Red One Canon 5D Video Test

[Rick Darge]

This was a rough and dirty test I shot yesterday afternoon. Results are interesting so I thought I'd share. The main purpose was to simply see how the 5D stacked up to my R1. I wanted to see if the 5D could hold up as a solid b-cam/mobile unit on shoots until Scarlet is released. I also really wanted to see how the Red sensor compares to the full frame Canon sensor when using the same glass. Since I'm going FF Epic, this will give an idea of what to expect with those still lenses.

My aim was to get both cameras as close to eachother as possible. I just went with natural light and rated accordingly. Initially everything was shot with REC709 320-800ISO. After I pulled the Red footage into color, I changed the gamma (redspace) and subsequently lowered the ISO when it was appropriate.

Tested out the trusted Nikon 17-35 f/2.8, Nikon ZF 50 f/1.4, Nikon AIS 85 f/2, Peleng 8mm f/16.

I chose Redspace on Red when testing since this gamma curve looked more similar to Canons' final image. Canon really crushes the blacks and boosts the saturation as you'll notice. I did a light white balance on the Red footage. Just adjusted kelvin a couple of times and ISO. No S curves on anything. I brought all the Red footage into Color to make sure gamma/colorspace was even across the board. Red was shot with 4k 16x9 RC28 29.97. Canon was 1920x1080 h264 30.








http://vimeo.com/5160288

[Noah Kadner]

You should install the new firmware for the Canon so you can go full manual:

http://web.canon.jp/imaging/eosd/firm-e/eos5dmk2/firmware.html

-Noah

[Rick Darge]

I did go full manual on the Canon. Latest firmware for both cameras. I kept the white balance on auto for the Canon.

[Raul Gonzo]

Hey Richard- thanks for doing these tests. I have been very curious about this for awhile. It would seem to be an acceptable B camera as long as you don't have to pan or tilt very quickly. (what JJ has been saying all along)
Also, due to the weight- it seems very vulnerable to any bumping. That of course could be fixed in post if necessary.
Again- thanks for taking the time.

[Phil Jordan]

Thanks for doing these tests. It confirms that Red has a much flatter negative-like image set up for many more grading options. The 5D's look is baked in and much more like a graded jpeg than raw like you get from shooting a still from the same camera. These are the same things I found out from my tests. Red for me. I am also waiting for Epic FF35.

[Grady Wilson]

Richard you must of boosted saturation and contrast on the 5D to max levels, there is no way the image can look that crushed out of the box. Or does it?

[Rick Darge]

Hey Grady, That's the Canon 'look' right out of the box. I brought the h264 files onto an FCP timeline. 2048x1152 Pro Res HQ codec. I sent the whole timeline to Color. I didn't touch the Canon files. Just spit them out to 29.97 Pro Res HQ.

[Phil Jordan]

Canon or the others can't compete with Red until they do compressed raw for motion instead of h264 or avchd. Did you use the new color science for these tests?

[Grady Wilson]

That is horrible. Can you fix it? I guess that is the power of a RAW workflow. If you screw something up even a little bit with the Canon you are pretty much screwed it looks like.

Did you shoot the RED in 24p? its surprising how much more filmic the RED looks than the Canon, if that makes any sense at all. Maybe that is what Phis is talking about.

[Rick Darge]

I did but am pretty sure that didn't translate over when I brought the Red footage into Color since we don't have the app update yet.

[Rick Darge]

That is horrible. Can you fix it? I guess that is the power of a RAW workflow. If you screw something up even a little bit with the Canon you are pretty much screwed it looks like.

Did you shoot the RED in 24p? its surprising how much more filmic the RED looks than the Canon, if that makes any sense at all. Maybe that is what Phis is talking about.

You have access to different picture settings with presets. You can dial up saturation, sharpness, contrast, color tone and white balance. The 5D was set to STANDARD mode for this test. I was just playing around with it and noticed the NEUTRAL setting looks more in line with the Redspace gamma curve. No doubt about RAW. I shot Red in 29.97 onto the hard drive. I know what you mean about filmic. Red has way more detail in those mid tones/shadows. Everything is nice and pleasing looking.

[Grady Wilson]

Thank you for doing this Richard. I must admit I was starting to eye that 5D and after this I am sticking with RED.

Out of curiosity, what was that test you did running towards the camera supposed to show?

[christopher witzke]

As an owner of 4 5d2's and a RED..... I can add the following:

The "baked" in settings are acceptable if your careful about your white balance and your exposure.... not much room for correction in post as it falls apart pretty easily. Also.... you'll get much better results if you create a "style" setting that pulls back on the contrast, sat, and sharpness.

My main problem with the 5d2's is the "alliasing" that appears... especially when shooting a scene with brick buildings..... If you shoot with a fast lens and have shallow DOF you will be able to hide it most of the time.

Another issue.... the rolling shutter makes it impossible to correct shaky with FCP's smoothcam.... just wont look good.

As for the RED.... fantastic! Obviously it's a bigger and heavier rig to work with.... but the imagery the RED creates is streets above the 5d2. I'm constantly impressed with the dynamic range of the footage.

