For this I filmed a resolution chart on my computer screen from a distance, using Quantum 50-100 settings which use quality settings that are already over the line for 2fps be it AVCHD or MJPEG. I loaded the MOV files into AVIDemux and extracted 100% lossless snapshots to compare the various modes. Shot on M Movie Mode, aperture and shutter speed were both fixed.

Note: I'm only showing the relevant 100% screen captures but all the files generated in this test (and I tested both HD and VGA modes) weight in at 38.5MB and can be d/l here: http://www.mediafire.com/?u2h8ycc2c6vmj0a

Unless marked as AVCHD all other files are MJPEG.

Myth 1 - MJPEG 1080p is MJPEG 720p up-scaled.

Result: Plausible.

It's too close to call, really! Even though there is evidence of a very very small resolution advantage to 1080p, it is hardly noticeable and it could be a consequence of the input-output workflow of 720 being done in a smaller resolution.

Evidence: (look at the second image, I can't delete the first which is similar)

Photo

Manufacturer specs

• 91Wh capacity
• 620g
• 25.4mm depth
• 12V cutoff voltage

Thermal behavior

Start

Depleted battery

OKS 15mm T2.0 + ND

OKS 20mm T1.9 + ND

Yashica 50mm f1.4 +ND

OKS 15mm T2.8

OKS 20mm T2.8

OKS 25mm T2.7

Yashica 28mm f2.8

Yashica 50mm f2.8

OKS 15mm T4

OKS 20mm T4

OKS 25mm T4

Yashica 28mm f4

Yashica 50mm f4

OKS 15mm T5.6

OKS 20mm T5.6

OKS 25mm T5.6

Yashica 28mm f5.6

Yashica 50mm f5.6

OKS 15mm T8

OKS 20mm T8

OKS 25mm T8

Yashica 28mm f8

Yashica 50mm f8

Lens Flare test. Added Yashica ML Zoom 35-105. Same as previous test all shoot in RAW, processed in DaVinci Resolve with ARRI LogC timeline and Alexa LogC to Video LUT. No any additional color correction.

OKS-15mm-T2.0

OKS-15mm-T2.8

OKS-15mm-T4

OKS-15mm-T5.6

OKS-15mm-T8

OKS-15mm-T11

OKS-15mm-T16

OKS-20mm-T1.9

OKS-20mm-T2.8

OKS-20mm-T4

OKS-20mm-T5.6

OKS-20mm-T8

OKS-20mm-T11

OKS-20mm-T16

OKS-25mm-T2.7

OKS-25mm-T4

OKS-25mm-T5.6

OKS-25mm-T8

OKS-25mm-T11

OKS-25mm-T16

Yashica-28mm-f2.8

Yashica-28mm-f4

Yashica-28mm-f5.6

Yashica-28mm-f8

Yashica-28mm-f11

Yashica-28mm-f16

Yashica-50mm-f1.7

Yashica-50mm-f2

Yashica-50mm-f2.8

Yashica-50mm-f4

Yashica-50mm-f5.6

Yashica-50mm-f8

Yashica-50mm-f11

Yashica-50mm-f16

Yashica-Zoom35-105mm-35mm-f3.5

Yashica-Zoom35-105mm-35mm-f5.6

Yashica-Zoom35-105mm-35mm-f8

Yashica-Zoom35-105mm-35mm-f11

Yashica-Zoom35-105mm-35mm-f16

Yashica-Zoom35-105mm-50mm-f3.5

Yashica-Zoom35-105mm-50mm-f5.6

Yashica-Zoom35-105mm-50mm-f8

Yashica-Zoom35-105mm-50mm-f11

Yashica-Zoom35-105mm-50mm-f16

!

Hope this helps, Cheers

**Some AC90 specs **

• 1080/60p, 60i, 30p, 24p, 480/60i. [50 Hz model]: 1080/50p, 50i, 25p, 576/50i.
• High sensitive BSI 3MOS sensor
• F1.5
• 29.8 mm wide angle / high powered zoom lens (12x)
• built in 5 axis hybrid OIS
• 3.5 inch lcd (touch) 1.150.000 dot resolution
• Dual SD Card Slots (support SDXC/SDHC/SD Memory Cards).
• 2 channel XLR audio inputs (MIC/LINE/+48V selectable).

• Audio CH1/CH2 with individual level controls and input selection.
• HDMI output, AV Multi output (D-connector, composite video and audio 2 channel), USB connector.
• Remote terminals enable remote operation of iris, focus, zoom, Rec Start/ Stop functions.

