VideoHelp Forum




+ Reply to Thread
Results 1 to 7 of 7
  1. Member
    Join Date
    Sep 2002
    Location
    Washington, D.C.
    Search Comp PM
    Good Morning to all.
    Please be patient I know this topic has been covered before, but I just am not getting it straight.
    First, in my own words, it seems to me that frame serving is using one program (capture) to feed another (convert) without creating the intermediate file. I understand that Vdub will write to a dummy intermediate frame serving file which Tempgenc will read. Hence frameserving.
    Now for my misunderstanding. I need to set up Vdub and the frame serve file and set up Tempgenc to read this file. Then, heres the sticky part, do I go back to Vdub and start the capture part??
    If so, then how does Tempgenc keep up with the real time cature of Vdub.?

    I have noticed that on 1-2 min captures it takes Tempgenc anywhere from 2-6 min to do the conversion to mpg. Even I can see that after a while Tempgenc will fall way behind.

    Or does one already have to have the avi file for Vdub to process to Tempgenc. If so, this seems kinda silly. Why not load tempgenc with the already available avi.???

    Thanks for your patience,
    Ed
    Quote Quote  
  2. Banned
    Join Date
    Apr 2001
    Location
    Calgary, Canada
    Search Comp PM
    Hi Ed,

    You use VirtualDub to capture the complete file first. VirtualDub's frameserver will not feed a live capture through to another application, you need an existing .avi file beforehand.

    After you have finished your capture, you are free to apply filters or cut frames or whatnot using VirtualDub.

    Then you start the frameserver (VirtualDub will create a small .vdf file) and load THAT into TMPGEnc.

    As for loading TMPGEnc with the original .avi file, VirtualDub is great for removing commercials and applying special filters (like DeLogo). If you wanted to load an .avi into TMPGEnc, you'd have to wait a couple hours (depending on the size of the file and the speed of your system) for VirtualDub to create a NEW .avi with all the filter/frame changes applied.

    Sure, TMPGEnc will have an actual .avi file to play with, but the couple hours it takes to create that file is a hassle (imo) -- just make your changes in VirtualDub, "Save Processing Settings" and fire up the frameserver... No middle-steps.

    And sure, using the frameserver takes CPU resources from TMPGEnc, but personally, I let my encodes run over night -- if it takes 7 or 8 hours using the frameserver vs. 3 or 5 hours without (PLUS the couple hours it takes to make the changed .avi), fine by me... I'm sleeping anyway

    Hope that helps
    Quote Quote  
  3. Member
    Join Date
    Sep 2002
    Location
    Washington, D.C.
    Search Comp PM
    Hilljack
    Thanks for the clarification on Frameserving. I have not yet used the editing features of Vdub but..... Thus, I understand the need for frame serving.
    Also, let me pick your brain. I am running into the FAT32 file size limit for a VHS movie that I want to convert. It is about 105 minutes long. I am capturing in Vdub using VCD stds and using Huffyuv 2.1.1. However, after about 40 minutes the process stops with the 4 Gb limit being reached. Somewhere ( I just cannot find it again) I seem to remember reading about segmented captures in some program. Maybe this is what I need to do.
    Any suggestions.

    Once again thanks,
    Ed
    Quote Quote  
  4. Banned
    Join Date
    Apr 2001
    Location
    Calgary, Canada
    Search Comp PM
    Hi Ed,

    Sorry about the delayed reply.

    VirtualDub does have the ability to "span" a large capture into several 2Gb chunks so you can overcome the 4Gb limit of FAT32. It's called a "Spill System" and you set it up by:

    1) Load up VirtualDub and go to "Capture AVI" as normal.
    2) Under the Capture menu is "Capture drives..." go there.
    3) Up pops a "Spill System Setup" dialog.

    It'll most likely be empty for you since you have not run it before. So, click the "Add spill drive" button and poof, a priority, threshold and Path appear. Leave everything alone, except click in the window area directly underneath the Path column. Here, you should type the path (like d:\ or d:\captures or whatever you want) to where you want VirtualDub to dump files. You can mess around with Priority and such, but that's only if you are using multiple physical drives.

    Once you set a path, leave the "50 megabytes" and the "1900 megabytes" options alone, and click OK.

