tpr

Most of the people in the comments disagree with you, so why on Earth would you assume that everyone else who read the post without commenting agrees with you? I don't think you can rely on your site statistics being an indication of how many unique visitors you have either. There are probably quite a few of us checking in regularly on the progress of this discussion.

If I thought for one second Clarkson endorsed bullying and abuse in offices I wouldn't be defending him.

I don't think anyone here is arguing that Clarkson positively endorses it. People have various ways of rationalizing their own behavior or excusing their own mistakes as exceptions, but if he's going to do things like this, there should be consequences. The actions of any individual are invariably explainable if you look closely enough at the circumstances of their lives, and you can have all the sympathy you like for someone who is going through personal issues, but it just isn't sensible to allow someone in a poor state of mind to hold a position of responsibility.

It should be possible to defend Clarkson whilst not being seen to be endorsing his behaviour.

When people say things like "I don't mean to be rude, but...", they generally follow it up with something rude. When they say "I'm not racist, but...", they generally follow it up with something racist. It's a familiar way of anticipating an objection and maintaining some deniability, so when you preface your comments by saying you consider what Clarkson did wrong, but go on to argue that he still shouldn't be sacked, you can maybe understand why some of us take that to be a less-than-convincing denunciation.

I don't think anyone here is arguing that you positively endorse his actions (that word again), but you are clearly downplaying their significance. Your explicit attempts to justify applying a double standard to talent vs non-talent are also deeply troubling to me.

If you like Top Gear, you should be angry at Clarkson for messing it up, not the BBC. What's happened is completely on him. The fact that Clarkson is despised by the left is irrelevant, the "PC police" can't be blamed for him punching the producer, and he should be held accountable for it like anyone else would be.

The same standards have to be applied universally. That means minor crew and it means talent such as Clarkson, David O. Russell, Christian Bale or anyone else. The fact that some talent have gotten away with vaguely similar things is no argument for allowing Clarkson to as well.

And why do people keep saying that punching someone isn't against the law? It is!

​I couldn't agree more with this statement.

As for Andrew's comments about how important Top Gear is to the BBC, the only relevant question is whether the allegations against Clarkson are true or not. The profitability of Top Gear is completely irrelevant unless you believe moral exceptions should be made when profits are at stake. Imagine someone making the same argument about Saville. It would be an absolute disgraceful position to take. The allegations against Clarkson aren't anything like as serious, but the argument about moral consistency still applies.

But colour distinctness is how I identified the raw image.

 

I wasn't making a technical analysis - I can just see that there is more colour information in B. Regardless of white balance or tonality. It's just there, plain as day. Not just in the red swatches, but everywhere. It doesn't really matter if the WB was off a bit. The hues will still be more varied and distinct.

 

I spent years in art school and grew up in a house full of painters so perhaps I'm more tuned in to colour than the average filmmaker, but I see it all the time in 8-bit video. The Canon C-line is the only 8-bit camera that I can't distinguish from 10 bit (they must do something very clever with compression). Put most 8-bit video next to Prores, RAW, etc, and you can see it straight away. It's not just about what can be done with the image in post, or dynamic range. You can see it as soon as you lift saturation. The colours look alive.

 

...

As jcs notes, both images are now 8-bit jpegs, so this talk of 8-bit versus 10-bit isn't entirely convincing.

Second, you'll notice that the blue color patches 6 and 8 appear more similar in the B image than in the A image, unlike the reds in 31 and 37 which, as you pointed out earlier, are more similar in A than B. It makes sense that the reds will be less distinct and the blues more distinct in an image biased towards a warmer white balance.

Look at colours 37 and 31 in image A (reds, 2nd line down). They look like almost the same colour, yes?

 

Now compare them in image B. I think that is what you call bit depth, yes?

Those colors may appear closer together in A because of the warmer white balance. I don't think distinctness of colors is a good way to identify a raw image. Think of a raw image in its flat unprocessed form. There is a lot more information there, but before processing, all the colors look very neutral and similar. The colors appear more distinct when tonal contrast is increased and we can't in principle know whether the contrast in a set of images was introduced via in-camera jpeg/video processing, color grading or processing on a raw image, so it's a bit hard to tell which was responsible.

