Ken Stone article on EX1 -> SD workflow in FCP

Craig Seeman · February 21st, 2008, 12:23 PM

Which DVD preset? Are you using Progressive source?

At some point I'm going to do a test/compare
iDVD from HD Progressive source (which I tried and didn't like - lots of line twitter)
iDVD following Rick Young's workflow (but using DVCPro50)
Compressor/DVDStudio Pro using preset you recommend
Compressor/DVDStudio Pro using Rick Young's workflow but using DVCPro50

I think the issue I saw was related to how iDVD was handling HD Progressive source and although it's using Compressor's "underbelly" it may not be using it "intelligently" with Progressive source.

Quote:

Originally Posted by Rob Collins

For DVD's I've had great results just exporting the sequence HD, then using Compressor's DVD preset to encode.

Tyler Franco · February 21st, 2008, 04:34 PM

Is this process designed for the purpose of ending up on SD DVD? I ask because I make SD dubs to DVCAM and I followed the outline discussed here. I put the FCP movie that I ended up with inside a standard DV NTSC easy setup timeline and then I just embedded the XDCAM EX timeline within the same timeline. The embedded XDCAM EX timeline looked better than the FCP movie I created using the outline discussed here. I thought it looked very nice actually. Or perhaps I'm doing something different.

IF this procedure is designed for the sole purpose of going to SD DVD, one should not master out as DV but rather DVCPro50.

I guess I'm just a little confused because my results are showing up different.

Craig Seeman · February 21st, 2008, 06:27 PM

Tyler can you please explain some things in your post.
"FCP Movie" what do you meant by that?

You added the "FCP Movie" to a "standard DV NTSC easy setup timeline"
You've lost me here. You're not using a 16:9 anamorphic DV NTSC timeline?
and then you "embedded the XDCAM EX timeline within the same timeline."

You've lost me here too. You embedded the XDCAM EX to the same timeline as what? To a Standard DV NTSC timeline with the "FCP Movie" also????

"looked very nice" on what? Computer screen, SD TV set, HDTV, burned to DVD, dubbed to DVCAM?

I couldn't begin to test your workflow. Heck I'm not even sure what you started with. HD Progressive? HD Interlace?

Quote:

Originally Posted by Tyler Franco

Is this process designed for the purpose of ending up on SD DVD? I ask because I make SD dubs to DVCAM and I followed the outline discussed here. I put the FCP movie that I ended up with inside a standard DV NTSC easy setup timeline and then I just embedded the XDCAM EX timeline within the same timeline. The embedded XDCAM EX timeline looked better than the FCP movie I created using the outline discussed here. I thought it looked very nice actually. Or perhaps I'm doing something different.

IF this procedure is designed for the sole purpose of going to SD DVD, one should not master out as DV but rather DVCPro50.

I guess I'm just a little confused because my results are showing up different.

Rob Collins · February 21st, 2008, 06:57 PM

Quote:

Originally Posted by Craig Seeman

Which DVD preset? Are you using Progressive source?

At some point I'm going to do a test/compare
iDVD from HD Progressive source (which I tried and didn't like - lots of line twitter)
iDVD following Rick Young's workflow (but using DVCPro50)
Compressor/DVDStudio Pro using preset you recommend
Compressor/DVDStudio Pro using Rick Young's workflow but using DVCPro50

I think the issue I saw was related to how iDVD was handling HD Progressive source and although it's using Compressor's "underbelly" it may not be using it "intelligently" with Progressive source.

That sounds right. iDVD gave me that line twitter too. I've exported several different progressive hd timelines and used compressor's best-quality DVD settings (my projects have been short).

Craig Seeman · February 21st, 2008, 07:31 PM

I have a short 1080p30 project. I think I should use various methods and put it on to one DVD using DVDSP and see what the differences might be.

I'd also try Rick Young's workflow to iDVD and see if that works better than just dropping the HD file in to a 16:9 anamorphic iDVD project.

It'll probably be a few days before I can run the tests though.

Quote:

Originally Posted by Rob Collins

That sounds right. iDVD gave me that line twitter too. I've exported several different progressive hd timelines and used compressor's best-quality DVD settings (my projects have been short).

