different output option for premiere cs5?

[krohm]

been editing various formats in premiere, different sequences depending on res and frame rate, all fine on output on most vids, i just select match sequence settings before exporting and it mostly exports everything to mpeg with really huge file sizes,

some issues though, queries, questions...

when using any bluray footage it always outputs to mpeg2, wanting to know if there is any loss of quality on output, media info shows the datarates to be the same, the video quality looks the same,

should i always use match sequence settings?

with loading wmv files i had to setup a custom sequence as cant seem to find the wmv sequence, this is strange as it now seems i have to reencode to f4v after editing, this seems to be a good codec to decrease file size somewhat, again i use match sequence settings when using this codec, how does this rate comparatively, are there other output options i should be considering?


everything is taking extremely long time to output, pretty much from what ive been advised prior, i need a gcard that has mercury playback to get this to load easier and encode quicker? what kind of differences should that make?, in indexing and conforming it will make a great difference aswell right?


also with some other settings, just confirming field order should be progressive for dvd bluray and all other internet video?
upper or lower for interlaced, which one for the interlaced footage?
also have optimising rendering performance to memory for maximum bit depth setting.. obv,

hope i can get some assistance, cheers

[poisondeathray]

when using any bluray footage it always outputs to mpeg2, wanting to know if there is any loss of quality on output, media info shows the datarates to be the same, the video quality looks the same,

Yes there is quality loss. It's re-encoded , so you incur 1 generation of lossy encoding. Even MPEG2 source blu-ray is re-encoded. PP doesn't smart render MPEG2 without external plugins, and most blu-ray sources are AVC these days

should i always use match sequence settings?

Not necessarily, depends on the project , and many factors .

with loading wmv files i had to setup a custom sequence as cant seem to find the wmv sequence, this is strange as it now seems i have to reencode to f4v after editing, this seems to be a good codec to decrease file size somewhat, again i use match sequence settings when using this codec, how does this rate comparatively, are there other output options i should be considering?

There is no WMV sequence setting, but you can import WMV files into mixed projects, or use "desktop" settings and enter custom values. F4V is a container (Adobe's version of MP4 for flash). It usually uses h264 compression. Again, generation loss, but if you use high enough bitrates, the quality loss might not be noticable

everything is taking extremely long time to output, pretty much from what ive been advised prior, i need a gcard that has mercury playback to get this to load easier and encode quicker? what kind of differences should that make?, in indexing and conforming it will make a great difference aswell right?

Everything is faster with CS5 if you have MPE enabled. The biggest gains are with CUDA accelerated effects - things like transitions are rendered in real time. When you have big projects, multiple layers , transitions, effects that can be very slow on CS4. The exact same project on the same hardware might be anywhere from 1.5-10x faster depending on your project than CS4 . Editing is waaaayy faster. That's the big difference IMO. In CS4 you have to render effects before you edit - very sluggish. Won't make a difference with conforming audio that sort of thing

also with some other settings, just confirming field order should be progressive for dvd bluray and all other internet video?

Yes for internet, probably for blu-ray and DVD . But blu-ray & DVD can be interlaced as well. If it's from a theatrical release, it's probably progressive

also have optimising rendering performance to memory for maximum bit depth setting.. obv,

maximum bit depth export setting is for 10-bit exports. You don't need that if your end goal is something like blu-ray (8-bit only). You would only use it for intermediate projects, exporting something like v210

[krohm]

it seems with wanting to output the wmv with the same quality or minimal quality loss and of similar file size this can't be done. i still dont know how to set the output options correctly, always seems either a massive loss of quality or fine quality and a 4gig 45 min mpeg2 output, what should i be setting it too, ive tried flv with matching sequence settings but it still wants me to add the bitrate, if i set the 600k or 800k from the source and encode the quality is terrible. what is my best output option with minimal quality loss and maintaining a similar file size

