Ken Stone article on EX1 -> SD workflow in FCP

[George Kroonder]

Originally Posted by Bill Davis

I consider it the digital video equivalent of re-screening a screened photo back in the print industry.

Hi Bill,

Sorry, but I can't agree to that. Being an ex-print-industry guy myself, I have to correct you on this. With re-screening you're trying to re-create information that is no longer there. In the case of a half-tone photo, screening leaves only black and white and nothing in between.

Let's leave print out of it and get back to the SD <-> HD conversion arena ;)

Shoehorning video into another format and make it look it's best can be difficult but at least we should have a lot more information to work with. Using the correct tools is a big part of the solution. Maybe that means recomposing some projects, re-wrapping others or using tools like Compressor to convert formats. It just depends on the project at hand.

George/

[Craig Seeman]

Bill, you're confusing a lot of things here.

First, SD to HD is going to have problems. It's going the OTHER way where you get the advantage of source resolution.

The poor image quality you see is often due to the broadcaster and how they're doing their compression. I've seen some local HD ONLY channels that look good. Channels that do both are hit and miss and miss more often than not. Don't blame that on the format though. I certainly would NOT hand in SD to broadcast on an HD channel when I can deliver HD.

I can certainly shoot in both SD and HD. Some of my clients are moving to HD with my encouragement AND THEY'RE PAYING MORE FOR IT. If you can show how it benefits them (and it certainly can even with SD delivery) they'll pay for it.

I haven't a clue what you mean by HD being "harder." Learning how to shoot, edit and even downconvert require new and improved skills. Getting good downconverts is about LEARNING not HARDER. All of us here are LEARNING. The actually process is VERY EASY. Discovering the best way may take some exploration though.

For ME, tapeless is SO MUCH EASIER that I'd rather learn to downconvert than continue shooting in SD.

BTW there certainly are many people shooting HDV and charging DV/SD prices. That's simply BAD business. Shooting with the EX1 can certainly differentiate oneself from those in the market.

If you don't think your customers will see or pay for it maybe you're not selling it right OR maybe your clients do NOT want to pay for it. That's OK.

Don't EVER buy gear that doesn't grow your business. I did NOT wait for my clients to ASK for HD. I bought EX1 because I thought I could sell the EX1's capabilities to my SD clients and . . . I have . . . even though I need to deliver SD for them.

Gosh maybe we need to start another thread on how to sell the EX1 to your SD clients.

[Bill Davis]

QUOTE:

I haven't a clue what you mean by HD being "harder." Learning how to shoot, edit and even downconvert require new and improved skills. Getting good downconverts is about LEARNING not HARDER. All of us here are LEARNING. The actually process is VERY EASY. Discovering the best way may take some exploration though.


Well, what I mean is that today, once I shoot and edit in today's (DV-NTSC) I maintain both the resolution and aspect ratio right onto the DVD. The compression is automatically applied during authoring and it's designed to fill a known raster.

When someone shoots with an EX-1 - they originate a 16x9 aspect ratio with a much higher pixel density.

Then you have to make a series of choices. Since the real world doesn't fully support EITHER that aspect ratio or the higher resolution of your origination, you're largely FORCED to resize and/or reformat and/or re-rez your work.

At least if you're going to deliver to the two largest target outlets for work today - 1- SD legacy sets and 2 - web and portable devices that have little use for an EX-1 size signal.

That's all I'm saying.

Where's the simple path?

And why are there so many "early adopters" of the EX-1 who seem to be having trouble delivering acceptable looking FULLY COMPATIBLE video into these lucrative streams.

Perhaps I am mistaking this. Perhaps folks here are churning out SD-DVDs for clients originated on EX-1s right and left.

But everything I've read so far talks about ADDING SIGNIFICANT TRANSCODING TIME AND EFFORT over and above what we do for SD originated work today.

Am I the only one concerned with this?

I'm not arguing about charging the same for HD work as SD. I'm arguing about inserting EX-1 acquisition in a perfectly acceptable current SD origination workflow that already simply and easily meets most video customers needs.

I'm curious about this question. Does this approach make things UNREASONABLY difficult if you aren't planning to do most of your delivery on HD ***RIGHT NOW!***

That affects (or at least it should affect) the speed and simplicity of the transition for many of us - not from SD to HD when we need HD. But from adopting and implementing EX-1s in a SD shop with the idea that we'll be ready to switch to HD when and if our customers ask us to.

Is that clearer?

[Tyler Franco]

I think Bill does have a point. Someone taping a training seminar that is going to be played at a corporation meeting in a month on a big standard def projector and then never played again does not have any need to go through the extra hassles of shooting HD.

However, I think most people want to be "future proof". I know if I was a wedding videographer I'd want an HD master of the ceremony to sell my client on Blu-Ray in five years.

As for what my company does, we produce children's programming that is supposed to have a reply value of about 10 years! So everything I shoot in SD now is something I won't be able to use in the future. So for many of us we just have to learn to jump through the hoops for a high quality downconversion process.