Tyler Franco · February 21st, 2008, 09:47 PM

Quote:

Originally Posted by Craig Seeman

Tyler can you please explain some things in your post.

Craig, I'm sorry. Re-reading my post I now realize how much it didn't make sense. I was at work and in a hurry when I typed it. I'll give some much greater detail.

Shooting with EX1 in 1080 30p HQ mode.

Ingest footage into Final Cut Pro and edit my project.

When project is finished I always export the entire timeline as a single Final Cut Pro Quicktime Movie file in the native format I'm editing. In this case, XDCAM EX 1080 30p. I do this by going to File>Export>Quicktime Movie and then leaving as current settings and checking "Make Movie Self Contained". I do this to save this file as my "Master". The quality of my program can't get any better than this file, so I will save this file to a DL-DVD and if I ever need the program again I can pull it off the DL-DVD and drop it right back on a timeline.

Now I need to get this project out to both DVCAM and Betacam SP for television broadcast. What I've been doing is taking that "Master" file and dropping it into a sequence using the DV-NTSC easy setup. It's not anamorphic so it puts my program in letterbox, which is what I want since the broadcast stations don't want anamorphic.

Now instead of that last step I followed Ken's workflow for down-converting and pasted everything into the sequence based on his settings. I then exported that as a final cut pro quicktime movie and placed that movie on my sequence setup with the DV-NTSC easy setup for out put to DVCAM and BetaSP.

To my eye while viewing on my Sony broadcast monitor the "Master (XDCAM EX)" file looks BETTER than DV file using Ken's sequence settings on the DV-NTSC timeline going to DVCAM and BetaSP.

Obviously this simplifies things for me as I don't have to do the extra steps that Ken posted. I just don't understand why though since so many folks have complained about the ability of FCP to simply down-convert the footage.

The only thing I can think of is that it's because my destination is not DVD and that where footage starts looking bad is the down-conversion to MPEG-2 for DVD, NOT just the downconversion to SD.

Craig Seeman · February 21st, 2008, 10:57 PM

This makes good sense.

When you drop into DV NTSC are you simply using typical lower field first interlace (not changing anything at all from Easy Setup)?

What Sony Broadcast Monitor are you using to check?

When using Ken's (actually Rick Young's) last step to DV are you taking into account the source is progressive as Rick does?

What I'm curious about is the conversion from Progressive to Interlace (Lower field with DV) and how that impact the video?

The problem I had going to DVD (admittedly with iDVD only seemed to be around line twitter and what I think is bad conversion of progressive to interlace.

In my case I'm only doing spots when I do broadcast and I'm using MPEG2 program streams with DG Fastchannel delivery. I need to look at HD for HD channel delivery as well as SD for those channels.

For client viewing I'm using SD DVD. In some cases clients want to do "paper" cuts in programs like iMovie so I need to consider DV vs DVD delivery for window dubs.

Tyler Franco · February 22nd, 2008, 12:22 AM

I'm not around the monitor at the moment, so I'm not sure of the model. It's by no means the top of the line though and my assessment is more "eyeballing" it than anything scientific.

I'm not changing anything with the DV-NTSC preset. Just using it like it is because I have to go out to DVCAM and changing anything messes with my firewire output to the DSR-40. I'm assuming Final Cut is switching field dominance as it sees fit and doing conversions automatically.

I've considered shooting in 1080 60i but I realize that progressive is the future and the programs I produce are for children and are meant to have replay value for at least 5-7 years as there are always new people moving into the age bracket we target. So far the progressive to interlaced translates very well and I can't complain. I see no twitter. If anything, it may just seem like it has a slight "stutter" but it basically looks like watching a film shot in progressive on an interlaced CRT.

