[markr041]

 http://www.eoshd.com/content/11982/video-quality-charts-february-2014


Examples: League 1 is the Alexa Studio; League 2 has four RAW-video cameras; and league 5 has the Sony VG900 (so this ranking is not perfectly correlated with expense).


The RX10, the GH's, the Canon DSLRs, and the RX100 are in there (no camcorders though).

 

[jogiba]

Whatever floats his boat but a BMCPC is a showstopper for me since I want to view my videos instantly with no PP and a 3x crop does not have the full frame look like the VG900 or shoots full frame 24mp jpeg/RAW stills .

http://farm9.staticflickr.com/8345/8198535296_efb2fcfe6f_o.jpg

 

[brunerww]

Neither does an Alexa. And it shot every motion picture nominated for best cinematography this year. Just sayin'

 

[Tugela]

Putting the RX10 at the bottom of the list does not make a whole lot of sense....certainly it is better than any of the Canon DSLRs for video. My guess is that they were considering lens selection above overall image quality, contrary to their stated objectives.

 

[thedest]



Quote:
Originally Posted by Tugela  /t/1517624/heres-a-link-bound-to-sp...ngs-of-cameras-by-video-quality#post_24353280


certainly it is better than any of the Canon DSLRs for video.

No, its not. The RX10, better than a 5D? The RX10 is simply nothing next to the 5D. Its a kids toy, next to the 5D.


All the listed Canon cameras are capable of raw video. If you look at the list, he is saying that the 5D, for example, when shooting raw, is on a different league, compared to the RX10. And when its using the ALL-I, its a lot better than the RX10.


What you guys must know is that the entire planet knows the advantages of better codecs, including raw. This forum is the only place where the majority still thinks that camcorders with a horrible codec are serious cameras.

 

[thedest]



Quote:
Originally Posted by brunerww  /t/1517624/heres-a-link-bound-to-sp...ngs-of-cameras-by-video-quality#post_24353148


Neither does an Alexa. And it shot every motion picture nominated for best cinematography this year. Just sayin'

And lets not forget another important thing. NONE of those movies were shot in 4k. Most of them were shot on a much lower resolution.


The Arri Group is not excited about 4K. And guess what? Everyone just LOVES the ALEXA. Their image quality is simply gorgeous.


And why? because the amount of pixels is not important as the quality of the pixels. The Hobbit was shot in 4k. Whats the difference between The Hobbit and the Oscar nominees? The difference is that it looks worse!

 

[Tugela]



Quote:
Originally Posted by thedest  /t/1517624/heres-a-link-bound-to-sp...ngs-of-cameras-by-video-quality#post_24353655


No, its not. The RX10, better than a 5D? The RX10 is simply nothing next to the 5D. Its a kids toy, next to the 5D.


All the listed Canon cameras are capable of raw video. If you look at the list, he is saying that the 5D, for example, when shooting raw, is on a different league, compared to the RX10. And when its using the ALL-I, its a lot better than the RX10.


What you guys must know is that the entire planet knows the advantages of better codecs, including raw. This forum is the only place where the majority still thinks that camcorders with a horrible codec are serious cameras.

The criteria wasn't raw/not raw or anything to do with codecs, or how well it edited, the criteria was the overall visual quality of the output.


I have a T3i with ML installed, and the video from that thing is horrible. The color is poor and the only way it is useable at all is if it is locked down on a tripod and the focus point doesn't change. I don't have a 5D, but I suspect that it would suffer from similar issues in it's stock formation.

 

[markr041]



Quote:
Originally Posted by Tugela  /t/1517624/heres-a-link-bound-to-sp...ngs-of-cameras-by-video-quality#post_24353816


The criteria wasn't raw/not raw or anything to do with codecs, or how well it edited, the criteria was the overall visual quality of the output.


I have a T3i with ML installed, and the video from that thing is horrible. The color is poor and the only way it is useable at all is if it is locked down on a tripod and the focus point doesn't change. I don't have a 5D, but I suspect that it would suffer from similar issues in it's stock formation.

Since the "visual quality," including the color, of video from shooting RAW is almost wholly dependent on the post processing in an editor, it mostly reflects the competence of the user not the camera. So, perhaps you should work on your editing skills with RAW video before jumping to conclusions based on your own single try at RAW video.


We have seen many really bad videos from RAW cameras because of poor processing by the user. I have shot RAW using ML on the Canon EOS M and it looks fine, even at 720P:

 

[thedest]



Quote:
Originally Posted by Tugela  /t/1517624/heres-a-link-bound-to-sp...ngs-of-cameras-by-video-quality#post_24353816


The criteria wasn't raw/not raw or anything to do with codecs, or how well it edited, the criteria was the overall visual quality of the output.


I have a T3i with ML installed, and the video from that thing is horrible. The color is poor and the only way it is useable at all is if it is locked down on a tripod and the focus point doesn't change. I don't have a 5D, but I suspect that it would suffer from similar issues in it's stock formation.