[jbeale]

My Canon 50D doesn't record video, but it has an HDMI output which works in "live view" mode. Looking at the HDMI output, it appears to be 30 fps at 1/30 sec shutter, which is certainly not the same motion as a 180 degree shutter. Does the 5DII work that way as well?

[Rick Darge]

Thank you for doing this Richard. I must admit I was starting to eye that 5D and after this I am sticking with RED.

Out of curiosity, what was that test you did running towards the camera supposed to show?

Just wanted a body in motion outside. No reason. :sifone:

[Andrew Walker]

Thanks for doing the test Richard. I still want to get a 5D2 but I've always wanted it just for night timelapse shots. I really have no interest if it shoots video. Seems cool and a nice tool to have in the bag. I'm more interested in the Nikon to Canon mount that your using to get these shots because the ZF glass that I bought for my Red would look cool on my 40D.

[David Swan]

No camera is set up correctly "out-of-the-box". I don't have a 5DII yet, but I do have a 5D, and find "Faithful" is a good starting point.
Another er..minor point....a working Red set-up is, what, 10x as much as a 5DII? And people are slating it for not being as good as a Red One? Get real!
Dave

[Tom Lowe]

Thanks for doing this test, Rick.

A couple of thoughts:

1) To be fair, this material should be presented at 1080p. 720p can hide a lot of resolution and other issues with the 5D2.

2) The color-correction work on the Red footage is very dull and flat. This is unfair, since the contrast, saturation, etc, are boosted through the roof on the 5D2.

3) The 5D2 seems to hold the sky a hell of a lot better than the R1. I don't know if this is a result of your lack-luster grading, though? I have heard people claiming that R1 has a clear "DR advantage" over the 5D2, but I never thought that was true. It has a million other advantages, for sure, but DR is not one of them, as far as I can tell.

[Rick Darge]

I would have put up a 1080 file but bandwidth issues. If someone wants to host it, I can send one over if people are interested. Just PM me. I kept the Red at defaults as well as the Canon. I'm using the STANDARD picture setting on the Canon so none of the SAT/CONTRAST levels are boosted. They're all zeroed out as far as I can tell. Once again, no s-curves applied to any of the Red or Canon footage. Rec709 has better dynamic range but I wanted to just get the imagery closer, gamma wise. I should do a retest when I have time to try and grade them both as best as they can be, rather than 'out of the box'

[Matt Gottshalk]

Glad I bought the 5DMKII. I can work within it's limitations and get some great footage at a fraction of the cost of a RED.

If I get some budgets that justify a RED I will certainly rent one.

Otherwise, my clients will more than love the look of the 5DMKII.

Check out this test with a Sigma 50mm 1.4:

http://www.theprogamer.com/daniel/sigma_50_14-stream.mov

[Jonathan Stevenson]

Another er..minor point....a working Red set-up is, what, 10x as much as a 5DII? And people are slating it for not being as good as a Red One? Get real!
Dave

I think this is the biggest thing that is overlooked in the Red/5D battle. Yes, hands down, the Red is a superior camera- because it costs $30k to get it running. If you want a slick, great looking camera for under 5K, the 5D is an awesome option. But the Red will always be better. :)

[Phil Jordan]

Another er..minor point....a working Red set-up is, what, 10x as much as a 5DII? And people are slating it for not being as good as a Red One? Get real!
Dave

And a Canon 5D2 is 10x the price of an Elf/IXUS which shoots video. You just have to decide what you are willing to pay for and what limitations you are willing to accept. I am sure that if Canon ever does step up and build a 24p raw motion camera we will have Red to thank.

[jimhare]

Yeah, same as when people like to compare the R1 to Genesis. The fact that you can even compare them is testament alone.

Difference here is the R1 can go against the Genesis feature by feature, and wins in many areas.

The 5D seems like it will always be a "if you get it exactly right on set and don't challenge it in post" scenario. A lot of pressure and extra time on set.

Still, a great thing to have in the kit, just don't know if it could ever be the main piece in your kit.


I think this is the biggest thing that is overlooked in the Red/5D battle. Yes, hands down, the Red is a superior camera- because it costs $30k to get it running. If you want a slick, great looking camera for under 5K, the 5D is an awesome option. But the Red will always be better. :)

[Tom Lowe]

Rick, you can upload 1080p videos to Vimeo, and then anyone logged in can download the unaltered 1080p video right from Vimeo. In other words, Vimeo is paying the bandwidth bill.

[Matt Gottshalk]

The 5D seems like it will always be a "if you get it exactly right on set and don't challenge it in post" scenario. A lot of pressure and extra time on set.



It's not like people weren't doing that for history of film making until relatively recently.....:couch:

[Rick Darge]

Rick, you can upload 1080p videos to Vimeo, and then anyone logged in can download the unaltered 1080p video right from Vimeo. In other words, Vimeo is paying the bandwidth bill.

Genius. I'll do that.

[Charles Angus]

It's not like people weren't doing that for history of film making until relatively recently.....:couch:

Great point. It's easy to forget how spoiled we are these days...

