gt3rs

For photography is almost a perfect camera on paper 12 fps mechanical 20 fps electronic at 45 Mpix,  if it has the DPAF of the 1Dx III it will be also very good for action and sports something that the 5d IV was never too good at. Buffer should be also very good as they can write 30 fps at 8K RAW probably at 20 fps 3:2 14 bit should be almost infinite at least in CRaw.... will probably not be blackout free, will have rolling shutter issue with the ES but for 95% of action would be good enough.

For video we need to see the recording limit (I'm betting 30min) and if and up to which fps 4k is super sampled (if it can do FF 4k 120fps super sampled it has a monster sensor and cpu) and DR. I tend to think that up to 4k 30fps is super sampled and then pixel binned but I hope to be wrong.

Not sure the R6 will be that good on video because at the Cinema event they only talked about the R5 but let see, hope to be wrong.

 

14 minutes ago, Trankilstef said:

I honestly don't have a lot of faith in Canon for having a great dynamic range. They were always behind in sensor tech in their stills camera, even though their Cinema EOS line are quite good at it. 

The sensor is basically on par with the A7R4: https://www.photonstophotos.net/Charts/PDR.htm#Canon EOS 1D X Mark III,Sony ILCE-7RM4
This does not mean that in RAW video it will have the same DR.... but for photo the sensor definitely has a really good DR

8 minutes ago, Patrick B. said:

yeah, I’ve only seen one quick and dirty test on YouTube for recovering highlights so it would be great to see some deeper dives.  I hate to believe the IDX is that much lower.

Also would be good that is done with RAW at iso 100, RAW with D+ at iso 200 and RAW with DLog at iso 400 to see witch settings is giving better result and then compare to 10big Log too

47 minutes ago, zerocool22 said:

Yeah I have some worries though. As the dynamic range of the 1DX III doesnt seem to be that great. (and when I compare to Zcam, their 6K has more Dynamic range then their 8K cameras). But fingers crossed. 

Where did you see a real dr test of 1dx iii ? 

As the sensor dr is quite good: https://www.canonnews.com/1dx-mark-iii-dynamic-range-tested-class-leading and so should be the 12 bit raw video.

Random youtube tests are worthless imo.

Some C300 III  4k 120fps sections in here

52 minutes ago, Andrew Reid said:

What prompted Canon to do a completely revelatory camera, in terms of the technology and completely u-turn on their previous strategy of holding back on DSLR video.

Was it seeing so many sales go to Sony?

Was it the need to ensure a new mirrorless lens mount succeeds on the market vs Nikon, Fuji, Panasonic & Sony?

Were they waiting for new manufacturing capabilities to be ready, new factories, to be able to produce such fast sensors and processors?

Did they realise cropped 4K on the EOS R was a bad business strategy?

I'd love to be a fly on the wall at Canon.

Probably a combination of all the above.

6 minutes ago, Trankilstef said:

Where did people see the IBIS thing is real internal sensor based stabilization? Because to me it seems it is more of an electronic satbilization + lens stabilization than real ibis.
Look this frame from the video posted above, doesn't seem to be real sensor shift IBIS, but lens based + ES...

The "IS mode" menu I think is the IBIS as on canon you cannot control lens IS from the menu or at least not on EF lenses....

51 minutes ago, heart0less said:

Yeah, something is definitely going on. I've seen some new lights that boast about 90W output from 384 LED beads.
Right now I get 40W of power from 660 LEDs, which is still plenty.

 

That's right, no magic inside.
Though I have to admit that those LEDs that Lupo used seem to be pretty efficient and bright, nonetheless.

Manfrotto used fresnels in their small LED lights back in the day and they could only achieve 550 Lx @ 1m
(Manfrotto Spectra 500F)

Yep on the newsshooter review is stated that the Lupo uses fresnel nothing wrong with this approach but is not due to led evolution.

