tugela
-
Posts
840 -
Joined
-
Last visited
Content Type
Profiles
Forums
Articles
Posts posted by tugela
-
-
Oh oh.... I thought the slew of slo-mo was insane after the fs700 launch.... Prepare for an avalanche.
Until it gets old. Current cameras can do up to 120 fps footage now, but other than the routine test clips people do, it does not appear to be used much.
-
With all due respect no you did say that Sir. That's exactly what you said, that it will not match the C line (which makes it lose one of it's major intended uses) and your logic in the post I quoted was that is had a different size sensor, pixel count, sensor technology (unknown fact) and that it has a different processor model. Enlighten me what did I miss?
My logic behind the counter argument: In the short film Trick shot it intercut perfectly with the 4K images out of the C300II (not even just C100/300), and the following facts:
-They are from the same company, the same sensor manufacturer.
-They have similar pixel count at 8mps (12mp in stills mode irrelevant as it's cropped in video mode to 8.3mp 16:9, identical to C100, C300, C100 MKII)
-They have the same colour science and Log Gamma, and WDR profile, meaning identical sharpening levels and sharpening algorithm, saturation, hue, contrast.
-They have XF-AVC codec at 4:2:2 and 305mbps which is closest to matching the C300II of any camera on the market, and HIGHER quality to matching the C100, C100 MKII and C300.
-It's 4:2:2 and higher than 50mbps for ability to be used for broadcast with the C300I, C300II with both having standard .MXF files
-It has the exact same DR of the C100, C100 MKII and C300 at 12 Stops with the exact same highlight/shadow distribution and the exact same ISO values
-It has a higher resolution than the C100, C100 MKII, and C300 and lower than C300II but we've seen even those intercut perfectly.
-It has the exact same downsampling method used in the C100, C300, C100II from 4K to HD (a properietry process that invloves no debayering and double green channel resolution)
-It has similar grain and noise structure (extremely fine and small) but of course it has more of at higher ISOs above 3200 ISO (although it's no slouch it's 5D3 level with a 1" sensor)
If you're into reading specs, it's clear the XC10 is the perfect C line match B camera, but I am not that much into reading specs, so I'll wait for side by side footage.
Yes you can but let's wait for an identical side by side tests just to make pixel peepers happy and sure.
The only thing that I am getting from the reaction to XC10 footage, is that viewers sadly are not familiar with how resolution differs from sharpness. They got used to internally tack-sharpened footage from Sony cameras and GH4 therefore see a high resolution non-sharpened image to be soft and sub-par. Look at 1DC 4K images in C-Log, absolutely no sharpness whatsoever because it adds zero internally, which is one of the reasons why that camera has a distinct filmic, non-video look. And while it looks softer than internally sharpened A7s 4K, when sharpened in post it shows the exact same sharpness and resolution detail.
The XC10 has downsides but resolution is not one of them, it's very high, just non sharpened internally in C-Log.
I think I am going to be sharpening the hell out of my footage for clients from now on since it's considered superior resolution. Really.
I said that the XC10 cannot match the C100/300 because the sensor is completely different, not any other of the extrapolations you are choosing to make. Mt statements to that effect were, are, and will remain true, irrespective of how you choose to twist it. It is a different sensor with different properties and characteristics dude.
FYI, an 8mp sensor inherently cannot resolve fully at 4K because of the effects of debeyering. At best it will have around 70% resolution. In order to maximise resolution you have to collect a larger pixel sample size, and none of these Canons can do that.
I know the difference between resolution and sharpening, and trust me, there is no way in hell that most of that HD footage is going to look anything but soft, sharpening or no sharpening. The detail simply isn't there. Look at the foliage in the clips, not the large objects. Those fool you, because your eye locks onto the macro object rather than the detail. Foliage is the acid test. With the XC10 it is mostly just green blobs. No sharpening can recover from that. The camera apparently is inadequate or the operators don't know what they are doing.
-
But we can say, because we've already seen it in the C300 II launch film!
I would rather see it from actual users than promotional material which has been heavily altered.
Certainly if we look at the HD user footage of the XC10 that has been put out so far, it is pretty clear that resolution is an issue in that mode.
