11.01.2020

Nearly four year old GH5 versus Nearly new Lumix S1H. Which is the more useful video camera? Well, I guess that all depends on what you're shooting...

Yeah. I know. You got a spreadsheet. You can prove you made the right choice.
But what if you get to make more than one choice?

I think it's so... cute... that many video content creators are forever searching for the "perfect" camera. You know, the ONE camera what will do EVERYTHING for them, and do all the stuff at the highest level. Of course those folks also presume that they can get all that for around $3,000. Maybe less. And off they'll go shooting miraculous stuff which they hope will buy them notoriety and/or fame on YouTube. Gosh, I wish life were so simple.

It seems to me that life and video both require a diverse collection of skills and gear with which to navigate well. Lately I am struck between my divided regard for two different cameras from the same company; the old GH5 and the newly arrived S1H camera, both from Panasonic.

You've read it everywhere before but the main differences between cameras with smaller, micro four thirds imaging sensors and full frame sensor cameras are these features: state of the art image stabilization, smaller size, lower weight, and (best of all) smaller and lighter lenses. The "smaller and lighter lenses" also means that the small sensor cameras are easier to design and build long lenses for. 

The full frame cameras feature: bigger sensors, the benefits of which are bigger pixels/lower noise at high ISOs, more control over depth of field (in one direction < blur) and more surface area on the bodies on which to put more and better physical contact and control points. 

Since the Lumix S1H is newer and higher priced camera it offers more state of the art, video-centric features; but not so many as to render the older camera totally obsolete. You have nicer screens and better EVFs on the S1H but in video a good external monitor quickly levels that playing field. Most of the improvements are in the realm of greater and greater flexibility for settings and codecs. The S1H has a full-fledged, professional V-Log codec while the GH5 has Panasonic's V-Log lite. 

I really like the S1H and use it in a lot of controlled interview shoots and in situations where I know we'll be working at ISO 3200 to 6400. That's the sweet spot in which the newer camera starts to show its advantages over the smaller, older tech sensor. The S1H also has a nicer set of audio features and more "dynamic range" in the sound output. Along with more audio controls.

The S1H is a camera I'll grow into. The GH5 is a camera I'll leverage to the max right now. 

Case in point. The S1H, and the entire Panasonic S1 system doesn't have a native lens longer than 200mm. I'm sure they'll get there but when they do it's going to be expensive. Maybe frightfully so. But with the GH5 there are two or three lenses right out of the gate that can get me the reach I sometimes need along with the quality I think I want. 

I've spent two Saturdays now trying to use the S1H as a higher magnification, follow camera for night time concerts at Zach Theatre. I like using native lenses on the cameras because I can set the camera to deliver linear focus and I can also choose the total angle of rotation for the focusing ring that works best when I'm manually focusing. I'm finding that the physical operation, in this instance, is much, much more important that that mythic last ounce of quality in the files. If the footage is jittery, out of focus, and hard to focus then all the magic of 4K and high data throughput is meaningless. Especially so to the viewer.

I'm set up in a fixed position at these outdoor concerts because I need to cover two areas with one follow camera. I can get great wide and medium shots with stationary cameras but I'd really like to get tight coverage as well. When I hit the limits of the 200mm long end of the lens on the S1H I have to switch to an APS-C crop from full frame to get some extra reach. Now instead of using the full frame I'm using a crop in the middle of the frame. If I want to get closer still I can switch to the pixel to pixel setting which gives me a 2X crop but when shooting in 4K the lens is basically reading the same area as a GH5 would but the GH5 is sampling from all 20 megapixels instead of 8. Lenses created for smaller format cameras are optimized to provide more resolution and performance than the lenses  designed to cover full frame sensors so, at least in theory, not only is the GH5 downsampling to 4K which should give us less noise but it's also delivering details to the pixels by way of a much higher resolving lens (if we're buying the good stuff from the lens offerings...). 

(I started writing this before I left the office to go and shoot video last night... what follows this sentence was written this morning...)

I have two previous weeks of data and observation to call on for an even tighter comparison between the hands-on video performance of the two cameras. Last night I used the GH5 as my main video camera. It's the one I operate during the entire show, following whichever performer is the soloist at the time and on whichever stage he or she happens to wonder off to. I used the 50-200mm f4.0 (which is a full frame angle of view equivalent of 100-400mm at f4). I immediately observed that the additional depth of field of the smaller sensor made fine focusing much more forgiving. Even without access to linear focusing or more generous lens throws I was easily able to use the lens with manual focusing; checking the results in real time on the Atomos Ninja V monitor. 

