We're heading into October so, of course, Zach Theatre is about to open our Fall, main stage production of "Dracula", directed by Stephen Dietz. The stage set is pretty magnificent and the lighting design is dramatic, wonderful, and filled with nerve racking (for the photographer) extremes between light and dark. The two rehearsal shoots I did were a text book experiment in using the shadow slider in Lightroom post production to bring detail into areas that read as mostly black in the camera previews. Definitely a play for which you'll want to use raw files, if for no other reason than to lift shadows with less noise...
As an additional experiment I used three different cameras during the two days of shooting. On the first day I used the Pentax K-1 and my small handful of lenses for just about everything. I did bring along a Fuji X-H1 with the 56mm f1.2 APD but it stayed mostly in the bag. Midday on Tuesday I found myself over at Precision Camera picking up a different full frame camera that I've been interested in trying out for both video and still photography work; the Panasonic Lumix s1, along with its companion 24-105mm f4.0 lens. I thought it would be interesting to see how the files from the actual stage show compared when all three cameras were confronted with the same lighting and scenery.
The Panasonic s1 is the newest of the three cameras and uses a new 24 megapixel, full frame (24x36mm) sensor. The Pentax uses a very well regarded, 36 megapixel, full frame sensor, while the Fuji is an APS-C format camera that uses a sensor that's about one generation back from the current state of the art. I literally had no idea which camera would emerge as the best, as far as image quality was concerned, and I had some idea of how much of a difference the lenses would play in the whole drama of photographing drama with different systems.
I guess I could just read all the reviews of the different cameras on the web and become somewhat vaguely expert on the subject but that seems a bit like cheating so I really did want to put the rubber on the road and hit the gas with all three candidates.
Each camera had its own set of advantages: the Fuji X-H1 was well served by lenses that were picked from the outset to be highly useful and selected for theater photography. The go to lenses for that system were the incredibly sharp and well balanced 50-140mm f2.8, and the really good all-arounder, the 16-55mm f2.8. Since both lenses are well corrected, even wide open, I found myself leaving them at, or around, their widest apertures. In this way I could get a one stop (at least) advantage over the other systems which would translate into a welcome drop in ISO, giving the Fuji a fighting chance at matching up with the other two. In fact, I shot the Fuji at 1600 ISO while I kept the other two cameras at 3200 ISO.
The Lumix was at the most disadvantage; I had never used the camera before and barely had the battery charged before I headed out the door to get to the theater in time. I even stuck the (thick and well composed) owner's manual in my camera bag; not that I'd have time to reference it in the darkened theater, but I might have consulted it during halftime (excuse me theater people, I meant intermission). The lens on the Lumix was good and I ended up sticking close to f4.0 (max. aperture) with that one as well. Several things about that camera were the absolutely amazing, the dual image stabilization that is purported to give nearly six stops of steadiness. It's about as close as I've seen to the benchmark Olympus OMD cameras! And that's very high praise. The second thing that worked in the s1's favor is that the (very rational) menus are much the same in style and layout as the ones in my beloved G9's. Getting up to speed in the half hour before the curtains went up was not nearly as taxing as I thought it might be. Finally, while the Lumix doesn't have the full on LED lighting prowess of exterior buttons and dials that the Pentax K1 features, the Lumix does have a function that lights up the most used buttons and controls on the back of the camera. A nice touch.
And then we come to the last camera in the mix, the full frame Pentax. I thought it would stumble on some issues like focusing since I'm mostly using older AF lenses that still rely on the less elegant whirling screwdriver method of auto focusing. A method given to a bit of hunting and sometimes, imprecision. I used two lenses for nearly all of the Pentax shots: the newest 28-105mm f3.5-5.6 zoom (latest FA - HD version) and the current 100mm f2.8 macro lens. The Pentax only got to play during the first, tech rehearsal because, as a traditional DSLR, it isn't very quiet and the second rehearsal included a "friends and family" audience and a few media guests. Not a good idea to blaze away with an old style mirror thrashing camera if the theatre is trying to be customer focused....
I won't keep you in suspense. While the Pentax and the Lumix would definitely outshine the Fuji at low ISOs and in typical use scenes where dynamic range and overall resolution are crucial, having good lenses for low light is critical for a holistic performance. The X-H1 with the 50-140mm f2.8 was by far the fastest and most assured at autofocusing all the scenes; from bright to stygian. It locked on quickly and, looking through the take from that camera there are NO missed focus shots. The Fuji zooms are bright and sharp, and easy to work with. The capper for me was the ability, during post processing, to insert the Eterna profile into the mix when developing the raw files. It added a much needed boost to the shadows and imparts a very well controlled highlight rendering to the files. Since I was able to shoot at a one stop (at least) advantage over the full frame cameras the noise rendering between all three cameras was more or less a toss up. I'm sure if I had to use all three at ISO 6400 the Lumix would have been the least noisy, followed by the Pentax and then the X-H1 in last place. But I will say that the "difference" between first place and third place would be, at the most, 5-10% different.
