referring to the thread here (opens in new window), discussion of mainstream movies and the DVD format, etc....

Let's say you have definately decided to make a short movie, 6 minutes, and you want to optimize for highest quality release DVD, HDTV, BigScreen.

For the sake of argument lets say you are satisfied you can render in Poser5 itself. You have a budget of 5 mintues per frame, have four seats of Poser on four strong PCs, and full duration render time is not a problem. (10,800 frames at 5Min/frame = 900 hours over 4 computers around the clock = 10 days, throw in down time pad, etc., still well within my project deadlines) and it is okay this is not a 'true' render farm.

Aspect ratio is 16:9.

What is the optimum target output? What should be the target size be for best results in the post-processing software. 1280x720 at 300DPI? Or?

Test renders confirm 5 minute budget is real, given 1280x720 300dpi, tif file size = 3.6MB

I am requesting opinions about post-processing/ripping software, not so much for 'effects' but for going from the 10,800 individual 3.6MB .tif files to finished video. I assume this will be progressive scan, but that is not an issue with the Poser render, that is a choice when compositing out the final from the post-processing software, right? NOTE: I have a loaded Mac G5 already running Logic just dandy. I am wondering if Final Cut Pro is a good candidate?

What are the target file sizes and formats out of Poser, headed for DVD, with an eye to big-screen projection and HD-TV as well.

I am looking for that crisp, smooth result described for the Shrek2 DVD discussed in the attached thread.

Thank you for any wisdom and experience and pertinent threads ...

::::: Opera :::::

http://heroinewarrior.com/cinelerra.php3 a tool designed exactly for this. and it's FREE. hardware requirements: dual opteron 2.5 GHz 4 GB Ram 200+ GB Hard Drive Space LINUX sorry but it won't work on winders.

interesting and sardonic page. I have to accomodate myself to the idea that once I render out 10,800 TIFF images out of Poser, the expectation seems to be that I have not even reached half-way yet, that I am facing another massive full render in order to make a movie. Now, if my Poser output is pretty clumsy and I am looking to massage the images themselves to save my life, I can understand it. Or...if I am adding FX, etc., I can understand it. But what if I just want to 'assemble?" I do it now with QuickTime Pro, which has a free partial-feature Sorensen Video compression, which I bet could be kicked up. Is this a path to "assembly" which can get me to the quality level I am looking for? ::::: Opera :::::

Meanwhile, even while thrashing through the decision about post-processing software, I hope I can discover that -- regardless of which path I go with -- there is some 'optimal' target image size/resolution that will be the best for all, so I can start test renders in Poser5 at those settings. ::::: Opera :::::

Obviously, the optimal target size is dependent upon the maximum "minimum" size for the desired targets. For DVD, the size is usually about 720x536 (NTSC), but Widescreen, HDTV have different specs (within the specification, that is). If you are going to have a Widescreen 16:9 version, that is one target complete. All movies are done in Widescreen and "accomodated" for Fullscreen to handle the 4:3 of TeeVee. ;) So, plan for that aspect ratio as default. Otherwise, you are going to need to check each target for its specifications. I had to do the same thing for DVD (and there are many various formats just for that - NTSC/PAL, Widescreen/Fullscreen, etc.). BTW, screw DPI. DPI = Dots Per Inch, which is only relevant for printing. Unless you plan on printing your movie - hmmm - DPI=1 is just as valid as DPI=1000000000.

Addendum: Premiere Pro is the best video editing software for Windows before you go truly 'studio-grade' professional (of course). QTPro is good, but it doesn't have the NLE capabilities of PP or FCP.

But what if I just want to 'assemble?" VirtualDub (Windows only) can assemble image sequences into AVI videos. It can't read TIFF files, unfortunately, but if you're willing to render to PNG instead, it can handle that. Use a lossless video codec like Huffyuv while assembling, editing, and postworking the video to insure maximum image quality before converting and compressing the finished video to MPEG-2 and burning it to DVD.

Addendum 2:

You want 'raw' footage. Raws and wrigglin', that's the ways we's likes it.

The master, for better or worse, death do you part your 200GB hard drive, should be non-lossy video. Why? Because formats like DVD will turn the video into compressed, lossy MPG2. Remember: Garbage In, Garbage Out. The better the video quality before turning into hard formats, the better the results. If you use a compressed/lossy format for input, the quality will go down from there.

