I have just moved over to FCS2 at the facilities where i work, using quad intel machines. We don't have the capacity to convert all our footage to pro-res as much as we'd like to and therefore have been working in HDV native codec. We have been delivering all our finished material on DVCAM or HDV SCQT's. (Which we are doing more so now than before.)

I have noticed/heard people mentioning using Pro-res as the render codec and wondered what (if any) benefits this would bring?

Any info much appreciated....

It's faster, and higher quality. ProRes is optimized for multi-processor systems, so it's a lot faster to work with it than it is to work with HDV, even if your source material is HDV and you're just rendering to ProRes.

ProRes is also really close to being lossless, so footage passed through the ProRes compressor multiple times will hold up better than nearly any other format.

ProRes is also pretty much essential as an intermediate format if you're shooting something very lossy, like HDV, and trying to grade it or do effects. You won't get any extra latitude by converting your footage to ProRes, but you will have more latitude if you convert to ProRes and stay there throughout your post pipeline, as opposed to compressing back to HDV over and over again as you work the show.

Thanks for the quick reply jeff, so it would be beneficial to change my render codec if everything else is HDV? will i have to change anything for the final product such as re-render in hdv or can i output it as a pro-res qt?

Yeah, that part of your post threw me. I don't know why anybody would ever export an HDV Quicktime. Only another Final Cut system can read a Quicktime in HDV format, and if you're going to another Final Cut system, there are so many better choices than HDV.

HDV is what we call an "acquisition format." It's a format in which you shoot, but it's not a format in which you work or deliver finished shows.

Not knowing anything about your deliverables, I'm not sure I can speak in an education fashion about what your delivery pipeline should look like.

we deliver to another fcp house where the stuff is converted to mpeg formats for the web and/or dvd. So i imagine i need to find out from them which they would prefer........ thanks jeff!

Yeah...I just started rendering grafix projects @ EA in ProRes and I'm likin' it. Nice tight file size in an almost lossless codec. If it supported embedded Apha Channels, it would probably be the "perfect codec"...but for now I still use Animation for embedded Alpha Channels. For final renders, you cannot beat ProRes. Our whole department is gonna be doing it come the new year as we ditch our PC Avids for shiny OctoCores.

When life gives you dilemmas...make dilemmanade.

...if they are a FCP house, they will like ProRes just fine

When life gives you dilemmas...make dilemmanade.

i <3 ProRes, i use it for everything. Alpha channel support would be super nice although i do all my graphics in Motion so i never actually have to export anything outside of FCP but i understand the need for AE and alpha channels in other workflows.

Hey Jeff, I know we've talked about this before, and I'm still a bit unsure.

If my acquisition format is HDV and my final destination is back to HDV tape...is there any advantage to using ProRes from a quality standpoint? And are we just talking about using ProRes as the render codec on an HDV timeline, or are we on a ProRes timeline using the ProRes render codec? Or are we capturing as ProRes as well?

Thanks!
Casey

ProRes doesn't support alpha channels because it's a packed-YUV format. I guess in theory it would be possible to do an RGBA equivalent of ProRes, but since it was meant for broadcast television production ? 4:2:2 YUV ? it's not hard to see why they didn't bother.

Besides, track mattes are easy in Final Cut. I just render out my After Effects shots with a ProRes fill and a ProRes alpha-only matte, and tell Final Cut to track-matte them for me. That's how I do chyrons and supers and stuff, when I'm finishing in Final Cut directly.

[Our whole department is gonna be doing it come the new year as we ditch our PC Avids for shiny OctoCores.]

OOooooooooo I like that. Although the new Mac Avid MC 3.0.5 is behaving well-- on an Intel, of course.

- Loren
Today's FCP keytip:

Apply your default audio transition instantly
with Command-Option - T !

Final Cut Studio 2 KeyGuide? Power Pack.
Now available at KeyGuide Central.
www.neotrondesign.com

Avid behaving has nothing to do with it. My Avid on my Quad at home runs just perfect - always has. The reason we are relieving ourselves of the Avids is not because of the interface. It's the industry's WORST Customer Care, the waiting for the software to always catch up to Mac OS updates and the "bend over - here comes another upgrade" $$$ for Hardware / Software "Support Assurance". Avid sux a$$ in these departments and it has hurt them immeasurably. Avid editors I know that once argued vehemently with me that FCP was a "kid's toy and will never make a splash in the industry" ask me my opinions now. Some of them are probably reading this & posting on the LAFCPUG under assumed names (you know who you are ).

