1080 24p native vs transcoding for editing?

P M

New member
Hello,

I'll admit I'm pretty new to the DVX-scene, but am pretty tech-savvy, so I've taken advantage of the pretty recent developments ptool has offered for the GH1.

I haven't really been able to find a solid answer on the forums and can't find anyone that I know personally that may be able to help me.

I've been running C settings on 32gb Sansdisk Class 10 with maybe 5% chance of writing errors.

My Question is this:

Is using the Native 1080 24p modification really worth it?



I'm really new to editing, but I'd like to be using the camera/workflow as efficiently as possible. I'm running an overclocked q6600@3.2ghz with 8gb ram in win7, and am using a combo of cs5 (buggy) and cs4 creative suites. I've tried http://ppbm4.com benchmarking and my results sit in the Q1 region..

Are the 1080 24p native .mts files efficient for editing in premiere pro and after effects? Or would I be better off not using 24p native files and instead transcoding the 1080 60i wrapped media to Cineform Neo Scene (which from what I read is more editing friendly for layered effects/renders)?

I thought the advantage of 1080 24p native files was you can straight-up just go to editing, but if the .mts files are not 'editing friendly' what is the real benefit of 1080 24p native files?

So far I'm seeing roughly a 5% chance of error writing with the SansDisk Class 10 due to rogue frames with 24p checked. I was also wondering if I uncheck native 24p files, if i could potentially bump the bitrate up to D settings, without a hiccup with the rogue frames that cause write errors in native 24p settings?

So please, can anyone clarify this for me? Time-wise, assuming I needed to use AfterEffects/Premiere Pro: 1080 24p Native files vs 1080 60i Wrapped files Transcoded to Neoscene?
 
Mate i don't know about Prem or AE but high bitrate GH1 native or wrapped footage struggles to be edited in Sony Vegas, so transcoding with Neoscene is a must for me. With 50i footage i let Neoscene convert but not allow it to deinterlace as it has a known bug when processing high motion footage. Vegas does a better job of delinterlacing and loves the Neoscene files.

For now i'm shooting 50i in 1080, 25p is too buggy for me.

Don't forget the Hack is still under development, it's not a finished product, just an option.

R
 
I edit native AVCHD /GH13 in CS5 all the time. Works very well. Recently I was having the lockups with static footage so now i use B settings with 24pN unchecked.. Then I import all the clips into TMPGEnc Xpress 4.0 and do the pulldown, export back to HDD as Mpeg-4 with 2 pass 22940 kb/s average and 26750 kb/s max VBR with audio at 160 kb/s.

It makes files roughly the same size... maybe a bit smaller and they work just like standard AVCHD 1080p/24 in CS5. The encoder makes use of my GTX260 (which is also added to work with CS5 with the hack) so it doesn't take too long to transcode to the same file type.
Plays like butter and looks fantastic!
 
Thanks for the feedback rambooc1 and ATL Media Group.

ATL, are you using a higher end SDHC card? I've also had issues with static video using the 32gb Sans Disk Class 10 in C settings and have discovered that formatting the card after every file-drop to comp/hdd has nulled out writing errors. Though I can't say it's 100% error-free.

I'll have to take a look at TMPGEnc Xpress. Are the qualities in pulldown comparable to each other? Let's say TMPGEnc Xpress vs Neo Scene vs After Effects?
 
Thanks for the feedback rambooc1 and ATL Media Group.

ATL, are you using a higher end SDHC card? I've also had issues with static video using the 32gb Sans Disk Class 10 in C settings and have discovered that formatting the card after every file-drop to comp/hdd has nulled out writing errors. Though I can't say it's 100% error-free.

I'll have to take a look at TMPGEnc Xpress. Are the qualities in pulldown comparable to each other? Let's say TMPGEnc Xpress vs Neo Scene vs After Effects?

I use 16GB Patriot LX Class 10 cards. The only way around the static error with 24pN is to use the SanDisk but it's still not 100% reliable, so recording 24p over 60i at higher bitrate does the trick.

Yes, the quality is the same for the actual pulldown. What I like about Xpress is you can choose the output format. Even if you just try the demo of NeoScene, the output codec appears in Xpress as an option so essentially you get Neo. I would only use Neo if I couldn't edit the mpeg-4 files natively as Neo .avi's are huge and I usually have several hours in each project and several projects on my HDDs at a time. With Neo it would fill my 2TB drives!
Plus, I can dump the pulled down 24p footage to a backup DVD/Blu-Ray easier if it's smaller mpeg-4 h.264 files.

Xpress does 2 pass VBR so essentially you can use an even lower bitrate with the same quality as the original files from the hacked GH13.
You can also do "Batch" encoding of an entire SDHC card. So, instead of copying the files to your HDD you can batch encode them at the same time to remove pulldown and copy to your HDD.
 
Hello again,

So it's looking like the person I'm working with prefers to edit in Vegas Studio Pro. What would be the ideal intermediate format to use with this assuming I just got TMPGEnc Xpress 4.0?

Thanks again for everyone's advice!
 
The neoscene is a must unless you have oodles of free time. You will save hours. But use with CS3.
Alternatively, get CS5 and an Nvidia card, like a GT 240 or better, and hack the ini file to use cuda in CS5. But I'd recommend getting the Neoscene anyway. Caveat: You really need CS3 or CS5, CS4 doesn't cut it with AVCHD or neoscene, IMHO, JM2C. There are some free utils for pulldown, like HDV split. Some ppl report sucess with After Effects, but I just dump the footage into CS5 now that I have cuda enabled.
 
The neoscene is a must unless you have oodles of free time. You will save hours. But use with CS3.
Alternatively, get CS5 and an Nvidia card, like a GT 240 or better, and hack the ini file to use cuda in CS5. But I'd recommend getting the Neoscene anyway. Caveat: You really need CS3 or CS5, CS4 doesn't cut it with AVCHD or neoscene, IMHO, JM2C. There are some free utils for pulldown, like HDV split. Some ppl report sucess with After Effects, but I just dump the footage into CS5 now that I have cuda enabled.


Just wondering... why is it that neoscene or avchd is not ideal for cs4 but works fine with cs3 and cs5?
 
The neoscene is a must unless you have oodles of free time. You will save hours. But use with CS3.
Alternatively, get CS5 and an Nvidia card, like a GT 240 or better, and hack the ini file to use cuda in CS5. But I'd recommend getting the Neoscene anyway. Caveat: You really need CS3 or CS5, CS4 doesn't cut it with AVCHD or neoscene, IMHO, JM2C. There are some free utils for pulldown, like HDV split. Some ppl report sucess with After Effects, but I just dump the footage into CS5 now that I have cuda enabled.

I went through some portion of the 'CUDA' checks, and it seems my Mac Book Pro graphics card does not have the requisite 768M of graphics memory to 'use CUDA'... Is there an upgrade nVidia graphics module with the memory on, or is there a 'some assembly required' solution to add memory to the existing graphics card.

The graphics module is: GeForce 8600M GT with 256M.
 
Unless you got a PHD in IC design, you're not going to be able to modify the memory on your gfx card.
 
i'm surprised more people don't know about this.....i use the FREE DNxHD codec from Avid:
http://www.avid.com/US/solutions/workflow/DNxHD-Codec

this allows lots of quality options (bitrates) and wraps the video in quicktime for easy offline previewing, easy PC-to-Mac transfer and no gamma shift when rendering out to h.264 or mp4 for online distro. the codec is available inside TMPGEnc when exporting to quicktime. just select the correct video size and the bitrate option. the frame rate doesn't matter, as that will be determined by your settings in TMPGEnc.
 
Back
Top