[albert rudnicki]

I am for one really surprised by the skew effect. Never tested my self...

[Roberto B]

who'd say it?.. the best 5d vs red1 comparative is yours old rick.. no charts.. real subjects.. rgdfilms is the messenger.. do not shoot the messenger btw.. ehehehehehe

[Andrew Walker]

Glad I bought the 5DMKII. I can work within it's limitations and get some great footage at a fraction of the cost of a RED.

If I get some budgets that justify a RED I will certainly rent one.

Otherwise, my clients will more than love the look of the 5DMKII.

Check out this test with a Sigma 50mm 1.4:

http://www.theprogamer.com/daniel/sigma_50_14-stream.mov

That was a sweet little clip Matt. Makes me want to get the camera even more mainly because I could grab some motion when I don't want to set my Red up. But still I want to mainly use it as a timelapse camera just as I have been doing with my 40D. Plus I don't think Getty would take video stock footage shot with the 5D2. But they will take timelapse shot with it. Looks like a fun camera to have around. Kind of reminds me of the days of the family 8mm film cameras, except much clearer.

[Charles Angus]

Looks like a fun camera to have around. Kind of reminds me of the days of the family 8mm film cameras, except much clearer.

Good comparison. And it's not like 8mm cam's haven't been used to great effect for any number of films.

[Roberto B]

okay.. red's higher rez but.. 5d 8mm?.. gimme a break lol

[albert rudnicki]

That was a sweet little clip Matt. Makes me want to get the camera even more mainly because I could grab some motion when I don't want to set my Red up. But still I want to mainly use it as a timelapse camera just as I have been doing with my 40D. Plus I don't think Getty would take video stock footage shot with the 5D2. But they will take timelapse shot with it. Looks like a fun camera to have around. Kind of reminds me of the days of the family 8mm film cameras, except much clearer.

Great shot is priceless, regardless the resolution.

[Kenneth Elkington]

okay.. red's higher rez but.. 5d 8mm?.. gimme a break lol

I think he's just comparing it to the convenience of old 8mm cameras.

[Joel Kaye]

I'm not really seeing the compression artificating in these shots. Am I missing something?

[Kenneth Elkington]

I'm not really seeing the compression artificating in these shots. Am I missing something?

Yeah, I'd like to see some high motion high detail shots, and some fences/brick walls/fabric patterns, so that we can have comparisons of "problem shots" that's likely where the more dramatic differences will become apparent.

[Rick Darge]

I have the 1080 version uploading to Vimeo right now. I'll post a link when that becomes available.

I also took an hour to shoot 4 more setups. This time I properly calibrated the 5D with better settings. Basically toned the contrast/saturation all the way down, turned the sharpening to zero and gave it a FAITHFUL picture setting. I wanted to see some super low light action so I pushed the 5D and the Red as far as possible, (2000iso / H1 iso) Results coming later. I am by no means trying to be picture perfect with these tests so take what you will from them. Still learning.

[Kholi Hicks]

Glad I bought the 5DMKII. I can work within it's limitations and get some great footage at a fraction of the cost of a RED.

If I get some budgets that justify a RED I will certainly rent one.

Otherwise, my clients will more than love the look of the 5DMKII.

Check out this test with a Sigma 50mm 1.4:

http://www.theprogamer.com/daniel/sigma_50_14-stream.mov

MKii's nice... but that video is a turn off from the camera itself. I know it's a sample "test" or whatever, but good grief hyper-shallow and oversaturated. And all those little bumps and jumps.

The MKII right now is pretty much a "sticks" only camera, or you ride forever on a Glidecam 4000.

I suspect Scarlet's gonna be the same way because of how light it is. This is what scares me the most.

Anyway, Thanks for the footages Richard. Indeed, you must use either Faithful or Neutral with settings dialed down to get a decent post-workable image out of the MKii. Or something close to it. Standard looks like crap.

And, as someone else mentioned, nailing White Balance is detrimental. But that's pretty much every camera that doesn't shoot RAW.

Oh and a word on ISO's if you dunno yet: 160, 320, 640, 1250, 2500 -- Outside of those is dangerous territory!

[Emanuel A.]

From this test, I think the most differences are about resolution indeed. At the wide end, RED ONE is frankly superior and the gap is visible. It does seem 4K means something, not only hype or sales talk. After all, for any reason there's the Nyquist sampling theorem.

We can't say exactly the same when enters the (shallow) DOF or noiseless properties in scene. Same would apply for lowlight performance. DR might be further tested. Beyond this or skew, RAW is a solid advantage on the RED ONE side but the relative fair bitrate becomes an attractive 5D Mark II feature despite its limitations or what can be said.

About size and weight is apples to oranges. Until the (upcoming) Scarlet release.

[Kholi Hicks]

Emanuel... do you realize you delete your post and repost it every time someone posts after you? At first I thought I was going crazy... but now I'm pretty damned sure that you're a bottom poster.


LoL

[Emanuel A.]