5 minutes ago, Avenger 2.0 said:

Certainly these: No dual card recording for video and recording time limit.

I'm not so sure as it could be RAW on CFexpress and 4k proxy on SD..... but we will see. What I'm sure  is that the 1Dx III even with two CFexpress you cannot record on both at the same time, only exception is 5.5 RAW + 4K "proxy". Recording limit will probably by 30 min for all fps and 7 min for 120 fps like the 1Dx III. 

Canon to complete the job add zebra and waveforms!

And include the ND filter EF adapter as goodie in the kit for free😀

4k 120 fps non cropped if it is supersampled is just incredible but probably it will be pixel binned but still 4k 120fps!

8k RAW I was sure it would be possible as the 1Dx III 5.5 k raw 60fps is in that data range.

1 hour ago, sanveer said:

The Lupo, is, surprisingly brighter. It also has higher CRI, arguably lasts longer (the Godox has 3 times the power draw, the Lupo uses NP-F-series batteries meaning that there are mmultiple capacity battery options (meaning you could go for hours on large capacity batteries), and is cheaper too $159 vs $349 (for the battery version of the Godox, the SLB60W). If the Godox is used with its fresnel attachment, it would be brighter.

The Lupo is much smaller and lighter. 

 

https://www.adorama.com/goslb60w.html

Lupo does not include battery, chargers or power adapter so it will be more than 159 but for sure less than 349. I find panel less flexible than COB lights but especially if you are doing interviews, corporates etc.. the new Lupo panel seems a better fit than the SL or SLB from godox.

Just that there is no magic in output, the lupo use fresnel lens inside this is why the beam is narrow you do the same with the godox and the power will be much more. 

11 minutes ago, zerocool22 said:

There is a press conference of Canon on 20 april, and one of Sony on the 30th(they were scheduled at the same time, but sony postponed theirs when they found out Canon was at the same time). I wonder why sony does that, is it because they will include extra features to compete with canon if the canon news is better? 

They will put a lot of *..  *coming in future firmware updates ...😂 On a serious note I doubt is for watching Canon, it is way too late to change things other than price....probably the idea is to be later so the press blogs etc will talk only about Sony announcement and not both at the same time...

Rumor strongly suggests is the C300 iii and Sony has already the FX9 in that price segment so I doubt that there is any overlapping. 

1 hour ago, heart0less said:

Hmm, it sounds almost too good to be true.. 🤔

It'd cost me slightly more than Godox 60W, but, allegedly, the final output would be higher. Not to mention the ease of use (weighs almost nothing and can be battery powered), transportation, etc.

It really got me thinking.

Thanks for sharing it, @sanveer!

The Lupo looks interesting but not sure you can compare with the Godox 60W in output as the Lupo has a beam angle of 40°, not sure with what Godox lux is measured but probably the standard reflector that has a much wider  beam angle. A SL-60W with Fresnel 12°-55° probably at 12° gives 25'000 lux at 1m so at 40° my guess it will be more than 10'000 lux.....  
But the portability and easy to use of the Lupo is hard to beat.

10 hours ago, thebrothersthre3 said:

The S1H and S1(even more so the S1) are just incredible for the price. 

For pro work though yeah the C500 MK2 looks amazing. Too bad Panasonic won't do phase detect AF. That the main reason people are going sony and canon over panasonic IMO. I also wonder why they didn't put a RAW option in the S1H. 

I am still sticking with my Fuji(and Ursa mini 🥰 😉 ) 

I am still waiting on Canon to release something more hmmm cheaper and better. 

The new 1DX is awesome but whats up with the old body with no tilt screen?? Really looking forward to their new mirrorless releases as the EOS R was a bummer imo. Also just saw a comparison between the S1H and 1DX3 and the Panasonic had a fair bit more dynamic range, even though the Canon was shooting RAW.

