic MFT cameras for video (and photo backup).  RAW beckons, however.


Nikon D600



DOF (large sensor, shallow depth of field)

Very large

Lock Exposure

Very heavy (seriously, a brick)


Lenses very expensive (and big)

Mic Input

Non-articulating LCD

Low Light (large sensor)

No video viewfinder focus

Fast Operations


Great Photo Stills

Remote Tethering






Limited Exposure Lock


Poor Color Saturation

Articulating Screen

No Mic Input

Inexpensive Lenses (with adapters)

Stabilization weak




Really cheap (used $100 body)

No viewfinder

Very light and small

Limited exposure lock

Takes Hacks

Low color saturation

Cheap lenses (with adapters)

No mic and weak stabilization