CineForm workflow questions
 

David,

Here's an idea for a potentially interesting workflow that provides a variety of distribution options.

Shoot in 25 fps cineframe mode ( either with an FX1e or a Z1)

Post at 24 fps ( the 4% slow down not being a problem) for tape to film - alterntively do 3:2 pulldown for broadcast in 60i countries

Run a version at 25 fps for broadcast in 50i counties.

Any general comments or issues with the Aspect/Prospect codecs?

Thanks,
David

We have though about an automatic 25p to 24p conversion during capture, the only issue is audio sync with pitch correction. So it is not technically a video issue. It may happen is a future release as this is just the sort of feature we plan to add to Aspect HD, keeping it fresh and doing things that native editing can't easily do.

Is it not possible just to set up a PPro project as 23.996 fps and import the 25 fps footage, accepting the 4% audo slowdown ? It's hardly notceable for most things.

Or alternatively, post the project at 25i and do the 24 fps and 3:2 pulldown after....

Doing the 25 to 24 later (for now) works better due to the way Premiere handles footage re-interperation. A 25P clip can be interperated as 23.976, but you have to manually do this one at a time -- clips don't automatically slow down when the project setting is a slower frame rate. Also doing it this way you also don't get the audio pitch correction. If we handle this during capture, these issues go away.

Sounds like a great idea. a couple questions:

if you convert 25p to 24, does that still involve slowing down the footage, or would Aspect compensate somehow?

the obvious next question is: if i record separately to a DAT (or anything other than the camera), would i still need to slow down the sound?

If you could convert to 24p without making me futz with the sound, you'd be doing us all a great favor.

thanks

Josef,

If you record to an external medium you can use just about any audio program to change speed without changing pitch. Samplitude offers a number of different algorithms for doing it. I'm sure there's a way in the Adobe sound app, too.

This also offers a way to deal with pitch sensitive recorded material.

David

Z1 PPro 1.5 workflow?
 

Hi,
My employers are considering purchasing the Z1U as a step toward HD. As an experiment, I suggested we rent one camera to use for shooting a presentation music video we have to do that has lots of greenscreen.

The idea is to improve acquisition for our shoots.

Anyway, we have a dual 3Ghz Xeon system with Premiere Pro, that I plan to use to do the edit. It has 1gb ram, but only one SCSI drive. Two monitors running 1280x1024, but no HD display.

Can I DL the Premiere Pro HDV plugin and capture to this system, or is an AspectHD type-tool required?

Either way, can I then edit the raw HDV files, or should I render out proxy footge, or perhaps capture DV and edit, then batch capture the final edit as HDV?

My final step, in the plan, is to render out full-size frame sequences to do the comp in AE, then downconvert to SD for distribution.

Should this work? Am I missing something?
Any suggestions?

Thanks.

Hi Carl,

that will work. I do so all the time since I got my Z1.
I capture the HDV Material with the Prem 1.5.1 update.
Then I recapture the same Clip as SD (sometimes there's another in and out point after capturing, make sure to check that after capturing the HDV) and use this SD for editing. Then when finished editing I break the connection to the sd files and reconnect them with HD Files. Very easy.

For Aftereffects work I export the HDV as a TARGA sequence. Seems to be the best format for all Programs, even when leaving the windows world.

The only thing I recomment in addition:

I did't find a good way to downscale the Material to SD with Software. The Z1 itself does a superb Downconversion in real time. So I output the whole, finished Project as HDV to the Camera (even for future use) and the get it back in with the Z1 downconversion.

All in all: It works and the results (especially for Keying) are simply astounding.

Greetings

Peter

<<<-- Originally posted by Peter Rixner : Hi Carl,

that will work. I do so all the time since I got my Z1.
I capture the HDV Material with the Prem 1.5.1 update.
Then I recapture the same Clip as SD (sometimes there's another in and out point after capturing, make sure to check that after capturing the HDV) and use this SD for editing. Then when finished editing I break the connection to the sd files and reconnect them with HD Files. Very easy.