Message edited on: 01/06/2005 03:57

Message edited on: 01/06/2005 03:58

Yes I am commited to highest-res non-lossy raw footage and thank you both for confirming that it is important to stay lossless as far into the process as possible. So far, that has been full-sized .tif out of Poser. I guess I could render to PNG as easily as TIFF. But I only want to render the movie out of Poser ONCE! So, my quest is to discover the optimal upper value, the one size and resolution for the Poser raw footage render-to-individual-frames, which will then universally be okay to 'downsize' from if I have to. I am satisfied I will NOT have to create a 4:3 version. If (ever) run on "Tee Vee" it would have to be letterboxed. Meanwhile, for video, HD-TV and whatever is required to go to actual 'movie screen', all will be 16:9. ::::: Opera :::::

"What is the optimum target output? What should be the target size be for best results in the post-processing software. 1280x720 at 300DPI? Or?" If I recall correctly, for HDTV (video), I usually output the render at 1920X1080, DVD-video (720x480 NTSC/720x756 PAL). DPI is irrelevant.

Hi maxx, was hoping you would see this thread and chime in. Geeze, those are two different ratios/sizes and the smaller one is not an exactly proportional reduction of the larger, because while 1920x1080 is 16:9, 720x480 is not, 720x405 would be 16:9. (720x756, meanwhile, is practically square and I won't be going there; that would be a compeletely different movie, shot-wise) Therefore, that would seem to call for two spearate renders out of Poser because I would have to believe that if I rendered out of Poser once at 1920x1080, 'downsizing' to 720x480 would produce mathematical garbage. Better to re-render out of Poser at the exact 720x480, I would think. Rats. I guess the compromise would be: 1) clip the outer edges off of a copy of raw 1920x1280 to the exact same ratio as 720x480 and then reduce it; or just reduce it and the hell with it, whatever you get. Thanks for that 1920x1080, someone else told me that was the optimal raw footage size, as well. ::::: Opera :::::

"So, my quest is to discover the optimal upper value, the one size and resolution for the Poser raw footage render-to-individual-frames, which will then universally be okay to 'downsize' from if I have to." Ohhhhhh. Uhm. Yikes. ;-) I'm not sure about that. I don't think there's any "universal" size that can be translated without distortion to every format you might need. Especially if you ever want to convert a PAL-compatible version.

On the 'dpi' thing...i agree it would be irrelevant, the exact pixel count is called out already. But why does Poser have a "resolution" setting on the render options expressed in 'pixels/inch'. What function does that perform? I just did a test, and no matter what you change it too, the render takes the same amount of time and gives a .tif of exactly the same size. ::::: Opera :::::

"Therefore, that would seem to call for two spearate renders out of Poser because I would have to believe that if I rendered out of Poser once at 1920x1080, 'downsizing' to 720x480 would produce mathematical garbage" Yes, it would. Here's some presets I just grabbed out of 3dsMax that (might) help you decide. Or, might confuse you even more. ;-) HDTV: 1920x1080 (optimal) 1280x720 490x270 320x180 (Image Aspect: 1.77778; Pixel Aspect: 1.0; Aperture Width (mm): 20.120) NTSC DV: 720x480 360x240 300x200 (Image Aspect: 1.35000; Pixel Aspect: 0.9000; Aperture Width (mm): 20.120) 512x341 (Image Aspect: 1.35132; Pixel Aspect: 0.9000; Aperture Width (mm): 20.120)

Clearly, rendering for DV (DVD) and HDTV will have to be done in two renders. The optimal output for each format results in a mathmatical incompatibility. You could resize in post, but it's going to result in quality issues.

Message edited on: 01/06/2005 05:43

Couple of points: At the end of post 10, where maxxxmodelz says "720x756 PAL", that's a typo and he actually means "720x576 PAL", so it's not as square as you think. TV pixels aren't square like computer pixels, which is why a lot of these sets of dimensions don't actually translate to 16 by 9. Some 3D apps (like Vue Pro) will render an image in non-square pixels for you, but Poser isn't one of them. With poser you'll have to make the images in a 16 by 9 size and then squash them down to the size required by the TV format you intend them for. I live in a PAL country so my plan (when I eventually get around to doing some DVD animation) is to render at 1024 by 576 and then squash the frames down to 720x576).

Thanks for confirming what I thought. I will drill down into those settings later, the main thing is facing the fact I will have to render twice. I accept that fate in the interest of quality. I wonder what is the path to 'the movies?' I don't mean the odd digital projector now beginning to appear in movie theaters in major cities, I mean if you create this great digital movie, what is the optimal quality path to 'film?' ::::: Opera :::::

I believe the standard NTSC has 525 scan lines, and HDTV is 720 & 1080i.

The native format of DVD is 480i, which is also the operational mode of all traditional televisions and video cameras. However, most DVDs are sourced from film, not digital video. Film is natively 24 frames per second, none-interlaced. When film is converted into a DVD, it's converted to 480i using a 3:2 "pulldown technique".

