This content has been marked as final. Show 40 replies
lossless and hdv dont go together.
Why do you want to rerender HDV to HDV?
Because I want to keep the footage I export at the same quality as it was captured. Premiere Pro 2.0 had a "HDV 1080i 30" export setting in the media encoder. 25mb MPEG2 saved as .mt2. Now CS4 doesnt have that feature. Looking to find what is comparable in CS4 from what I used in 2.0.
If you want to export "export at the same quality as it was captured" render to a lossless or near lossless codect. What your doing is generation loss. Like recording a VHS to to another AHS recorder your gonna lose a good bit of quality.
But that's not what you want to hear. If your running the demo, it has no MPEG support so no HDV presets. If you bought it then you can easily make a custom preset.
I had the same issue initially. I used Liquid previously that allowed me to fuse a HDV sequence back to a m2t.
I don't understand why no one seems to understand why someone might want to do an intermediate render of a complex sequence (e.g., multiple tracks, multiple effects, etc.) Suppose you build, in HDV a complex animated title sequence. Why not render it once and not have to wait for it to re-render each time you use it in a new project. This is not all that different from Premiere's capability to use a sequence as if it were a single clip. The big difference is (1) not having to wait for multiple effects to re-render when you apply another effect on top--you only have to render the one new effect. and (2) by saving it as a HDV file, you can import it into another project whereas sequences can't be imported into a new project. My previous NLE had a FUSE tool that would fuse a sequence or section of a sequence to a mt2.
Based on other posts, I have come to learn that the Blu Ray preset in Adobe Media encoder creates a file that is compatible with a native HDV file. The only difficultly I've experienced is that I get an error message when I try to multiplex the audio with the video--as such I get a video and a wav audio file from AME.
I question your claim that the .m2t export in 2.0 was adding any more loss to the original HDV footage. HDV itself is highly compressed 25mb mpeg2. The "HDV 1080i 30" export in 2.0 was the same rate 25mb and was quick. Similar to exporting DV from DV source. Footage looked great.
I have the real paid version. In the mean time, I've been messing with QuickTime h.264 exports after installing the QuickTime Pro h.264 codec.
You don't have to install H264 codec!? Not sure where you got that from. Link maybe or do you mean you just installed the latest version of quicktime?
Exporting DV to DV is a pretty lossy. That's a fact. It's a redundant rerender.
I'm really at a loss. The whole reason I even use Adobe is so I don't have to render to an intermediate codec. You edit your HDV or DV then render out to DVD or H264 for flash or blueray. Uncompressed if you must send it to a non Adobe program.
Did you at least try to create a custom export preset?
I have 2.0 on one machine and CS4 on another. In looking at the HDV .m2t export in 2.0, I see it's MPEG2, 1440x1080i, UFF, 29.97 df, 25mb CBR. I guess I can manually recreate this config on CS4 and save it as a preset.
When looking at the BlueRay MPEG2 export in CS4, I see it's similar to this, but variable bitrate with min 25mb, target 30mb, and max 35mb. Seems kinda sensless to target 30mb MPEG2 when the source is only 25mb MPEG2 (i.e. HDV stream from camera is 25mb.) Still, what you are telling me confirms some other stuff I read. I might just use the BlueRay preset and be done with it. Thanks for a useable answer Steve. Regards, Brian
Apples QuickTime h.264 codec is way better than the main concept one. You get their codec when you buy QuickTime Pro for $29.
Double long GOP compression lossless? Gimmi a break. For preview purposes I understand, for a final product it would be fine for home movies or youtube. For professional work like discovery channel, commercials or feature film it's joke. Sorry but that's a fact.
If your just doing home movies I apologize. If I'm wrong bet your *** that Harm, Craig, Jeff or one of them would set me straight right quick.
I've been selling quite a bit of HD ENG footage to a national HD cable network using export with QT Pro's h.264 at 15mb. Not exactly home movies or youtube. Stuff looks fantastic and at 15mb, it can be FTP'd quick. h.264 is roughly comparable to MPEG2 at 3 times the data rate.