When life gives you dilemmas...make dilemmanade.

To get back on topic, if your delivering an HDV tape for DVD authoring then there is little to no advantage to using ProRes, either as the render codec or in a ProRes timeline.

In order to print to HDV tape it must be conformed from an HDV sequence or your Sony deck will not recognize it.

The reason some DVD authoring facilities want HDV tape is because they do the downconvert in the tape deck to capture SD and MPEG encode that, which is probably the worst way to do this but the few I've argued with about this workflow just say "that's how everyone else is doing it."

>... if your delivering an HDV tape for DVD authoring then there is little to no advantage to using ProRes, either as the render codec or in a ProRes timeline.

I'm not sure I'd agree with you on that one Chuck. I'd argue that if nothing else there is a definitive advantage in the form of better/speedier handling of intermediate renders when working with ProRes as the render codec, and even if one creates a HDV master for a client that is no real reason to limit your edit master to that format. Certainly many have also argued that their DVD's encode better from a ProRes intermediate than from native HDV or similar, and it rather sounds like you've argued this latter one yourself with facilities who have said they want an HDV master from which to dub to DVD .... something you'd want to consider if you were working with such a facility but I don't think thats a reason in and of itself to counsel that there's actually no real advantage to using ProRes.

Just my 2c

That's actually slightly misleading, Chuck. What a lot of people miss is the fact that "conforming" to HDV is more than just rearranging the encoded media. Like all GOP formats, HDV consists of spatially compressed I-frames and temporally compressed P- and B-frames. What Final Cut calls "conforming" is really a process of going through the timeline from the start looking for a broken GOP, then recompressing all the frames after that to conform to a standard HDV GOP structure.

In effect, all the media in your Final Cut timeline after the first edit gets heavily compressed twice. First when it's recorded in-camera, and second when you "conform" it for lay-back to tape.

This isn't really anything new. Whenever you lay off your timeline via SDI to any analog or compressed digital format ? and that includes formats like Digibeta and HDCAM SR ? the media is compressed a second time. The difference is that Digibeta and SR are just barely compressed, while HDV is compressed a lot. So that second hit is significant.

You can never make your footage any better, objectively, than it was when it was recorded. But you can structure your workflow in such a way that you avoid making it objectively worse during the post-production process. This includes doing things like asking how your DVD place is going to make their master. If they ask you for an HDV tape, that's an automatic second compression pass that you have to inflict on your footage before it leaves your facility and goes to theirs. If they, in turn, play out your HDV tape over something like component analog, or worse, downconvert it in the VTR to DV before encoding it, those are additional generations of loss.

There was an article in American Cinematographer about ten years ago that talked about how James Cameron was able to get such good-looking prints out of Super 35 negatives. It goes into a lot of detail about the steps that were taken on shows like True Lies and Titanic to make that relatively small Super 35 negative hold up throughout post. It goes into detail like which lens was used on the optical printer for making the interpositives from the original camera negative.

Digital post has the same considerations. We don't deal with film stock and processing very often, of course, but we have to be aware of which steps in our workflow are degrading our footage, and do our best to route around those steps in order to produce the best-looking and -sounding product we can.

So if you're starting with HDV and the final output is HDV, is there an advantage to using ProRes in the middle?

If I'm understanding this right, there is and advantage because there is less conforming that has to be done...just at the end when you're going to lay it to tape, rather than conforming multiple times while editing and rendering, right?

Thanks,
Casey

My opinion...stay in the target format throughout. Final out to HDV tape? Edit HDV. No benefit in converting to ProRes in that case.

When life gives you dilemmas...make dilemmanade.

Thanks Joe, that's all I needed to know!

Again, the missed point Casey is that during the edit process you will likely enjoy a more responsive system with faster rendering if you use ProRes as an intermediate, rather than working in the more processor intensive native codec.

Thanks Andy,

That hasn't been my experience however, the opposite has been reality for me...everything takes significantly longer when I've used ProRes.