[Craig Seeman]

If my clients see an advantage to shooting HD and are willing to pay extra for it than I'll certainly have the time to do the downconvert given the extra money.

Many people get EXCELLENT results downconverting HD to SD. Keep in mind that downconverting Progressive HD may have a different workflow than downconverting Interlace HD.

The downconvert takes time. It's worth that time if clients are willing to pay for HD. It's not worth that time if clients aren't.

The ability to use HD in SD timeline and do fake track and dolly shots is a good example why HD MIGHT be worth it. It can be less expensive than actually buying/renting/setting up tracking and dolly shots and you have complete freedom to alter and improve such movement in post.

16:9 with letter box in SD is not only acceptable but desirable by many clients. Of course you could get an HDV camera that shoots 16:9 SD and there's nothing wrong with going that route if that's what your clients want and what they're willing to pay for.

Of course I can hand a client a DL-DVD with the raw clips from the EX1 and Sony's Mac and Windows Clip Browser and my clients can view the Master clips without the need of either a DV or HDV deck.

[Peter Kraft]

I am a little bit astonished you are talking 34 messages about which workflow might yield a better (the best?) result to convert from HD to SD and nobody mentions the basic math you have to do first to get the best possible result.

Let's assume you have original footage in 1920x1080. The best downresszed
format will be in 960x540 or 480x270. It's always a full figure division, be it 1, 2, 3 or 4. We are working in the digital domain and digital means that only full figure divisions are possible to get a decent result. Anything else can only yield to more or less mediocre results.

As I am living in a PAL country, I can only argue from that point on now.
Please bare with me...

960 x 540 is 16x9, as is 480x270. I then open the file in the player typing Command-J and changing 540 in the resolution dialog box into 576 ... and voilà ...
960 turns into 1024. Save that and you have an SD file, which can be easily converted into 720x576 (PAL anamorphic) in outstanding qualty.

You may de-interlace at any point after down-rezzing from HD to SD or start with a progressive file from the start.

Hope this helps.

[Chris Hurd]

Originally Posted by Peter Kraft

It's always a full figure division, be it 1, 2, 3 or 4. We are working in the digital domain and digital means that only full figure divisions are possible to get a decent result. Anything else can only yield to more or less mediocre results.

I am sorry but this is absolutely false. The notion that decent resizing / recompression results can be achieved only by whole-number divisions "because it's digital" is purely a myth and there is no truth in it whatsoever. Or to put it a better way, this may be true only for a lousy and inefficient resizing or recompression algorithm.

[Bill Davis]

Originally Posted by Craig Seeman

If my clients see an advantage to shooting HD and are willing to pay extra for it than I'll certainly have the time to do the downconvert given the extra money.

SNIP

The ability to use HD in SD timeline and do fake track and dolly shots is a good example why HD MIGHT be worth it. It can be less expensive than actually buying/renting/setting up tracking and dolly shots and you have complete freedom to alter and improve such movement in post.

Craig,

First and foremost, I don't see any of this as my clients paying EXTRA for anything. The VAST majority of the money they're paying me is for my knowledge and experience. For my decision making. The cost of the equipment is and should be incidental in my view.

Don't get me wrong, it's nice that cameras are getting both better and cheaper. IN SOME WAYS. In others not so much...

For instance, I think everyone here should stop and realize that one potential downside to camera's like the EX-1 is that their low financial entry point mean that there will be a LOT of competition suddenly showing up in the "high def" production realm. Supply and demand realities mean that this very camera will put HUGE downward price pressure on "high def" production prices.

I'm all for competition, since like I said above, I don't base my business on equipment.

But it sure wouldn't surprise me at all if most of us find lots of beginners flooding the market with "low cost HD production" based on cameras like the EX-1. It's just how things go.

So if you're planning to charging MORE for EX-1 "high def" I'd suggest you do it while you can. A year from now, I bet EX-1 shooters won't be getting a penny more than decent SD shooters get today - likely less.

Hopefully by then, we'll all have absorbed the cost of working with HD footage natively and can handle the lower competitive rates.


As to your other contention...

I'm sorry but this just seems inherently WEIRD to me.

I can't imagine looking at a monitor and trying to decide - not if the framing is RIGHT - but if I *might* want to use some part of the whole later.

At least in my brain, rather than showing the proper relationship between the characters or visuals, the idea that somebody will have to go in after the fact and RE-FRAME the work sends chills up my spine.

It might seem like a "bonus" capability on the surface, but I see it as a HUGE time waster in all but some exceptionally unusual cirtumstances.

After all, isn't that precisely the problem with pan and scan films re-aspected for TV?

Instead of seeing what someone with a decent eye framed, you're got to compromise via pan and scan in virtually every scene!

Yikes.

And as to your "virtual dolly shot idea", that escapes me as well.

If the shot needs to be a dolly shot - I can't even FATHOM trying to fake that with the 15% left and right you'd get from EX-1.

If it needs to be a dolly shot - it needs to be a dolly shot. Period. Thinking about providing one audience with a pesudo dolly shot and possibly providing another with a fixed original framing in my mind guarantees that you'll never get the damn shot RIGHT for either.