While I found myself, earlier in the week, needing some extra ISO for stills that's a function of the need to use shutter speeds that can do a better job at freezing action. With photographs that tends to be in the range of 1/200th to 1/250th of a second. Even with the f4.0 lenses used wide open the need for more ISO is evident.

But it's not at all the same when shooting video. My theater client (and editor) likes to work in 24 fps which means we're setting the cameras for 1/48th of a second shutter speeds (180°). Some will do the exact frame rate while others need to be set to 1/50th of a second because that's the closest we can get. But that's still two or two and a half stops slower than our required photographic shutter speeds so I'm able to use the camera, when shooting video, at around ISO 800, a setting which is easily handled by any of our front line camera units. 

While the 70-200mm f4.0 S-Pro is a great performer, wide open, on a full frame camera; and a very able performer on APS-C video, by the time I have to crop down to a pixel-to-pixel magnification it's no great shakes. The 50-200mm lens for the m4:3 format is using the full frame at the same magnification and benefits from both the downsampling benefits and also having a design optimized to deliver great results with the small pixels.

Here's another thing I noticed when comparing my experience with both systems used in the same set up: The first week I was a sucker and bought into the "wisdom of the web" and turned off the image stabilization because the camera and lens were mounted on a tripod. What that really meant at the highest magnification was that I was giving up a lot of quality when actually handling the camera to make focus or zoom changes. While you might get artifacts when using the image stabilization with still photographs at certain shutter speeds the issues don't necessarily translate to tripod mounted cameras used in video mode. 

Last week I turned on the lens and body image stabilization and was rewarded with much less camera jitter in those times when I had to handle the camera or lens. And the stabilization in the S1H worked well. But the dual stabilization in the GH5 is better! And when you add in the e-stabilization you can get away with a lot more camera handling without obvious penalties. You will go through batteries quicker but even with all the I.S. engaged I finished up the hour and fifteen minutes of performance with two bars out of five left on the battery indicator on the GH5. For comparison, I used an S1 with all the extra features turned off, as a stationary camera, and at the end of the same time period the battery indicator in that camera read 57% remaining. Nice. 

If I were to base my video-oriented camera preferences solely on how well they shoot high magnification, moving performers from a stand-off distance I'd choose the GH5. Even more so when you calculate that you could buy three of them for the price of one S1H; and with the idea that creative content, in this instance, is even more important than any small difference in overall quality. Having three cameras gives you the chance to shoot three angles of events simultaneously which will make a lot of difference, in a nice way, when you sit down to edit the project.

But you knew a caveat was coming so here it is: While the GH5 is still totally relevant today (Yes, you can shoot high quality 60 fps at high data rates with no crop) the S1H has the potential to deliver higher quality content. The audio pre-amplifiers are cleaner, nicer to listen to and have more dynamic range. With the right subject matter the full frame video is sharper, cleaner and has more dynamic range. I also like the colors right out of the S1H better than those from the GH5 which I believe is the result of a new generation of color science versus the last gen. The version of V-Log in the S1H is better. The EVF and rear screen on the S1H are much better. And all of the setting features, meters and professional tools in the newer camera make it easier to get great results with. It's the camera I'd always bring to do corporate interviews or project on which the video footage has to be as good as I can get it. 

And now the S1H has a bigger quality differentiator in the ability to shoot Pro Res Raw video in conjunction with the Atomos Ninja V. We've tested the footage shot that way and while it sucks up memory space it has the potential to be really, really good. Especially on those times when I've slowed down enough to get my initial settings right. In the best of both worlds I'd have at least one of each camera so I could match them to the kinds of project on which they would excel. And since we live in the best of both worlds that's pretty much what I've done. 

But be aware that this is a moving target and my calculations might change if I test the Sigma 100-400 and find that it's great. Or, even better, Panasonic comes out with a long, fast native zoom for the S1 system. But for now I'll choose the GH5 and the 50-200mm lens for all the stuff that needs to be shot at a standoff distance and pull out the S1H when I can take complete control of the shooting parameters.

So, what kind of craziness did I engineer last night?

I was getting bored just shooting the Saturday concerts (a different cast and theme each week) with only three cameras so I added a forth one. I had one camera on a tripod with a long lens that could pan and cover the whole complex of stages. I had a second camera directly in front of the main stage and set to 35mm to cover the main action. I had a third camera on the "B" stage. That stages gets about 15% of the total play time that the main stage does but we need a camera there for wide coverage. Then, I added a fourth camera off at an angle to the main stage fitted with 16mm lens (32mm ff, eq.).  Here they are in their configurations.