All that being said though, the downconversion from the EX1 camera via component absolutely blows away any downconversion I have been able to do in software. Which is exactly why I wish I could send my edited project BACK to the SxS card and then make dubs via component to DVCAM and BetaSP. I guess I'll just keep dreaming. Or I'll get a new Mac Pro with lots of memory and hard drive space, and a fancy capture card. I'm dreaming either way. :)

Tyler Franco · February 23rd, 2008, 12:03 AM

Well, after much testing today I think I've determined how to get the best SD downconversion to BetaSP. After the completion of my edit, I drop my final edited sequence on an HDV 1080 60i sequence. I then render and send that back out firewire and record it back to the EX1. I then hook up via component to my deck and let the camera do the downconversion. This method really produces great results, even with progressive footage that has been converted.

Ideally, Sony will give us a way to put the XDCAM EX BACK on the SxS card so we could skip the entire conversion to HDV. It would save a ton of time and maintain a higher quality.

Of course none of this solves anything for those who aren't going back out of their computer and want to burn DVD's.

Bill Davis · February 23rd, 2008, 01:04 AM

Quote:

Originally Posted by Greg Boston

I agree Craig. If down converting HD, why not stop at a 50 mb codec that will allow for more chroma resolution.

-gb-

Greg,

Because sometimes more resolution isn't necessarily a good thing.

I have more corporate clients asking me to compress and downrez work for the web or corporate intranets than in asking me to deliver HD content - a LOT more.

They aren't even asking for HD origination for large venue stuff like corporate conference and trade show display. Most forward looking companies are in belt tightening mode with recession/current stock market anxiety and don't want to pay for "frills" right now. So HD origination is a total non-starter.

Current SD output is more than good enough to populate a 320x240 15fps window in a Flash based training module.
Wait, strike that. I don't even need 320x240! I'm doing mostly green screen work where I can knock out a live talking character to composite with a still background to make compression easier and further reduce bandwidth. So the actual moving video window might be a 100 x 240 live video area!

For that, the EX-1 is an OVERREZZED solution. But so affordable and future proof that it commands a close look.

But only if there's a smooth workflow for using it NOW.

And that's what I'm looking for.

Craig Seeman · February 23rd, 2008, 09:37 AM

Bill, I think you're making a few mistakes in some of your assumptions.
More rez is better. Just look at the Apple movie trailers . . . even the small versions. Coming from "feature film" resolution helps even if you're going small.

Progressive is always best for the web. No need to deal with de-interlace issues.

More rez can get you a better key. More tightly knit samples even at 4:2:0. Depending on the workflow (and I'm not recommending any particular one since I haven't tested) going into 4:2:2 color space will help keys and graphics. That means DVCPro50 4:2:2 is a good thing. One can also take 10 bit uncompressed 4:2:2 straight from the camera head if you have the record infrastructure for it.

Best bet would be to shoot 1080p30 out of camera HD-SDI 10 bit Uncompressed 4:2:2. great for keying. Progressive. Downconvert gives you plenty of room to reposition shots, do fake dolly and tracking shots.

Otherwise I suspect putting that 1080p30 material in a DVCPro50 (or 8 bit uncompressed or ProRez in FCP?) timeline will really help keys and graphics going to the "smallest" screen.

Greg Boston · February 23rd, 2008, 01:11 PM

Quote:

Originally Posted by Craig Seeman

Downconvert gives you plenty of room to reposition shots, do fake dolly and tracking shots.

Exactly, that's one of the often overlooked reasons to originate in HD even for SD delivery. You give yourself a lot of 'wiggle room' within the smaller output frame. Keyframing camera moves can save an otherwise unusable shot.

I also feel that the finished image when down converting to a smaller size, will benefit from a better camera in terms of dynamic range and color reproduction. IOW, giving the encoder more to work with going in, yields a better down convert IMHO.

-gb-

Bill Davis · February 23rd, 2008, 05:01 PM

Guys,

I'll spot you the general principal that of "more origination rez is better"

HOWEVER,

as a practical matter, starting with one raster and delivering on another - has ALWAYS been problematic - precisely because that very process can easily yield poor results.

As evidence, just look at the glut of SD content currently being shoehorned into HD cable channels. The work doesn't look NEARLY as good as it does on an SD set - precisely because they're re-mapping a high def raster onto a lower def one.

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

EVERYPLACE I watch HD on SD displays *** OR *** SD ON HD displays I see more POOR IMAGE QUALITY than I ever did when we had a single raster standard to work with. Hard to read type. Color fringing. etc. etc.