For Aftereffects work I export the HDV as a TARGA sequence. Seems to be the best format for all Programs, even when leaving the windows world.

The only thing I recomment in addition:

I did't find a good way to downscale the Material to SD with Software. The Z1 itself does a superb Downconversion in real time. So I output the whole, finished Project as HDV to the Camera (even for future use) and the get it back in with the Z1 downconversion.

All in all: It works and the results (especially for Keying) are simply astounding.

Greetings

Peter -->>>

Hi, Peter.
About the recapture. You said you relink to the HDV?
Does this mean you capture twice with the same clip names?
Is it possible to capture DV, edit, and then only recapture the clips used in the Timeline at HDV resolution?

Finally, I read in the readme notes that Premiere will reset the inpoint to the nearest i-frame to account for the GOP, is this what causes the in-point change?

Thanks for sharing your expertise.

-Carl

The batch capturing of the only-needed clips would be great, but seems not to work.

But as the datarate of HDV is as small as SD I keep both files on disk and simply change the link in Premiere.

But great to read from you, that there is a explanation for that timecodeshift I experienced.

As for this timecode change to the i-frames or whatever, it is not possible with just capturing the used clips after editing SD.

Let's see with Premiere 2.0 :)

Peter

Workflow with CFHD/Uncompressed
 

Is this Workflow makes sense??

1. Capture HDV or the new ProHD format to Cineform Avi (CFHD Codec) in 1280x720/24p. Edit the entire movie real time in Premiere Pro (Aspect HD) in CFHD (Cineform High Definition codec) and when editing finished, export as uncompressed 720/24p HD with Blackmagic codec.

2. Do Color Correction in Premire Pro 1.5.1(Color Finesse) in Uncompressed 720/24p Blackmagic HD and export as again uncompressed 720/24p HD with Blackmagic codec.

3. Import this file to After effects do the final touches of the editing here and again export it as uncompressed 720/24p HD with Blackmagic codec.

4. Import this file into Abode Audition and do all audio editing with or without surround. Export completed movie now with complete audio track as uncompressed 720/24p HD with Blackmagic codec.

5. Import this file into Encore DVD and export it as various formats like: mpeg2-ts 1920x1080 to D-VHS, SD-Mpeg2 for DVD and HD-DVD with WM9 codec.


Does this make sense or not????? Any help is greatly appreciated.

How? To my knowledge, Cineform's HDLink does not have capability to re-scale the original 1440x1080 signal (as comes from FX1)?

Alex, no rescale is needed as the ProHD format (JVC) is 1280x720 at a true 24p.

Gabor, we don't see the need for working in uncompressed at any stage during the workflow. If you after additional quality, using Prospect HD will give a 10bit workflow with 10bit color correction using Color Finesse within PPro (CF directly works with PHD's deep color format.) Without PHD Color Finesse within PPro is 8bit, even on uncompressed data. Plus Step 2 would be best done on the orginal CFHD timeline as at least your edits are still intact, allowing you to easily color correct on edit boundaries.

David as soon as Prospect falls from the minimum package cost of $2500 USD, we can talk ;>)
Untill then Gabors work flow is OK if he changes his CC to being done in After Effects Pro, so he can take advantage of 10bit CC.

Daivd- "Without PHD Color Finesse within PPro is 8bit, even on uncompressed data. "

How about in After Effects Pro?

Yes, if he color corrects in After Effect he gets 10bit support, however I would still use CFHD files as the source materials (not uncompressed) as there is no quality impact to import the Premiere project with CineForm AVIs. To get 10bit output using only Aspect HD with require the export to be uncompressed from AE. Staying with CFHD for step 2 will help the workflow.

I'm not following you David. Are you saying import CFHD files into AE and CC in 8-bit, then export to uncompressed 10-bit? Or are you saying somehow CFHD files can be CC in 10-bit as long as they are exported uncompressed?

