@mccrash the h264 file needs to be a.mpv (elementary stream ie without a container). There wasnt much use for elementary mpeg-4 streams til now so nothing really makes them (there isnt really much playback support).

I have a feeling proper demuxers will start sprouting up soon. [ps] hopefully more mpeg-4 encoders will start including a program stream output optionthe file from Malow is a.264 elementary stream (btw.264 is the only official extension for raw avc streams, as created by the reference encoder) its easy to create such elementary streams, be it from mpg ts or ps or mp4 with most free tools widely used (ffmpeg, mplayer, mp4box, mp4creator.) you can also create elementary streams with x264. @mcrash you dont replace those numbers. You have to insert the numbers in front (so you would have 00 00 00 01 09 10 00 00 00 01 41 9a.) your hex editer has to have an insert function.those are the access unit delimiters i was talking about. At the start and between every frame there needs to be this AUD when placing avc in mpg ps or ts when placing avc in other containers, like mp4 or mkv, you dont need those AUDs, therefore not every avc stream/encoder has those in x264 you can enable/disabling writing AUDs into the stream Is x264 unable to produce such a stream?it is surely able to produce such a stream. Caloric Ultra Ray Manual. The only problem is we dont know the hddvd specs to know what settings are allowed and which ones arent. PS: a vc1 elementary stream would be nice also, if that ever existed.there is none.

Sonic scenarist accept only elementary stream: -.264,.h264. For MPEG4 AVC -.m2v,.m2p. For MPEG2 -.vc1 or something like that. For VC1 x264 produce perfect compliant stream but there are not good information in the bitstream. Actualy you can't use x264 for HDDVD/BD multiplexing.:-(.

Simply demuxing the wmv9 video stream from.wmv to raw doesnt seem to produce a useable stream (at least till now). If scenarist accept the resulting stream that's enough for me for the moment. So i demuxed a transport stream and then used avc2avi but x264 does not want to use the avi as input. It says 'avis [error]: unsupported input format (h264)':eek: avi should not be used with avc anymore and hardly anyone does so can you play the original.ts file in a directshow player? If no, install the haali media splitter and ffdshow. Create an.avs script loading the.ts file (no need to demux to raw.264) feed this.avs into x264 and encode to.264.

You can try to use megui for encoding with x264 which includes a hddvd profile that should make x264 use the features supported by hddvd try to feed the resulting.264 to scenarist. Well well, after a few attempts with x264 and megui, i decided to give mainconcept encoder a try.

The encoder picks up directshow but it crashes on my TS stream when i go into the advanced settings, so i gave it my h264 avi made with avc2avi instead and encoded it to elementary stream (mpv) with high profile. The resulting mpv was not accepted by scenarist on my first attempt because of a gop size problem that had to be 30 or lower. So i decreased the keyframe interval in the advanced settings to 15 and voila, scenarist loaded the file. I then builded the project and a nice HDDVD_TS folder was born. The resulting mpv was not accepted by scenarist on my first attempt because of a gop size problem that had to be 30 or lower. So i decreased the keyframe interval in the advanced settings to 15 and voila, scenarist loaded the file.. Thats the difference between the other files.

Files out put by mainconcept are being fully scanned at least(which takes a few seconds for these very short files). Then it will either say accepted, or it will tell you why it wasnt. With the other raw files or h264 files or ts files it is failing immediately. Seems it is reading a few bytes, decides it has no clue what its reading and gives up. Well well, after a few attempts with x264 and megui, i decided to give mainconcept encoder a try.

The encoder picks up directshow but it crashes on my TS stream when i go into the advanced settings, so i gave it my h264 avi made with avc2avi instead and encoded it to elementary stream (mpv) with high profile. The resulting mpv was not accepted by scenarist on my first attempt because of a gop size problem that had to be 30 or lower. So i decreased the keyframe interval in the advanced settings to 15 and voila, scenarist loaded the file. I then builded the project and a nice HDDVD_TS folder was born.how did you encode with x264? Did you take into account what cruncher wrote? How did you encode with x264? Did you take into account what cruncher wrote?