That's makes way more sense. Go from your HD ENG (you mean HDV? or a pro i frame codect like DVCPROHD) to quicktime h.264. Not HDV to HDV to H264 qt. That's all I was saying.
> Similar to exporting DV from DV source.
Not really. HDV has aggressive temporal compression. You want to get your footage out of it, not back into it.
That's what I meant Curt. I was hoping that the DV to DV comment might help him understand why going from HDV to HDV is a bad idea. 'aggressive temporal compression' might be to much for him to understand so I used a simpler concept.
But DV to DV is not that big a deal, Josh. HDV to HDV is a much bigger deal. That's why Curt was citing the comment as inaccurate.
Josh knows that; he was saying the same thing in different terms.
I was citing one of Brian's comments earlier on without reading all the subsequent posts. I think its been beaten well enough now that i took time to read all the later posts.
Brian, See I told you I'd get jumped on if I was wrong. HDV to HDV is a plain bad idea. My comparison of dv to dv wasn't even close to the image loss you get from HDV to HDV.
Still I wouldn't recompress dv to dv anyway even it does give you better multi-generation than HDV to HDV.
I took the bullet for ya Brian if I had not opened my big mouth (er fingers) these guys would've explained in better terms why HDV to HDV is a really dumb way go about getting a final product that you send to 'a national HD cable network'
>I wouldn't recompress dv to dv anyway
That's what I'm saying. Doing so is not that big a deal. It's a perfectly fine method of going from Premiere to Encore and letting Encore do the Automatic transcoding, for example.
This is separate than the main issue of going from HDV to HDV, which I agree is not recommended.
Since everyone is already on this train, rather than start a new thread, perhaps I can get a simple, straightforward answer.
I don't want to do any recompression. I just want to get out what I put in. e.g., If I import a 218MB avi file and place it on a timeline, how do I get a 218MB avi file back out of Premiere. I don't want to change anything about it, I just want the same quality, compression ratio, file size, etc., that I started with.
What are the settings to use and where are they located???
>I don't want to do any recompression. I just want to get out what I put in
You can't do that with HDV. You can do it with DV as such becomes simply a file copy process. But with HDV, it'll be recompressed during export.
> I don't want to change anything about it, I just want the same quality, compression ratio, file size, etc., that I started with.
Use Explorer to copy the file. You don't need PR for that.
> I don't want to do any recompression.
Then you definitely dont want to export to HDV.
The closest thing you can do and meet your requirements of small file size and "Less" compression is to use the cineform or similar codec to create intermediate files that edit fast, have fairly small file size and are what some consider "observably lossless". Cineform is also significantly compressed but in a way less observable and higher quality that hdv recompression.
If we take your request literally; "No compression" then save as uncompressed. But the file size will be huge.
To summarize the dv <> HDV comparison discussions; Many folks believe that DV pojects may be saved as DV files even a few generations before a loss is observable. However, most folk believe HDV compression is observable after a single generation, ergo few if any recommend saving to an hdv format or exporting back to hdv tape.
Thank you Harm for your (once again) "insightful comments." The sum total of your contributions to this forum have been to consistently put down people asking simple questions or shout "OE."
Since you obviously cannot understand what I'm asking, I'll try again and if all you can contribute is another wise@$$ answer then just spare us all the depth of your wisdom.
I am not working in DV as DV is 720x480 SD not 1280x720p HD as I mentioned before. I am also not working in HDV which is another tape-based format. I am working in a tapeless format that writes HD avi files.
An *.avi file is a wrapper, the same as a *.mov file is a wrapper, and a wide variety of codecs can write to and access those wrappers:
What I am trying to understand is why there needs to be any change to the original imported footage when it is exported?
If I import a 215MB avi file that was created using any codec you care to name, place it on the timeline, cut three frames off the head and tail (so Harm can finally grasp the concept here) and then export that timeline what are the settings that will simply create a new file that is exactly the same as the original (minus, of course, the six frames)???