It is, Tugela. The criteria has everything to do with codecs.


As you can see, the Canon 5D appears two times.

Quote:
League 2


1. Canon 5D Mark III (Raw Video)

Quote:
League 3


1. Canon 5D Mark III (stock ALL-I codec) and 1D X

As you can see, the 5D is on 2 different leagues, depending on the codecs.


And im an ex-owner of a 5D and a 60D. And let me tell you, the video quality on the 5D mark 3 is in another league compared to the other Canon DSLRs. Even when shooting videos without the ML hack.


I think that, at the moment, the most limiting factor on most cameras available is the codec. Most cameras have great sensors and decent lenses.

 

[markr041]

This is not quite correct: "I think that, at the moment, the most limiting factor on most cameras available is the codec. Most cameras have great sensors and decent lenses."


How the high-resolution sensor is downrezzed to 1080 video also matters a lot. That processing alone (pixel binning, etc.) accounts for why the GH3 video is far superior in resolution to that of all the Sony (alpha, NEX) and Canon large-sensor cameras shooting video, and why the Sony RX10 video is miles sharper (in the good sense) than the Sony RX100 II, which uses essentially the same sensor, and rivals if not beats the GH3 in real 1080 resolution. Doing the downrez correctly requires a fast processor chip, which the Canons and the older Sonys do not have.

 

[thedest]

Yep. You are not wrong. A good downsampling is VERY important. The 5D also has a great downsampling, and it performs very similarly to the GH3. The same doesnt happen with other Canon DSLRs.


A bad downsampling brings a bad resolution and some artifacts. A bad codec also brings a bad resolution and even more artifacs. And bad colors. And bad gradations. And a bad dynamic range.


Anyway, both of them are important. I was just choosing one that I think its worse.


Now we are entering the 4k era. 4k requires a lot of pixels, so cameras like the GH4 are shooting in crop mode, using all the pixels for the final video, with no downsampling needed. I do think that downsampling wont be a major problem in the near future.

 

[flintyplus]



Quote:
Originally Posted by Tugela  /t/1517624/heres-a-link-bound-to-sp...ngs-of-cameras-by-video-quality#post_24353816


The criteria wasn't raw/not raw or anything to do with codecs, or how well it edited, the criteria was the overall visual quality of the output.


I have a T3i with ML installed, and the video from that thing is horrible. The color is poor and the only way it is useable at all is if it is locked down on a tripod and the focus point doesn't change. I don't have a 5D, but I suspect that it would suffer from similar issues in it's stock formation.

Maybe its good we dont all agree about colour

i never realy liked the colour on my GH2 after my previous 550D

The EOS M i now use at times for video has the same colour i like

i am Talking colour only it cant do 4K or very large dynamic range like some of course

 

[MTyson]

There's no way that the Red Dragon is that low. It shoots upto 6k, has 16.5 stops of NATIVE dynamic range exceeding that of the best film stocks, has been said to have better highlight roll off than film, native 2000 ISO, shoots up to 6k resolution, high & low frame rates, new and improved color science and shoots upto 21 stops DR with HDR-x. I'd take Red Dragon over everything. It has it all, amazing DR, good color science, high resolution, high & low frame rates, superb low light, expanded DR in times where even 16.5+ stops is not enough. Seems like there is some bias against Red, because Red Dragon is a leap over Mysterium-X. The Red Monochrome ranking over the Red Dragon is a joke.

 

[brunerww]

Specs aren't everything. Only two pictures have been shot with the DRAGON to date: http://shotonwhat.com/camera/red-epic-dragon-camera


Compare that to 283 for Alexa (not including Alexa Plus, XT, etc.): http://shotonwhat.com/camera/arri-alexa-camera


Overall, I agree with Andrew's rank order. Over the past few years, I have read his hands-on comparisons between most of the cameras in Leagues 2 through 5, and have pretty much agreed with his conclusions.

 

[Tugela]



Quote:
Originally Posted by flintyplus  /t/1517624/heres-a-link-bound-to-sp...ngs-of-cameras-by-video-quality#post_24354175


Maybe its good we dont all agree about colour

i never realy liked the colour on my GH2 after my previous 550D

The EOS M i now use at times for video has the same colour i like

i am Talking colour only it cant do 4K or very large dynamic range like some of course

Yup, that is the sort of horrible video I am talking about

(nothing personal, its the camera).


At the time I got my T3i I used an XT for stills and a HF S10 for video. I had a lot of issues with the XT and dust (it has no dust management system), so when the T3i came out I decided to upgrade. No regrets about that, not only did the T3i substantially solve the dust problem (essentially eliminated it actually), it also had far fewer issues with sensor light scattering (purple fringing). So the overall image quality was noticeably better.