[krohm]

just set it to 4mb flv as a custom output, and although the file size is comparable, the quality isnt, checked bitviewer, image below,

shows the original wmv on the left, and the 4mb output of flv, same res and frame rate, similar file size output;,

http://imageshack.us/f/812/20110609173402.png

what do i need to output this to for minimal file loss? a higher data rate obviously but then it will give a much larger file size, is there any other option, what is my best way of outputting not only this but other formats that have a higher than 25mbps data rate, 30mbps etc, what are my output options considering this,

[krohm]

bump

[poisondeathray]

either bite the bullet, or use a higher bitrate

probably I-frame h.264 using x264, using fast decode settings would be easy to edit and decent filesize

or use a proxy / offline edit workflow (swap in the high quality versions on final render)

[krohm]

atm im using match sequence settings, and it gives large output files but the quality seems comparable to the original, is this my ill informed inexperienced video eyes talking,

it seems i need a new output 'workflow' right? or is using match sequence settings is fine if i dont care about file sizing? i can just store all the mpegs whereever, and than compile all of them together and trim them, edit them together etc, into a final project, and it will all be just one generation loss from the original formats, how different will it be? from there i just output again , which i assume will again incur quality loss, a second generation of loss, to convert this mpeg final product to mkv mp4,?

this or i need a way to work with the material losslessly, before outputting to mp4 mkv etc etc. and is this able to be done losslessly? or is this where avisynth comes in, to use it to allow importing all formats to premiere, and outputting losslessly to whatever container/res ild like, what steps am i missing here, i need all the material edited up and outputted before i can put it all together to create the final product. so i need outputted material to work with, hope this is somewhat understandable...

[budwzr]

... is this where avisynth comes in, to use it to allow importing all formats to premiere, and outputting losslessly to whatever container/res ild like, what steps am i missing here, i need all the material edited up and outputted before i can put it all together to create the final product. so i need outputted material to work with, hope this is somewhat understandable...

You should have bought Sony Vegas Pro. You just throw everything on the magic timeline, and you're good to go. That's a real problem with Premiere, in that it just doesn't play well with others.

[poisondeathray]

You should have bought Sony Vegas Pro. You just throw everything on the magic timeline, and you're good to go. That's a real problem with Premiere, in that it just doesn't play well with others.

It's not a problem with Premiere; it handles many formats like Vegas. You can throw everything on the timeline as well. I should know, because I use both.

it seems i need a new output 'workflow' right? or is using match sequence settings is fine if i dont care about file sizing? i can just store all the mpegs whereever, and than compile all of them together and trim them, edit them together etc, into a final project, and it will all be just one generation loss from the original formats, how different will it be? from there i just output again , which i assume will again incur quality loss, a second generation of loss, to convert this mpeg final product to mkv mp4,?

this or i need a way to work with the material losslessly, before outputting to mp4 mkv etc etc. and is this able to be done losslessly? or is this where avisynth comes in, to use it to allow importing all formats to premiere, and outputting losslessly to whatever container/res ild like, what steps am i missing here, i need all the material edited up and outputted before i can put it all together to create the final product. so i need outputted material to work with, hope this is somewhat understandable...

If you're concerned about compression generation losses, use a lossless format for your master export e.g. lagarith, huffyuv, ut video codec . Then you can encode and resize that master to fit whatever target formats (e.g. you might want a web version, or a blu-ray version, or a phone version etc....)

[krohm]

but i will need to output to a lossless codec (huffyub, ut or lagarith) while clipping up all the video first, create all the clips losslessly first, ?? then..

edit up all the clips into one final master project to output again losslessly and then encode with ripbot or something to whatever formats (bd, mkv, mp4 etc)

so this is the workflow i would be needing...?


im updating my pc, finding out whether i will need to change over motherboard ram and cpu aswell as graphics card, or just replace my graphics card to allow for real time editing. obviously will need more hd space, thats no problem. running premiere pro (and windows) on a ssd drive and outputting to another ssd drive is a lot faster than outputting to a sata drive yes?