But maybe I'm just not bright enough to see the benefit in trying to frame for two purposes at once.

FWIW.

[Greg Boston]

Originally Posted by Bill Davis

If the shot needs to be a dolly shot - I can't even FATHOM trying to fake that with the 15% left and right you'd get from EX-1.

It's really a bit more than that. You can fit 4 standard definition frames in the size of a high definition frame. If you were to place an SD frame in the center of an HD frame, you could move 50% in each direction. Not that you would be thinking that during the production phase, but it could save a shot, or make one a bit more interesting.

If you don't feel HD is right for your business model at the moment, that's okay. There's still a lot of BetaSP cameras finding work at the moment. Even if you choose not to shoot HD, I'd still encourage you to start shooting widescreen format as soon as possible.

-gb-

[Bill Davis]

Fair comment, Greg.

But since you're shooting an F-350 and have the OPTION of shooting SD or HD - how about your take on which format your camera spends more time in?

Are your clients clamoring for widescreen? How are they showing it and to what audience?

I'm seriously interested.

[Peter Kraft]

Originally Posted by Chris Hurd

I am sorry but this is absolutely false. The notion that decent resizing / recompression results can be achieved only by whole-number divisions "because it's digital" is purely a myth and there is no truth in it whatsoever. Or to put it a better way, this may be true only for a lousy and inefficient resizing or recompression algorithm.

Chris, with all due respect, I am not of your oppinion.
First, I am not talking about recompression but resizing, which is a different process.
Second, the whole-number divisions thing is not a myth but something people are taught at the university, so it can't be THAT false.
Third, try yourself: To resize 1920/1080 to PAL 720/576 would require you to reduce the HD original by factor 0,53333333. Compare the result with the same original HD footage you resized by factor 0,50. You will end up with very small black bars on top and bottom of the resulting video, kind of "fake letterboxing". BUT it will be significantly sharper then the first one. Tried that several times, went to the lab, evaluated the result and always came to the same judgement - it is sharper.

Bottom line: It do not say the a non-whole-number division would not be possible. All I want to say is, if you strive for the sharpest possible result, only believe in maths not myths.

With all due respect
Peter

[Greg Boston]

Originally Posted by Bill Davis

Fair comment, Greg.

But since you're shooting an F-350 and have the OPTION of shooting SD or HD - how about your take on which format your camera spends more time in?

Are your clients clamoring for widescreen? How are they showing it and to what audience?

I'm seriously interested.

Bill, I've pretty much settled on being just a camera owner/operator at this point. Technically, I know how to edit, but it just doesn't give me the same satisfaction as the act of acquiring the images. That being said, I find myself shooting HD for the most part. Doing live events, I either supply the HDSDI output, or, the SD composite output to the client, but if rolling (er... spinning), the camera is laying to disc in HD for posterity.

An example was in December where I was hooked to the live truck of a local tv station via composite SD. Although the program was broadcast in HD with the main cameras, the remotes were fed with SD widescreen because the trucks haven't been upgraded yet. This same station uses BetaSX in widescreen for local news and the upconvert looks pretty nice, not HD, but still pretty darn good. But the parts I recorded locally to disc were in HD so that the client can cut promos for next year's event in HD. I've noticed more television ads showing up in HD on my set lately. Other freelancers say they are getting more calls for HD shoots too.

I've had a client shoot HD to disc for SD DVD delivery. I can count on one hand the number of times I've actually had to use DVCAM mode on the camera. Once was for intercutting my camera with a PD-170 in post.

The photographer side of me says to acquire images in the highest possible quality and that's why I'm such a proponent of HD. I want as much exposure latitude and resolution that my budget will allow. That's my creative side speaking out.

-gb-

[Bill Davis]

Originally Posted by Greg Boston

Bill, I've pretty much settled on being just a camera owner/operator at this point. Technically, I know how to edit, but it SNIP
hat my budget will allow. That's my creative side speaking out.

-gb-

Okay, I get that totally. If I was shooting for others, the clear path to success is to shoot whatever format and onto whatever media they choose. And if I was primarily shooting for TV Stations that are transitioning to HD I'd be crazy delighted if I could work with something like an EX-1 for delivery.

That's a lot different from program origination for corporate and industrial clients.

I can see the pressure to deliver on high def for broadcast. But I've still got to contend that for every dollar spent to create video content for broadcast - I think there's $100 (if not a thousand!) spent on standard business video.

After all there are a whole lot more businesses out there than TV stations, cable companies, or even home high def screens - at least right now.

Perhaps in time, direct HD video sales to set users via something like iTunes might change the market. But until then, I think cameras exclusively shooting in High Def like the EX-1 will only generate significant returns for niche players - not the wider market. And precisely because of it's modest price point, I think that market will be REALLY well served by current shooters adding that capability rather than opening up scads of opportunities for early adopters and beginners.

Let's both hope I'm wrong. It will make for much more interesting business if I am!