When I got the camera I thought that I would be able to retire the HF S10, after all, the T3i had a bigger sensor and recorded at a higher bit rate, and we are led to believe that these are good things. Boy was I wrong!!! The video from the T3i is a lot less resolved than the S10, a lot of artifacts can be seen and any sort of camera motion results in jerking video. The only advantage the T3i had over the S10 was rapid motion by subject matter. The S10 records interlaced footage, so that sort of motion results in a stuttering effect on monitors (it is ok on TV sets). The only way to get anything approaching decent footage from the T3i was to lock it down on a tripod and ensure that the camera did not move at all.


Nowdays I use a G30 for video. It is much better camcorder than the S10, both in video quality and flexibility of use. I no longer see chromatic aberration, although sensor light scattering is still there (which, suspect, is an issue with any small sensor camera). I would never even think about using the T3i for video now. When I first got the G30 I did side by side comparisons with the S10 and the T3i, which reminded me how bad it was.


I have seen footage from the R10, and IMO it is better than the G30, and better than any other consumer/prosumer camcorder I am aware of. Footage from MLRAW-5D3 mostly (as far as I can tell) is locked down tripod stuff. It is not horrible, but neither is it all that impressive either. And I have doubts that it is much more capable of dealing with dynamic video any more than my T3i was. Neither camera was designed for video beyond being able to record it. When you see products like the R10 appear on the market, you know that old design philosophy is in the process of change!

 

[Tugela]

double post

 

[flintyplus]



Quote:
Originally Posted by Tugela  /t/1517624/heres-a-link-bound-to-sp...ngs-of-cameras-by-video-quality#post_24355514


Yup, that is the sort of horrible video I am talking about

(nothing personal, its the camera).


At the time I got my T3i I used an XT for stills and a HF S10 for video. I had a lot of issues with the XT and dust (it has no dust management system), so when the T3i came out I decided to upgrade. No regrets about that, not only did the T3i substantially solve the dust problem (essentially eliminated it actually), it also had far fewer issues with sensor light scattering (purple fringing). So the overall image quality was noticeably better.


When I got the camera I thought that I would be able to retire the HF S10, after all, the T3i had a bigger sensor and recorded at a higher bit rate, and we are led to believe that these are good things. Boy was I wrong!!! The video from the T3i is a lot less resolved than the S10, a lot of artifacts can be seen and any sort of camera motion results in jerking video. The only advantage the T3i had over the S10 was rapid motion by subject matter. The S10 records interlaced footage, so that sort of motion results in a stuttering effect on monitors (it is ok on TV sets). The only way to get anything approaching decent footage from the T3i was to lock it down on a tripod and ensure that the camera did not move at all.


Nowdays I use a G30 for video. It is much better camcorder than the S10, both in video quality and flexibility of use. I no longer see chromatic aberration, although sensor light scattering is still there (which, suspect, is an issue with any small sensor camera). I would never even think about using the T3i for video now. When I first got the G30 I did side by side comparisons with the S10 and the T3i, which reminded me how bad it was.


I have seen footage from the R10, and IMO it is better than the G30, and better than any other consumer/prosumer camcorder I am aware of. Footage from MLRAW-5D3 mostly (as far as I can tell) is locked down tripod stuff. It is not horrible, but neither is it all that impressive either. And I have doubts that it is much more capable of dealing with dynamic video any more than my T3i was. Neither camera was designed for video beyond being able to record it. When you see products like the R10 appear on the market, you know that old design philosophy is in the process of change!

Mm i have the HF-G30 as well,it is far better for filming video than the EOS M but i prefer the colour and overal video from the EOS M if care is taken.

The 550D videos shown were from HDV down samples so far from perfect,550D footage i have on BD is good and if you think the colour is horrible you must use different opticions to me.

You are right getting decent footage from DSLRs does require a tripod for me, they are certainly not run and gun in my case,the 70D film was filmed without a tripod though.

You say the G30 you have is a better cam than the S10 in video and flexibility and shows how bad the T3i was,so we will never agree.

Yes i realy like my G30 and would like a 4K upgrade one day but will not agree about footage from Canon DSLRs,far preferring it to the artificial looking footage my GH2 gave

The 60D & 70D can make good films as well the 5D 111

 

[gunleik]


Hi.


I find this list a bit confusing.


I took part in the CML/UWE test of most of the current cinema/RAW cameras .

 

And I cannot say that that test really supports a list like the one posted here.

 

I have also started to do some complimentary analysis of the cameras here .

 

I cannot say that I get the same results.

 

Also you have the recent DxO test, here

 

Where Dragon is pitted against all the current still-cameras, including PhaseOne and D800.

 

I cannot say I get how this list was assembled...

 

Best

 

Gunleik Groven

---

My Blog

 

[Ken Ross]

It's one person's opinion, not worth getting crazy about. The list could have been arranged 1,000 different ways by 1,000 different videographers. Yawn.