I would still put my money on A being the video and B being the raw, and for essentially the same reasons that other people are using to justify the opposite conclusion: B has less sharpening, B has uncorrected chromatic operation, B has less contrast. An unprocessed raw file would have all of those attributes, so my guess is that Andrew has been less heavy handed with his raw processing than whatever workflow produced the video still.

Live stream discussion about the Samsung NX1 from Photokina:

 

You might as well join the fun on the new bright side and change the name of the blog from EOSHD to something closer to today's reality.

"Non-EOS 4K - Directed by Mirrorless Filmmakers"?

A fairly obvious question missing from both lists is if they have any 4K capable cameras in the works that will be priced within the reach of non-professionals. In the same vein, you could ask whether we should expect to see other video-oriented features like higher frame rates, zebras, peaking, image stabilization, quiet operation, aperture control on Nikons, and so on, which could lead to an interesting discussion of why projects like Magic Lantern exist and the competing demands of stills and video on a hybrid camera.

I'd also be interested in whether they are planning to enter the higher-end the mirrorless market.

In general, I'm much more interested in finding out what to expect in the future than badgering them about the lack of emphasis they've placed on video in the past, and I agree with what others have said about the tone of many of these questions. I think you could ask the same questions in a much more positive way. They can be just as difficult, but I think it's better to ask them in a way that makes it sound like you genuinely want to know the answers rather than simply telling them off.

I mean... This doesn't really look that bad, lol.

It's hard to tell what it would look like without youtube's compression, but there are some issues:

1. Moire at 54s.

2. Rolling shutter at 8m18s.

3. It's basically your fault if you don't succeed even if the economic system is rigged to direct more wealth to the wealthiest at 2m39s.

According to the press release, they've added zebras! Woohoo!

Anyone know when the official release will be?

This mechanism on metal legs looks like something out of The War of the Worlds.

You don't need an ftp client to access the Arri Alexa files. Just use your browser: ftp://ftp-footage.arri.de/ and plug in the user name and password.

There are some panning shots in this AX100 clip from around the 7:40 mark, but there doesn't seem to be anywhere near as much rolling shutter as in the other video:

 

 

The GH4 really does offer everything doesn’t it?

I'm sure a few people will be put off the base unit when they realise it needs to be powered from an external source.

Where can I find these "video" stop numbers you are referencing? 

 

Any info on the Nikon D610 and D800?

One place you can find them is DxOMark. They score sensors and lenses according to various metrics. The sensors in the D800 and D610 share the very top position of DxoMark's ranking for dynamic range at 14.4 Evs, with the D5200, D5300 and D7000 being the highest ranking among APS-C format cameras at 13.9 Evs. In general, the sensors in Nikon models currently dominate DxoMark's rankings, especially for dynamic range.

http://www.dxomark.com/Cameras/Ratings/Landscape

(those values apply to stills, but it is possible that part of the dynamic range is truncated for video)

The color depth of a given image can never be increased -- not without introducing something artificial.

I would think that you could increase the color depth of an image if you reduce its viewing size without changing its resolution, just taking a step backwards for instance.

If I understand correctly, a 2K image downscaled from 4K via the method described in this thread will have greater color depth than either (1) a 2K image downscaled via a less clever method, or (2) a 2K crop of the original 4K image, but I'm not entirely sure I understand how color depth is quantified.

BTW do you know for sure that readings from a still will translate accurately to video mode?

It's a good question. I just tested by taking a video and a still of the same scene using the neutral profile. Their respective histograms have peaks in the same places, but the video's histogram looks smoother. That makes sense given that the video was 8-bit and the still was 14-bit raw.

Do you have any D5300 stuff online yet? There's so little online it would be great to see some ...

Nothing yet. Only got my hands on it last month, but I'll let you know when I put something up.

Are you sure about Active-D? I thought I'd read about needing to turn that off somewhere.

 

It's never worked with video on any Nikon. Pretty sure that's still the case with the D5300.