By rationalizing my technique over the three cameras, and by also setting up white balance identically between the three cameras, I was able to use all of the resulting photographs more or less interchangeably. You'd be hard pressed to tell which files came from which camera if you are not cued by the look/character of the lens and the way the backgrounds are rendered. But, again, the difference in f-stops closed that gap somewhat, as well.
To be quite honest I entered into the test with a small preference for the Lumix s1. I wanted it to win by a wide margin. It's a wonderful hunk of camera with one of the nicest sounding/feeling shutters I have ever experienced. I also had high hopes for it because of the many reviews I've read which lauded the noise performance and dynamic range of the shutter over and over again. And, in a fair test, it may actually be a better sensor than the rest of the 24 megapixel, full frame sensors on the market now. But in the end it all comes down to system performance. In a month or so Panasonic is supposed to introduce their 70-200mm f2.8 zoom and then maybe the entire theater shooting calculus will change and it may be the preferred system. Maybe.
I could say the same things about the Pentax K-1. In this test it was hobbled by lenses that were not optimum for the shooting task. But, surprisingly, the files look quite good. I could rush out and buy the currently available 70-200mm f2.8 for that system and do competent work but I'm not so sure it would be a good long term investment. The noise of the shutter and kinetic mirror already mitigate against the camera for use in the kind of theatrical assignments I find myself and, well, let's just say that system support for Pentax right now is a bit less than transparent. I'd hate to splash out nearly $1,800 for a lens, in a secondary system, which may or may not have continued camera support for full frame.
My takeaway? Familiarity breeds fluidity and control. The X-H1 has one of the softest shutter sounds of any camera I've used and that's important because you can't always default to electronic shutters when shooting under light sources that can cause flicker and uneven illumination of the photographic frames. I've done a good job selecting just the right lenses to make my jobs photographing live theater as perfect as I can get it.
There are still come challenges to theatrical photography that even the newest and best sensors can't overcome. The difference in light intensity between an actor standing in a bright spot light, and the much less bright accents lights on much darker parts of the set or the background, is too much for any sensor to bridge. Since blown highlights are irrevocably lost the faustian bargain is to expose for the highlights and then try to lift the shadows as much as you can without the resulting noise becoming excessive. One would like to think that a bigger, newer sensor would make child's play of this issue, compared to a smaller sensor, one generation removed, but my "real world" tests show me that the difference in shadow performance is not that great. When comparing all three cameras with a three stop shadow boost I found that all of them started to generate small, white speckles randomly throughout the darkest areas of the images.
And while we talk about this I want to make sure readers understand that these artifacts are not coming from long exposure noise since all the samples compared were shot somewhere between 1/125th of a second and 1/320th. It's just pure shadow noise.
I've seen tests from websites that review cameras showing pushes of up to five stops which look much better than the samples I was able to pull from my tests. I can only conjecture that the tests are making full use of additional noise reduction somewhere in the process. After I compared all of the files I too turned to noise reduction techniques offered by Adobe Lightroom in order to deliver a better product to the clients. But it's always a compromise between overall detail and lower noise....
I'm sure there are advanced programs that would help me fine tune my noise reduction results but most are not compatible with the big batch processing routines that I normally need to use for quick turnaround of hundreds and hundreds of images.
That the Fuji camera was right there in the race with the two full frame cameras is interesting; and refreshing. We spend so much of our time accepting what we read on sites that seem authoritative but we never really spend time vetting the information we seem to take for granted. As I said, I headed into this "experiment" with the prejudice of wanting the Lumix camera to be "head and shoulders" better than its two competitors but wasn't the case.
Sure, if I fine-tuned all of the Lumix camera's advantages and coupled the camera with lenses costing $3,000+ each I am certain that the advantages of the sensor would become more obvious, less clouded by the deficiencies of smaller aperture and lens quality. The same goes for the Pentax camera. But it's nice to see that an optimized APS-C approach to a job can deliver really good, competent and competitive results. And with lenses that are less expensive than the ones I would need for either other system.