I was just thinking doing color correction in uncompressed becasue it takes the heavy color correcting better than the CFHD wont it ????
Prospect HD is great David but $2500 is too much for me at this time.Especially that I own Adobe Video Collection so no point for me to purchase it one more time !!!
Gabor

Yes to the last sentence. Put AE is 16bit mode, import CFHD files, CC in deep color (even though the orginal data was 8 bit) and output to uncompressed 10bit.

David.

No doesn't work that way. When you take 8bit compressed data and export as 10bit uncompressed you still have 8 bit data only it is 10 times bigger -- that is a pain without value. Take the orginal CFHD file into AE (in 16-bit mode) and you have extractly to same quality as if you converted it to uncompressed. The CC is performed in deep color, then you can export the results to 10bit uncompressed. Same output quality without the conversion step (until you need it.)

So If I CC in After Effects (16bit) and finish my video editing there, what is the point of exporting as uncompressed 10 bit???

Why cant I just export it the way it is in CFHD and import it to Audition for the audio editing and when that is also done I just

Import this CFHD audio and video file into Encore and make a HD-WM9, SD-DVD and D-VHS version!!!

You can do this very thing. A good trick though is to add 0.4% noise to the final render. The deep color will be clipped back to 8bit when you export, the 0.4% noise with dither the 16bit signal so banding doesn't jump out in the 8bit final.

Interesting thread, the aspect hd file is 8bit, but you can deep cc in after effects at 10 bit ( are you saying just because the file says 8 bit it is actually 10?) and then export to 8 bit? What is the advantage? Why would I want Prospect HD if I have AE pro?


Thanks
Mike

Mike,

The techniques I described above do extend the usefullness of 8bit though the additon of noise, this trick has been known for a long time but it is hardly a replacement for 10bit work-flows. Only Prospect HD offers a true 10bit work-flow allowing deep color correction within Premiere rather than the more awkward environment of AE. Banding due to 8bit also commonly happens in dissolves and fades, and it is a pain to do them all in 16bit AE adding noise to cover 8bit limitations (note adding noise in standard 8bit Premiere will not work the same.) The Aspect HD compression only has 8bit I/O (per channel), such RGB 24 and YUV 16, whereas Prospect HD compression supports 16bit per channel I/O for very deep color. In Prospect HD processing is performed is 16bit per channel and sent to the codec that way, no 8bit truncation occurs.

Thanks David, will a prospect hd trial be available soon/at all to compare the 2 (aspect/prospect)


Mike

HDV workflow
 

Okay, I am finally starting to get into this now, and of course seem to have issues. So for you folks who are made things work, got some questions for you. So, I did a wedding yesterday and have now inputted the HDV data into PPro 1.51. I took a 2 minute piece and am trying output conversions to see what happens. So, I first tried wmv9 1080i anamarophic. It first took like 45 minutes to encode for this 2 minute clip. Ouch. But, when I went to watch this on my computer monitor, it really broke up on anything that was moving. So, I then did a microsoft dv avi output of the same click and it works just fine.

So, I am just starting this output and really have no idea what I must be messing up. Any help to get me point in the right direction would be great.
I am looking at giving them two outputs. One would be SD DVD. The other I thought would be MS HD for use in an I/O Data DVD player.

thanks in advance.

Dave

Are you sure that's not interlace artifacts? Did you deinterlace the footage for output to the WMV file?

Mark, it probably is interlacing.

Is there an option for this? All these different outputs.
The 720p seemed to work on the monitor. Not sure if I have a decent enough
video card either. So many darn things to learn.

Dave

Did you use upperfield first for output? That can cause the problems you describe.

I can not seem to find an upper field option in PPro output.

I did see in Procoder but it will not let me select it.

Man, this who thing with are you reviewing on a LCD panel, vs CRT.
So, are LCD panels always progressive?

Dave

To Flash - how?
 

