Hello, in our company we have a Abekas 8150 production switcher. It has a framestore option so you can save framegrabs to a .yuv file. The format of this file is very simple: (i can email a sample if someone wants it) 720x576 Pixel (PAL): Y(8bit), U(8bit), Y(8bit),V(8bit)... and so on. As you can see its 4:2:2. Now my question: is there a way to import these files (lets say i have 250 grabs = 10sec) into avisynth and serve it directly to tmpenc without any colorspace conversions and create a mpeg2 file? There is also the possibility to save .y10 files witch are 10bit instead of 8! But i think mpeg2 is always 8bit (am i right?) thanks hans runz
--------------------- OBD1M52:S52cams:M50intakemanifold:3.38LSD:AAtuning :Eisenmannexhaust:Cosmosintake:Z3steeringrack:Mroa dsterSSK:KoniSA: EibachPro::ReRSM:Xbrace:MillimigliaMM2:Vaders:Alpi ne:Original MARB parts Porsche 1997 993 911: Stock 2003 Mini Cooper S: Superchar
did you try ImageSequence.dll ? It's the only plugin around to import image files. Is there no way to change the output format? (we have a similar task)
--------------------- "If at first you don't succeed, then skydiving definitely isn't for you."
The fourcc code us YUY2. If you attach a frame or if you can give more specific information about this raw data and how it is produced and someone might look at is out of curiousity, anonymously. It is at least interesting as a question. Stephen
Hello, @scmccarthy: i gave information about this format in my first question. a PAL frame (720x576): row1: Y001 U001 Y002 V001 Y003 U002 Y004 V002 ... Y719 U360 Y720 V360 row2: Y721 U361 Y722 V361 ... etc. it is produced with an Abekas 8150 production switcher () or with a photoshop in-/export plugin: Hans Runz
@ hanfrnz it's very easy tocode. you just need to read YUY2 from yourfile and memcpy to new frames. BTW, AFAIK, TMPEGEnc only works with RGB24 input. so you'll need a convertion. if you attach a file with 2-3 320x240 frames i can code that in speed. <EDIT> not a good idea : i would't be able upload it :-/ so i've just 2 question : - how are consecutive frames stored ? (behind eachother i thing) - is there any header, or you need to enter size manualy.
@ MarkFD answer 1: each frame is stored in a seperate file! like t001.yuv, t002.yuv ... answer 2: there is no header in these .yuv files and the resolution is always 720x576 (PAL) and 720x480 (NTSC) is there a mpeg2 coder which uses yuv2 input? I always thought an mpeg2 file is yuv (4:2:0) or is it rgb??? <edit> if you give me your email-adress i can send you a sample image. i tried to attach it to this reply, but is doesn work... hans runz
>answer 1: each frame is stored in a seperate file! like t001.yuv, >t002.yuv ... >answer 2: there is no header in these .yuv files and the resolution is >always 720x576 (PAL) and 720x480 (NTSC) the source filter will need 4 args (path,maxframes,width,height) and will output YUY2. that's okay ? >is there a mpeg2 coder which uses yuv2 input? I always thought an mpeg2 >file is yuv (4:2:0) or is it rgb??? MPEG is YV12 (4:2:0) YV12 support is recent in Avisynth (2.5 alpha), before we needed to upsample YV12 to YUY2.