So I'm not fully convinced that TODAY, dumping all of our SD origination isn't asking for a lot of trouble. Not so long as it's still largely an SD viewing world

In fact, isn't that what this thread is all about? Dialing in on the "magic" formula to translate the EX-1s high density raster into something that will translate well to the real world SD and SMALLER than SD rasters that we're all forced to work with?

Unless I miss something in my understanding of what's involved, we're talking about a lot of digital number voodoo that's the functional equivalent of the "convolution" and "alpha edge masking" that has always been necessary to make something created in one rez look acceptable in another rez.

The mere fact that the early adopters who are reporting here are saying that they're trying workflows that give them TERRIBLE results, indicates that this isn't nearly as simple as having an "output to SD" choice in FCP that does it right every time.

Isn't that the whole point of this discussion?

I'd be interested in hearing from those who are having an easy time of this.

How are you taking your EX-1 footage and delivering it to clients via SD DVD? Is there a simple way that always looks great? Understandably not as great as HD on Blue Ray with all the original rez, but at LEAST as good as what we have now via standard SD rez to DVD.

Cuz I'm a little reluctant to tell clients I'm shooting in HD so that I can deliver POORER quality on the discs they use right now in their current SD kiosk traiining systems.

The minute I can get this answered to my complete satisfaction - I'll order a couple of EX-1s and never look back.

But NOT if it's gonna add another messy step in my production chain for extra resolution that right now, NOBODY in my customer base is asking for.

Bill Davis · February 23rd, 2008, 05:22 PM

[QUOTE=Craig Seeman;831380]Bill, I think you're making a few mistakes in some of your assumptions.

Craig,

With due respect I don't think so.

My only assumptions are as follows.

A) Without a ready market for what any gizmo produces, investing in it MUST be questioned if you want to foster a sustainable business.
B) Workflow changes that make things HARDER to accomplish must return MORE profit to justify the grief.

And most importantly,

C) It's not what I want that makes my business grow. It's understanding what my clients actually need AND WILL PAY FOR!

And for most of them, they're willing to pay to accomplish solid business communications with customers, employees and occasionally the general public. Very few of which are spending much time watching HD content RIGHT NOW.

Go out and look for yourself. Where is the HD stuff? It's pretty much limited to retail displays - from mall signage to bar televisions - and a pretty small subset of home televisions which can display the native rez of the EX-1 properly. (Granted that number is growing, but it's still VERY, VERY small compared to SD viewing.)

For EVERYTHING else, down-rezzing is either clearly advantageous or downright necessary!

So I don't think it's foolish to address the down-rez necessity before I retire my DSR gear and switch to EX kit.

Don't get me wrong, I can see that I'll need to do it at the right moment.

I just don't want that moment to be after I've spent a YEAR watching the gear depreciate without anyone actually PAYING me for the higher resolution potential it represents.

That's not just an assumption. I kinda think it rises to the level of BELIEF. :)

Mark OConnell · February 23rd, 2008, 05:36 PM

Quote:

Originally Posted by Craig Seeman

I posted this a few days ago and few read it

http://dvinfo.net/conf/showpost.php?...12&postcount=1

My question would be why DV? Why not DVCPro50 or some higher quality SD? It would be easy enough to figure that out based on this article. The important thing is it shows how to downconvert HD Progressive.

I agree. If you want an SD DVD converting to DV first is crazy. I had good results yesterday making an SD DVD from a 1280x720 30p project by rendering to ProRes, then just dragging that to Compressor and selecting 90 min best quality-

February 21st, 2008, 04:34 PM	#17
Tyler Franco Regular Crew Join Date: Feb 2008 Location: Columbus, OH Posts: 172	Is this process designed for the purpose of ending up on SD DVD? I ask because I make SD dubs to DVCAM and I followed the outline discussed here. I put the FCP movie that I ended up with inside a standard DV NTSC easy setup timeline and then I just embedded the XDCAM EX timeline within the same timeline. The embedded XDCAM EX timeline looked better than the FCP movie I created using the outline discussed here. I thought it looked very nice actually. Or perhaps I'm doing something different. IF this procedure is designed for the sole purpose of going to SD DVD, one should not master out as DV but rather DVCPro50. I guess I'm just a little confused because my results are showing up different.