Why should the exported file be any different than the original in terms of compression, quality, etc other than a slightly smaller file size due to the six missing frames.
My understanding is that all of today's NLEs do not actually do anything to the original footage, therefore why should there be any change when the footage is "exported?"
I've have tried to do this using numerous different settings and there is always a major change in either filesize or quality.
I will not be exporting to tape, I will not be exporting to HDV or DV or H.264 or MPEG-2 or any other compressed format. I simply need to edit and then get back out of PPro with the same quality, compression, filesize, framesize and framerate as I started with.
How do I do this???
Sorry William. This thread is about exporting to HDV so i assumed thats also what you wanted to do.
To answer your question requires more info. AVI is a wrapper; you could have any codec going on inside it. Not all codecs are appropriate for editing or re-compressing. So simply guessing without knowing what codec your source material is in; is its not a codec adobe supports or you have installed for edting. There are many codecs that are great for display but not for editing or re-compression.
>If I import a 215MB avi file that was created using any codec you care to name
There's the rub. Whether or not you can do what you want to do will vary with the codec in the file.
>Why should the exported file be any different than the original in terms of compression, quality, etc other than a slightly smaller file size due to the six missing frames.
It's just the nature of compression and the way Premiere works.
>My understanding is that all of today's NLEs do not actually do anything to the original footage, therefore why should there be any change when the footage is "exported?"
Because some types of codecs are recompressed on output. Nothing you can do about it.
>How do I do this?
Again, back to codec used. What are you starting with? What codec in inside the AVI wrapper?
MJPEG & LEAD MCMP|MJPEG codec. Problem is that AME (updated to 4.0.1 or 4.0.789 or whatever it is) does not complete an export. Goes about a minute and then simply stops (this is an improvement over version 4.0 when it would only crash).
Reports that either there is a hardware or codec problem or sometimes that the disk is full. However, there is over 700GB free on the hard drive - RAID 10 - and I can successfully export from After Effects using the same file, codec, hardware, etc.
New computer, updated nVidia driver, 8gb RAM, quad processor, nothing else installed except CS4.
Trying to come up with a workaround that gives me close to the same filesize with the same quality since I need to run the files through two more apps to prepare them for stereo projection.
So what codec should I use in Premiere that will simply pass the video along without changing it.
Uncompressed avi UYVY makes humongous files and they don't even look that good. Sheer makes humongous mov files but they look pretty good. Almost everything else either looks bad or takes way too long to be viable. These are not projects that can take hours to re-render after the client has made changes.
What do you have to "pass it on " to? If AE, use DL; no intermediate at all. ie 0 file size.
There are two additional apps that take a left and right avi file and create a single side-by-side file that is then used by a player app to split the file into two distinct DVI streams for individual projectors. Because it is polarized projection, any compression differences in the files during the conversions lead to "swimming" and destruction of the stereo imaging when projected to 30-foot widths. It's hard to explain, but "you know it when you see it" to paraphrase. That is why I need the least modification of the original captured media during the entire editing and postproduction workflow.
Does this mean I have I have to put on my 3D glasses to understand this thread?
What other programs are using for the is file, why can't you just use the same file twice or duplicate it in the 3D program and just curious what kind of 3d project are making?
>what codec should I use in Premiere that will simply pass the video along without changing it.
The only way to do this is to export to the same format as your original, and even then it will depend on the original. In this case, I'd say you just can't do it. (Maybe if you can get AME to work with MJPEG...)
What is your source? Perhaps you can convert it to something like Lagarith instead of MJPEG before importing into Premiere. That should allow identical quality output.
To clarify, my orignal question is along the same intent that William Meredith is asking his question. I capture footage from my HDV tape camera, simply cut it into ordered clips and add title headers to them. No effects, rendering. I just want to archive these clips to hard drive, and do so in a format that is no worse than the HDV stream I captured from the camera.