For clarity sake, could you answer the following as to how you get faster results:

Captured in HDV or ProRes?
Sequence settings - HDV or ProRes?
Sequence settings - Render Control - Same as sequence or ProRes?

I know all these settings can technically work in conjunction with each other, maybe I had one that was wrong in my test.

Here's what I found on my test from several months ago (capturing ProRes on a ProRes sequence using the ProRes render control):

Scanning through the footage was slower...I couldn't play back at 2x smoothly enough to hear the audio.

Rendering time on practically everything was significantly slower (slow mo, filters, etc).

File sizes are 3-4 times larger, which is not fun if you're capturing on a different computer and copying the files over to the edit computer.

It took 150 minutes to prepare for a Print to Video on a 25 minute clip on a MacBook Pro, but it only took 15 minutes to prepare for a Print to Video on a 15 minute clip on a Mac Pro with 2 dual core processors.


Thanks!
Casey

Quote

So if you're starting with HDV and the final output is HDV?

?then you're either doing something wrong, or something very unusual.

HDV is an acquisition format. It comes out of the camera. It's not generally ? not generally ? a delivery format. It's too highly compressed for that.

If there's a specific situation in which you must deliver as HDV, that's fine, but understand that that's going to be an unusual circumstance. Be careful not to say to yourself "Well, I shot in HDV, so obviously I'll finish in HDV." Because that's usually not the best option.

That's what I'm trying to figure out...if I'm doing something wrong.

For me, the final output HDV isn't my delivery format, it's my edit master so in case any customer wants anything higher than standard DVD as an output, we have the HDV master to use.

I know you've recommended using ProRes the whole way through, and keeping a ProRes master on your hard drive...that's the part I'm not so sure about...I have about 2 hours of edit master footage, and in ProRes, that would make huge files, and for the quantity I do, I don't think using hard drives is a reasonable solution...especially since as of today, 1% of our customers have actually done anything with the footage in HD (a few HD-DVDs and hopefully soon we'll be offering BluRay)...but case they do anything, we're ready. Maybe we shouldn't be editing everything in HD, I don't know

I must be doing something wrong because in my ProRes test, everything was slower, not like 5% slower, but 200% slower...and that goes for the whole process from capturing to handing the customer the finished product, not just for rendering alone.

Thanks for your patience in trying to explain this to me!

Casey

Quote

Andy Mees said:
Again, the missed point Casey is that during the edit process you will likely enjoy a more responsive system with faster rendering if you use ProRes as an intermediate, rather than working in the more processor intensive native codec.

Not true IMHO. HDV editing is native to FCP. It's very fast as long as you don't have to conform it for a different output.

When life gives you dilemmas...make dilemmanade.

Or, y'know, render anything.

So what do you think I might be doing wrong not to see the speed difference? I've described my process in my test. I have a Mac Pro with 2 x 3 GHz Dual-Core Intel Xeon processors with 2 gigs of memory...shouldn't that be enough?

I don't know. What you reported doesn't make any sense. Capturing takes longer? Capturing HDV is capturing HDV is capturing HDV, regardless of what your timeline settings are. It's a real-time process. The math on that just doesn't add up.

Andy in theory I think your right, but after doing more than 100 HDV projects I have not noticed any benefit to having the render set to ProRes. Of coarse that make me a hypocrite becuase my render settings are always set to ProRes but we have other workstations that aren't and they work just as fast as mine.

Jeff, I think you'd be surprised at how many projects that are finished on SD DVD are delivered from HDV tape. Also, if you are not printing to HDV tape then there is not conforming. Just render SCQT and your done.

Casey I'm not sure I understand what your saying. If your converting to ProRes while capturing from HDV tape, remember that the file sizes are almost six times larger and that should be edited on a ProRes timeline. The biggest bottleneck to working with ProRes is I/O bandwidth. If your editing multiple video tracks from a single FW400 drive you may experience what appears to be slower performance.

What you said about the time it took preparing to print to tape sounds like you were attempting to print an HDV timeline to tape which is has to conform. I believe the need to conform has to do with the tape deck. There is no conforming to export ProRes but you can't export ProRes to HDV tape.

>Not true IMHO. HDV editing is native to FCP. It's very fast as long as you don't have to conform it for a different output.

Hey Joey.

