MattH

I wasn't sure whether these cards would make a difference when I read about them on DPreview yesterday. http://***URL removed***/news/2013/09/12/sandisk-offers-new-256gb-compactflash-card?utm_campaign=internal-link&utm_source=news-list&utm_medium=text&ref=title_0_4     In their page it says: "The card delivers minimum sustained write speeds of 65MB/s2"   The higher numbers of 140MB/s and 160MB/s are given as transfer speeds:   "The 256GB SanDisk Extreme Pro CompactFlash card has industry-leading transfer speeds of up to 160MB/s4, designed to dramatically improve workflow efficiencies, and delivers fast shot speeds of up to 140MB/s4 for advanced functions such as fast action and continuous burst mode shots."   Does this mean that they will only write at these speeds for short bursts or/and only when transferring to computer via a reader?

  It actually is equivalent in light gathering terms.  The light per unit of area is different by a factor of 4, but the area of full frame is 4 times greater, so the total amount of light energy is the same.   On equally efficient sensors the noise will also be the same even with the full frame at a higher iso to compensate.   Micro four thirds generally still has a weight advantage overall though.

The highlights in the dngs look harsh with this camera.  They look clipped even when they aren't.  Even in the wide shot of the park which looks exposed to the left, the highlights on the bike, the table cloth and the guys back looked clipped.  It doesn't even change if you drop the exposure down to minus 4, the brightness and contrast to 0 and recovery all the way up: they are still there beaming away like solid blocks of bright.   I will be very interested to see if the Exposure Index Mode you speak of can be utilised to alleviate this.

This footage is good.  It shows you what you get straight out of the camera on the street in normal light with no stabilisation and no extra lighting.   That is what you want from a camera sample!   Over produced graded footage is pointless.   The image is what I expected.  This was never going to be an ultra sharp camera and looking at the previous videos full screen showed that plainly.  The sensor is like a 100% crop of a 18 megapixel full frame camera.  Get a full resolution still sample from a 1dx or 5d shot at 1/50th a second and zoom in to 100% and tell me if it looks sharp.   I don't understand the people saying were is the dynamic range and that the highlights aren't recoverable. The computer i'm on now doesn't have video editing software but all I did was play the files in quicktime, take a screen grab and open in photoshop.  Looking at the histogram on a levels adjustment layer shows loads of room above and bellow the data.  Its all there. (On second look the highlights are clipped but displayed at less than white, but if you leave them alone they are gradual enough)  Just move the left slider to choose the black level and all the highlights are maintained.  The only other thing I did with this frame is increase the saturation by 25% and apply 80% very low radius 0 threshold sharpening to his face using a mask.  sharpening the whole image exaggerates the noise.You probably wouldn't bother doing the sharpening for video, but I think this shows it is good enough for still images.   http://www.flickr.com/photos/99922584@N05/9453075099/sizes/o/in/photostream/   No colour correction was needed for this naturalistic grade.  So if this is straight from camera then the pocket cam outputs very balanced colour.  No green tinge like its big brother.

Just thought I'd let people know that the GX7 in body sensor stabilisation will not work with video according to imaging resource.   I can't confirm it but the article seems pretty confident in its wording:   "The system will, incidentally, work with third-party lenses, or lenses mounted via an adapter, but you'll need to manually enter the focal length in this case. It's also important to note that the body-based stabilization system is not used for video capture."   This appears about half way down the page next to an illustration on how the sensor works:   http://www.imaging-resource.com/PRODS/panasonic-gx7/panasonic-gx7A.HTM   This is a pretty big deal for video as it removes the main selling point of this camera.   I guess its there to compete with the OMD EM5 for stills but disabled for video to not compete with the GH3.   Unless this is proven wrong then I am no longer interested.

  Whoa just read from that link:   "It's also important to note that the body-based stabilization system is not used for video capture."   Piece of shit!

Thanks for the post.  It was interesting to show what a lens coating does.  And it does more than I expected.  A little flare is ok in certain circumstances, but the type of flare illustrated in the article is excessive to my taste.  It shows how important a lens hood would have been in the past.

This is the camera I have always wanted panasonic to create.  If this had come out when the omd em5 came out people would be going mental.  But peoples interest has been diluted by the other possible options.   Black magic pocket camera being the main one.   But for now that is still vapour ware due to black magic's reputation in delivering products. (Although im sure you'll get one soon and for you it makes sense to get all these cams. But for people who can only really afford one $1000+ pocket camera these are in direct competition.)   but even if the poket cam ships in the millions, I think the GX7 beats it anyway on paper.   The black magic should have larger dynamic range, which is an important thing, but that is the only advantage. The GX7 has a larger sensor for shallow depth of field at medium angle of view. Wide angle will be a pain in the ass for the pocket camera: fast wide micro four thirds lenses will not be wide, and ultra wide lenses will only be normal wide and slow. The Pocket camera doesn't have a veiwfinder which is important for stability and focusing. pocket camera doesn't have image stabilization unless you use i.s. lenses.  the widest of which is 35mm equiv. with the GX7 you can use fast primes for handheld video. And not to forget the high megapixel still capability.   I would rather get this and an active 4/3rds mount BMCC.   Of course the stills shooters are intrigued by the possibility of NEX full frame, and I don't blame them.   But even if Sony put sensor image stabilisation in a full frame NEX body.  The chances are that the video quality would still suck.  They don't want to canibalise the FS700 etc.