-
So you've concluded the XC10 will not be a matching B camera for the C100/300 images as it has a:
1 -Different sensor size and pixel count therefore different sensor technology (which is 100% unknown to anybody outside Canon Semiconductors)
2- Different processor model?
Okay.
Instead of listing the hundred things they do have in common I'll say we wait to see actual side by sides to know whether it will work as a B camera for the C line or not.
I didn't say anything of the sort. Try reading my response again, you will find that it is accurate. Don't extrapolate to things that were not said.
-
No, but having the exact same gamma, color, processing, CFA, and sensor tech means it'll intercut pretty well with its bigger brothers. I don't have the link on hand, but the C300 II launch film had a couple XC10 shots in it that looked just fine. It'll stick out less than GoPro footage in Hollywood flicks--I couldn't believe the pov shots in the second Hobbit flick actually made the cut.
I still don't think the XC10 has a lot of appeal for people in our price bracket, but YMMV.It doesn't have the same sensor tech as the C100/300. Those have a larger sensor and lower pixel count than the one on the XC10 (I am guessing that the C300/100M2 use the same sensor). They also have DPAF, and there is no mention of that in the XC10, so it probably doesn't. The sensor performance and characteristics of the two will be completely different.
Also, the XC10 has a single DV5 processor, whereas the C300 has dual DV5 processors. That means it will have superior processing capability, which has to affect the IQ otherwise there would be no point in the C300 having dual processors. The C100 uses a single DV4 processor (which is the older generation) running at a different bit rate and codec. So it is different from the other two in that respect as well.
-
-Codec is broadcast approved as in it's 4:2:2 (the pre-requisite for broadcast acquisition, and 50mbps and higher (up to 305mbps). NX1, RX10, RX10II, FZ1000, RX100IV, GH4, A7s, A7r, Nikons, all are 4:2:0 h.264 and therefore will never be used by broadcasters. This is a major selling point and Canon thought this out carefully when they gave it a 4:2:2 codec that's higher than 50mbps, it opens an entirely different market for the camera,
Also remember the XF-AVC codec is currently supported everywhere, it's a normal .MXF file that imports natively Premiere, Final cut, Vegas, Avid, everything. It's not a new unsupported codec that needs transcoding as many seem to think.
-About C100/300 compatibility, I highly believe this is wrong, it does match the C line perfectly from what I've seen so far, haven't seen side by sides yet (which will determine . but the resolution is the same (higher than C100/C300 MKI/C100 MKII and close to C300 MKII), the colour science is exactly the same as well as the gamma curve in C-Log, the codec is the higher than the C100 MK & II and perfectly matching the C300II (XF-AVC), it lacks aliasing/moire, has similar low rolling shutter, lowlight performance is great up to 6400 ISO with the same grain structure, it has the exact same Dynamic range at 12 stops and and the same highlight/shadow distribution at 800 ISO, The camera even uses the exact same way of downsampling 4K to 1080p as in the C100/300 (no de-bayering, what they call super-green sampling). It's clear it was specifically designed to match the C100/300. It's most probable this is the closest you will get to matching the C100/300.
Let's wait for the side by sides with the C cameras though before drawing conclusions.
You can output 4:2:0 as 4:2:2 if you want to. After all, this is what pretty much every camera that has a HDMI output does. There is nothing special about it. What it does do is give you more latitude when grading since there is more color information, but if you are not doing that then 4:2:0 is fine. This camera is being marketed as a portable news camera, and that sort of footage is used as is after transcoding.
XF-AVC isn't supported everywhere. Some people have to transcode it until their NLEs support it. Maybe that will happen soon but apparently it hasn't happened yet in some cases.
I am not that familiar with C100/300 footage, but if the resolution is as bad as the first clips from the XC10 are, then I would be very surprised.
There is no way that it has the same performance as the senior cameras, since it is a different sensor size and pixel pitch. The only thing the same between the cameras is the codec, and I don't consider that to be "compatibility".
-
Why are you so angry at this camera?