This was my main/follow camera. It's the GH5 outfitted with the new (to me)
50-200mm Panasonic/Leica lens, the Panasonic audio interface with 
a feed from the main sound board, and an Atomos Ninja V monitor
which allows me to punch in to 1:1 or 2:1 while shooting to 
fine tune focus. Something you can't do on the camera's own monitors.
A project saving accessory. And yes, it works with stills.

This is my camera used looking straight into the main stage. It's an S1 (with the upgrade) outfitted 
with a Sigma 35mm Art lens and sitting on a Sirui tripod.
Do you see all the white tape everywhere?  It's on the tripod, lens, on the camera and 
on the lens hood That's so crew doesn't miss seeing the camera 
as they move through the set and run into it. The camera is parked next to a railing 
but you can never be too careful.  

This little pup is a Panasonic GX8.
I've outfitted it with a Sigma 16mm f1.4 lens 
used at f5.6. It's sitting on a Leica table top tripod
and that's sitting on the ground just next to a wall. On the far side of the wall
is the audience area. In front of the camera, at a 45° angle, is 
the main stage. I just wanted something a bit different
and when I discovered that the "old" GX8 could 
shoot 4K (in M4P) for an unlimited time I pressed it into service.
The files actually look great!!!

Finally, I'm using the S1H with a 24-105mm S lens as 
the wide camera for the secondary stage area. 
This camera got demoted when I discovered how good 
the GH5+50-200mm lens could be. 

It did a great job though.

Sitting on top of an old Gitzo tripod and a Manfrotto ball head. 

One thing I did differently last night: There had been a lot of lighting changes since I last shot the show here but unlike a highly rehearsed and practiced stage show the lighting doesn't change during an individual performance of these smaller productions. I remembered to take along a favorite incident light meter and I carefully measured the actual light in the areas covered by all three of the fixed position cameras and then set them very exactly. The result is a bunch of technically better files to work with today. 

On the main/follow camera I fine-tuned the exposure with a waveform reading on the screen of the Ninja, setting darker skin tone to 45-50%. It also produced technically perfect files for our use. The meter was much better than eye-balling stuff. And I can't remember the last time I saw a professional whip out a meter and use it. I bet I looked so cool. (That's a joke --- a program note for visitors from the humorless site). 

So, "No good deed goes unpunished." Yep, we finally nailed the technical aspects of perfectly shooting a live stage show. The theater uses the finished, edited videos to make some extra cash by offering them behind a paywall on Vimeo. This should have been the video for this particular production. But, as fate would have it, the female performer was having some issues with her voice and two of her songs were scrubbed at the last minute. The creative team punted by adding a song each for the two male performers. I'm sure though that the artistic director had/has the original playlist burned into his brain and we'll be shooting this show again in two weeks when they hold it over for a three week run before Thanksgiving. Dammit. I was so close.

Funny. We say that cameras are just tools. But then we expect one camera to do everything. Like hammering framing nails with a Swiss Army knife. I really do believe that cameras are visual construction tools and like every other endeavor we do need different tools for different applications. Few people would be obtuse enough to try and put a four by five inch, technical view camera onto a drone. You might not want to shoot highly detailed material destined for monster big enlargements with an iPhone 5S camera. The fun thing about doing artistic work is that you get to choose the tools (plural) with which you'd like to work. And you aren't limited to having only one. 

On the other hand, last night sure would have been interesting if all four of the camera had been GH5s...





 

10 comments:

James said...

Next week dare I say the Sigma fp will show up as a fifth camera?

scott kirkpatrick said...

How do you do sound, which might be really important when the shows are edited and rebroadcast? Are you getting sound tracks from each of your 3, 4, maybe 5 cameras at the different stages? At your suggestion, I spent some time listening to Curtis Judd on microphones and preamps. He's skeptical of ever getting the control over sound that he likes without using a separate preamp, like a MixPre3 or 6. Then you could have both line input that the theater has captured as well as your best mikes, without any camera handling noise.

Raymond Charette said...

Some people might call you foolhardy to try to turn what used to be a 4 person television program live recording into a one-man does-it-all shoot.
I can only admire your technical skill and willingness to learn.
Best of luck!
P.S. Keep telling us about your adventures in videoland - I'm sure we're all learning.

Kirk, Photographer/Writer said...

Hi Scott, Great question. There is a live band as well as three singers for last night and tonight's shows. It would be functionally impossible for me to record good sound via camera microphones. All the singers and musicians are miked by the same crack audio staff that run the usual shows. There is a person at a mixing board sitting at the rear of the audience monitoring levels and mixing down the music and singers. This is feed to the speakers that deliver the audio to the audience. And, by the way, these folks do much, much better sound than you'd get in a typical club or casual concert.