This is the problem with the D5300 for me with my un-chipped primes - I have no metering, let alone a histogram! 

You can still view a histogram even when using a lens that doesn't give you metering. Just take a test still or video, then in playback mode, press the up arrow to cycle through the information views you have enabled. By default the histogram view isn't enabled, so you have to change the settings under MENU->PLAYBACK MENU->Playback display options. I have it set up like this:

I can't claim that these are THE best settings, but this is what has been working best for me so far:

 

In the menu:

Picture Control (profile): I'm currently of the opinion that a Neutral profile is best most of the time. I don't think Flaat is necessary, but I do think you loose out on a lot of perfectly useable detail in the shadows/highlights by shooting with the Standard profile - and for no good reason unless you don't want to grade. You can always put contrast back in post.

 

I set all of the following to OFF in the menu: Active D-Lighting, HDR, Long exposure NR, High ISO NR and Auto ISO sensitivity.

 

...

 

I'm using Auto WB quite a lot.

Active D-Lighting, HDR and long exposure NR don't have any effect in video mode so you may as well leave them at whatever settings you would otherwise have them set to for stills. High ISO NR does affect video though. I'm not sure about auto ISO. I wouldn't use auto WB. It's bound to make the WB vary somewhat from shot to shot depending on the colors that appear in the frame, so could make it harder to match shots in editing.

As for picture controls, I think all this talk about flat profiles is a bit off the mark. What makes more sense to me when dealing with an 8-bit codec is using a profile that spreads the tones that contain the detail you're most interested in as widely as possible over the histogram. Normally, that should mean getting it to look as close to how you want it in the final grade. For a very low contrast scene though, you could give yourself more room to play with by using a higher contrast profile than usual. If you have to increase contrast in some tonal range in the grade, you'll run out of in-between tonal values and end up with banding effects, but when reducing contrast, you won't have that problem.

I think the word 'contrast' is a source of confusion here, so I think it's worth saying something about that. What people think of as a low contrast image is really a trade-off in contrast between the mid tones on the one hand versus the shadows and highlights on the other. In other words, a 'low contrast' image corresponds to a tone curve that is steep in the shadows and highlights, and less steep in the middle. If you use a picture profile with this kind of tone curve, it will produce a file that records more of the tonal gradations the sensor differentiates in the shadows and highlights than in the mid tones, which isn't necessarily a good idea if the mid tones are much more important than the shadows or highlights because you leave yourself with fewer tonal gradations where you want them.

Rather than just use one ideal picture profile, I think it's better to choose different profiles for different shooting conditions. I've made a few of my own using the ViewNX2 software that comes with the D5300.

I'm pretty impressed by this footage. The detail is amazing.

Does this mean we could also convert 8-bit to 9-bit by downscaling 1080 footage to 720?

If so, would we gain anything by delaying the color grading process until after scaling down to the resolution that the footage will be displayed at?

 

Slow motion test shot with D5300 this weekend.

Very nice stuff!

So I'm asking you to stop picking on the little guys and get comfortable with the idea you've outgrown a certain demographic.

The standards Andrew is applying are constantly raising higher, but that's understandable. Andrew is maturing as a filmmaker and the cameras available at the moment are much more advanced than they were a couple of years ago. If he were judging the D5300 by the standards he applied when the 5D mk3 came out, I'm sure it would be a completely different review. That could easily come off as inconsistent if not properly explained.

For those who do enjoy the D5300, feel free to let your views be heard. I don't have anything against it. It just doesn't really turn me on. Especially not as a 5D Mark III owner, as even the stock Canon ALL-I video mode on that is better than the D5300 and it's full frame.

When you reviewed the D5200, you tested it alongside the 5d mk3 and posted some comparison video that impressed me at the time. If I remember correctly, your opinion was that they were almost on a par for image quality. I've now tested the D5200 and D5300 alongside each other and the improvement in the D5300's low light capability is quite obvious to me, so I wouldn't be surprised if it surpasses the 5D mk3 with original firmware. Of course the 5D mk3 has a greater feature set and all that, but I'm just talking about image quality.