MKii's nice... but that video is a turn off from the camera itself. I know it's a sample "test" or whatever, but good grief hyper-shallow and oversaturated. And all those little bumps and jumps.

The MKII right now is pretty much a "sticks" only camera, or you ride forever on a Glidecam 4000.

I suspect Scarlet's gonna be the same way because of how light it is. This is what scares me the most.Kholi,

With a proper glass solution such as an IS feature, it is far from being a problem. PM me and I'll provide you a link if you don't believe it.

E. :-)

[Emanuel A.]

Emanuel... do you realize you delete your post and repost it every time someone posts after you? At first I thought I was going crazy... but now I'm pretty damned sure that you're a bottom poster.


LoLLMAO. Editing is into my blood, I'm addicted to. :)

Now a little bit more seriously, once someone really smart upgraded the browser for a newer version, the 'save' buttom doesn't apply at times for some mysterious reason -- perhaps an old slow computer should be the reason (you see, it happened again). Only the 'delete message' and 'submit reply' buttoms work, the computer freezes.

EDIT -- 'buttom poster' seems nice, though. ;-)

[Kholi Hicks]

Kholi,

With a proper glass solution such as an IS feature, it is far from being a problem. PM me and I'll provide you a link if you don't believe it.

E. :-)

IS will be good. In fact, the one thing that might keep an MKii in my possession is a Sigma 18-250 Zoom. Does that linky link of yours feature this lens? =D I'd sure like to see what I'm about to waste my money on. LoL

[Emanuel A.]

Canon L 24-105mm f/4. AFAIK the Sigma is not full frame.

[Kholi Hicks]

Canon L 24-105mm f/4. AFAIK the Sigma is not full frame.

Yeah I don't think it covers until after 24 either. Haven't found any confirmations though.

[Emanuel A.]

Indeed. BTW, a Canon L lens class is, well, a L lens class :-) optically and mechanically. PM sent as requested. Check your PM box. Hope it helps.

[Rick Darge]

Vimeo version is up. There is a link to the 1080 file. 500MB

http://vimeo.com/5160288

[Tom Lowe]

That was a sweet little clip Matt. Makes me want to get the camera even more mainly because I could grab some motion when I don't want to set my Red up. But still I want to mainly use it as a timelapse camera just as I have been doing with my 40D. Plus I don't think Getty would take video stock footage shot with the 5D2. But they will take timelapse shot with it.

I'm pretty sure Getty is taking 5D2 1080 HD video clips. I'm not talking timelapse, I'm talking video. I think 5D2 clips are on a "shot by shot" basis with Getty, but then again, all shots are that way really. They have to pass editorial review. Sometimes the 30p 5D2 clips look nice transcoded to 24p, to play slightly slo-mo. So I guess you can submit at 30p or 24p, depending on the clip.

[Emanuel A.]

Vimeo version is up. There is a link to the 1080 file. 500MB

http://vimeo.com/5160288Same clip, yet?

[Rick Darge]

Yes same exact clip. Although now a much higher bitrate and 1080P file for you guys to download.

[Rick Darge]

I'm compressing another test clip. It's 5 more quick setups. This time I dialed the Canon in a little better and kept the kelvin at 5000 for both cameras.

Off the cuff evaluation at this point is that Red is superior when you feed it light. By far the cleanest, highest dynamic range picture with a 'thick' negative feel to it. Duh... That's obvious though.. Canon trumps the Red in low light, no doubt. The results were pretty outstanding when I pushed both cameras as far as they could go using the Nikon 17-35 f/2.8 lens. I would really love it if Red can match or sprint pass the low light Canon with the Epic/Scarlet because the night performance on the 5D is just amazing.

[Sanjin Jukic]

The results were pretty outstanding when I pushed both cameras as far as they could go using the Nikon 17-35 f/2.8 lens.

I would really love it if Red can match or sprint pass the low light Canon with the Epic/Scarlet because the night performance on the 5D is just amazing.

Richard,

obviously you never "crossed" a border between f/1.4 and f/1.2 or f/0.95 on RED1.

Further "the night performance on the 5D is just amazing" is just a fake statement based

on "crippled" HD video suitable only for web application that you can get out of it.

DSLRs like 5DM2 or DX3 have "outstanding" night performance just because of "boosting" higher ISO

that's not possible with moving images that are bigger then HD.

Just have a look more carefully at 4K night shots by 5DM2 and see all those artifacts.

That's why exist the super-fast lenses like a Master Primes that you wouldn't need higher ISO (with all those artifacts)

on digital or you would need appropriate film stock with analog acquisition.

5DM2 is a perfect camera for so called video "advanced" tourists but right now there is even better camera available for that sort of shooters:

Panasonic GH1 that can get PL mount and Master Primes.