But x264 don't produce HDDVD/BD compliant elementary stream because: - vbv work correctly only in CBR mode for x264, vbv in multipass mode fail and crf mode too. - x264 don't write HRD parameters in the bitstream. Actually you can't use x264 for make HDDVD/BD compliant stream.:-( yeah. I looked at my x264 output and then the mainconcept output in the hex viewer.

And heres what i see.Reference h264 es (mainconcept) 0x00000001 gop header 09 10?? At the beginning of reference elementary h264 stream - no other data in header 67 64?? Second header in reference elementary h264 stream - 40 bytes data 68 e8??

Third header in reference stream 4 bytes 06 00?? Fourth header in reference stream 18 bytes 65 88?? Fifth header in reference stream large (picture header?) picture 09 30?? After large segment (0x65) - no other data 06 01?? 14 bytes 41 9a?? Large (picture?) or picture 09 30?? After large segment (0x65) - no other data 06 01??

14 bytes 01 9e?? Large (picture?) above headers repeat AVC I made test_avc_video.raw (x264 demulltiplex with yamb) 0x00000001 gop 06 05?? At the beginning. Contains open text of the encoder used and parameters used 67 4d?? 21 bytes 68 ce??

3 bytes 65 88?? Data segment (picture?) picture 41 9a??

Data segment (picture?) or picture 01 9e?? Data segment (picture?) repeat above (either 41 or 01)the x264 has the extra 0x06 header (which is just text of the settings used) which wasnt in the elementary stream. I dont think that hurts anything. But its missing headers info in the group header area and missing headers in the picture area. Im pretty sure once what is missing is under stood, that the file can just be updated with the extra header info and a reencoding would not be necessary. This post describes some of whats missing.

Thats what i get for not keeping up with x264 threads.:rolleyes: i tried to hex edit the type access unit info (0x09) with search and replace into my entire test file and didnt help, so it appears i need the sei (0x06) info too. The sei data is a tad more meaningful than the access unit info i dont know if i will have time to really figure it out or even if the data is recreatable from the original stream. So still hoping i find something that already exists or someone else makes it. (or support is added to x264:) ).

Hi all, First, x264 can not make compliant H264 file (play fine on HD DVD player from Toshiba or X360). There is some problem with VBV, SEI. I test it the last summer (you can see my post this forum and x264 developper known it). Second, there is lot of H264 software compliant with HD DVD: - Mainconcept software (because Sonic with CineVision use Mainconcept SDK) - Ateme software because you have access at all options - KDDI software (KDDI company work with MemoryTech) - Tiger AVC from Thomson (developped for HDDVD and BD) - Fraunhofer software (all H264 options) Third, there is only, now, PEP encoder from Microsoft which can make compliant VC1 video file.

All other solutions from Microsoft can not make compliant VC1 file. All VC1 video file in HD DVD sell on the market was encoded with Microsoft software (may be 1/100 or 2/100 was made with Inlet tool, I'm not sure).

Last, all HD DVD on market (HD DVD with HDi function) was made with MemoryTech software (98/100, may be 99/100). There were lot of problem with Sonic tool (correct now) to do compliant HD DVD. To test the compliance of video file you can use DCA software. And to test HD DVD disk compliance, you can use MemoryTech/Toshiba software. Here the compliant setting for HDDVD/BD compliant stream: HDDVD use MPEG4 AVC HP@L4.1: - for the profil: CABAC, 8*8dct, inloop, CQM, wpred, Pref, Bref - for the level: Max GOP at 0.6006 sec, max adaptative bframe at 2, max reference frame at 4 for P and 3 for B, max motion vector lenght at 512 pixels, max bitrate at 29.4 Mbps, cpb at 30000 Kbits Sonic scenarist accept only elementary stream: -.264,.h264. For MPEG4 AVC -.m2v,.m2p. For MPEG2 -.vc1 or something like that.