Most HDTV's out there are able to perform a reverse 3:2 pulldown automatically if it detects film-source material.

Message edited on: 01/06/2005 05:57

According to something I read recently, digital film is rendered at 2048 by 1536.

"At the end of post 10, where maxxxmodelz says "720x756 PAL", that's a typo and he actually means "720x576 PAL", so it's not as square as you think." Ooops. Yep, you're right. ;-P

Ajax, when you say "TV" pixels, do you mean SDTV, HDTV and also DVD intended to be played on a SD/HD set (as opposed to computer monitor pixels?)

I am getting a little confused by your post...you are making reference to the 'pixel shape' but then also the overall aspect ratio. Are these two things related?

Are you saying that when you intend to squash down from 1024x576 to 720x576 you are doing it to compensate for Poser's non-round output...that the squashing will 'fix' the non-round poser output into the round output needed by TV?

::::: Opera :::::

Message edited on: 01/06/2005 06:02

On a computer, a pixel is a square. On a TV a pixel is a rectange with a different height to its width. I think the deal is that NSTC TVs have pixels which are 0.9 times as wide as they are high and I think PAL pixels are wider than they are high but I don't remember the exact dimensions. I don't understand what you mean by "round output" but on a PAL TV, 720 by 576 actually is a 16:9 ratio because the pixels are wider than they are high, whereas on a computer where the pixels are square, 1024 by 576 is a 16:9 ratio. So if you take a 1024x576 render and squash it down to 720x576 then it will look horizontally squashed on your computer but when you put it on a DVD and show it on a 16:9 PAL TV it will look normal.

One thought: Do you have Adobe Photoshop? If so, you could avoid rendering twice by using a batch routine to resize all your larger renders to the additional, smaller size render set you need.

"One thought: Do you have Adobe Photoshop? If so, you could avoid rendering twice by using a batch routine to resize all your larger renders to the additional, smaller size render set you need." Well, if he has a video editor, he won't need Photoshop to do that. Resizing is a bad idea anyway unless the downsize is mathematically compatible with the format. Operaguy, if you're seriously considering making a feature-length movie of this caliber, I think you might really need to consider making a financial investment in some higher end software and hardware that can also work with Poser figures/scenes/animations. It's obvious you're looking to take your skills to the next level, so I'd be looking hard at something that can take advantage of dual or quad processors, and/or network rendering. Or perhaps outsourcing your project to a "pay-per-hour" renderfarm resource, like Respower.com. It's admirable you want to tackle something like this in Poser by itself, but I think you're setting yourself up for some self-induced torture. ;-)

Yes, you may be right Max. No one has ever made even a short serious film in Poser. This six-minute preview will be my trial by fire. If I encounter that "whoops, I am working with a toy" feeling, I will be ready to bail. I alreay have VuePro4 setting in a box right here on my desk. It's cute! If on the day I finish this preview I can truly say "I have a movie..." then I will look at options. But if I can do it in Poser by scaling up my "CondoFarm" idea, I just may do it... the simplicity of being in only one piece of software, etc., and I already have the budget to buy another four or so PCs. [note: on Final Fantasy, their time budget was 90 MINUTES per frame and they had 1000 CPUs under Rendrman on that project.] Now that I see I am faced with multiple complete renders for various formats....that is intimidating. I will be mulling that over as I produce this first short. Just to put one concern I know you have away...there is no client pressure on this project. I am the client. Pushiest b'tard I ever worked for. ::::: Opera :::::

If you are creating a dvd for normal television the image size is 720 x 480 x 16bit color The file on DVD is an mpeg 2 file format so your combined image files will be effectively rerendered through the appropriate codec. The dpi is for printing use, does not affect the render time at all (you are rendering a number of pixels x color depth). Changing or setting the DPI will change the actual print size in most paint programs. If you want some suggestions try newsgroup rec.video.production at Google groups.

"Well, if he has a video editor, he won't need Photoshop to do that. Resizing is a bad idea anyway unless the downsize is mathematically compatible with the format."

Fair enough. I had just found that Photoshop worked for me when I wanted to downsize some larger images to use in a DVD stream. By using a black background, I was able to resize then cut and repaste, maintaining the original aspect ratio by making it a letterbox format to fit the TV screen. And Photoshop seemed to do a very good job with the resizing even though the original resolution dimensions weren't whole number multipliers of the target dimensions (if that's what you meant by "mathematically compatible"). But I have also had bad outcomes using resizing routines in VirtualDub, QTPro, etc so maybe he is better off doing multiple renders. [Anyway, I hope it works for you operaguy.]