Have a look:

http://philipbloom.co.uk/wp-content/uploads/2009/05/still4-1024x576.jpg
Shot on Panasonic GH1 with Zeiss Master Prime 85mm (click here to get a lager size (http://philipbloom.co.uk/wp-content/uploads/2009/05/still4.jpg))

http://philipbloom.co.uk/wp-content/uploads/2009/05/still5-1024x576.jpg
Shot on Panasonic GH1 with Zeiss Master Prime 85mm (click here to get a lager size (http://philipbloom.co.uk/wp-content/uploads/2009/05/still5.jpg))

More>> (http://philipbloom.co.uk/2009/05/12/panasonic-lumix-gh1-in-joshua-tree-and-pl-lenses/#more-2953)

[Rick Darge]

Further "the night performance on the 5D is just amazing" is just a fake statement based

on "crippled" HD video suitable only for web application that you can get out of it.


How is my comment 'fake'? I use both cameras. I own both. They both have different purposes. One is clearly A LOT brighter in the dark than the other. There's nothing fake about that. :couch:

[Sanjin Jukic]

Richard,

sorry and maybe "fake" was a hard word that also cannot explain what I wanted to say.

Simply it's about a "bluff"!

Take a simple logic:

If Canon 5Dm2 would make "usable" video at higher ISO then Canon will for sure jump in a business

of making "digital cinema cameras" for Hollywood and "kill" all the competition in a digital even at analog (film) acquisition.

After that scenario nobody would need super-fast lenses like a Master Primes.

That's sort of future still does not exist.

And that why is Canon 5Dm2 with enabled HD video a DSLR for sort of "advanced-video-tourist" shooting only.

BTW, Canon 5DM2 is an excellent digital photo camera that with additional little HD video "bluff" is probably selling pretty good and can stay in a competition with other DSLRs of that type.

That's all.

It cannot compare with RAW moving images in a low/available light when you bring Leica Noctilux-M f/0.95 type of super-fast lenses to RED1.

[Rick Darge]

Hahaha. Sanjin ¥ou're crazy but I love you man. I'm just basing this off of what my eyeballs see and so far, the 5D looks a lot better in low light when compared to Red using the same 17mm f/2.8 Nikkor zoom lens. I'm all about Superspeeds on Red for low light. That's why I bought a set of those Canon K-35s for that very reason. I can never go back to shooting T3 on the Red at night with very little light. Too ugly.

[Andrew Walker]

I think he's just comparing it to the convenience of old 8mm cameras.

Yes that is what I was comparing it to. I know resolution wise it is better looking than 8mm film any day of the week. But it seems with the choice of lenses to use on the camera and the look, its very similar to the old film cameras that families would have with them. Its nice to see good quality images for almost everyone.


I'm pretty sure Getty is taking 5D2 1080 HD video clips. I'm not talking timelapse, I'm talking video. I think 5D2 clips are on a "shot by shot" basis with Getty, but then again, all shots are that way really. They have to pass editorial review. Sometimes the 30p 5D2 clips look nice transcoded to 24p, to play slightly slo-mo. So I guess you can submit at 30p or 24p, depending on the clip.

Your right Tom, I was just saying based on the compression of the clips and the baked in look it might be a little harder to get it by their QC department. I'm sure if you have something that is very marketable that they would want it. But its not going to be like Red footage...at least I don't think so because the whole digital negative aspect of the R3D file doesn't exist with the 5D2 footage.

Richard, the other thing I liked about these tests you did was to see the difference in the coverage between a Super 35 frame and a FF35 frame using still glass. Kind of a good reference to have when thinking about going up to the FF35 Epic.

[Sanjin Jukic]

I'm all about Superspeeds on Red for low light.

Great Richard,

also the same with a film acquisition!!!

[Rick Darge]

Video disabled for the moment. (165MB - 960x540 h264)

Ok. Sleep time. I just started the upload for my quick&dirty ROUND 2 tests. It should be up there 40 minutes from the time of this posting.

Red was shot on the Red-raid again, 4k 16x9 RC28. Redspace Gamma and Color. No S curves. 5000k for every shot. 1/60 for all but the last 2 shots. Sat/Contrast were not touched. Rendered out of Color again.

Canon was kept at 5000k for every shot. (5000k on a Canon is noticeable warmer when compared to Red's 5000k setting) Picture setting was set to FAITHFUL this time. Sharpness turned down to ZERO. Saturation and Contrast were dialed down to -4. Color tone was at ZERO.

Just used the Nikon 17-35 f/2.8 lens. Kept it at 17mm for both cameras and all shots this time. Fully open for every single shot.

Enjoy.

[Pawel Junak]

And that why is Canon 5Dm2 with enabled HD video a DSLR for sort of "advanced-video-tourist" shooting only.


recently I found this:
http://vimeo.com/5000009

I'm not saying that everything is possible on canon, but for sure it's more than "advanced-video-tourist shooting only"

[Sanjin Jukic]

recently I found this:
http://vimeo.com/5000009

I'm not saying that everything is possible on canon, but for sure it's more than "advanced-video-tourist shooting only"

To repeat again that Canon 5DMk2 is quite good photo camera for web video compression apps (web commercials, touristic, wedding video, etc...) but not for serious HD application, feature or so...

[Matt Gottshalk]

To repeat again that Canon 5DMk2 is quite good photo camera for web video compression apps (web commercials, touristic, wedding video, etc...) but not for serious HD application, feature or so...

Says you.