For VC1 x264 produce perfect compliant stream but there are not good information in the bitstream. Actualy you can't use x264 for HDDVD/BD multiplexing.:-(. So I've done some test encodes and I've noticed that playback in windvd is choppy and weird.

When I mux to mp4 it's fine though (as opposed to scenarist). I wonder if it's windvd or scenarist. Edit: Not sure if its cause it was a 24 fps source and I converted it to 29.97 or what. 25 fps h264 (reencoded) from european broadcast is perfect at 1080p. Edit 2: 29.97 source worked fine. =/ Edit 3: Ok the original source I was talking about worked better with tomsmocomp plugin instead of anything else. I need to try a whole movie though before I can comment for sure.

So I've done some test encodes and I've noticed that playback in windvd is choppy and weird. When I mux to mp4 it's fine though (as opposed to scenarist). I wonder if it's windvd or scenarist. I've just encoded some content with MainConcept 2.1, and finally managed to get all the settings correct for a succesful import into Scenarist 4.01 SCA. My footage was captured from NTSC Digi, and so encoded as 720x480 video, with an average of 6Mbps and a max of 8Mbps. Audio was 192Kbps AC-3 Files play back fine on the PC, as elementary streams, but I get the weird choppy playback like chadamir when playing back on a Toshiba HD-A1 with latest firmware.

I've done two encodes to rule out a field dominance issue. Bitrate is low enough that it shouldn't be causing problems.

Could Scenarist be only fully compatible with the Cinevision encoder? Could this be some kind of funky muxing issue with Scenarist?

Any other thoughts? What frame rate?

If the footage is interlaced do not deinterlace it, rather encode it as interlaced. The footage was 29.97, interlaced so I tried both top field and bottom field dominance on the encoding. No De-Interlacing. What did you used to burn?

In which burning mode? I used Nero Burning Rom >DVD VIDEO with a dummy IFO / BUP in the VIDEO_TS folder. I'm not in work now, but I will try the process you listed on tuesday. I'm running Scenarist 4.01, and I'm wondering if there is an update because this does seem like some strange muxing bug.

Does anyone know what the latest version is? Here if you want a 1080p HDDVD image demo at only 6Mbps: Encoding with H264 (Elecard beta) and quality is really high (more than 47 dB for OPSNR) Audio is simple DD at 448 Kbps. Mux by Scenarist (but not my work for muxing part. I don't have Scenarist).

With H264 it will be possible to put compliant HDDVD structure file on simple DVD9 with very high quality ('HDDVD like quality'): + Video: H264 1080p at 6-8 Mbps + Audio: Multiple DD 5.1 at 448 Kbps or DD+ 7.1 at 640 Kbps + avdanced HDDVD menu, chapters, subtitles. Here if you want a 1080p HDDVD image demo at only 6Mbps: Encoding with H264 (Elecard beta) and quality is really high (more than 47 dB for OPSNR) Audio is simple DD at 448 Kbps. Mux by Scenarist (but not my work for muxing part.

I don't have Scenarist). With H264 it will be possible to put compliant HDDVD structure file on simple DVD9 with very high quality ('HDDVD like quality'): + Video: H264 1080p at 6-8 Mbps + Audio: Multiple DD 5.1 at 448 Kbps or DD+ 7.1 at 640 Kbps + avdanced HDDVD menu, chapters, subtitles I cannot extract image files to my hard drive:confused: file corrupt? Or it's me (downloaded twice) Edited: I was using wrong soft,:). Hi Sagittaire, Thanks for the sample. It plays fine on the Xbox 360 Addon and on PDVD 6.5. This is really cool. You brought us big step forward in HD-DVD authoring.