Good Idea about updating the original BMCC.   But what they need more than anything is an anti aliasing filter.   That rainbow moire is just too much.   A large viewfinder with focus peaking would help too.  Basically a small camcorder form factor or a maybe a bit like ikonoskop (but ambidextrous) since it doesn't have to appeal to stills people.   But I cant see this happening, and I cant see either the pocket or the 4k coming out in any real numbers this year.

I wish I'd known about this sooner.  This is right in my interest zone.  And I can make it to Manchester easily.  Oh well, I will be sure to catch the documentary when it comes out in full.

Lenses NOT fully compatible with the new AF system EF14mm f/2.8L USM EF24mm f/1.4L USM EF100mm f/2.8 Macro EF400mm f/2.8L USM EF500mm f/4.5L USM EF600mm f/4L USM EF1200mm f/5.6L USM EF16-35mm f/2.8L USM EF17-35mm f/2.8L USM EF20-35mm f/2.8L EF24-85mm f/3.5-4.5 USM EF28-70mm f/2.8L USM EF28-70mm f/3.5-4.5 EF28-70mm f/3.5-4.5 II EF28-80mm f/2.8-4L USM EF28-80mm f/3.5-5.6 EF28-80mm f/3.5-5.6 USM EF28-80mm f/3.5-5.6 II EF28-80mm f/3.5-5.6 II USM EF28-80mm f/3.5-5.6 III USM EF28-80mm f/3.5-5.6 IV USM EF28-80mm f/3.5-5.6 V USM EF28-105mm f/3.5-4.5 USM EF28-105mm f/3.5-4.5 II USM EF28-105mm f/4-5.6 EF28-105mm f/4-5.6 USM EF35-70mm f/3.5-4.5 EF35-70mm f/3.5-4.5A EF35-80mm f/4-5.6 EF35-80mm f/4-5.6 PZ EF35-80mm f/4-5.6 USM EF35-80mm f/4-5.6 II EF35-80mm f/4-5.6 III EF35-105mm f/3.5-4.5 EF35-105mm f/4.5-5.6 EF35-105mm f/4.5-5.6 USM EF35-135mm f/3.5-4.5 EF35-135mm f/4-5.6 USM EF70-210mm f/3.5-4.5 USM EF70-210mm f/4 EF75-300mm f/4-5.6 EF80-200mm f/2.8L EF80-200mm f/4.5-5.6 EF80-200mm f/4.5-5.6 USM EF80-200mm f/4.5-5.6 II EF90-300mm f/4.5-5.6 EF90-300mm f/4.5-5.6 USM EF100-200mm f/4.5A EF100-300mm f/4.5-5.6 USM EF100-300mm f/5.6 EF100-300mm f/5.6L That's a lot of lenses that don't work.  I wonder why they don't.  And what about third party lenses.  I guess it will be a case of trial and error.

Ok, so there are different explanations as to what is happening from the sensor to what is recorded. I will use my limited understanding to try and suppose what might be happening. The 5d mk3 is interesting in that its native horizontal photosite resolution is 5760 which happens to be 3 times 1920. which means that 9 photosites are being used (or should be being used) to get the values for one output pixel. The image bellow shows how this would apply to a small section of the sensor. 5d mk3 bayer pattern to 1920.bmp As you can see there are 4 possible pixel types. Thomas could be right in saying that the output is YUV. However RGB to YUV conversion could happen regardless of how the RGB values are obtained from the sensor in the first place. There are 3 rough methods I can think of for obtaining the RGB values: Method 1. The most information would be obtained by reading all photosites of a particular colour and averaging them to obtain a colour value for that output pixel. All photosites would be used. So for example in the top left group of 9 photosites, 5 green samples would be averaged to for the green value, and 2 blue and 2 red samples would be averaged for their respective values. Method 2. If however Andrew is correct that no calculation is done on camera, then this would mean only 1 photosite from each group of 9 would be used for each output colour value. Meaning you are throwing away 2 thirds of the sensor information and therefore 2 thirds of light gathering ability / noise performance. Method 3. Or if it turns out the final image still needs debayering then this means only one colour is being recorded for each group of 9 photosites. Which means only 1 photosite out of 9 is used: 8/9ths of the information thrown away! If Method 3 is the case then I would say this is quite poor. A 4K raw frame from the 4kBMPC from would offer 4 times the resolution (even after downscaling the BMPC) and 3.1 times the light gathering ability (taking into account sensor area difference of 2.9* but not photosite size) If method 2 is true then the 5D MK3 and 4K BMPC will be fairly evenly matched in terms of resolution and light gathering ability. If the Method 1 true, which is probably unlikely, then the 5D would be simply beastly. It will be interesting to see what the case is.