[poisondeathray]

but i will need to output to a lossless codec (huffyub, ut or lagarith) while clipping up all the video first, create all the clips losslessly first, ?? then..

edit up all the clips into one final master project to output again losslessly and then encode with ripbot or something to whatever formats (bd, mkv, mp4 etc)

so this is the workflow i would be needing...?

There is more than 1 way to do things

You should be able to edit everything as is, then output to some master format. If you only have 1 format to export, then use that final format instead of the mater

im updating my pc, finding out whether i will need to change over motherboard ram and cpu aswell as graphics card, or just replace my graphics card to allow for real time editing. obviously will need more hd space, thats no problem. running premiere pro (and windows) on a ssd drive and outputting to another ssd drive is a lot faster than outputting to a sata drive yes?

Separate drives is a good idea, even if they are sata

[budwzr]

You should have bought Sony Vegas Pro. You just throw everything on the magic timeline, and you're good to go. That's a real problem with Premiere, in that it just doesn't play well with others.

It's not a problem with Premiere; it handles many formats like Vegas. You can throw everything on the timeline as well. I should know, because I use both.

Why doesn't the OP just do that, then? I'm puzzled.

[krohm]

what if i have thousands of clips from 100s of sources, premiere would be sourcing terrabytes and terrbytes of source material...

[budwzr]

what if i have thousands of clips from 100s of sources, premiere would be sourcing terrabytes and terrbytes of source material...

This kind of editing is done in chunks. In Vegas it's called "Nested Veggies", but Premiere probably has something similar.

[krohm]

so obv i need this new system to be able to work competently with this material, so i will be able to load 25gb bluray files, and have 100s of clips from these blurays, +other media, to premiere with a new system, and it will all render in real time, with no bs... then i just need to work out a way to output the master losslessly, rather than clipping up everything first...

[poisondeathray]

You can do it that way, but it might be easier to make rough clips first before you edit. E.g. tssniper for blu-ray

[krohm]

tssniper wont allow me to frame accurately edit and have the control i have in premiere to be able to edit exactly the scene/clip im after, by the frame

so im wanting to be able to output these final edited clips losslessly, ready for easy insertion into the master with no editing required.

thats why it would seem it would be best to make the rough edits lossless, then i can bring them back into the project later rather than dealing with so much data and not outputting anything at all, thats not an option as i need to see how it plays etc,

how can i go about outputting them losslessly with premiere?

if i want to output to lagairth or huffy ill be able to do so quickly with new system, so could just be outputting losslessly as i go through all the material, and then collaborate all of that material at a later stage.

atm premiere outputs match sequence settings mpegs, but if i do it this way obviously will incur a second generation loss when outputting the master, and then a third when outputting to other formats. if i work with lagarith or huffy for all clips, there would be only the loss from the master to outputting to the other formats, what type of differences are there...

ive tssnipered a clip from a bluray source,

http://www.mediafire.com/?11rwpkkah7bijdd

and this is how the same edit of the source, outputted to mpeg match sequence settings for premiere,

http://www.mediafire.com/?3jsjs3hj9ta8c2b

there is a noticeable difference, so if im putting that clip back in and i output again, its just incurring more quality loss.. what type of workflow am i going to need, am i going about this the wrong way, any further feedback appreciated

[poisondeathray]

tssniper wont allow me to frame accurately edit and have the control i have in premiere to be able to edit exactly the scene/clip im after, by the frame

so im wanting to be able to output these final edited clips losslessly, ready for easy insertion into the master with no editing required.

There's many ways to do this, but that's probably not the "right" way . You want rough cuts with "handles" on both sides, otherwise you won't be able to do transitions properly . You do your frame accurate cuts , timing , and transitions in premiere.

If you're just putting clips end to end, not effects or transitions or "real" editing, premiere is way overkill for this. Just put it together in avisynth. A+B+C+D etc....

how can i go about outputting them losslessly with premiere?

if i want to output to lagairth or huffy ill be able to do so quickly with new system, so could just be outputting losslessly as i go through all the material, and then collaborate all of that material at a later stage.

atm premiere outputs match sequence settings mpegs, but if i do it this way obviously will incur a second generation loss when outputting the master, and then a third when outputting to other formats. if i work with lagarith or huffy for all clips, there would be only the loss from the master to outputting to the other formats, what type of differences are there...

