-
HDV to SD DVD, QT vs direct to Compressor
I captured and edited in hdv using FCP 6. Will there be any quality difference if I use a self contained QT from fcp timeline and import to compressor vs. exporting direct from fcp TL to compressor.
I exported direct to compressor but saw that estimated compression time was 35 hours for this 90 minute video using apple’s 90 min. preset and dolby audio.
Although the time estimate i know is not exact, it must be at least in the ball park, and I don’t want to send 35 approx. hours on the compression.
Now I’m trying a QT self-contained instead. So far, compressor is saying the QT version will take 17 hours.
So, looks like it’s less compression time using the QT, but is there any quality loss vs. direct to compressor from fcp?
-
17 Replies
-
The problem of sending to Compressor from the FCs time line is that when you make a Two Passes compression, everything that have to be rendered is rendered twice during the process. Even if you have everything already rendered in FC.
I will suggest you a different path:
– Open a new sequence SD 10b Unc Anamorphic (PAL or NTSC depending of your movie).
– Drop your Self Contained HDV movie on the sequence.
– Set “Render all YUV material in high precission” (Sequence Setting).
– Set Motion Rendering to BEST.
– Export as a Self Contained and bring it to Compressor.
The downresizing will be done by FC. But don’t worry, if you set the rendering as I told you, FC will do a perfect job und much faster that Compressor.
Compressor will only need to transcode from .mov to MPG2. -
I’ll try that, but what’s better, should I bring the hdv QT and drop that on the new sequence and render, or just copy my hdv TL footage and paste into the new 10b unc seq. and render?
-
In the seq. settings, should I use “Render 10 bit in high precision . . .” or Render all YUV in high precision…”?
My hdv footage is 8 bit.
-
Hi Chris,
As long as you have your movie already exported as Self contained, the easiest thing is that you use that movie.
If your HDV sequence have many effects or things to render, could be worth to drop the HDV sequence (no the movie exported) in the 10 sequence. resizing and effects would be rendered all at once instead of in two steps.
This is you who have to decide.
You have to set RENDER ALL YUV…
You can not set “Render 10 bit in high precision..” because you have no 10b footage.
rafael -
OK, but I want to clarify a couple confusing things. Even though I’m starting with 8 bit footage, you say I drop it into a 10 bit sd unc sequence. Why is that?
I see that it also has to change all the footage to lower field from the upper field as it goes from hdv to sd. I suppose that’s necessary, but maybe that’s what adds to render time.
The 10b unc in fcp is taking 15 hours to render this 90 minute video. (even though all the footage was fully rendered in the hdv seq).
At this point I’m not at the final step of going to compressor, so I can’t see how much time compressor will take on this new QT movie from the 10 bit unc.
You think it saves enough time (while maintaining quality) to make it worth the additional long render in fcp?
-
I think I screwed up when doing the “General” seq. settings (maybe i can blame in on sleep deprivation).
In the gen. tab, top few lines, what do I choose for frame size, pixel aspect ratio?
I was unsure about the above, but for other stuff under this tab I chose Anamorphic though I questioned if it was necessary, field dominance at lower/even, QT/compressor set to Unc, 10 bit.
(My edited footage was hdv 1080i)
-
[chris gorman] “OK, but I want to clarify a couple confusing things. Even though I’m starting with 8 bit footage, you say I drop it into a 10 bit sd unc sequence. Why is that? “
When you work in 10b and force High Precision rendering, the render is done in 32b Floating point. Better quality downscaling.[chris gorman] “I see that it also has to change all the footage to lower field from the upper field as it goes from hdv to sd. I suppose that’s necessary, but maybe that’s what adds to render time. “
Yes de picture is shifted one field up or down to match the new field order. This won’t really increse much the render time because you are doing that at the same time than the downscaling.[chris gorman] “The 10b unc in fcp is taking 15 hours to render this 90 minute video”
Chris, 90 minutes is a VERY LONG movie. I saw those rendering times for few seconds long clips when applying certain effects.
[chris gorman] “You think it saves enough time (while maintaining quality) to make it worth the additional long render in fcp?”