I guess that's why Rodney Charters, the DP for the tv series "24", used a total of FIVE, 5D mkII's on the set of the new CBS pilot, "Washington Field".

He must not be serious.

[Daniel Lipats]

To repeat again that Canon 5DMk2 is quite good photo camera for web video compression apps (web commercials, touristic, wedding video, etc...) but not for serious HD application, feature or so...

There are many who insist that the 5D Mark II video and codec is better than the EX1/EX3 and the HPX300. I'm not convinced myself, but at 42Mbps I have a hard time remaining skeptical. Are they all only good for "sophisticated home videos" and web commercials?

You also leave out the fact that the 5D has already been used in theatrical feature films. Not to mention major commercials, TV shows, music videos, documentaries, etc...

[David Nardini]

Says you.

I guess that's why Rodney Charters, the DP for the tv series "24", used a total of FIVE, 5D mkII's on the set of the new CBS pilot, "Washington Field".

He must not be serious.

I'd like to know how he controls the moire / aliasing the 5Dii it's so fond in producing (eg : visible in the Richard MOV download on the Fu's Palace scene and the overhead electrical wires). Perhaps he keeps all potential bad subject matter out of focus ;-)

As for the baked in 5Dii 'look', there are plenty of pf2 files on the net to help control the hard blacks the 5Dii defaults to. NeatVideo seems to do a pretty good job of cleaning up the compression artifacts.

Biggest issue I have with the 5Dii files is the aliasing/moire ... if the shot is locked off there is always the option of blending in a high-res still ;-)

Does anybody have any links to full res GH1 files ... I'm curious to see if this cam produces moire like the 5Dii or is it better behaved ?

[Tom Lowe]

"Advanced tourist"? :smilielol5: Sanjin, you crack me up with these declarations!

[Kholi Hicks]

Probably just ignored the Aliasing. I don't know how else you could get around it, like the slow pan jello, other than making yourself feel better by saying "well if people are seeing that then the story must be bad".

=B

[Daniel Lipats]

Probably just ignored the Aliasing. I don't know how else you could get around it
Fixing the aliasing has been addressed hundreds of times across multiple forums, including this one. I'm surprised it keeps coming up.

1) Anti-moire filter removes aliasing without reducing resolution.
2) Diffusion will remove aliasing but reduce resolution.
3) Put the problem slightly out of focus. With a full frame sensor this is easy.

[David Nardini]

Fixing the aliasing has been addressed hundreds of times across multiple forums, including this one. I'm surprised it keeps coming up.

1) Anti-moire filter removes aliasing without reducing resolution.
2) Diffusion will remove aliasing but reduce resolution.
3) Put the problem slightly out of focus. With a full frame sensor this is easy.

Care to share more detail around (1) ... ideally with some before and after examples ?

I've tried the stills approach (guassian blur or median on layer and blend using colour mode) ... kind of fixes sometimes but not always ... hence the search for a consistent approach.

[Sanjin Jukic]

Says you.

I guess that's why Rodney Charters, the DP for the tv series "24", used a total of FIVE, 5D mkII's on the set of the new CBS pilot, "Washington Field".

He must not be serious.

"24" TV series need so often by its style, concept and genre always a bit "weird" video look that cameras like Canon 5Dmk2 or Pansonic GH1 fits perfect...


"Advanced tourist"? :smilielol5: Sanjin, you crack me up with these declarations!

Some of "advanced video tourists" (like me) don't care about a moire made by 5Dmk2:

http://homepage.mac.com/sanjinjukic/RED/Canon5dmk2_moire.jpg
Canon 5DMK2 with visible moire at test charts, find more at the link below

LINK>>> (http://screeningroom.smugmug.com/gallery/8501805_LEtDE#559500586_95RU8)

but some another "advanced video tourists" care and have got Easyrig 3 cinema for easy handheld shooting

on mountains (Alps) this summer.

Easyrig 3 cinema costs about the same as Canon 5DMK2 but I decided to shoot with RED1 because it makes better pictures, it's a RAW,

shoots 4K, 3K, 2K, HD, 24 fps, 25 fps, 23.98 fps,...more, has slow motion 120 fps, more,... also long shooting possible with 320 GB drive, has CF card slot,

much easy to avoid camera shake at handheld, better sound, etc,...

http://homepage.mac.com/sanjinjukic/RED/EasyRig3.JPG
My Easyrig just arrived!!! :)

[Daniel Lipats]

Care to share more detail around (1) ... ideally with some before and after examples ?
I can't say if it will remove all of it completely, but people have reported success with Caprock Anti-Moire filters. They come in 67mm and 4x4" with different strengths. Have not tested them all, or tested in detail but it does improve the image by removing both moire and aliasing. Waiting to get my hands on the 4.0 version, which should be the maximum strength.

http://www.vistek.ca/store/CameraFilters/229815/caprock-filters-antimoire-4x4-filter-40.aspx

[David Nardini]

I can't say if it will remove all of it completely, but people have reported success with Caprock Anti-Moire filters. They come in 67mm and 4x4" with different strengths. Have not tested them all, or tested in detail but it does improve the image by removing both moire and aliasing. Waiting to get my hands on the 4.0 version, which should be the maximum strength.

http://www.vistek.ca/store/CameraFilters/229815/caprock-filters-antimoire-4x4-filter-40.aspx

Ahhhhh well ... I have one of those right in my hands (4x4 - 4.0 strength) and it's not working for me ... brick work still looks like s**t (and yes, it does not affect sharpness that I can see).
Speaking with them there are stronger ones (2.x I believe) but unclear if it will work ... awaiting to hear back. I had hope you had another magic wand ;-)

And yes Sanjiin, my R1 is outstanding, absolutely happy with it ... however the 5Dii has it's uses ;-) ...