In the end all three cameras have their charms and all three are fun to shoot with. I'm pretty sure that for my live performance work I'll keep using the Fuji cameras, for now. The Panasonic seems like it would be the optimum studio camera but we're working on comparing the s1 to the Pentax K-1 using flash for a studio project today. I have a sinking feeling that the one area in which the Pentax will be a clear winner is in the studio --- because of the direct viewing of the finder. But, having drunk the mirrorless Kool-Aid(tm) I am equally certain that I'll miss stuff like pre-chimp previews with all imaging parameters baked in as well as good face detection AF. But again, that's what the testing is for.
Posted by Kirk, Photographer/Writer at 17:51
Posted by Kirk, Photographer/Writer at 15:27
A three way camera system brawl with two clear winners and one consolation prize. Lumix vs Pentax vs Fuji.
Our wonderful vocalist, Paul Sanchez, with our youngest cast member, Illiana. .
I am usually mercenary and want to get paid for the photographs I shoot that aren't specifically my art. But I make exceptions... One client gets a good amount of shoots for free but there is an underlying and valuable return on my time since I get to use the free shoots I do for them as testing grounds for new camera bodies, new lenses and complete new systems.
On Friday and again on Sunday afternoon I went over to the Zach Theatre rehearsal stage to take candid photos of the rehearsals for the upcoming production of "Christmas Carol." On each day I took along the new camera and lens system (the Lumix stuff) and one of my other, more seasoned systems. On Friday it was a match up between the Fuji and the Lumix; and more specifically, between the S1 + the two Panasonic zoom lenses versus the Fuji X-T3 and the 56 f1.2 (APD) and the 90mm f2.0. I'm giving that first round to the Fuji stuff because the speed of the primes (and the optical bite and color) were a bit more competition that the Lumix could handle. Also, it was my first time really trying to use the (mostly new to me) Lumix under dire lighting conditions, with fast moving subjects, while trying to go for the shallowest depth of field I could achieve (because the aesthetics of the space are nothing to write home about).
While the 56mm f1.2 is no speed demon when it comes to auto focus my familiarity with the camera and lens saved my butt over and over again. The 90mm is pretty much perfectly behaved and was the standout favorite of the day.
I was photographing moving actors practicing choreography as well as actors blocking out stage moves while they sang and practiced lines. Unlike a well lit stage the space we worked in was a bit like a big warehouse; high ceilings, dim lighting only from the ceilings and bare white walls on three two sides with overhead doors on the third side and a wall of mirrors on the fourth.
I expected more out of the Lumix system but it's my fault for trying to toss two zoom lenses that each max out at f4.0 into the ring against two very, very good and fast primes. Wanting to see the real potential of the S1 I decided to slant the playing field in the other direction and supplement the pokey zooms with an advanced rocket ship of a fast lens in the form of the L-mount version of the Sigma Art series 85mm f1.4 lens. It's optically wonderful, incredibly sharp, but at 2.2 pounds the darn thing is a monster. It needs its own set of wheels if you intend to walk around with it.....
Any camera system that is new to you has it's own peccadillos when it comes to the autofocus systems and I'll confess that I have not done a deep dive in the details (and potential) of the Lumix S system. It's close enough to the G9s I owned that I thought I'd wing it with the default settings for now. That may be misguided.
So, on Sunday I showed up with three camera set ups, the Pentax K-1 with a 50mm f1.4 AF lens, the Fujifilm X-T3 with the 56mm f1.2, and the Lumix S1 with the above-mentioned 85mm. The slowest focusing of all three cameras was NOT one of the mirrorless models. The Pentax is just not a good tool (at least when combined with that old screwdriver AF lens) for fast moving subjects under perilously low light..... (How low was it? How about 1/125th of a second at f2.0 with ISO 800-1,600. And super-icky color balance courtesy the banks of ancient florescent lights high up in the tall ceilings....).
I shot as much as I could with the Pentax and then, during a break, tossed it in the back of the car. Then it was down to Lumix versus Fuji. And here I found an interesting weak point for camera face/eye detect focusing and people with darker skin. If I had a person with dark skin, who was in the middle of the frame, fairly close to camera, the Lumix camera would create boxes around all the faces it could find in the frame and then locked focus on the face with the most contrast -- light skin against dark eyes and eyebrows, etc. Since many in the cast are African-American I mostly gave up on face detect AF on the Lumix and switched to single spot AF. The Fuji wasn't fooled as often but the trade off there, which drove me away from their face AF is the slow, slow response and focusing of the 56mm f1.2. It's much faster with single point AF and even a bit more sure with full manual focus and focus peaking.