If you use high enough bitrate MPEG2, it will be visually lossless. It depends how picky you are. If you've done effects, transitions, those will be in RGB, and really once you are in RGB, the only lossless format is an RGB export (e.g. ut video codec in RGB mode, or lagarith in RGB mode, etc....)

ive tssnipered a clip from a bluray source,

http://www.mediafire.com/?11rwpkkah7bijdd

and this is how the same edit of the source, outputted to mpeg match sequence settings for premiere,

http://www.mediafire.com/?3jsjs3hj9ta8c2b

there is a noticeable difference, so if im putting that clip back in and i output again, its just incurring more quality loss.. what type of workflow am i going to need, am i going about this the wrong way, any further feedback appreciated

I haven't looked at this, but MPEG2 is lossy. If you use high enough bitrates (e.g. 100Mbit/s for long GOP, 250+ Mbit/s for I-frame), then it's close to visually lossless. You are also incurring colorspace losses if you add transitions, effects. Y'CbCr is converted to RGB. Once you're in RGB , stay in RGB until the final format . If you're worried about it use something else truly lossless, or a "pro" DI format like cineform or dnxhd

[krohm]

probably not the "right" way . You want rough cuts with "handles" on both sides, otherwise you won't be able to do transitions properly

yeah i will be needing handles on either side, but mostly i cut an entire scene and i will only be using a part of it, so can just rough cut entire scene that has the footage im after, output that losslessly, and then when all the material is together i can start to storyboard it before doing the editing process. not sure if this is the best way to go about it, but it seems like it from my perspective, which admittedly is somewhat uninformed.

Once you're in RGB , stay in RGB until the final format

im probably not going to export if it gets 'out of RGB' and incurs quality loss if im wanting to reedit the material. so DI is digital intermediate i take it.. cineform is a lossless codec,? why cant i just use ut or lagarith to export and then work with that material, that would be easier no,?

maybe i will need cineform, so it converts whatever input to a cineform file? to be worked in and outputs to a lossless cineform format that i can then convert/encode to whatever format, container i want..

seems both are effectively the same thing, considering this using lagarith and ut are the better option, considering new pc etc, are there some guides or some places i can start to look at regarding how to use these codecs with premiere..

i feel like , as i am, enveloping myself into a slow progressive understanding of a field without the right knowledge or direction. i dont really know what im doing, dont really know how to proceed from here, i am undertaking this with the hope to perhaps work in the field. i have a lot of time in the next couple of months to really get involved with this on a daily basis but need a lossless workflow to be able to do so, im almost there, but at this point im still going to need to learn how to put a story together essentially, an entirely new field in of itself, where do i turn to to begin to undertake a process good enough to be able to put this together. this is a little offtopic maybe i should repost this elsewhere.. , at this point it seems i would need to mostly rough cut the story i have in mind, and take it to professional editors to assist me with the process.

[poisondeathray]

probably not the "right" way . You want rough cuts with "handles" on both sides, otherwise you won't be able to do transitions properly

yeah i will be needing handles on either side, but mostly i cut an entire scene and i will only be using a part of it, so can just rough cut entire scene that has the footage im after, output that losslessly, and then when all the material is together i can start to storyboard it before doing the editing process. not sure if this is the best way to go about it, but it seems like it from my perspective, which admittedly is somewhat uninformed.

This is the way I would do it. If you only need 1min clip from a 2hr movie , why import the 2hr movie ? It will take ages conforming audio that you're not even going to use.

Once you're in RGB , stay in RGB until the final format

im probably not going to export if it gets 'out of RGB' and incurs quality loss if im wanting to reedit the material. so DI is digital intermediate i take it.. cineform is a lossless codec,? why cant i just use ut or lagarith to export and then work with that material, that would be easier no,?