Are the encoding parameters those you have posted earlier? MPEG4 AVC HP@L4.1: - for the profil: CABAC, 8*8dct, inloop, wpred, Pref - for the level: Max GOP at 0.6006 sec (15 for PAL), max adaptative bframe at 2, max reference frame at 4 for P and 0 for B, max motion vector lenght at 512 pixels, max bitrate at 20.0 Mbps, cpb at 14745 Kbits Highest possible quality mode for Elecard (done same metric result than x264 at max quality). I think that quality is really good for only 6 Mbps (IMO really better than HDTV quality) and this movie trailer is really difficult source. Scenaruist says: Error: AVC profile or level or their combination are wrong.:mad: Malow, I get this same error with anything encoded by Elecard except 720p59.940. UPDATE: OK, so I switched to MainConcept 2.1 and am getting much better results - encodings of 480p and 1080p film sources with 3:2 pulldown applied by the encoder for a target frame rate of 29.97 are now accepted by Scenarist. I encoded to 6000 kbps (max 24000 kbps), High profile at Level 4.1, max GOP 12, 4 reference frames, 2 B-frames, and NO B-slices as references. Strangely, Mainconcept sets the VBV buffer size (bytes!) to 11,718,656; that can't be right, can it - it must really be using bits, not bytes?

Here is where i am. The program parses a raw mpeg4 avc now (like what mpeg4ip does except without failing at the first picture) it will also recreate the same avc exactly (using the parsed info, not straight bit copying). Although this is assuming it only has the data that mpeg4ip parses ( aud sps pps sei(00,01) ) and also the slice header info. Although i dont decode the picture itself (of course). That is blind copied. The h.264.es.mpv is recreated perfectly with the parsed info. I then used a x264 avc i made and used the tool to put in a proper AUD and dummy HRD and sei values and write them in aud-sps-pps-sei-slice order.

(i encoded my avc with the wrong level so i forced it from 51 to 41 ). The file is excepted now although it complains about gop being too big. So i need to make and find some example streams with proper gop structure but improper headers and see if my fixed version plays back ok once compiled this is still early on and ultimately it may be better just to encode it right in the first place as (with the HRD-x264 version), as i stated. The gop length has to be small (.6 secs ).

Also forcing the level value may have unwanted concequences if it wasnt set correctly during the original decode. (hopefully it doesnt) anyways. Will keep posted. *ps* the reason mpeg4ip and my tool only do the aud sps pps sei and slice is thats all thats in most streams. If i can find out info about the other types i may add it in but for now it doesnt seem needed. (the reference stream is accepted by the reference compiler).

Btw i dont write the sei 05 (it is freeform text, x264 uses it basically to write text of your command line )to the output stream as it doesnt seem needed. It is not in the reference stream.

I have done a test encoding of my ******** HD recording with x264. Converting the Original ts (h264) to HD-DVD compatible h264 and syncing with the DVD was a real pain in the ass. Synced it with the DVD so I was getting the Dolby Track from the DVD. Finally putted it all together in Scenarist and created an ISO.

Here it is (50MB, the first 50seconds in 1080, 25fps) *** Removed the Link, if U wanna Test, PM me *** Since its only the 20thCF and Lucas Intro i hope its not a problem this posting here. Can somebody with PowerDVD or an Toshiba A2/E1 test this? Especially because its being said, that the A2/E1 don´t support 25fps ATM. @Chefkoch_ico I tested your iso in the xbox360 and powerdvd and it worked fine. Can you tell me exactly how you built the iso in Scenarist since my iso locks up my 360? Thx for testing.

Good to hear. Interesting, a user over on avsforums tested on a Toshiba E1 and the disc didn´t even load.

------- This disc is not DVD format. Cannot play the disc. ------- He burned MPEG2 HD-DVDs with Ulead Movie factory before.