Admittedly I don't often have HDV to work with but I do work in XDCAM HD's 35Mbs variation of the format. Yes editing is native but if you need to render anything then you are render both from and to a complex Long GOP format. In my experience rendering from XDCAM HD to Apple ProRes 422 is significantly faster than from XDCAM HD to XDCAM HD. When I was first proposing upgrading the FCP version on our ENG laptops I ran a few tests on one of our stock laptops (first gen 17" MBP's, so 2.16 GHz Core Duo w/ 2GB RAM). The Capture Scratch and Render directories were on a FW800 connected 1TB GRAID, the house format is 1080i50 so the tests are centered around that codec:

10? cross dissolve using native 1080i50 HQ
Render To ProRes: 51?
Render To Native: 1?43?

10? of 1080i60 HQ footage in a 1080i50 HQ sequence
Render To ProRes: 26?
Render To Native: 1?26?

10? of DV PAL footage in a 1080i50 HQ sequence
Render To ProRes: 18?
Render To Native: 1?03?

10? of DV NTSC footage in a 1080i50 HQ sequence
Render To ProRes: 19?
Render To Native: 1?16?

10? of HDV 60i footage in a 1080i50 HQ sequence
Render To ProRes: 26?
Render To Native: 1?22?

10? of HDV 50i footage in a 1080i50 HQ sequence
Render To ProRes: 22?
Render To Native: 1?14?

As a reference, I ran the exact same tests on a (then) brand new MBP with 4 GB RAM ... and got the same basic results throughout excepting that all timings were approximately halved.

Pretty much ever since then, I've always recommended that when editing such sources natively its definitely worthwhile to set your render codec to ProRes. If you don't render anything and work entirely within RT, and the auto render never kicks in, then having set renders to ProRes won't affect anything (you gain nothing and you lose nothing). If you do render ... its faster.

Caveat: My recommendation to use ProRes is always assuming that whoever is doing it has a reasonably sensible setup ie one that wouldn't choke on ProRes.

Anyhoo, thats my reasoning.
Andy

Lots of stuff here. Okay. First, HDV is NOT a "master", just as DV is NOT a "master". Neither are DVDs, VCDs, VHS, H.264 compressed for web, etc...

If you are going back out to HDV (which is rarely the case), it's better work native as the footage will still have to conformed back to long GOP to go back to tape. You may want to render to ProRes while you're editing, to allow you faster render times to preview the cuts/effects. For straight cuts only edits, you may want to stay in HDV.

A long while ago, the word was to get rid of the GOP as soon as you can. But today, there are benefits to keeping the GOP for a while longer, eg. insufficient storage for ProRes. I haven't heard of much issues with having the footage as HDV and rendering/MM it out to ProRes, but I hardly work with HDV. However, this seems becoming of an eventless workflow that ensures storage efficiency and better generational quality.

As Jeff pointed out, just a simple cut, will degrade the footage, if you choose to render back out to HDV.


>"that's how everyone else is doing it."

They may mean everyone in their facility. It's a warning. Run out screaming! No seriously. If you do a proper down conversion from a ProRes MASTER you get miles better quality than pumping out a HDV "master". Quality in, quality out.



www.strypesinpost.com

Wow, this is pretty complicated...even for people who are using this stuff every day...what's Apple's stance on this? It seems that everyone has a different opinion of how this thing works...I would think that a concept like ProRes would be more like believing in Spain's existence rather than Santa Claus'.

Capturing is the only thing that didn't take longer. I captured using the ProRes preset, edited on a ProRes sequence with the ProRes render codec. ProRes seemed okay as long as I didn't apply any filters or speed changes. Once I did that, the render times jumped.

Andy, are you saying capture HDV, edit on a HDV sequence with the ProRes render codec?

Chuck, the project in question was HDV tape captured as ProRes, edited on a ProRes sequence with the ProRes render codec. I did indeed do a Print to Video from the ProRes sequence, which converted (conformed, rendered, whatever) the video to HDV before it actually recorded on the deck. And maybe what I was experiencing with speed is just a bandwidth issue, since the files were 6 times larger and the computer had to deal with that. I do have 2 external FW800 drives, so that should be better than a single FW400.

I'll try not to get any brain matter on the keyboard when my head explodes from all this!