Available at:

• http://www.amazon.com/Panasonic-AG-AC90-Full-HD-Handheld-Camcorder/dp/B00AA2P9UE

In all the excitement with the GH3 (some 1200 replies in 5 days here only ) I wanted to start a new discussion, because I feel we (or at least some of us) have to face the reality. Should we buy the GH3 or start filming again with a camera that is made for it ?!

I have the GH2 for 2 years, have 3 different lenses and tried several hacks, I bought the Tascam DR-40 for sound (and other stuff). I needed a rig to hold everything, and now it looks stupid. If you are not doing fiction, (always) filming on a tripod, or in a studio and you also want to be fast and flexible (which is often necessary) it just doesn't work! (for me ;-)

Therefore I am considering a 'normal' video camera again. But which one? Next month the AG-AC90 will be available for only € 1800,= . I got it yesterday to try it out for 1 day. As there is no information (or footage) about the camera, I wanted to do some video comparison.

I compare footage of the AC90 with the GH2 (As we do not have the GH3 yet, but I don't expect it to be a miracle anyway ;-). It was not my intention to make good shots, but to put both cameras in 'hard' circumstances, such as high contrast, back light, movement, moire, rolling shutter effects etc. I did different tests and tried to keep equal settings, same shooting moments and movements, tripod / handheld, slow motion, rack focus / auto focus, daylight / low-light / dark, menu / settings. I only had the camera for a day (without a manual), so I don't know it well. Most images in standard settings were a bit too bright (but this can be adjusted in the camera).

The AC90 has a smaller sensor so less DOF (but in daylight can be achieved with a ND filter.) I didn't have one during testing (but I did use a ND in some shots with the GH2, so that is not a fair comparison). I expected the AC90 to do worse in lowlight but I am actually quite surprised!

The video is definitely not a 'scientific' pixel tweaking test. I know little about how to test GOP, I / P/ B structure, 4:4:2, peaking, dynamic range, AVCHD profiles, etc. But you guys do! @Vitaly_Kiselev @Driftwood @Ralph_B @cbrandin and others ?! I can put some original files online for some deeper analyses. What can you tell about the footage? GOP, I/P/B, etc? Is it hackable (or worth hacking ;-) to give it a bit more Mbs? etc?

I hope to open up the discussion (from the large sensor cameras such as the GH3) and look forward to hear everybody's opinion!

Ps1 Some shots in the end are only with the AC90 because it was too dark for the GH2 (with the 14-140 mm lens)

Ps2 I know YouTube is not good, but I also think it is only showing HD 25P as I rendered it 50P (and the orginal looks better and more smooth)

Ps3 Sorry for the stupid music (without sound was even more boring)

I dont know if its just me... but I see a lot of noise and not such great quality especially on the C4K madden video which is weird. All shots are straight from the camera.. no color grading.. I know the editing and shots arent so great... but i just wanted to upload these to get your opinion... I purchased my gh4 off ebay through what seems to be a camera shop's account and they were running a promotion on the GH4 for $1,399.99 and I am using a Panasonic Lumix G Vario 14-42mm f/3.5 - 5.6 lens. Please let me know what you think... is it fine and im just overreacting? should I try to return the camera to get a new one?

MADDEN (C4K, 24p):

This little plastic fellow, using h.265 maybe the best video to come out of a dslr - maybe on par with the Red Epic. Look for yourself.

Although I have worked with still photography and video operating I haven't had much experience with this more artistic film work so I would love some feedback. Most of the shots I was on a very wide aperture, around 0.95-1.4 and kept my ISO very low. I haven't really done any colour correction and I am not entirely happy with the edit. I did it more as a test and rushed the edit a little whilst at a friends house (so I couldn't concentrate 100%). One thing I have noticed is that there is a "judder" effect going on constantly throughout, like it is missing some frames. Not sure if this is because I am using a cracked version of premiere or its the way I exported it.

Anyway here it is (its kind of depressing).