What is the best workflow to go from hdv to Flash movie and keep as much quality as possible? Using PPro and Aspect HD.

Poorman's HDV workflow
 

Hi,i m interested about using freeware on HDV with a slow pc.

I use CapDVhs on XP to capture whole hr of HDV into m2t file.
read the CapDvhs carefully if u intend to use Virtualdub mpeg2.
then use virtualdubMpeg2 open it,it takes 8minutes on my celeron 850.
apply smart deinterlace and timestamp plug-in. which i think always nice to work with time code, it's at the bottom but only 10-15pixels high.
i stay at 1440 x 1080 but it converts to Progressive.
testing cineform and this takes about 2 hrs to convert 1 hr of footage.
for 2x of size u can have a workable format to edit on, the audio also convert to wav format too.

i also try others such as targa but seems crash always. may be the image sequences are too big.

HUFFY codec doesn't play although it can encode, so far only cineform works.
or go for uncompress or image sequences.
it's slow but the goodies are hard disk is cheaper and cheaper, and i can avoid many frame loss in batch capture, so far no good confirm that is oktouse batch capture in HDV,as i don;t know how to avoid chopping off key frames,may be a sec. of handle on both end will help.

My question is ;
is there any way to convert to cineform on the fly, so i can eliminate the re encode.
i know Vasst has a gearshift, but they use sony YUV codec, which i don;t have, so is this the way to do similar job with virtualdubMpeg2 and cineform?

thanks
JY

it's 2days not 2hrs
 

my apology,
it's rendering at 2days,i mistook it as 2hrs.
i run another test ,10sec on cineform is about 6minutes for deinterlace, and 3 min for just plain encoding.

cheers
JY

clarification?
 

John, what are you trying to accomiplish with this workflow?
Is it just designed to work on older computers?

Thanks for the post,
Daniel

Workflow question
 

Although I'm still yet to purchase a copy of Cineform Connect as I'll hopefully be purchasing a new computer and don't want it registered solely on one of my current machines, I've got a bit of a question regarding deinterlacing and conversion.

Now, in the John Jackman dv.com article linked from Cineform main page, he mentions that he's found the best method of converting PAL 50i footage to 24P, is to firstly capture the footage to the Cineform codec at 24P and then batch deinterlace all the shots using DV Film Maker. Now I've tried DV Film Maker, and it does quite a decent job. Although what I was wondering, is if it would be simpler just to capture the footage in the native HDV format, then batch deinterlace using DV Film Maker, but save the files in the Cineform codec from there whilst converting them to 24P in the process? If that's even possible? Or is the HD Link app the only way to do the conversion? The demo version of Connect HD I had has already expired, so I can't test this myself.

Glenn

Glenn,

I think you idea is sound. You can use HDLink to capture only the M2T files then import those into DV Film Maker, from there directly encode to 24p CineForm AVIs. That will work great.

Hi David,

Sounds great. So the conversion from 4.2.0 to 4.2.2 will still be carried out aswell during the process?

I've got a couple of hardware questions aswell, but they're probably more appropriate for the Dual core thread.

Glenn

Ah, the catch. We do the upconversion filtering only within HDLink and the Premiere capture interface. Within external tools like DV Film Maker, we have to assume that chroma filtering has been performed as it outputs 4:4:4 RGB data (before is it compressed as 4:2:2.) It don't know whether DV Film Maker 4:2:0 to 4:2:2 chroma filtering is as good CineForm's, that will be up to you to judge.

Ah I see. That's probably why John Jackman captures to Cineform first, and then deinterlaces using DV Film Maker. Although I assume not much would be lost anyway judging by the clip on your website's quality tests page?

By the way, how does the 4:2:0 to 4:2:2 chroma filtering work in the Cineform codec? Is it similar to blurring the chroma channel in Vegas using the chroma blur plugin? There's no need to answer this if it's a guarded secret :-)

Glenn

It is a little more complex that that. :)