I am not angry at it. I am disappointed because of what it signifies. Although I had no intention of buying one because of the form factor, the implementation reinforces the idea that Canon simply do not take consumer/prosumer video seriously. The DIGIC in the camera is the next generation of processors that are going to be used in all Canon consumer products for the next 2 years. In other words every consumer camera that Canon releases in the next 2 years is going to have the same soft crappy video they are so famous for.
That makes me sad. It should make you sad as well.
-
It isn't a second camera for a C300/500 since the image quality is dramatically inferior. That pretty much destroys it's supposed main advantage.
Shooting in HD limits you to 50 mbps at 60p, and 35 mbps at 30p. Something like the NX1 can shoot HD at 80mbps at 60p, and 60 mbps at 30p, using a more efficient codec. Some of the bandwidth of the XC10 will be sacrificed to support 10 bits and 4:2:2 as well, so relatively speaking even more resolution information is lost. The Canon codec isn't "broadcast approved" either, it is simply a format that can be readily transcoded, that is all. But so can all of the other codecs out there.
-
Guess you're not an editor then...
BTW, last time I asked you to back up your post about all the purple fringing with links you ignored. So I will assume it's most likely BS.
Sigh.....look at this then, the backlit trees, especially the "graded" material around 1:20 which enhances it:
-
How do you know what the final edit looks like, can you share it?
No amount of editing is going to fix that dude.
Getting back to the quality of the footage. It is soft, just like all the other examples of HD footage that is showing up over the last few weeks from new users of the camera. It isn't YouTube that is doing this, it isn't the shooters that is doing this, it is the camera. Full stop. I was quite surprised about that. While I had serious doubts about the form factor and design of the camera, I never the less expected that the footage at least would meet modern standards, but the evidence so far is that it clearly does not. I have given this some thought and I believe I know what the problem is. The new Digic processor can do 4K at high bit rates, but for HD it is probably still using the old encoding engine from the Digic 5/6 family, which as we well know produces soft footage that is effectively 720p resolution. The slightly higher bit rate accommodates the 4:2:2 output, but the resolution is no better than Canon's previous consumer/prosumer camcorders, in other words lacking by todays standards. More shortcuts and lack of competitiveness from Canon it seems.
You want to judge the camera by your prior expectation that it was going to "pleasantly surprise" us. I prefer to judge it based on the footage people are producing with it. For me, it is "surprising", but not in a pleasant way. We have not seen to much 4K for this camera yet, but the HD output that has appeared so far falls below todays standards, it is last generation quality while the world has moved on.
To be blunt, there are much better options than this thing considering it's price point and market. IMO the RX10M2 is probably going to outperform it, based on what we know about the RX10 and extrapolating that. Time will tell with camera as well of course, but I think it is going to prove superior to the XC10.
-
Excuse the last post, but the editor on this web site is horribly bugged.
Anyway, that clip is VERY soft. I guess if you are used to using Canon cameras it might seem ok because most of them shoot what is effectively 720p resolution footage, but compared to modern cameras from most other manufacturers it is lacking.
One more thing, don't wave the camera around like that, it is terrible to watch.
Most of the XC10 clips I have seen on YouTube, both the 4K and especially the HD ones, seem to have poor resolution.
-
-
I've had and posted examples of moire (which is a form of aliasing) in 4k. And downscaling will only make it worse.
It's not a big deal, not often, I just didn't agree that downscaling removed it.
I did a series of shots a few weeks back with all of my old cameras capable of doing some sort of video, with a mind to making a youtube movie showing the issues with them and how video has evolved over the years. The last two sets of clips were the NX1 in HD mode and 4K outputted as HD in post. All of the clips were shot at default settings. Aliasing was common in the HD clips, but the same scenes shot in 4K had none. Other than that both sets looked similar when conformed to HD.
So, the advice I would give is shoot in 4K and conform that on output to HD, and it should solve the issue the earlier poster was whining about.
-
Yes it is.
Hasn't been a problem for me. There is aliasing when you shoot in HD, but the same shot in 4K downsized in post to HD shows none of that.
-
This has been my experience too after some months with the NX1. Worst of all is frequent aliasing. I simply cannot recommend this camera if your intent is video. NX1 seems to under perform its specs and I think it's a bust in terms of video capability. Stills is another story. For 3 grand you can get a way better video camera (body plus 16-50 package).