The screen tearing looked more of a problem though. I cant see how that was happening in daylight either.

I'm a bit confused about the resolution thing. I presume the "full frame modes" use the full height or width of the sensor (or both depending on the aspect ratio), That would mean that the camera is down scaling from the full resolution of the sensor to the output resolution. But that would mean the data is not raw in the sense that it has to be debayered. Does this mean you can only have video output in the full frame modes, or is it kind of like 14 bit bitmap frames with a blue, red and green component for every pixel? In relation to this, your list of resolutions also confuses me: 1280 x 1280 (16:9) how can that be 16 by 9? Is that a mistake?

      I didn't find the "smearing" unpleasant.  You could easily call it blooming and say it was attractive.     What i DID find unpleasant was the fact that you could see the joins between the sensor quadrants.   That totally negates the advantages of this cam really.   Edit:  Just watched again and realized that the bloom itself only appears in certain quadrants and you were probably talking about this.   If it was one big sensor the blooming might not be too bad.  but it is,  and even without the blooming you can see the joins.

How would you change the aperture with the mft version?  You could have electronic control, but that wont work with the bmcc.   They should make it with a full manual aperture control build in.

      Ive just worked it out.  with a full frame to mft it will make the bmcc APS-H.   Metabones will squeeze the full frame image to the full image circle of micro four thirds which is utilised by the gh2 in its 16:9 mode.  And the crop factor between bmcc and gh2 16:9 is 1.205.  The crop factor between full frame and aps-h is 1.254.   Hopefully you're right that they will make different "squeeze" options.

If black magic made an e-mount version of the bmcc, then this very adapter would make it essentially a super 35mm camera.   As the crop factor of the bmcc compared to super 35 is 1.6

I think people are looking for a term that is analogous to "photographer":   Someone who uses a camera but shoots motion instead of stills.   The reason I don't like "Videographer" is it can give the impression (whether fairly or not) of someone walking round with an automatic camcorder just pointing it dumbly at things.   Whereas "Photographer" has inbuilt connotations of artistry.   The good thing about "photographer" too is that it doesn't imply pro.   You can be a amateur photographer or an enthusiast photographer.  To imply pro, you would probably say "pro photographer".   You don’t have to be a high earning pro just to take part in the activity of photography.  So shouldn't the same apply for cinematography?   Perhaps "Motion Photographer" would be a good term for most enthusiast video people?  : more artistic sounding than "videographer" but less pretentious and assuming than "cinematographer".

"The cards have various amounts of video memory from 1GB to 4GB. You can edit Blackmagic Cinema Camera raw with 1.25GB of video RAM but running Resolve and Premiere at the same time will require 3GB."   How do you know this please?  Is it from personal experience or someone else’s personal experience that you know?  Did you/they try running resolve and premiere simultaneously on both a 2GB and 4GB graphics card?  What happened with the 2GB version.? Did it just fail or run more slowly?   I'm currently specking out a pc and up to now I'd kind of concluded that 2GB would be enough.  Most tests are bassed on games and consistantly prove no benefit to a 4GB card.  But for the sake of future proofing I may be inclined to go for 4GB if the evidence is there. As its seems to be around £100 extra for a 4GB version of a graphics card that would otherwise offer no benefit it’s a pretty significant question for me at the moment.   Any info or a forum thread url would be really cool.

Just been looking into this a bit and apparently running SLI does not increase the amount of ram. The memory is instead mirrored. In a game it boosts the performance so much because it renders frames alternately so both cards are running the same game and so have the same data in their ram. So it looks for video editing that SLI is not worth it.

WTF? The GTX 680MX is a mobile GPU. These imacs aren't desktops, they are high end laptops on a fancy stand.

Boy, that apple fanboy demon left your body with one last bitter sarcastic hissy-fit of a struggle, but hopefully we’ve seen the back of it now.

Cool interview. He answered everything extreamly well and to the point. A bit irritating that the headphone adapter runs throug the evf connector. Maybe they can provide an EVF with a headphone socket, Or a stereo mic. (He says as if he can actually afford the thing) EOSHD: There’s a coding on modern Leica M lenses for vignetting, is that compensation done in video mode on the new Leica M? It won’t need it as the vast majority if not all vignetting will fall outside the 16:9 crop.

[quote name='OverCranked' timestamp='1348186225' post='18773'] Comedy is regional. Tragedy is global. You consider it a joke calling a senior engineer a " LIER " in public forum. You consider me as a person with lack of sense of humour for pointing out to your possible ignorance to Japanese culture. This is not British Culture. This is you joking around with disregard for others' honours and opinions. Hey c'mon you have sense of humour you should be able to take a little criticism. What's all this agitation about !? [/quote] I didn't call him a liar. I made him a character who called himself a liar. It was a comedy sketch by definition. You're right I don't know a lot about Japanese culture. I know a bit but it doesn't interest me that much.I certainly don't subscribe to the concept of honour. Also, what agitation?