When you will send your Self Contained 10b movie, Compressor needs only transcode from .mov to MPG2. With a fast Mac this can be done in one hour or so. With my MBP the 90 minute can take the same time for a Double Pass compression.– “I think I screwed up when doing the “General” seq. settings (maybe i can blame in on sleep deprivation).
In the gen. tab, top few lines, what do I choose for frame size, pixel aspect ratio?
I was unsure about the above, but for other stuff under this tab I chose Anamorphic though I questioned if it was necessary, field dominance at lower/even, QT/compressor set to Unc, 10 bit.
(My edited footage was hdv 1080i)To make your 10b Unc sequence, go to Sequence Presets and choose10b Unc NTSC. Open a new sequence and this one will be 10b Unc NTSC. Just open the Sequence Setting and check Anamorphic. You don’t have o worry about size, pixels or field order.
When you drop the HDV (Upper) on top of the 10b Unc Sequence (Lower), FC will add the necessary Shift-fields filter.
When you will bring your SD movie to Compressor, you will only need to apply the 90 minutes preset.
Let everything in auto. Compressor will do the rest.
Rafael -
I am having this same issue for a client who brought in a final HDV FCP project for me to tweak in FCP and Compressor for their DVD master. Using the HDV self-contained QT directly in compressor yields absolutely horrendous results, even at highest quality. I actually asked in the Kona forum if anyone has had success rendering to DVCPROHD, Prores or uncompressed HD/SD with acceptable results and haven’t had a response yet. If I output and downconvert the HDV project via AJA LH component directly, it looks absolutely awesome, but of course the stand alone recorder doesn’t allow for menu and chapter customization like DVDSTPRO does. So unless someone else has done this successfully with high quality compression results versus the mangled mess Compressor does with HDV footage, I may just have to get back with you with the best method after several days of testing.
Dual 2.5 G5, IO, Kona LH, IO, Medea Raid, UL4D, NVidia 6800, 4Gig RAM
Octocore 8 GB Ram, Radeon card, MBP, MXO
Windows XP Adobe Studio CS3, Vegas 8.0, Lightwave 9.2, Sound Forge 9, Acid Pro 6, Continuum 5, Boris Red 4, Combustion 2008, Sapphire Effects -
Hi Christopher,
Compressor can make a good but you need to set Frame Control ON. In this case will work in 32b Floating point. But this means time.
Since few months I’m working with an EX-1 (basically the same than HDV).
I end up in SD or web videos. I’m doing everything (deinterlacing, downscaling, pixels aspect,.) in FC. If you set it properly, FC will makes a perfect job and in a fraction of time compared with Compressor.
Only the final transcoding I make it in Compressor.
rafael -
Rafael,
I guess my main question is which codec works best for the highest res encoding of HDV to standard DVD. I read all the past postings on the subject, and at times you suggest going the Uncompressed SD route, at times the ProRes route, and I even saw (other peoples’)posts suggesting using the DVCPRO50 SD route. I have seen both referenced and self-contained QT methods mentioned, and encoding from within FCP (using Quicktime conversion), or going to Compressor with your self-contained or reference QT movie. Have you seen any big differences going with one method or the other?? I am also wondering if anyone has tried rendering/transcoding the HDV project to either uncompressed HD or DVCPROHD and had a successful transcode from those formats in Compressor and therefore a better looking final SD DVD? That is my real question. I never work in HDV myself, and am just finishing up someone’s two year old project that has gone through several editors and organizational nightmares that I inherited and am finally down to the mastering step! I am really spoiled by having a component DVD recorder that takes the Kona LH signal and downconverts it in real time to a stunning letterboxed DVD. The problem is of course the client wants a DVD with motion menus etc, so I am stuck with this DVDSTPRO encoding conundrum.
Dual 2.5 G5, IO, Kona LH, IO, Medea Raid, UL4D, NVidia 6800, 4Gig RAM
Octocore 8 GB Ram, Radeon card, MBP, MXO
Windows XP Adobe Studio CS3, Vegas 8.0, Lightwave 9.2, Sound Forge 9, Acid Pro 6, Continuum 5, Boris Red 4, Combustion 2008, Sapphire Effects
Reply to this Discussion! Login or Sign Up