February 21st, 2008, 10:57 PM	#22
Craig Seeman Inner Circle Join Date: Feb 2004 Location: Brooklyn, NY, USA Posts: 3,841	This makes good sense. When you drop into DV NTSC are you simply using typical lower field first interlace (not changing anything at all from Easy Setup)? What Sony Broadcast Monitor are you using to check? When using Ken's (actually Rick Young's) last step to DV are you taking into account the source is progressive as Rick does? What I'm curious about is the conversion from Progressive to Interlace (Lower field with DV) and how that impact the video? The problem I had going to DVD (admittedly with iDVD only seemed to be around line twitter and what I think is bad conversion of progressive to interlace. In my case I'm only doing spots when I do broadcast and I'm using MPEG2 program streams with DG Fastchannel delivery. I need to look at HD for HD channel delivery as well as SD for those channels. For client viewing I'm using SD DVD. In some cases clients want to do "paper" cuts in programs like iMovie so I need to consider DV vs DVD delivery for window dubs.

February 22nd, 2008, 12:22 AM	#23
Tyler Franco Regular Crew Join Date: Feb 2008 Location: Columbus, OH Posts: 172	I'm not around the monitor at the moment, so I'm not sure of the model. It's by no means the top of the line though and my assessment is more "eyeballing" it than anything scientific. I'm not changing anything with the DV-NTSC preset. Just using it like it is because I have to go out to DVCAM and changing anything messes with my firewire output to the DSR-40. I'm assuming Final Cut is switching field dominance as it sees fit and doing conversions automatically. I've considered shooting in 1080 60i but I realize that progressive is the future and the programs I produce are for children and are meant to have replay value for at least 5-7 years as there are always new people moving into the age bracket we target. So far the progressive to interlaced translates very well and I can't complain. I see no twitter. If anything, it may just seem like it has a slight "stutter" but it basically looks like watching a film shot in progressive on an interlaced CRT. All that being said though, the downconversion from the EX1 camera via component absolutely blows away any downconversion I have been able to do in software. Which is exactly why I wish I could send my edited project BACK to the SxS card and then make dubs via component to DVCAM and BetaSP. I guess I'll just keep dreaming. Or I'll get a new Mac Pro with lots of memory and hard drive space, and a fancy capture card. I'm dreaming either way. :)

February 23rd, 2008, 12:03 AM	#24
Tyler Franco Regular Crew Join Date: Feb 2008 Location: Columbus, OH Posts: 172	Well, after much testing today I think I've determined how to get the best SD downconversion to BetaSP. After the completion of my edit, I drop my final edited sequence on an HDV 1080 60i sequence. I then render and send that back out firewire and record it back to the EX1. I then hook up via component to my deck and let the camera do the downconversion. This method really produces great results, even with progressive footage that has been converted. Ideally, Sony will give us a way to put the XDCAM EX BACK on the SxS card so we could skip the entire conversion to HDV. It would save a ton of time and maintain a higher quality. Of course none of this solves anything for those who aren't going back out of their computer and want to burn DVD's.

February 23rd, 2008, 09:37 AM	#26
Craig Seeman Inner Circle Join Date: Feb 2004 Location: Brooklyn, NY, USA Posts: 3,841	Bill, I think you're making a few mistakes in some of your assumptions. More rez is better. Just look at the Apple movie trailers . . . even the small versions. Coming from "feature film" resolution helps even if you're going small. Progressive is always best for the web. No need to deal with de-interlace issues. More rez can get you a better key. More tightly knit samples even at 4:2:0. Depending on the workflow (and I'm not recommending any particular one since I haven't tested) going into 4:2:2 color space will help keys and graphics. That means DVCPro50 4:2:2 is a good thing. One can also take 10 bit uncompressed 4:2:2 straight from the camera head if you have the record infrastructure for it. Best bet would be to shoot 1080p30 out of camera HD-SDI 10 bit Uncompressed 4:2:2. great for keying. Progressive. Downconvert gives you plenty of room to reposition shots, do fake dolly and tracking shots. Otherwise I suspect putting that 1080p30 material in a DVCPro50 (or 8 bit uncompressed or ProRez in FCP?) timeline will really help keys and graphics going to the "smallest" screen.