[Daniel Lipats]

Ahhhhh well ... I have one of those right in my hands (4x4 - 4.0 strength) and it's not working for me ... brick work still looks like s**t (and yes, it does not affect sharpness that I can see).
Speaking with them there are stronger ones (2.x I believe) but unclear if it will work ... awaiting to hear back.
Thanks for letting me know! Do you know if it's much better than the 5.6? I'll see if I can find a stronger one too.

[David Nardini]

Thanks for letting me know! Do you know if it's much better is it than 5.6? I'll see if I can find a stronger one too.

No idea, I simply aimed for the strongest 'as advertised' (5.6 is going the wrong way) it clearly needs to be less than 4.0. I spoke with www.caprockdev.com the other evening and we concluded to aim for the 4.0 and see what the outcome is before going lower

PS : I did warn him that if it works he may have a few enquiries ;-)

[Tom Lowe]

The 5D2's aliasing is not going to be fixed with simple filters. It's a major, major problem. I have had a number of amazing shots ruined by moire and various artifacts. Trying to shoot water, on a lake or ocean, during the golden part of magic hour is a disaster. I also had some nice Saguaro cactus shots ruined by moire.

[David Nardini]

The 5D2's aliasing is not going to be fixed with simple filters. It's a major, major problem. I have had a number of amazing shots ruined by moire and various artifacts. Trying to shoot water, on a lake or ocean, during the golden part of magic hour is a disaster. I also had some nice Saguaro cactus shots ruined by moire.

I agree 100% Tom ... this is my last 'push' to try a few options. Soft focus 'stuff' / no sharp edges and you're ok ... everything else BEWARE.

More importantly NOBODY seems to know if the GH-1 suffers in the same way ?!?! ...

[C'on RED lets have those small compact FF35 ;-) and put an end to our suffering]

[Daniel Lipats]

You can already completely remove aliasing with diffusion filters, the problem is they blur everything all at once.

Once you remove detail at a certain frequency (that the 5D can't resolve anyway) the moire is gone. Caprock filters at a high enough power should provide the solution.

GH-1 also has aliasing but to a lower degree.. Not to mention a collection of its own unique problems including "mud".

[Roberto B]

The 5D2's aliasing is not going to be fixed with simple filters. It's a major, major problem. I have had a number of amazing shots ruined by moire and various artifacts. Trying to shoot water, on a lake or ocean, during the golden part of magic hour is a disaster. I also had some nice Saguaro cactus shots ruined by moire.man no water no cactus.. enough said?

[Ben Holmes]

I'm still trying to take in the insanity of 8 pages of discussion of web-compressed video being compared to RED footage. Based on web compressed footage.

The 5DII shoots h264. The RED shoots compressed RAW footage. As someone who is not a cinematographer, but an editor and QC specialist for broadcast, I cannot possibly imagine what you hope to achieve by comparing the two. In the post world, it's like comparing a tricycle with a Ferrari.

Please - can we accept this is a 'horses for courses' type of discussion and move on? It's just not a fair or useful comparison. It reminds me of a friend who was shooting Digibeta being harangued by a 'videographer' shooting on DVCAM. If you can't see the difference between the two acquisition formats, then by all means, get yourself a DVCAM camera...

[Rick Darge]

Did you download the 1080 file off of the Vimeo site?

That file is almost identical to the original file I spit out of FCP.

And yes, I find it a very useful comparison for those wanting to know just how the 5D cuts in with the Red. Seriously though, why do so many people knock the 5D as if it were some kind of threat? I find it amusing calling it a 'web camera' or a 'advanced tourist cam'. It shoots 1080p with DOF and dynamic range comparable to Red for 1/10 the cost! Nobody is saying it's BETTER. I just wanted to see how the two flavors compare to one another when you throw them in the mix.

I could have grabbed some $500 DVCAM and did the same comparison, but that would have made no sense. The 5D is very similar to the Red in a lot of ways. For one, both cameras can share the same Nikon lenses. Thus this test..

[albert rudnicki]

I Greatly appreciate your test Richard, I am in need for a low light capable "b" camera in upcoming weeks and the 5D being not "perfect" as it is, makes a Perfect cam for what I need.
Thanks again

[Rakesh Jacob]

Seriously though, why do so many people knock the 5D as if it were some kind of threat? I find it amusing calling it a 'web camera' or a 'advanced tourist cam'. It shoots 1080p with DOF and dynamic range comparable to Red for 1/10 the cost!