Tguyus, I appreciate both your first post and that you came back and adressed legitimate objections with concise positive knowlege. My mind DID gravitate to your solution, because I did it once on a small scale. I had a folder of 300 frames, .tif, and I wanted to test an idea for "guerilla post-processing." So, I opened one frame in my graphics software and establsihed the change I wanted. It amounted to a desaturation of about 50 frames in one place, and the overlay of a layer starting in a certain place and going to the end. In both cases the exact same edit was needed on each subsequent frame. I was able to create a macro. I set it loose on the folder. It took a LOT of fiddling, but I DID get it to work. My software is Canvas. (Yes, I know, no one in this forum has ever heard of it and howcome operaguy does not have Photoshop. Insert smiley face here.) Now...is there a substantive difference in manipulating the frame with a great 2D editor and running a macro (if the change is consistent, such as a re-size or crop) then there would be with top-flight video editing software? You never know. There COULD BE. Perhaps even the best video editing software would fall short in some respect (2D graphics pros -- I am not one -- can do insane magic - and it may be memorizeable)...and there might be a window of opportunity to take a copy of the master folder and unleash the most godawful macroization ever attemptd on it. Might need one of those dual opteron screamers jacqui mentioned above! I appreciate you placing two posts here on this issue, especially the important information that you DID achieve superior results in Photoshop than in a few mediumly fine editing apps. I will definitely give this out-of-the-box idea due respect. Thanks. ::::: Opera :::::

markschum, >> If you are creating a dvd for normal television the image size is 720 x 480 x 16bit color << That's NTSC, right? And 16:9, right? (even though it is not mathematically 16:9 because of the rectangular pixel thing.) This agress with what others are saying. And is that "common denominator" NTSC or "the best you ever want to do because anything better is wasted on DVD?" >>>The file on DVD is an mpeg 2 file format so your combined image files will be effectively rerendered through the appropriate codec.<<< Is it most appropriate for an indie artist to turn the master folder of lossless TIFF or PNG images over to a pro house at that point and let them DVD-up? Or...what is your recommendation for video editing software? >>>If you want some suggestions try newsgroup rec.video.production at Google groups. <<< I will visit. Thanks. ::::: Opera :::::

Hi Opera, May I ask how you have arranged for voice work for your film?? as an animator and aspiring(albiet frustrated) Digital filmaker I find my biggest abstacle is getting Voice Dialog recorded to use with Mimic, as i am somewhat of a rugged individualist who is not really taken to depending on others. Do you have a small team of reliable people Male &female to provide you with dialog??? at this point I am considering making my animated film silent with subtitles :-/

i have my male lead. He is great. He is "inside" my circle, so I have great flexibility and interactiviy from him. My female lead...not yet, I have some feelers out; but I am not ready to pull the trigger on her yet. Here in Los Angeles, it does not take more than five minutes to put out the word and get gigantically flooded with prospects, especially for voice work only. Just imagine...all the spectacular acting talent in this town that can't get work because of the cruel 'appearance' paradigm of "the Biz." I have my own voice equipment for recording on a level that will be satisfactory for a big majority of this project. That consists of Mac G5 LogicPro Mackie Mixer DFX12 AKG C 3000 B Condenser Mic Mackie HR824 Audio Monitors Various sound retardant and acoustical panels. As you can imagine, there is NO shortage of professional recording studio space in this town, when needed. Did you know you acquire voice work virtually? I found sites...you can audition actors right online, then submit your script and get a quote...they will exectue and send you the final raw audio file! Google for a few minutes and you will find them. Here's one: http://www.voicemodels.com/ ::::: Opera :::::

operaguy--

thanks for the kind words on my earlier posts...

On this point: "Is it most appropriate for an indie artist to turn the master folder of lossless TIFF or PNG images over to a pro house at that point and let them DVD-up? Or...what is your recommendation for video editing software?"

In my limited experience, I have achieved the best picture quality results for DVD by using one of two approaches, depending on whether I need to convert my individual frames to an AVI stream along the way:

Approach 1
a. Render individual frames as PSD files in 640x480x72ppi

b. (optional step:) process as needed in Photoshop using batch automation

c. Import image sequence into Quicktime Pro 6 and save as self-contained .MOV

d. Import .MOV file into Ulead Movie Factory 2 (one of the few DVD authoring programs I've found which import MOV files and which I could afford)

Approach 2 (if working with an AVI file):
use TMPGEnc to encode MPEG-2 file from AVI then import to DVD Authoring program. TMPGEnc is the best MPEG-2 encoder I've found, and it is significantly better graphics quality than native encoders packaged with Ulead MF2, DVD Complete (which became DVD X Maker then disappeared when 321 Studios folded), and MyDVD. These are the only DVD Authoring programs I have. Perhaps Adobe Premiere or some other program has a native encoder as good as or better than TMPGEnc but I have never tried these or any others besides the three programs I own.