I pulled out an my old weather worn GH2 today and did some home vids. Looked way way way cleaner and artifact free than the NX1. Samsung needs to get serious about optimizing this cams alleged potential, these firmware updates are just nibbles along the edges, the camera just isn't that good at video, it needs nearly a perfect storm of small lighting ratios and carefully chosen subjects that won't alias to get a good image in video.
Shoot in 4K and deliver in HD. Aliasing isn't a problem then.
-
I modded my Canon 17-55mm 2.8 by ripping off the back. I use it on an EF to BMPCC speedbooster and it works great. I know you are a BMPCC fan. Did you happen to do the mod on your Canon EF-S 10-18mm f/4.5-5.6 and use it with the EF to BMPCC speedbooster? The reason I ask is I need to know whether the modded lens will hit the front optical element of the speedbooster at the extremes of zooming.
I didn’t realize the Samsung 15-50mm S is more like the Canon 16-35mm F2.8 L. Thanks for the information. But if the Samsung 15-50 is more like the 16-35mm 2.8 L why would you prefer it over the Canon EF-S 17-55mm 2.8? The Canon EF-S 17-55mm 2.8 is sharper overall than the Canon 16-35mm 2.8 L. Why would you pay $500 more for a lens that is less sharp? This is my problem. Everyone points out what a great deal the Samsung NX-1 body is but they fail to mention you have to swap out an $800 lens and pay $500 more. And now you are telling us the lens that costs $500 is less sharp.
I doubt it’s noticeably better. Do you have a link to this information? My problem is any company that is a competent lens designer can come up with an excellent 50mm 1.8 lens. Fast zooms are a different animal. 50mm prime lenses are one of the best understood and least complex lenses to design. You see this time and time again on internet forums. People assume that the optical quality of the Canon 50mm 1.8 STM, 1.4 and 1.2 L is vastly different. Well if you stop down a little they are virtually identical when viewing the final product (print or video). The big difference between them is build quality, focusing system, and some performance wide open. I just don’t think at f/2.8 and f/4 any normal person is going to be able to tell them apart. And don’t get me started on the 4k myth. I have routinely used Canon 50mm lenses on APS-C bodies with 18+ megapixels for years. Now 4k is supposed to be stressing these lenses? Seriously?
Overall my problem with Samsung and a lot of second tier camera makers is the lack of choice. I mean it’s great if you want to shell out $500 more for build quality and less than a stop faster lens that isn’t as sharp but what if I don’t… or more realistically can’t? It’s nice to have a cheapy zoom, something in the $800 range and something substantially north of $1,000 on offer. Considering my BMPCC cost less than $500, $500 is a lot of money to me. So if you want to spend over $250 on a “cheap” nifty fifty that’s your choice. But I prefer to have the option of paying something closer to the traditional $100… like the $125 Canon 50mm 1.8 MK II. I tell every person who has the slightest interest in photography to buy a Canon 50mm 1.8 STM. At $125 there is no excuse not to own it. But $256? That is not a casual purchase.
For the Samsung 45mm F1.8:
Resolution:
Chromatic aberration:
For the Canon 50mm F1.8:
Resolution:
Chromatic aberration:
The Samsung is clearly the winner.
If you are buying a ~$1500 body, an extra $100 on a lens is trivial.
-
Specs sound near perfect on the A7rII but I really wish sony would redesign the form factor. I really dislike the handling of the A7s and find it's short battery life and awkward ergonomics beyond frustrating. The form factor of the GH4 is much more to my liking but Panasonic needs to make a mirrorless camera with a larger sensor.
Sony fix the form factor or Panasonic offer something larger than MFT, whoever moves first gets my money. Canon and Nikon can keep living in the past.
Yup, the form factor is the biggest turn-off with the A7 series.
-
I think what you are beginning to realize is the Samsung NX1 isn’t a free lunch. They can knock $200-$300 off the retail price but they more than make it back when you have to buy their lenses. I have a BMPCC and BMPCC EF Speedbooster. I could go out and sell them both and add $100-$200 and get a Samsung NX1, but I would end up hemorrhaging cash buying Samsung specific lenses.