They run a balanced line about 50 feet to my primary camera position. I plug the XLR connector for this feed into my DMW-XLR1 audio interface. I set the controls to "line in" and then, at the sound check before the show, I listen in with headphones, watch the meters on the camera and set the levels. I try to never let the levels get above minus 6 Db and to never, ever let them clip. The audio from the board, through my interface is applied to both channels in the camera and recorded along with the video.

While I could "set and forget" after the sound check I prefer to continue checking the meters and listening in the headphones to make sure nothing goes awry.

The three other cameras are set to "auto level control" and "limited on" and we record the audio coming through the air so we can use the "scratch" tracks on these three cameras to sync all the video together in post for easier editing.

We can't duplicate the audio set up with our own microphones. We'd be running about 9 different inputs simultaneously and I doubt we'd get anywhere as clean a signal as the audio engineers are delivering directly to my main camera for free.

I think Curtis is discussing recording sound in situations like interviews in which super clean audio is critical because the environment is so quiet you'd be able to hear the noise floor or hiss from the camera pre-amp circuits. Most of that noise is well covered by a live music concert. As long as we match levels we're in very good shape. His techniques are more applicable to narrative projects like movies and drama. For concert audio nothing beats coming right off the boards.

Also, with hand-held microphones the mics become part of the show. Not just visually but they provide their own sound style.

Finally, as many reviewers have pointed out, while the camera pre-amps in most cameras aren't that great even Curtis Judd has give high marks to the pre-amps in the DMW-XLR1 interface. That piece of gear sends the audio signal directly to the video mix and does and end run on the camera's internal pre-amps. And, most of the reviewers have acknowledged that the camera pre-amps in the S1H are superb.....at least compared to everyone else's.

Good question. Hope this is a good answer.

Kirk, Photographer/Writer said...

James, I just bought the adapter that allows for external power in the FP. You can be nearly certain that it will be in the mix next Saturday. That will be our final show in the series (unless they extend) and I'll be playing with all the toys by then. I haven't used it yet because I'm not sure the internal battery will go for an hour and fifteen minutes and I can't go and change batteries on the peripheral cameras during the live show. Once I hook an Anker battery pack to the camera it should run for at least a day. Not that I'll need that kind of stamina. But, yes! I'm dying to weave it in.

Kirk, Photographer/Writer said...

Hi Raymond, I'm doing all this for two reasons. First, I'm having a blast seeing just how complex I can make recording an event and delivering usable stuff to my editor. Every camera we add makes the show more multidimensional and makes it seem like a more elegant production. But, secondly, the theater can use all the help we can deliver. The better we make the programming the more likely they are to be able to generate much needed income from selling the viewing rights from behind a paywall.

Since we're not doing lighting changes as we would do during a regular season show, inside, with a computer lighting board, the unmanned cameras don't take a lot of work for what they provide. I set them up, tweak them each five minutes before the show and then hit record on each of them. The only camera that gets personal attention and coddling is the follow camera. That's also the one that gets the clean, happy audio feed. At times there's nothing going on at the second stage but since the camera is inaccessible during the show it just runs and runs and records the same wall.

But it's really fun to see people's eye pop open when you start setting up multiple cameras all over the place. Most people are used to one or two. Fun to be different.

MikeR said...

Having fun here, going along for the ride. Thanks.

Ronman said...

Really enjoying your sharing of experiences, Kirk. Great detail. Thank you.

Kirk, Photographer/Writer said...

I forgot to mention that I used four different Panasonic cameras last night and that each one ran for one hour and fifteen minutes, in 4K video, and not a single one of them ever so much as showed a heat warning sign. They were flawless. Even the oldest, the GX8 was a trooper. And the footage looked great.

It's one thing to have a camera with the latest specs but it's quite another thing to have camera that's both fun to use and also as reliable as a Lexus.

Ronman said...

Like so many others, I often think newer technology is used to attract sales. It certainly makes sense as a business model. And sometimes newer technology becomes the expected standard by consumers. But I also believe long term users, particularly those more interested in the craft than in gear, tend to value gear more on the basis of its role in being a part of the creative process. Often we hear people speak of its functionality as not "getting in the way", and this being a very tangible value. As my skills have evolved and advanced I find my interest in newer technology has evolved as well. I'm still fascinated by it, but find my interest in acquiring it has become far more pragmatic. It's an interesting dynamic, perhaps due to a maturing view of how I see myself and my equipment in being tools for the creative process, and a realization that it really is the content which matters most. The more I interest myself in the content the less I find interest in the technology. Just my thoughts after reading your post. Thanks again for sharing your experience and insights.