All the DSLRs have drawbacks...but considering that they are just add-ons for the main function of producing still images, some of the cameras are well capable of also producing pro-class video results if placed in the right hands. Have a look at this stunning clip from the 5D Mk II:

http://www.dvinfo.net/forum/canon-eo...wonderful.html

I suppose it all depends on your own requirements and needs. I'm sure that more positive 'video' controls and enhanced built-in handling options and frame rates with less compression/moire/artefacts etc., will be available in the upgraded models during the next couple of years...but the present options from major companies already offer viable alternatives to normal video cameras for use in certain situations.

To be honest, I laugh to think that in the real world, the majority of TV programs in 2010 are still broadcasting in SD and even all the HD programs and feature films are compressed to lower rez before the viewer actually watches them on their large flat-screen HD TVs. The same of course can be said about almost everything viewed on the internet.

Tony,

Some very pretty pictures undoubtably. Although all of these cameras look great down-sampled to that resolution. I've just been trying to quantify for myself what the actual real capture quality of these cameras are like right now and whether i consider any good enough to get. The answer is possibly although i would love to see full res source at it's worst first!

cheers
paul

I've posted several examples of that here in the past. Not sure which part of the forum though. If I get a chance, I'll look for it today.

that shot I posted earlier, that you presumed was locked off, was not as I alluded to in my posting. That shot was taken on a shoulder mount.

Paul,
It sounds like you're doing a lot of research on these - certainly more than I did before picking up a 5d2, and I don't know what your situation or application is.

You should rent one and try it out. Most of the time I'm still blown away by the images I get. However, there are scenes where aliasing artifacts persist (horizontal blinds, striped shirts) etc.

I'm not sure what you're filming. But I tend to treat the 5d2 like a film camera, which is mostly on a tripod, rarely if ever hand held, though I've used it more lately with a glidecam, with good results.

These cameras are by no means perfect, and for the price I wouldn't expect it. I have more issues with audio than with the image quality, but then I now record audio seperately (again, just like film). I just ordered an variable ND filter as well as I tend to blow out highlights when striving for bokeh.

I think you really need to consider your application, and rent one and try it out in that situation. If this crop of DSLR aren't good enough yet, then certainly in the next few years things will improve... and there's always the Scarlet.

7D video compression
 

Hello Paul, i am the author of this video. The original file have aproximately 800MB this is a mp4 compression format that i upload to vimeo, and have a grong color corretion, (testing filters in magic looks) for tests purposes. The original footage is superior than my EX1 sure. I will post some comparision tests EX1 vs 7D later.

Any questions feel free.

Cortez

Cortez,

That's very kind of you and would be fantastic to see some source samples.

I may still get a 7D because the form factor is ideal for a particular type of work my wife would be doing.

After a bit of research, the main concern is the very low real resolution due to the line skipping (and line skipping a debayer too) and how the compression affect things in the real world.

- shimmering due to the relatively low line resolution of the camera.

- how the compression deals with slow moving detail.

To test this i'd love to see a range of source frames sequentially, a second or so in their original format from this sequence.

Also AVC compression tends to turn subtle details into a watercolour effect, i don't think the level of compression for the 7D is too bad and i can see from Perrones sample that flesh tones seem pretty good however any wide shots with people and detail in would be a real test.

If you get the time for some EX1 vs 7D side by sides then some wide shots with detail and subtle movement would be a fantastic test!

Thank you so much,
paul

honestly, it depends on what you intend to shoot...

I used a 5dMKII for a rave and it fell on its face. The CMOS failed so bad it isnt funny. The footage is usable however when analysing it, i personally do not believe it lives up to the hype.

Combing effects from huge LED panels, strobe lights and scanners as well as simple motion and panning break the imagery to a point of dissapointment.

I had it running at 1/125 with a high ISO to alleviate the issue of slow shutter blur.
For the most it was fine, but when it came to hiting the lights, forget it

by all accounts, the canon 1DmkIV is getting pretty close.... very little jello, selectable frame rates, amazing low light capabilities... but it still falls down when it comes to aliasing/moire/sound and actual resolution.

The fact is, the vDSLRs are not there yet... within 2010 though, I expect some outstanding releases that fix alot of the problems.

I think the resolution/moire issue is the biggest hurdle. I've seen a handful a nice pieces but a lot of examples that really don't look HD at all. The shallow DOF hides that very well sometimes.

I don't really see why they can't scale down in hardware, i wouldn't have thought that it would be that much extra processing. Even a simple scaling algorithm.

Makes me wonder how intentional it is - after all canon make some nice prosumer video cameras too. If dSLR did actually produce real 1080p images with an APSC sensor and lens options it would surely kill off the prosumer line. Afterall that's most peoples holy grail.

So i'm not so sure we'll see that much improvement unless of course canon put these sensors into their prosumer cameras - canon don't have a high end to protect unlike sony and panasonic.

cheers

Let's not forget, they have never come out and said "this video feature is Pro/Broadcast quality"..... it was tagged on to some great DSLRs and the chips were not designed for pro video. They do amazingly well considering.

I think the next set of chips may well be built with pro video (as well as pro imagery) in mind. The Digic V chip is said to have video specific processing power... Sony have also said that IF they do get on board, it will be when the video they can output is pro standard.

I'm sure they will put this technology into a video camera body too... as you suggest. It is a logical move and would also steal some thunder from RED if Scarlet launches this year.

I think they will keep progressing with video DSLRs too though.... the demand for B cam/small format video cameras seem pretty big.

If someone released a 7D type camera that outputs: raw/great codec... s35/FF, 1080p/1fps-72fps.. no jello/moire/aliasing/good sound/hdmi out.... $5000 or under and just about everyone in the industry would snap one (or ten) up.

These steps all seem doable within the next year. Maybe in the form of a video add-on, so that photographers don't go crazy.

Next year, possibly... this year, no. Here's why: Canon Reveals Their Next Pro Video Cam at DVInfo.net

Lord amighty. All this is now available in a portable package that costs only 500 clams, and people are COMPLAINING? The video image doesn't quite measure up to cameras costing orders of magnitude more? Color me shocked.

Add to that the fact that this camera will happily use any K-mount, or with an $11 adapter, any M42-mount lens ever made. Never mind that it's an excellent DSLR to boot.

I know where my next 500 bucks is going. For that price, grab this creative tool and figure out how to use it to its best advantage (I figure B-roll for my HD100) and run with it.

That article is pure speculation and is in no way related to DSLRs anyway.

Of course it's related. You brought it up yourself:

I've addressed exactly this point in the article that I've linked to... specifically the last portion of it which discusses exactly what you're talking about here, which is D-SLR image sensor technology in a video camera body. And I've carefully explained why it's not going to happen this year. I wouldn't have posted the link if I didn't think it was relevant to this conversation.

As far as it being "pure speculation," I have already pointed that out in my article, but anyone who knows me and knows my history is fully aware that my "speculation" is highly informed and carries an awful lot of weight. In other words, I'm usually right about this sort of thing.

Besides, Michael Galvan has already confirmed my prediction -- but he's just the first of many folks who are finding out on their own that what I've said in that article is true. Hope this helps,

I'm already there and for precisely the reasons you state:

#1: Big upgrade from my previous still camera AS A STILL CAMERA. Low light performance is orders of magnitude better.

#2: Excellent learning tool at an unbeatable pricepoint. So the footage isn't perfect. It is still an excellent learning tool so I can have a clue as to what I'm doing.