When I worked with DV, if you captured DV, and exported to DV AVI files, you pretty much got what you started with from what I was told by the self-proclaimed experts. From what I read here, if you capture HDV, cut up your vid in Premiere Pro, then export to a .m2t file (i.e HDV export in PP 2.0) you actually lose significant quality. Is that the actual truth? I'm look to graciously hear from people who really know what they are talking about (sans Josh Townsend who is nothing more than a lonely stroker/tool.)
What file format is best to archive your HDV native 1080i footage after you cut up in Premiere Pro CS4? NOTE: Totally uncompressed QuickTime YUV isn't the acceptable answer here. One minute of that is 7GB, which is ridiculously huge and overkill. It seems there should be an optimal codec and settings to use to at least *minimize* generation loss and not eat up the hard drive space when you need to archive your edited native HDV clips back to files.
Thanks in advance,
There are two distinct and separate files that were shot with two matched cameras so there is a left and a right perspective with different image information, so you can't use the same file twice.
Every conversion accentuates the differences between the two original camera files and ultimately degrades the stereo perception. That is why I'm struggling to keep the workflow output as close to the originals as possible. If AME would simply work properly with the LEAD codec - it is after all just another plug in and works fine with everything except AME - the problems would be nonexistent, but I can't tell my clients they'll just have to wait until Adobe figures out what the problem with AME is!
What are the two original camera files, William? What camera? What file specs does it record?
I'd say use Lagarith Brian. It's visually lossles. Might be larger than the originals, but you shouldn't lose much if any quality.
Thanks for the lead on Lagarith. I looked at the cineform product, but was bent a bit that it would cost me another $500 to go that route. I'll give the Lagarith codec a try.
CS3 had the option to export your project without reencoding. When you selected "export movie" you could uncheck "recompress". The render takes a fraction of the normal time to finish, and it dosn't use more space than the original. No generation loss. No need to save as uncompressed. That option is very usefull for workfloow, when you forinstance want to merge down your finished cut and then apply filters to the whole thing such as boradcast safe. CS4 dose not have that option now that everything goes through AME.
Not for HDV. Even the simplest edit (one cut) changes the GOP structure ergo recompresses.
"Not for HDV. Even the simplest edit (one cut) changes the GOP structure ergo recompresses"
This one statement, if understood would have eliminated the need for most of the posts in this thread. When PP "captures" the HDV stream, it must first remove the MPEG from the transport wrapper. Before you can do ANY editing, the MPEG must be decompressed, since under the HDV format, there is one complete image and a series (15) of partial images that only show the changes from the complete image. You can't edit at any of the partial images since you won't get a complete image. Thus, the capture process converts the complete image and the partial images (i.e., "group of pictures, or GOP") back into uncompressed frames so that you can edit on any frame. Otherwise you could only edit on the complete picture and lose the following 15 partials--which would leave a giant frame gap in your film. To export the edited sequence, PP must recompress the sequence back to a series of GOPs. If you're exporting to tape, PP must also place the MPEG in the M2t wrapper. There is no other way. This is the nature of compressed streams.
I understand that there are MPEG editors that will cut --but you still can't do frame by frame edits because of the GOP structure.
> the capture process converts the complete image and the partial images (i.e., "group of pictures, or GOP") back into uncompressed frames
It actually doesn't. The data is read from tape exactly as is, compression and all. The 'decompression' you're talking about happens on the fly as you edit, which is why editing HDV requires a beefier system than DV, and why AVCHD (with much higher compression then HDV) is a bear to edit even on a top line rig.
Other editors, such as Avid and Final Cut, may use intermediary codecs and do their decompression during capture, but Premiere edits natively. You work with the original unaltered media.
It should also be mentioned that it is possible to export out without recompression, or with compressing only the new GOPs that have changed. This 'feature' has in the past come from a plug-in, and was not a part of Premiere's native capabilities. But it can be done.