Cineform isn't lossless, it's visually lossless, but very easy to edit (very fast decode speed and low latency) . There are different versions of cineform, at the highest levels it is 4:4:4 12-bit, but those are the more expensive forms. The commonly used one is 10-bit 4:2:2 and has different quality levels as well. Not free, but it's used by many pros because of editing performance

Lagarith is very slow to edit. It's not an editing format. It's an "archiving" format. Use Ut Video Codec, it's easily the fastest multithreaded lossless codec, faster than huffyuv . Lagarith is more compressed (smaller filesizes) . In general, lossless codecs aren't meant for editing - you don't get faster than realtime performance, scrubbing is slow as molasses. UT is the only one that can with HD footage, if you have a fast computer

i feel like , as i am, enveloping myself into a slow progressive understanding of a field without the right knowledge or direction. i dont really know what im doing, dont really know how to proceed from here, i am undertaking this with the hope to perhaps work in the field. i have a lot of time in the next couple of months to really get involved with this on a daily basis but need a lossless workflow to be able to do so, im almost there, but at this point im still going to need to learn how to put a story together essentially, an entirely new field in of itself, where do i turn to to begin to undertake a process good enough to be able to put this together. this is a little offtopic maybe i should repost this elsewhere.. , at this point it seems i would need to mostly rough cut the story i have in mind, and take it to professional editors to assist me with the process. can anyone provide any feedback regarding, wanting to essentially undertake editing and storyboarding to create a documentary based film, i have all video voice over , images im wanting to use, some txt even from books perhaps, minimal, and i have no idea on how to make this into a coherant story that is not going to look like an amateurish piece of sht google video doco, this is a little offtopic maybe i should repost this elsewhere..

If you 've never edited before, start looking at tutorials. Youtube or lynda.com has some beginner tutorial series

[krohm]

If you only need 1min clip from a 2hr movie , why import the 2hr movie ? It will take ages conforming audio that you're not even going to use.

na im not, i use tsmuxer to grab the sections i want to edit, split the m2ts into sections, atm thats what im doing, might use tsmuxer to remux them back to ts and use more tssniper seems more useful than i last remembered

Use Ut Video Codec, it's easily the fastest multithreaded lossless codec, faster than huffyuv .

ill get onto ut, with new beast system ill work with that.

[krohm]

If you 've never edited before, start looking at tutorials. Youtube or lynda.com has some beginner tutorial series

did the lynda tutorials for cs5, did 20 out of the 25 or whatever, so yea im somewhat confident i can put some nice scenes together with the material i have, and storyboard it somewhat, but still im completely shooting in the dark, dont know where to turn read a lot of websites they all seem to be talking about filming etc, not sure how to actually put this together the creation of it, anyways ill look into that further myself that was more a side note.

thanks pdr with my goals of the way in which i want to approach this. seems like new pc + ut = project underway..

[poisondeathray]

Yes, if you get the editing mechanics down, then you can actually work on the storytelling part.

But there are so many different "styles" and ways to do convey your message and a lot of it is subjective - there is no "right" way to do it. e.g. Things like type of transitions chosen, grading can affect the mood of the piece. For full control, you need to do the original shooting as well , and plan this FAR in advance - because you are already limited in the way you tell your story, if you use a preshot asset like a retail blu-ray.

[krohm]

if you get the editing mechanics down, then you can actually work on the storytelling part.

yeah one step at a time, already thinking ahead , as im sure everyone does.. ill put in the volume when i have new pc and take it from there

[budwzr]

i feel like , as i am, enveloping myself into a slow progressive understanding of a field without the right knowledge or direction. i dont really know what im doing, dont really know how to proceed from here, i am undertaking this with the hope to perhaps work in the field.

Get a copy of "The Filmmakers Handbook".

It's impossible to make anything professional-looking without understanding how an audience has been trained to think by Hollywood.

[budwzr]

tssniper wont allow me to frame accurately edit and have the control i have in premiere to be able to edit exactly the scene/clip im after, by the frame

PDR recommended using TSsniper to make ROUGH cuts as a way to break down the media into easier to manage clips, but you failed to grasp the concept.