The test of S1H vs 1DX III it is not scientifically done, not even the iso where matching... even then are the iso really matching between cameras (and we know it is not), was he using the same lens yelling the same exposure etc..., was he using dlog enabled (high light protection) in 1DX RAW or not, was the exposure measured with a light meter and not with the camera meter etc.... it was a quick non scientific test just focus on the highlight and not the full DR that proves nothing... classical youtube quick and dirty comparison.... Maybe still the S1H has better DR but until is tested scientifically is just a random guy doing a random test.... Hopefully cinema5d will test the DR of the 1DX III (also there you could argue about the methodology but is definitely much better than random tests)

2 hours ago, Kisaha said:

..the Sony costs significantly less also (as you mention).

For how long people will be paying the Canon-premium? Almost everything they have in the market is overpriced.

Only thing from Canon I happily overpaid for, was the last 70-200mm 4f. It is a great piece of equipment, nothing else justifies its price for me, especially after the recent releases in hybrids/cine cameras (they also have some very competitive dead cheap EF-S lenses that are ok).

The FX9 seem a very good camera but It cost less as it offer less.... the C500 II can do 6K RAW internal at 60 FF (zero crop) with DPAF, Sony caps at 4k 10bit 30fps FF or when they will ever release the firmware 4k 10bit? 60 fps Cropped ca. 1.25..... Not really the same.....

At 16K there is no camera that can do FF 60 FPS RAW internal with good AF 4k or above.... I would not call it overpriced. For most people in this forum yes but so is the FX9.... 

I'm quite a fan of enabling EIS on post based on gyro and other metadata.

Insta360 allows you to turn it on and off in post so you could decide what is best. For example if you have the camera on a roll-cage in a car you want to have it off but on a helmet you want to have it on. With GoPro's you need to select it first as it does on the camera itself on insta it does it on post. Also in post it could use more processing power of a PC, disadvantage you need to process the files before using it. 

It is very cool on the FX9 that you can do it in post, I hope Canon will follow as they simply need to add the metadata (maybe is already there). 

4 hours ago, Leica50mm said:

Doesn’t do Dpaf in 5.5 raw, right?

Sure it does, I would never be able to track somebody at 560mm or at 50 1.4 while doing actions. The limitation is max 30fps at 60fps is manual focus only. 

Some frame grabs at 5.5k 

Grab from 5.5k 60fps 16mm 2.8 at 1/1000

Grab from 5.5k 25fps 1/50, 560 mm using DPAF

Grab from 5.5k 25fps, 1/50, 50mm 1.4 on Ronin S

8 hours ago, barefoot_dp said:

If all of that is actually true, I'd expect them to price it pretty close to the C500mkII.

Instead of offering clearly tiered products, they'll essentially have two flagships - neither of which does all of what you want, and you'll be forced to choose between 6K/FF vs s35/4K120p. Kind of like the old days of the 1d/1ds before they merged them into the 1dx.

My guess is somewhere in between C200 and C500 II maybe at 11k if you want the FF look mostly for movie you get the 500 II if more doc action etc C300 III with s35..... c200 is the entry level... let see

On 4/5/2020 at 1:02 PM, ade towell said:

 

I don't like to criticize camera comparisons because it takes a lot of effort and is basically impossible to get it right. 

Just a couple of things are puzzling me: 

Kind of a wired way to test DR.... by just looking at the highlights..... also how did he check that the exposure is exactly the same (iso don't match already but also iso between models and brands are also not 100% matching)?  Highlight recovery is not the same as DR.
Also the stabilization test where it is evident (as it should be) that the Panny is better, it is completely ruined by the AF... for something like this I take EIS + working AF than OIS but crappy AF.... as I can put it on a gimbal and solve the issue, much harder to keep the focus.... 
Would have liked to see also the comparison in HFR between 1Dx III 120 fps vs 180 of Panny..... instead he does not even mention the 120fps of the 1Dx....

Maybe I missed it but he should mention that Panny has as an advantage the flippy screen and many more video features like waveform/zebra, etc....  