DOUCE CHINE: Technical detail for every shot (Camera/lens/Aperture)

Birth of the “Douce Chine” Project: Twice a year I am teaching a 10 weeks (25 hours) filmmaking course in Shanghai. www.shanghai-classes.com/filmmaking For the 2014 spring term, I decided to help one of my friends, by shooting a music video for his new Album: Douce Chine (China Delight). This was an interesting project for my students as well, as we had time to work on lighting set up and camera positioning. Song is about a guy who is used to travel from France to China, mixing up jet lag feeling with night life, good wine and wave of exciting emotion spicing up his life in China. This 4 minutes video was integrally shot in 10 hours (2 afternoons) in our classroom which is only 25m2. That means we could only walk 5 steps max in the different directions of the room.

Why multi cameras shoot? Out of my teaching work in photography and filmmaking, I am both a pro photographer and DP/cameraman. I am having lots of gears and this shooting was a good place to put all of my cameras at work for the good of the students; Nikon D800, Lumix GH2, Lumix GH3 and Panasonic AF100. I hired one of my Photography students (Richard) to be my main assistant and he brought along his Leica M240 he had never tried in video mode so we decided to give it a try on that project. Music video is the place for multi camera shooting as you cannot ask the singer to repeat the exact same performance many times in one afternoon. Despite the lack of space in our classroom, we managed to position the camera to cover as many angle as possible in our 3 different sets of background: Simple black back drop, (fake) Bed and Chinese curtain, Window with (fake) street light reflection. There is no use to get many cameras in action if their position and angle of view are not both different and matching.

Image bank from extra cameras On the image shot for the music video, I add a layer of images I shot in the previous years in France and China. They were shot with GH1 (night shots and on board shots) Panasonic HVX200 (club and slomo of people walking) and HVX200 with Letus Ultimate (Water shot, airplane, traditional places…). I followed up with the HD evolution but all my HVX shots (720p 50fps) still looks real good even blew over scaled to 1080p. Great cams always existed and any day, I can still grab my HVX200 if it matches my need on a new project.

Which lenses? I am having some great lenses and I used some of my favorite on that shooting; Voigtlander 25mm f0.95, Nikon AIS 50mm f1.2, Nikon 85mm f1.4 AFD, Computar 75mm f1.4 (a rare but cheap CCTV lens great for portrait) Prior to the shoot, I bought a 35mm f1.8G (DX version) which is doing very well on the D800 even in FX mode as I knew I would add some vignette in post. I am always taking advantage of this type of experimental project to test new things without pressure. The GH2 was hanged to the ceiling on top of the fan with the Voigtlander 25mm which was wide enough to get the establishment shot at once. The D800 shared the 35mm and 50mm at f2. The GH3 always used a Speed Booster (great tool that I highly recommend) with the Nikon 85mm 1.4 which is an amazing lens (older AFD version half price than the new G version). The AF103 mainly used the Computar (200 USD lens). For the Leica, Richard brought along his 50mm f0.95, 35mm f1.4 and 90mm f2. Those 3 lens costs more than my complete equipment, but they are already sharp fully open and get their own character which makes their full frame image looks more like Medium format camera.

Light and camera settings. This is not a proper camera test as the aim was to complete the music video shooting. Therefore, we didn’t try different set up on camera or aperture on lenses to compare. Instead, we kept everything pretty much the same. I am used to set up a lighting to be able to shoot at f2 on all camera at ISO200 (base iso for GH2, GH3 and AF100). This Aperture offers great sharpness on all my lenses and it also allows me to use lenses of any angle of view from 12mm to 180mm at f2(with speed booster) on M43 (which means 24mm to 320mm on FX). I am also working at f2 most of the time as I know the combination of lights able to deliver this result which is time and budget saving. There is just some slight correction according camera position versus light position. Playing a bit on the ISO or the aperture can handle that. On the other hand, Richard wanted to go shallow depth of field (f1.4) on the Leica and sets the ISO at 100. Sometime, he went too much on that direction (while being close to the subject) and you can notice that some of Leica’s shots are a bit out of focus as depth of field is too limited. I guess that Richard went as far as f0.95 few times on the 50mm due to noticeable overexposure on some shots. Bokeh Otaku!