    So now, you've told VirtualDub you want < 2Gb segments and WHERE you want it to dump the segments, now before you start the actual capture, you need to "Enable multisegment capture" under the Capture menu. Once you set that, VirtualDub will capture the first file and when it reaches 1.9Gb in size, it'll automagically move onto the next file, until it reaches 1.9Gb... and so on... Until your capture is complete (or you run out of drive space). The segmenting is transparent to the end product...

    A segmented capture in VirtualDub is really no different than one large capture on a different file system. You use the files the same way as you would one large file -- except when you want to edit your ENTIRE capture, you load the FIRST file of the set, and VirtualDub will automagically load the others. It's really quite slick imo

    One final note -- Assuming you are using VirtualDub 1.4.9 (there is a 1.4.10 but there are problems with framserving, so 1.4.9 is the best version imo), you can save the segmented file settings (the spill locations) for later use, but should you exit VirtualDub and then reload wanting to do another capture, you will need to "Enable multisegment capture" again -- for some reason, it remembers the settings, but won't remember if its enabled or not

    I hope I helped,
    good luck
    Quote Quote  
  5. Member
    Join Date
    Sep 2002
    Location
    Washington, D.C.
    Search Comp PM
    HillJack,
    Great.!!! I have frameserving working and have accomplished my first multisegment capture. It went as smoothly as you had described. I even managed to make some edits with Vdub. And frame serving to TEMPGENC went just as smoothly. However, if I could impose on your knowledge just a little bit more.
    I think I understand the process but maybe you can verify my logic here. If my goal is to create a standard compliant VCD. I will need to cut my original source up into approximately 80 minute segments. So to be safe lets say I want two 75 minute mpgs. I load up the first captured segment into Vdub, do a section select from frame 1 to a frame that is about 75 minutes into the movie. Then frame serve that to Tempgenc so it can create the first 75 minute mpg. Then repeat the process for the second half. Being sure to provide a little bit of overlap for the second half.
    I do know that the frame serving is an overnight job. However, is it possible to batch up Vdub for the second half and then include the frame serve into the batch job as well.???

    Once again, thanks for the helping hand.

    Ed Smith
    Quote Quote  
  6. Banned
    Join Date
    Apr 2001
    Location
    Calgary, Canada
    Search Comp PM
    Hi Ed,

    I do know that the frame serving is an overnight job. However, is it possible to batch up Vdub for the second half and then include the frame serve into the batch job as well.???
    You bet it's possible! You just need to have that many different copies of VirtualDub running their own framserver session. Pertaining to the framserver, VirtualDub will batch everything EXCEPT framserver sessions. So, lets say you've installed VirtualDub to...

    c:\Program Files\VirtualDub

    That one location will let you run one framserver at any one time. You want to batch up X more? You can run any program multiple times, but the safest bet is to run that program from different locations; Make copies of your VirtualDub directory...

    c:\Program Files\VirtualDub1
    c:\Program Files\VirtualDub2
    ...
    c:\Program Files\VirtualDubX

    VirtualDub is tiny, so copy as many as you think you'd ever need (plus a couple for good luck

    Now, I personally have only 5 VirtualDub# directories. At most, I'll ever run 3 framserver sessions at once. 3-5 should be sufficient imo.

    Now, the key: Lets say you want to framserve two separate captures, using two different framserver sessions (from two separate VirtualDub dirs of course...): You know that box that appears when you first hit "Start framserver..." asking you for a name//file type of deal? (just before you specify the actual .vdf filename?). For each instance of VirtualDub, make sure THAT info is different for each one. Otherwise, even though you've loaded up two separate instances and two separate files, your encoder will encode the same stream each time The .vdf, make sure those are different for each instance too

    Now, once all your framserver sessions are fired up, load them up into TMPGEnc just as you would separate AVI files using TMPGEnc's Batch Encode or whatever its called -- I havent used TMPGEnc in awhile (CCE for me), and process as normal

    I hope that makes some sense
    Good luck!
    Quote Quote  
  7. Member
    Join Date
    Sep 2002
    Location
    Washington, D.C.
    Search Comp PM
    HillJack,
    FANTASTIC !!!!! Just the answer I was hoping for and presented in a manner that I actually understood. Without having to repeat it manually in order to grasp what was going on. Boy, that will save hours of work.

    Thanks for your EXCELLENT advice and patience with a new comer.
    This will is going to be fun. Once I understand the language and the techniques behind the processes, that is.

    Gotta go and encode about 70Gb.

    THanks,
    Ed Smith
    Quote Quote  



Similar Threads

Visit our sponsor! Try DVDFab and backup Blu-rays!