Another issue across all three cameras, and one that photographers and review sites rarely discuss, is the big variations from frame to frame that are caused by old florescent lighting banks. They run on a 60 cycle set up which means they tend to get brighter and dimmer as they go through a cycle. It mostly affects images taken at shutter speeds above 1/60th to 1/125th of a second. Below 1/60 you can shoot all day long and never see the effect. At a 1/60th you can usually do just fine, and 1/125th is generally usable unless the florescent systems are on their last legs and their ballasts are giving out. But!!! But!!! as you get to faster and faster shutter speeds you see whole frames that are a stop or two dark followed by one or two normal frames followed by a bright frame or two. Since the EVFs in most new cameras also refresh at 60hz or 120hz you rarely see the effect in live view. You only see it in the post review.
I would have loved to have shot all the dance movement at 1/500th of a second; especially with the Lumix camera, which does not seem to fear high ISO settings, but I was getting some banding and lots of dark frames as I shot through the day. My workaround was to settle in at 1/125th of a second with all the cameras and then work hard to anticipate the peak of action at which things seem to become more or less stationary for a piece of a second. Capture it and you can stop some action. For most of the shots I ignored potentially blurry hands and feet and worked to make sure faces were sharp. There's no real workaround for working under A/C driven florescent lights. In the film industry the good cameras have VFR (variable frame rates) which allow camera operators to fine tune shutter speeds to eliminate flicker from ambient light sources that are known to flicker. These would include CRTs, some A/C LED fixtures, most industrial fluorescents and neon.
Good LED lights tend to have power boxes that convert electricity from A/C to D/C to drive the LEDs without flicker. But even tungstens have flicker, it's just that it's usually fast enough so the temperature decay and glow from the filament doesn't have time to cool off enough in the cycle to change the light output... I only worry about A/C LED spotlights and bad, commercial florescent fixtures for the most part. But this is by way of explaining why, even with cameras that can shoot at 51,800 ISO you may be constrained to shoot at 1/125th of a second. And if you have to shoot at 1/125th of a second to prevent flicker you'll probably want to use a lower ISO so you can still have the advantage of shooting wide apertures with longer lenses in order to drop backgrounds out of focus.
Back to our main (commercial free) program:
I used the two camera systems interchangeably and mostly concentrated on getting shots in which one person was in focus and the rest of the people in the shot, in front or behind my main subject, go out of focus. Both lenses (Fuji and Sigma) are sharp wide open, and more so nearly wide open, but I have to give the nod to the Sigma Art 85mm. When it nails focus on eyes, for example, the amount of sharply rendered detail at the plane of focus is amazing. If I had never used the Sigma I'd be well pleased with the Fuji 56mm as its performance is close. Very close.
When it comes to how the photographs look on the screen both the X-T3 and the Lumix S1 are keepers. With the Fuji I can reliably go to 1600 ISO and get great files with low noise. In Jpegs or Raw files. With the Lumix I get the same kind of performance at ISO 4000. The Fuji cameras and lenses are easier to carry, easier to use and the lenses are a fraction of the size and weight of the Lumix system stuff.
Whatever will I do? Well, I went out today and bought a second Lumix S1 body because it's just a natural system for the theatrical documentation work I'm doing for several theaters now. While most photographic disciplines don't really require two cameras (other than for back-up) shooting a non-stop dress rehearsal requires me to use a wide range of focal lengths combined with a certain fast enough f-stop. I normally use two zoom lenses to cover the range I need. Something that starts at 24mm equivalent and goes to a full frame 85mm equivalent for wide shots that encompass the entire stage along with a traditional 70-200mm zoom that allows me to crop in on smaller groups, shots of two people together and even single person shots. There's no time to change back and forth between lenses so I go in with a different zoom on each camera, set them identically and go with the flow of the show. My choice now is to shoot with the Lumix S1 and the two zooms wide open at f4.0 (where they are plenty sharp enough) or to shoot with the Fuji X-T3's with the 16-55mm f2.8 at 4.0 (it needs one stop down to match sharpness...) and the 50-140mm f2.8 wide open.
On the dress and tech rehearsals we'll be doing in a couple of weeks I'll probably use one system for the tech and the other system for the dress and see which one yields the best results. I better practice with the Lumix S1 in the meantime. It's complicated. But at the same time it feels familiar. The Fuji is like a nicely worn flannel shirt on a chilly day. Comfortable.
An embarrassment of riches. A nice time in which to be a photographer.
Posted by Kirk, Photographer/Writer at 11:23