The Canon 18-55mm 2.8 is $800. Samsung doesn’t have anything comparable. You have to spend over 60% more for the Samsung 16-50mm 2-2.8! Want a cheap OEM nifty fifty? Canon 50mm 1.8 mk II is $125… Samsung 45mm 1.8 $256! Over 100% more!
If you use only manual lenses and have no need for things like aperture control, autofocus, and IS then the NX1 could be for you. But if you want to replicate Canon functions like IS for a comparable price you are out of luck.
Samsung has made a truly revolutionary product but they haven’t delivered the total package. Those lens prices need to come way down.
The Samsung 15-50mm S lens is only a few hundred more. Remember, it is a faster lens than the Canon, and faster lenses increase in price very rapidly. The S lenses are comparable to Canon's L lens range, and those cost in the region of $1.5-2K. The closest equivalent lenses optically for comparative purposes are the 16-35mm F2.8 L or 24-70mm F2.8 L, and those lenses cost a lot more than the Samsung.
Regarding the 45mm F.8, it is optically a much better lens than the Canon 50mm version (an important consideration when you are using a high resolution sensor such as that on the NX1), and it has better build quality. An extra $100 is not a deal breaker, you get what you pay for in that particular case.
When you compare equivalent lenses, the Samsung versions are generally cheaper than the Canon analogs, or offer more features for about the same price.
-
Interesting.
It seems there is a huge portion of people that feel if you don't shoot 4K now in 2015 (on a narrative feature), you're in trouble.
Including HOSTEL horror director Eli Roth:
Watch below
I'm guessing because what you shoot today might only be used a few years from now, and the expectations will be for the standards then, not the standards now.
-
Here is the link to the original 1080120p H265 clip in which at 1:22 in the Youtube video there is a frame drop:
There don't appear to be dropped frames when I step through the original as far as I can tell.
-
Anything that uses a mirror in its light path is probably going to be less than ideal for video, so I don't see the 5D4 can possibly win this contest as a video camera. An A7s mark II will kick the stuffing out of it.
-
----- I still miss --- you can suggest more if you want
- EVF switch in video mode
- magnification help for MF in video mode
- with Samsung lenses: camera does not remember focus settings between modes - eg. manual focus should remains manual focus even I switch between photo/video mode.
- crop in video mode- higher bitrate for 4K and 120fps1080p
- 240 fps 1080p (at least in crop mode)
- RAW videoI doubt the camera can shoot at 240 fps in 1080p. The sensor read rate is 240 fps, but that would leave no time from processing, so the maximum practical read rate is probably 120fps at 1080p and less at 4K. Since 4K would need ~4X the processing, it would work out to 120/4 = 30, or what we currently get). That is likely the limit of what the hardware can do short of dropping significant quality. It might be possible with a crop though, since the camera can effectively disregard a large chunk of the data.
4K60p will require the next generation of hardware IMO.
-
APSC camera oustells FF because it is cheaper, and the mass sales come from standard people, not the passionates or the pros.
You guys are very funny. Say a system is not dead because it ranks good on Amazon. LOL . . .
Steve Jobs would be shocked in his tombstone.When I say APSC is dead I am not talking about Amazon ranking, I am talking about THE DIRECTION photo and video is taking, and this direction is towards FF and not APSC which sets the death of APSC.
Of course it will sell on Amazon this year, the next year and year after. You are very funny ....No it isn't. The direction photography is taking for most people is lighter and smaller, which is NOT the direction FF takes them. You are confusing what YOU want with what most people want. It is not the same thing.
Unless you want razor thin DOF or extreme wide angles (and most people neither want nor need that) FF does not offer significant advantages over crop sensors. Crops sensors on the other hand have better reach, something which most users would find more valuable. For the average user something like the APS-C size is the ideal compromise between large and small sensors, not FF.
-
The Nx1 says you're wrong...
APS-C cameras outsell FF versions by a large margin, and for good reason. The idea that APS-C is dead is absurd.
Some new Canon XC10 footage
In: Cameras
Posted
Maybe you should get glasses then.