So I think he knows, what he is doing. Burned it was with ImgBurn 2.1. Also he stated, that with PowerDVD-HD 6.5 the image looks distorted. Oberon, which Version of PowerDVD did U test? Maybe another explanation could be, that u should use the xBox compatibility flag on burning in Nero. Or a 29,7 fps source could do the trick.

I am using mainconcept to encode some 720p [59.94] material that I shot using a panasonic varicam. I exported a quicktime using the AJA Kona 10-bit uncompressed codec. This file makes mainconcept crash, so, I converted it to a super-high bitrate h.264 which mainconcept accepted.

After encoding this 3.5 minute clip using a couple different settings. They import into scenarist 4.1 just fine, but when I multiplex and play the EVO's in powerDVD 6.5, they play choppy--like skipping around weird. I went back and played the AVC file that mainconcept encoded in windows media player and that doesnt look completely right either. Not skipping around like the EVO but slower and seems like its missing frames. I am using mainconcept to encode some 720p [59.94] material that I shot using a panasonic varicam. I exported a quicktime using the AJA Kona 10-bit uncompressed codec. This file makes mainconcept crash, so, I converted it to a super-high bitrate h.264 which mainconcept accepted.

After encoding this 3.5 minute clip using a couple different settings. They import into scenarist 4.1 just fine, but when I multiplex and play the EVO's in powerDVD 6.5, they play choppy--like skipping around weird. I went back and played the AVC file that mainconcept encoded in windows media player and that doesnt look completely right either. Not skipping around like the EVO but slower and seems like its missing frames. Was it filmed at that framerate? Did you try it in a hardware player?

Interresting: I have 2 h264 Files. The first created with x264, the second one with Elecard Converter Studio beta. Both files created from the same AVS with same resolution and Bitrate.

Both load fine in Scenarist. But when I play the files in PowerDVD HD 7.2, the x264 file plays very choppy and CPU-usage is 100%. The Elecard one plays fine with 70% CPU usage.

Far more interesting: When I author the x264 file in Scenarist and afterwards play the authored content, its not choppy anymore and its behaving like the Elecard encoded file concerning CPU usage. Hi, I´ve downloaded the Mainconcept demo but still only get choppy results or bad clip length with 24fps vc-1 sources via avs. Sagittaire, could you please post the settings you used with Mainconcept to feed it with a 23.976fps source and get a perfect authored HDDVD? Would be greatly appreciated. Has someone had any positive results with re-encoding 24fps vc-1 sources? Nas Hello Nasgul, I've experienced exactly the same problem encoding film sources with Mainconcept 2.1 (demo version). Choppy playback (using PowerDVD Ultra) and clip length is wrong (too long).

Sagittaire, please enlighten us!:o. Hi, Here little sample for test vbv limit with your Toshiba SAP's (I don't know the buffer limit and it's important for quality) Haven´t had much time today but I just did a quick burn of the three 1080p samples and put them in the Toshiba. The first two (1100+1715) play fine.

The third (3450) had a strange discoloration in the tunnel scene while playing for the first time, but I couldn´t reproduce it playing it again. It may have to do with the rewritable I used (was another brand), I have to check that again tomorrow/today. Apart from that the 3450 also played fine.

They all look fantastic btw! You can find the setting for HDDVD profil in the rar files with the sample. Will try that tomorrow. Scenarist doesn´t complain about the 23.97fps elementaries??? I´ve tried that before, but only got 'Invalid framerate'.

Also, when trying to set the pulldown 3:2 on 23.97fps source, my Mainconcept demo gives me an error saying the target fps has to be 29.97, 30, 59.94 or 60. The screens look different, too, I guess I have a different version.

Hi, Sagittaire, I´ve just tried your 480p sample - works as well in the Toshiba (the picture is quite soft, though). I also tried your HDDVD settings with an older Mainconcept demo I found (2.0.15) and it almost works now. At least it accepts the pulldown flags with the 23.976fps target and loads afterwards. The playback in the Toshiba is ALMOST smooth, but I still have slight stuttering and the audio (tried a dd+ track) shuts down after about 4-5 seconds. Sagittaire, could you do me a very big favour and try your script/settings/scenarist combination with a 24fps VC-1 source? Here´s a sample video (and audio (That would be great!

