Camera Upgrade for Web Series?

[Barry Gribble]

All,

I am getting ready to shoot a sit-com that I'm going to distribute on the web. Most people will see it on something like YouTube - that quality at least. The style is the pseudo-reality look like Arrested Development or The Office.

I have a Canon Xl1 that puts out a pretty good image. I have been thinking about whether to rent or something else for the shoot, but now I’m thinking that for a web audience I might be doing fine with that camera.

Certainly there is some chance of selling it on DVD down the road, or of trying to pitch it to someone, and that factors in a little bit. I also may shoot the pilot on something better just to have it as a portfolio piece.

My thoughts are this:

If I shoot with an HVX200 or a HD100 and compress it down to web size, it isn’t going to look too much different.

If I shoot with an SDX900 or DigiBeta I think I will get richer colors that will translate.

So my question is:

If I am going to distribute on the web, is it worth it at all to use a better camera than I have?

Any input is appreciated.

Thanks.

Barry

(PS - this isn't so much about the XL1 against any specific camera, it is about lower-quality source images versus higher quality ones for web delivery)

[Jon Fairhurst]

I'm looking at a similar situation - only I don't have an XL1.

Personally, I think an XL1 is everything you need for this application, assuming that things are well lit, well staged, nicely processed, etc. The one caveat would be if you're chroma-keying. If so, keying, say, from uncompressed HD, and banging it down to web resolution would make any edges absolutely disappear. Keying with the low chroma resolution, OTOH, won't be as effective.

But if you're not chromakeying, I might invest in other aspects of your kit to compliment your XL1.

[Barry Gribble]

Jon,

Thanks for the input. Yeah, no chromakey... just basic shooting.

What is your similar situattion?

Barry

[Jon Fairhurst]

We're making video content for the web; however, chromakeys are a big part of our vision. Right now we're getting by (barely) with a 1-chip camera, and we're looking into the HDV, and the new Canon with HD-SDI out.

You can check out our http://colcrush.com website. I've learned enough from that version that I'm completely redesigning it with a raft of new features and a more accessible presentation. We've been creating new content for steady releases after the new design is up.

I sure wish we had a bigger equipment budget. I worked for Grass Valley as a design engineer for ten years, and now in the digital video department of a research lab for the most recent ten. I've worked on/with some amazing equipment - but none of it is mine. So far the website brings more expense than income, but we plan to turn that around in the next year...

[Mathieu Ghekiere]

I personally think, for the web, an XL1 is more then fine enough.
Why all the extra resolution if your audience is only gonna see a marginal increase of image quality?

But that's only my opinion. There surely are better cameras on the market, but for the web, I don't see those cameras as a real 'must'.

[Jon Fairhurst]

Aside from chromakeys, the HVX is also on our shortlist because of the variable frame rates. Also, the P2 workflow would work well for our quick takes.

For sure, doing a fight scene recorded at 20 fps and played back at 24 fps will give a different look than just speeding up 24 fps capture. Also, being able to slap together a quick edit while on the set is nice insurance. If something is wrong, we can fix it before we pack up.

The bottom line for me isn't that I need HD. It's that I want clean chromakeys, variable frame rates and a tapeless workflow. If HD gives a cleaner picture, that's just gravy. (Note that the HVX has one of the lowest resolutions of the 3CCD 1/3" HD camera out there. As long as our keys look good, that may be acceptible.)

Another consideration is the filesize and workflow. With DV you just injest and edit. With HDV you need to convert the file to an editable format. (Some software supports this during injest, if you have enough CPU.) Also HD takes more bandwidth and fills you hard drives more quickly. And rendering is slower.

I figure that raw 720p at true 24fps is only 2.56 times heavier than raw 480i (considering resolution and lower chroma sample rates for DV). And 720p @ 24pN using DVCPro HD is only twice as heavy as DV (50mbps vs 25mbps), so the burden isn't unreasonable.

On the other hand, to go for 1080p at 60 fps would eat CPU ahd HD space like crazy - especially if uncompressed. You really need some big toys in post to go that route and render you work by morning.

Certainly, if you can stick with DV you can avoid many back-end problems. But for me the chromakeys, variable frame rate, tapeless workflow and reasonable filesizes make the HVX very attractive for web work - in theory, at least!

[Barry Gribble]

The 24 fps question is another good one...

With web delivery, is it going to matter if I am shooting 30f or 24p? Can people read that difference? Because the XL1 limits me there too.

[Greg Boston]

Originally Posted by Barry Gribble

The 24 fps question is another good one...

With web delivery, is it going to matter if I am shooting 30f or 24p? Can people read that difference? Because the XL1 limits me there too.

In that case Barry, I see two options. Stay with the XL1, shoot in 60i, then de-interlace and convert to 24p in post. There are many good plug-ins for that method these days.

Or, pick-up a second hand XL2 and get the 24P and native 16:9 if that's important to you. I just recently sold both my XL2 cameras here on our classifieds but I continue to see others selling theirs.

Just some food for thought.

=gb=

[Barry Gribble]

Greg,

Thanks... yeah, I have two issues on this that I am tossing around...

1. Does 24p even matter for this format? Even on DVDs I'm not the best judge of what is 24p and what is 30p. With a YouTube-style viewing, who knows... Do they even make it up the 30 fps in that format?

2. If I'm going to buy another camera, I might as well go with a top-of-the-prosumer-line version and get an HD.

[Greg Boston]

Originally Posted by Barry Gribble

Greg,

Thanks... yeah, I have two issues on this that I am tossing around...

1. Does 24p even matter for this format? Even on DVDs I'm not the best judge of what is 24p and what is 30p. With a YouTube-style viewing, who knows... Do they even make it up the 30 fps in that format?

2. If I'm going to buy another camera, I might as well go with a top-of-the-prosumer-line version and get an HD.

I'm certainly not opposed to #2, it's up to you and what your budget will allow. My goal in recommending the XL2 was to get you where you wanted to go at the least possible price.

One thing you would gain with a true 24p workflow all the way through is smaller output filesize. And many web downloads are actually down around 12-15 fps to save even more space.

-gb-

[Jon Fairhurst]

One thing about video for the web is that you have no idea what framerate people are really running. It might be 60Hz (yuk, I see too much flicker, at least on CRTs), 72 Hz, 75 Hz or whatever. That means you can't predict the judder that the end user will see at any given frame rate. Bottom line: don't sweat it.

As Greg pointed out, low framerates save file size and processing for you in post and bandwidth on the web. Sure, you can shoot and edit at 60i, and then downconvert to, say 15 fps, but this only works well for slower material. If you are showing sports or action scenes with, say, a one frame (not field) muzzle flash, going from 60i to 15p has a 50% chance of missing that frame.

That's why I like 24fps native. You don't need any frame rate conversions from start to finish on the web. You can subsample to 12fps for slow stuff if you're stingy, but you can stick with 24 fps for the fast stuff.