When I got my license for it, TMPGEnc was only about $25-30 as I recall. When last I checked, you could download the free version and MPEG-2 encoding would work for about 11 days if you want to try it out.

Note: some might point to apparent inconsistency in my advice re Ulead MF2 encoding given that I use that program to import MOV files. For some reason, 640x480 Poser-based MOV files look great when imported to Ulead MF2 whereas MF2 does a much poorer job importing and encoding 720x480 DVD-compliant AVI streams. I don't know why.

Again, good luck with the project.... cheers

Tguyus,

I will always render out of Poser5 to individual frames, the best lossless images I can afford (render-time-wise) and always wide format, 16:9. So I am thinking through your Approach 1.

NOTE: the need for some sort of Web format is minor here, absolutely the tail of the dog. I am intent on the path to the very highest reproduction for DVD (letterboxed on 4:3 television sets) HDTV (obviously 16:9) and for providing the best output needed to go to "film" for the big screen.

1. are PSD files just as lossless as TIFF and PNG?

2. if not, couldn't Photoshop open up TIFF or PNG files for batch processing? I definitely want my folder of raw images to be in the most universal format with the highest possible amount of detail.

3. after batch processing my raw through Photoshop...am I still at very highest resolution and detail, or have I now applied (along with whatever cropping or filtering or effects I wanted PS to do) compression or lossy of some sort?

4. that frame size of 640x480...it won't take me to 16:9 aspect ratio, right? I need to shoot this movie in wide format.

5. making a .MOV before going into the DVD authoring software...isn't there a compression invoived there? Can't Ulead work with a folder of imdividual images?

Note: as anyone reading this thread can tell, I am not a professional in this realm (that's an understatement) and I am on the lookout now for my postprocessing genius who will be on board to move me through. But I DO need to get a lot of this stuff soaked into my brain so I can interact intelligently with him/her.

However, I have to start animating and rendering soon, and setting up shots for this 6-minute preview. So I need to settle on my first render size and ratio. From what I am learning on this thread, the width seems destined to be 720. However, I am still not clear on the heigth. My calculator tells me 405, but I have never actually seen that number anywhere! It's the non-square pixel issue. I need to re-read this entire thread, get serious with what my first target output is vectored for, and then perhaps ask more questions.

Thank you all for your contributions.

::::: Opera :::::

Hi... sorry the delay in responding. I also have to run off to a meeting, but here are my quick replies:

1. are PSD files just as lossless as TIFF and PNG? AFAIK, yes.

2. if not, couldn't Photoshop open up TIFF or PNG files for batch processing? I definitely want my folder of raw images to be in the most universal format with the highest possible amount of detail. Photoshop can import TIFF and PNG files too, so I would think any of these formats would do fine.

3. after batch processing my raw through Photoshop...am I still at very highest resolution and detail, or have I now applied (along with whatever cropping or filtering or effects I wanted PS to do) compression or lossy of some sort? As long as you save back to a lossless format you shouldn't lose anything.

4. that frame size of 640x480...it won't take me to 16:9 aspect ratio, right? I need to shoot this movie in wide format. Not an expert here, but if you're going to DVD, you have to wind up somehow in a 720x480 DVD-compliant format, whether you do the conversion or your DVD authoring program does it. Whether or not the clip, once converted, displays correctly on a 4:3 or 16:9 display is determined by the original render dimensions (i.e., whether the images were rendered in a 4:3 format like 640x480 or a 16:9 format like 852x480 prior to being converted to the DVD-compatible format). NOTE: For much more detailed and nuanced info on the various video formats, see http://wiki.helpware.net/index.php?page=Video

5. making a .MOV before going into the DVD authoring software...isn't there a compression invoived there? Can't Ulead work with a folder of imdividual images? I don't believe Ulead can import individual images (at least I haven't found such a routine) though there are programs which can (such as, I believe, Adobe Premiere... which I don't have). AFAIK, QTPro does not impose any compression when you load a lossless image sequence then save it as a MOV. When imported to Ulead, though, Ulead adds compression. It's not perfect either, since I can see some jagginess on diagonal, high-contrast lines; but it's better than anything else I've found.

If you find a better way, let us know! Or perhaps one of the more knowledgable folks here will chime in.

Cheers...