MPEG4 AVC HP@L4.1: - for the profil: CABAC, 8*8dct, inloop, wpred, Pref - for the level: Max GOP at 0.6006 sec (15 for PAL), max adaptative bframe at 2, max reference frame at 4 for P and 0 for B, max motion vector lenght at 512 pixels, max bitrate at 20.0 Mbps, cpb at 14745 Kbits Highest possible quality mode for Elecard (done same metric result than x264 at max quality). I think that quality is really good for only 6 Mbps (IMO really better than HDTV quality) and this movie trailer is really difficult source. Anyway you can post screenshots of your config pages in elecard? I am trying to interpret these soecs and I can quite figure it out! I have been doing some tests with varied results.

I have been testing 1080p QuickTime files, encoded with Mainconcept's H264 encoder and then muxed with Scenarist 4.1 SCA. Initially I tried using Sagittaire's Mainconcept settings in the demo of version 2.1, but 2.1 will not allow a 3.2 pulldown to be applied to a 23.976 source! On closer inspection of Sagittaires's settings and screen shots, it appears he is using version 2.0 as the layout matches and it is possible to apply the pulldown. In 2.1 I just select 29.97 as the output frame rate and it produces a stream that plays smoothly and has the same running time (is it applying pulldown automatically?) Both 2.0 and 2.1 produce streams that Scenarist is happy to accept although it refuses to mux streams with constant bitrates higher than about 10,000 kbps! Throwing the error: ErrorVideo or Audio Buffer underflow: (Dts: 1634865, SCR 1634883) Errordvd_mux: DoMux Multiplexing Error ErrorTerminated Multiplexing (GrindHouse-300-t_t.evo).

Error Total bitrate is too HIGH. Please reduce the stream bitrate or the number of stream. Error Multiplex is failed. ErrorMultiplexer failed '300-t' One trailer for the Grindhouse will not accept anything above about 6,000 kbps. Is this being caused by my VBV? It has me a little confused. I need to invest in a good h264 encoder and seeing as the Mainconcept one is several thousand pounds cheaper than the Elecard one, I would obviously prefer to take that route.

I just want to make sure that it is fully compatable at this stage. Oh I almost forgot. Jumping between VTSs appears to freeze a DVD5 test disc on an Xbox360 HD drive. @ Trahald The problem is 2.1 throws an error in the right hand pane and if I try to apply pulldown to a 23.976 or 24 source.

Only 2.0 seems to allow it for me. @ Sergey Are you saying that my muxing error is caused by changing the framerate without using pulldown. Thus corrupting the buffer? Could pulldown problem be caused by the source?

Nope, I said that changing SEI messages in existent AVC stream will broke buffer (as timing has been changed after encoding). Changing frame rate for source before encoding has no effect on buffer. This will only change output frame rate. Any idea what the muxing problem might be caused? I don't know what I should be setting the VBV to.

All I know is that it changes the DTS and SCR figures within the Scenarist error: Error Video or Audio Buffer underflow: (Dts: 1634865, SCR 1634883) depending on how high I set it. I am also still getting the error: C045:H.264 Validation Error: Invalid target frame_rate/pulldown combination.

Must be 29.97, 30, 59.94, or 60. Every time I try to set pulldown on a 23.976 source Very Confused!:confused: My company is ready to dip it's toe into the HD arena. We can produce compliant MPEG2 files with procoder, but we would like to move over to the more efficient H264. Untill we can see it working properly, that just isn't going to happen:(. Any idea what the muxing problem might be caused? I don't know what I should be setting the VBV to. All I know is that it changes the DTS and SCR figures within the Scenarist error: Error Video or Audio Buffer underflow: (Dts: 1634865, SCR 1634883) depending on how high I set it.