I use Resolve Studio (paied) to add motion blur for time lapse and works quite good if you have a good GPU, did not try yet with normal scene... you need the payed version for this feature.

19 hours ago, Trek of Joy said:

The press release specifically says broadcast and cinema, this is likely what would have been the NAB announcement. The R line is not a broadcast or cinema line. The R5 will get its own announcement day so the C300mkIII or whatever else they announce doesn't get drowned out by the R5 hype. That's my bet.

Chris

From a marketing prospective it make no sense to have both announcement on the same event. Two announcements means more press coverage and more important spread over time... So I agree with you will be Cinema stuff only.

Some more observations:

I did try out RAW quite a bit and is quite impressive also in Resolve is easy to work with. I will probably use RAW more than what I first thought I would. I wish RAW 50/60 would still have AF but you cannot have it all…

Other than for  shutter speed you can pull a picture from the RAW file and it will be almost indistinguishable from a RAW photo (12 vs 14 bit), in some cases I will use a compromised shutter speed so I can get both video and stills.

On Canon web site you can download the LUTs: https://www.usa.canon.com/internet/portal/us/home/support/details/cameras/eos-dslr-and-mirrorless-cameras/dslr/eos-1d-x-mark-iii/?subtab=downloads-software

For RAW you can decide how much highlight protection by either enabling D+ or Log (yes is confusing), be aware that if you record proxy to this setting will affect the proxy:

As already said the menu with all the resolution/codec is super messy I wish you could enable only certain modes…. like disabling all the UHD modes, IBP etc…. Also the 100/120fps is another separate setting and once you disable it put the camera to FHD….

I find very good the AF that has various option for size like small, surrounding, vertical or horizontal zone etc.. a big improvement over the 1Dx II for video.

I did try again the Active Track on Ronin S and if you set the tilt stiffness to >80 it has no problem with an iPhone on the hot shoe (1Dx III + 50 1.2 + iPhone X max)

Some interesting comparison that I found on youtube (again these cameras are not really targeted at the same audience but still interesting to see):

1Dx III vs C500 II https://www.youtube.com/watch?v=Ab2pj8CEJzE

1Dx III vs RED Scarlet W https://www.youtube.com/watch?v=LUYGQFXlwXY

For video a mirrorless version with flippy screen and the RF ND adapter would make an amazing camera…..

To summary up the findings so far around Resolve Studio 16.2 and 1Dx III files:

- HEVC ALL-I and IPB 10bit 4:2:2  25fps neither Quick Sync or Nvidia 20xx can HW decode them. From our test seems all done in the CPU. 

- A top class desktop CPU 9900K can software decode in realtime both HEVC ALL-I and IPB 10bit 4:2:2 25 fps

- A top class Notebook CPU 8750H can software decode in real-time but really at the limit (not really usable on a timeline without setting half resolution) HEVC but only IPB 10bit 4:2:2 25 fps not ALL-I (this is a strange one as IPB are more computationally intensive, probably the software decoder is more optimized for IPB).

- A top class desktop CPU 9900K with a Nvidia 2080ti can realtime playback RAW 24fps. Probably even at 60fps I would need to upload and send a file to @Trankilstef

-  A top class Notebook CPU 8750H but with a medium gpu 1050ti cannot real-time playback RAW 24fps, CPU is at 70% and GPU at 100% so maybe a gaming notebook with 8750H or 9750H with a 2070 or 2080 is able to do real-time RAW playback. Hope to find it out soon

- A top class Notebook 9750H with 2070 cannot playback real-time HEVC ALL-I 10bit 4:2:2  25fps. We would need to test IPB

I just cannot imagine if the R5 at 8K will also be HVEC 10bit 4:2:2 if there is a machine that can playback those files. Hopefully there will be HW decoding in the future for this file type.

Thanks a lot to @KnightsFan @Trankilstef and @Matthew19 for the tests and contribution