LOL because it is a threat, maybe not now but the next gen and the one after that, and because it IS 1/10th of the cost. "Hey some jackass on the interwebs thinks his $2700 toy can hang with my RED ONE? Who the fuck does he think he is?! Ima professional gol-dangit!!!!!"
:head_explode:

[Rob Castiglione]

Ben's answer is definitive. The fundamental problem lies in the limitations of the codec.

[Cristina S]

To repeat again that Canon 5DMk2 is quite good photo camera for web video compression apps (web commercials, touristic, wedding video, etc...) but not for serious HD application, feature or so...How can you say it without fearing to look like a dork?

[Rick Darge]

Sanjin is fearless. That's why...

He's a true pirate..

[Matt Gottshalk]

LOL because it is a threat, maybe not now but the next gen and the one after that, and because it IS 1/10th of the cost. "Hey some jackass on the interwebs thinks his $2700 toy can hang with my RED ONE? Who the fuck does he think he is?! Ima professional gol-dangit!!!!!"
:head_explode:

Exactly.

Thanks again for the tests sir.

[Obin Olson]

That is horrible. Can you fix it? I guess that is the power of a RAW workflow. If you screw something up even a little bit with the Canon you are pretty much screwed it looks like.

Did you shoot the RED in 24p? its surprising how much more filmic the RED looks than the Canon, if that makes any sense at all. Maybe that is what Phis is talking about.

and that my friend is the power of RAW.

[Rick Darge]

I will be putting up another comparison video within the week pitting up the Canon 5D + Nikon Glass & the Red One + K-35 Primes. Did a job this weekend with my K-35 set and the quality is amazing. The 5D could never achieve what I'm seeing here but I'd still like to see a side by side, if not for me, then for my clients. Going to try to make this test more controlled too. Lit subject, etc.

[I Bloom]

I just shot a fashion video on the 5DmkII. I chose it in part to experiment, n part because my Red is on a feature and in part because I was shooting all night exteriors in available light.

It's kind of great having such a light weight camera.. but one big problem is that the viewfinder is closed in video mode, so unless you have an elaborate rig you need to use the LCD to operate, very awkward.

The image quality in video mode, simply doesn't compare to Red. You don't need to see a comparison. Red is light years ahead in terms of the final image that you get... except one thing: The lowlight capability of the 5D is kind of amazing.

At the end of the day the 5D is a great still camera, and the video feature is a fun thing and very useful. But it made me really want the DSMC more than anything else. I want to shoot compressed RAW on a camera that feels like a DSLR, that's an exciting prospect.

IBloom

[Joseph Ward]

I just shot a fashion video on the 5DmkII. I chose it in part to experiment, n part because my Red is on a feature and in part because I was shooting all night exteriors in available light.

It's kind of great having such a light weight camera.. but one big problem is that the viewfinder is closed in video mode, so unless you have an elaborate rig you need to use the LCD to operate, very awkward.

The image quality in video mode, simply doesn't compare to Red. You don't need to see a comparison. Red is light years ahead in terms of the final image that you get... except one thing: The lowlight capability of the 5D is kind of amazing.

At the end of the day the 5D is a great still camera, and the video feature is a fun thing and very useful. But it made me really want the DSMC more than anything else. I want to shoot compressed RAW on a camera that feels like a DSLR, that's an exciting prospect.

IBloom

Got to shoot in Japan with Red DSMC. :hurray:

[David Swan]

Ok, so the 5DII isn't as good as the Red One. Would you expect it to be? I wouldn't.
The Red was designed from the ground up as a kick-ass cine camera and it is.
The 5DII was designed as a mid-high enthusiast DSLR with live-view, and is damn good at that, then the video was added as an after-thought as a by-product of the live-view system, hence the 30 fps frame-rate. That it's even as good as it is is pure serendipity.
So where now? fortunately my finances pretty much dictate waiting, so I'll see what transpires, I may end up with a 5DII (Or its successor) as my walk-about do-it-all cam and a 2/3" Scarlet for more serious cine (FF35 Scarlet is out of my budget)
But remember, Canon make their own sensors in their own plant, heck they even melt their own glass, so they could pull off something amazing (Anyone remember when the 300 f/2.8 Fluorite hit the news?), whether they will or not is up to them.
Dave

[Emanuel A.]

Richard,

Did you use FCP or Adobe Premiere?

Someone posted this:

http://www.dvxuser.com/V6/showpost.php?p=1670501&postcount=39

E. :-)

[Rick Darge]

I used FCP 6

[Emanuel A.]

Richard,

Take a glance @ dvxuser:

http://www.dvxuser.com/V6/showthread.php?p=1670540#post1670540

'cause there that's going wild.

E. :-)

EDIT -- Not better @ 5D Mark II forums (the rumors are like a plague):

http://cinema5d.com/viewtopic.php?f=14&t=3381#p22989

Thanks for your tests in any way whatsoever. We're still waiting for your most recent findings. Didn't you have more stuff?