Main differences between cameras - GH2 and GH3 are classic, great tools, the most efficient, my favorite and they get even better with Voigtlander lenses or Speed Booster + Nikon. They are easy to set for a flat image for post processing purpose. Still, I think the GH1 was the best in term of cinematic colors out of the box. Image is sharp and it is up to your treatment to smooth them up. Those cameras are easy to handle due to their size, articulated screen, electronic viewfinder with all needed information minus peaking. - AF100 (AF103): Often described as a weak product by people who don’t own one and just judge on technical data, this camera is a workhorse, the one I pick up first for my paid job due to its form factor and ergonomics. Great for sound recording and flexible with its ND filters (the feature I am missing the most on GH series). Image quality is good and it had never been an issue for me or my client. Interesting point, it is the most accurate camera of the bunch in matters of color rendition. Green is much more right on the AF100 than on the Gh2 or Gh3 for instance. I love this cam and it is far better than its reputation. - D800: I used it once for a very short movie just for a test. (

Final words: Camera model, Full frame or M43 does it matters? This was not a camera test, but a real world shooting with very different gears. As far as I am concerned the possibility to get a flat image out of the camera, working on a nice lighting set up, taking care of camera positioning and making a wise choice of quality Lenses are more important than camera type, model or brand. Cinematic look is not about crazy shallow depth of field. Watching great classic films help to reset some mind on that matter. Pilling up image with crazy smooth out of focus area is sometime so obvious that it takes over the story itself most of the time… The real advantage of FX is to be able to offer more separation between element with wider lens (even with an f2.8 zoom) and at longer distance from subject. If one day I am trying to shoot a fiction with the leica, I am going to write a story with that idea in mind and I might shoot in flat light, colorless environment to compensate the high saturation of the footage.

Franc Péret French Filmmaker/Cameraman/Photographer/Teacher based in Shanghai cn.linkedin.com/in/francperet View my video on (need a VPN from China): http://vimeo.com/channels/659428/videos Filmmaking and Photography classes www.shanghai-classes.com

4 biuli

SOUNDTRACK

André Tasso

This is simply a handfull of shots put together for testing 3 things:
• a new mini rig set - though it's a bit pretentious to call that a rig :P
• a tweaked to taste Vashi Nedomansky's VashiMorphic40 template
• and incorporating LUTs to my workflow, in this case Fujifilm Eterna

Thanks to my dear gifted André Tasso =)
also thanks to Vashi and @TrackZillas

gashô

OT: CB 'Oh God, now I will do best better photos'.

from vimeo description: carrying on my TL experiments - see previous - I fitted my shitty android phone (running cyanogenmod 7) in the window and took 301 pics with camera FV-5. processed them in lightroom and exported to panolapse, which is as simple as an application gets. still tweaked the yaw, roll and zoom and exported as a deflickered quicktime movie. added an spiced John Cage's Dream loop and some hisss. voilà!

you all have nice weekend (a la goddard)

gashô

I'm still developing/experimenting a workflow I feel good about: one within the tools I have and know, that delivers quality enough and time wise optimized.

As always all feedback, comments, thoughts and secrets are most welcome.

gashô

Credit to Hunter Richards' Tri-X B&W trick - http://tinyurl.com/hunterrichardsTRIX

maybe it's not necessary to reiterate, but vimeo just obliterated (macroblocks the size of satellites) the fine grain on this video test. It's ok... I can live with that. all feedback, comments and thoughts are welcome.

gashô

IMHO , love the Nokton at all , nikkor got strong bokeh cool colors and is sharp , Fujinon got very nice pearly bokeh but is very soft .

hope You enjoy cats ; )

This is the trailer for The Crazy Life. It's feature film being shot with hacked GH2.

http://multimedia.fnac.com/multimedia/editorial/labo/hybrides_et_reflex_2013.pdf

I managed to use my school's dolly to pull off some test shots for smooth camera movements. Take note that I used and controlled everything singularly: I pulled focus, performed pans, and physically pushed the dolly itself. It was difficult, but I managed to get some usable footage. With one or two extra grips I can easily pull off great looking, smooth shots. I'll be shooting more with the dolly this week. I like grain.

• GH2 w/ Driftwood Hack
• Canon FD 50mm f/1.4
• Sigma 30mm f/1.4
• 16mm grain overlay

Brett Plank 2012

Just for fun to play with my new wide lens. No special grading.

Camera : Panasonic GH2 AVC-Intra (Driftwood VY Canis Majoris Day PM) Lens : Tokina AT-X 116 Pro DX AF 11-16mm f2.8 Gear : Sachtler FSB-4 75 CF, Gini Rig Follow Focus, Lilliput field monitor Tools: FCPX

Locations: Paimpont forest, sometimes said to be the Arthurian Brocéliande, is in the French commune of Paimpont, near the city of Rennes in Brittany. As Brocéliande it had a reputation in the Medieval imagination as a place of magic and mystery. It is the setting of a number of adventures in Arthurian legend, notably Chrétien de Troyes's Yvain, the Knight of the Lion, and locals[citation needed] claim the tree in which the Lady of the Lake supposedly imprisoned Merlin can still be seen today.