February 23rd, 2008, 05:01 PM	#28
Bill Davis Trustee Join Date: Jan 2004 Location: Scottsdale, AZ 85260 Posts: 1,538	Guys, I'll spot you the general principal that of "more origination rez is better" HOWEVER, as a practical matter, starting with one raster and delivering on another - has ALWAYS been problematic - precisely because that very process can easily yield poor results. As evidence, just look at the glut of SD content currently being shoehorned into HD cable channels. The work doesn't look NEARLY as good as it does on an SD set - precisely because they're re-mapping a high def raster onto a lower def one. I consider it the digital video equivalent of re-screening a screened photo back in the print industry. EVERYPLACE I watch HD on SD displays * OR * SD ON HD displays I see more POOR IMAGE QUALITY than I ever did when we had a single raster standard to work with. Hard to read type. Color fringing. etc. etc. So I'm not fully convinced that TODAY, dumping all of our SD origination isn't asking for a lot of trouble. Not so long as it's still largely an SD viewing world In fact, isn't that what this thread is all about? Dialing in on the "magic" formula to translate the EX-1s high density raster into something that will translate well to the real world SD and SMALLER than SD rasters that we're all forced to work with? Unless I miss something in my understanding of what's involved, we're talking about a lot of digital number voodoo that's the functional equivalent of the "convolution" and "alpha edge masking" that has always been necessary to make something created in one rez look acceptable in another rez. The mere fact that the early adopters who are reporting here are saying that they're trying workflows that give them TERRIBLE results, indicates that this isn't nearly as simple as having an "output to SD" choice in FCP that does it right every time. Isn't that the whole point of this discussion? I'd be interested in hearing from those who are having an easy time of this. How are you taking your EX-1 footage and delivering it to clients via SD DVD? Is there a simple way that always looks great? Understandably not as great as HD on Blue Ray with all the original rez, but at LEAST as good as what we have now via standard SD rez to DVD. Cuz I'm a little reluctant to tell clients I'm shooting in HD so that I can deliver POORER quality on the discs they use right now in their current SD kiosk traiining systems. The minute I can get this answered to my complete satisfaction - I'll order a couple of EX-1s and never look back. But NOT if it's gonna add another messy step in my production chain for extra resolution that right now, NOBODY in my customer base is asking for.

February 23rd, 2008, 05:22 PM	#29
Bill Davis Trustee Join Date: Jan 2004 Location: Scottsdale, AZ 85260 Posts: 1,538	[QUOTE=Craig Seeman;831380]Bill, I think you're making a few mistakes in some of your assumptions. Craig, With due respect I don't think so. My only assumptions are as follows. A) Without a ready market for what any gizmo produces, investing in it MUST be questioned if you want to foster a sustainable business. B) Workflow changes that make things HARDER to accomplish must return MORE profit to justify the grief. And most importantly, C) It's not what I want that makes my business grow. It's understanding what my clients actually need AND WILL PAY FOR! And for most of them, they're willing to pay to accomplish solid business communications with customers, employees and occasionally the general public. Very few of which are spending much time watching HD content RIGHT NOW. Go out and look for yourself. Where is the HD stuff? It's pretty much limited to retail displays - from mall signage to bar televisions - and a pretty small subset of home televisions which can display the native rez of the EX-1 properly. (Granted that number is growing, but it's still VERY, VERY small compared to SD viewing.) For EVERYTHING else, down-rezzing is either clearly advantageous or downright necessary! So I don't think it's foolish to address the down-rez necessity before I retire my DSR gear and switch to EX kit. Don't get me wrong, I can see that I'll need to do it at the right moment. I just don't want that moment to be after I've spent a YEAR watching the gear depreciate without anyone actually PAYING me for the higher resolution potential it represents. That's not just an assumption. I kinda think it rises to the level of BELIEF. :)