Mont Saint-Michel (Saint Michael's Mount) is a rocky tidal island and a commune in Normandy, France. It is located approximately one kilometre (just over half a mile) off the country's north-western coast, at the mouth of the Couesnon River near Avranches. The population of the island is 41, as of 2006. The Mont-Saint-Michel and its bay are part of the UNESCO list of World Heritage Sites.

==MY TEST==

I went on a shoot this weekend, and was using Sandisk Extreme 16gb 30mb/s cards. I wanted to ensure both spanning and playback in camera for the shoot, so I went with driftwoods "spanmybitchup" patch. I did just a few short preliminary tests: the test I did was to shoot the papasarts star chart (which I've attached--I also attached the more extreme codec test chart v2 2010), and let it run. I waved my hand around in the frame from time to time and also had some talking to get audio. Everything seemed good: I was getting spanning and smooth playback in camera with audio. (EDIT: also attaching Stray's death charts with color information!)

==MY SHOOT==

So during the shoot, after a few shots, I was getting the message "motion image cannot be played" for all my clips (or something like that I don't remember exactly)--my heart sunk! An actor asked me to check to see if we recorded something on a previous take, I had to say that "yes it was" without checking. I was worried about requests for playback--and this was only with one actor: imagine this on shoot with clients! For certain shots, I wanted to be sure that something was recorded since I wasn't at the camera, and I couldn't do this. The scene I was shooting was an actor in a car at night--he was lit and most everything else was dark. That's when I got the messages. I put in a new Sandisk card, and formatted it. Reshot the same scene and I was getting the same error.

==LESSONS==

1) In camera playback is important at least for some of us, and for some, like me, it is even more important than low gop. For me, I am willing to forego a very very high quality low GOP setting (awesome work driftwood!!) that doesn't have playback in camera on good SD cards for an improved long GOP with ultra reliable in camera playback. I'm NOT saying that everyone wants this or should want it (especially if you have the cash for more super high end SD cards or for extra harddrive space). But I'm guessing some people are in my situation as well.

2) It seems there's a relatively standard procedure for patch development (using streamparser, tweaking settings incrementally, etc etc). BUT there seems to be no standard procedure for testing spanning and in-camera playback. You might not think that makes a difference, but it seems that it does: it seems to matter what you shoot when you are testing things like spanning and in-camera playback. Or, at least the number of clips you shoot or the ISO you shoot at makes a difference: otherwise, I don't know how to explain the divergence between my pre-shoot tests and what happened on the shoot. I welcome some constructive thinking about standardizing spanning and camera playback tests.

==DESIRES==

The ideal for me would be to get a hack that improves quality over the stock settings, that has very very reliable spanning and in-camera playback with audio, and works on good quality cards like the Sandisck 30 mb/s (16gb). I know some people are getting the SandDisk extreme pro 64gb cards... but a lot of people (like me) already have a stash of good quality cards and don't want to fork over even more money to get these super high end cards.

I did test w/ @cbrandin 's (Chris's) 66M patch. We know that this doesn't span (but getting around 10 or 11 minutes is okay for me, although I'd love spanning). But what I did notice is that I couldn't get in camera playback with this either. I'm going to try the 44M patch.

Thoughts?

I just wanted to share the FOV of the Panny 100-300. There was alot of wind so there was excessive RS at the telephoto end. I'll probably do one more of these tests with no wind to see if it lessens the severity of the 'vapor' artifacts. I think that will always be there but it depends on the weather. I don't believe OIS or hacks can take care of it. If you guys know of how to minimize it, please share. I may be asking too much of the cam cause the reach is an insane 1,560mm (w/ETC) :)

Basically just went out with camera and shot. Nothing planned.

Smooth -2

Lenses:

Tokina 11-16mm 2.8 Zeiss ZF 50mm 1.4

Singh-Ray Vari-N-Duo Vari-ND and Circular polarizer in One

ps: I use rgrain for grain, but youtube just kills the grain so that is the werid stuff you see in the sky, looks much better in original!

Here is the Vimeo version where you can download the original if you like.

Here are some before and after stills from grading:

Random fun test with GH2, Olympus 38mm 1.8 and Sedna AQ1 at 2am in Silverlake after a film shoot.

Graded in Colorista II