I am also still getting the error: C045:H.264 Validation Error: Invalid target frame_rate/pulldown combination. Must be 29.97, 30, 59.94, or 60. Every time I try to set pulldown on a 23.976 source Very Confused!:confused: My company is ready to dip it's toe into the HD arena.

We can produce compliant MPEG2 files with procoder, but we would like to move over to the more efficient H264. Untill we can see it working properly, that just isn't going to happen:( if this is commercial interest I don't think that this is a good place to discuss your problems. If you are intersted in our program and have a problems you can contact our techsupport for the answers. I used 2.1 and i set turn pulldown on in hte advance settings->misc->general->film mode framerate conv 3:2. This will apply pulldown to the timecode sei..the tool i made will add 3:2 to an existing file but slow as molasses.

Doing some speedups first and i'll make it available If you make that you must rewrite HRD parameters in the stream. IMO the best way is pehaps to make complete flags rewrite tools: ->input is basic H264 raw stream - tool write Sequence End Code flag - tool write SEI flag - tool write HRD flag - tool write pulldown flag etc etc etc ->output is compliant HDDVD/BD stream. If you make that you must rewrite HRD parameters in the stream. IMO the best way is pehaps to make complete flags rewrite tools: ->input is basic H264 raw stream - tool write Sequence End Code flag - tool write SEI flag - tool write HRD flag - tool write pulldown flag etc etc etc ->output is compliant HDDVD/BD stream wrong. Encoder was used given buffer model with given timing - it will be a luck to change these parameters and still keep buffer model correct.

No one could give a guarantee for this without buffer analyzing and adjusting HRD parameters for new timing. Sorry Sergey, I didn't mean to be rude. I am just used to solving my video related problems on my favourite video related forum:) I have just gone to the mainconcept website and sent an email to technical support.

Hopefully between the two places I will get an answer to my problems. At least before my boss tears me a new one! @Sagittaire Can you confirm which version of the encoder you have been using? Also as i see you too have been using QT trailers, have you managed any at high constant bitrates, Specifically in the 20,000 kbps range? Thanks again guys. Ok, update: I have tested the same source material with some variable encodes. I can push the max bitrate all the way to 29 mbps without problems, but the average cannot pass the same limits I was restricted to in the constant bitrate encodes.

The curious thing is, the limitation varies between sources. It seems to be between 6000 and 10000 kbps (at least on everything I have checked) I am going to test out the Elecard demo today and see if the same restrictions apply. Can anyone else confirm this bitrate limitation issue? Ok, update: I have tested the same source material with some variable encodes.

I can push the max bitrate all the way to 29 mbps without problems, but the average cannot pass the same limits I was restricted to in the constant bitrate encodes. The curious thing is, the limitation varies between sources. It seems to be between 6000 and 10000 kbps (at least on everything I have checked) I am going to test out the Elecard demo today and see if the same restrictions apply. Can anyone else confirm this bitrate limitation issue?

Try with differents values for the buffer. 1 150 000 bytes 1 843 000 bytes 3 750 000 bytes. Sagittaire, can you tell me how you worked out the correct VBV values? The default is set to 11,718656 which is so drastically different it's not funny.

As I will be purchasing this encoder for commerial purposes, I want to make sure I don't get caught out like this again. 1 150 000 bytes is the official value for HDDVD with MPEG2 MP@HL 1 843 000 bytes is the official value for HDDVD with VC1 AP@L3 3 750 000 bytes is the official value for HDDVD with H264 HP@4.1 I don't know the particular value for sonic authoring but 1150000 must work in all case. Higher buffer values imply potentially higher values for local peak bitrate and more constant quality in complexe part. It's particulary true if your average bitrate is close to your max bitrate. If average bitrate = max bitrate and if buffer = 0 then Rate Control will be strict CBR If average bitrate = max bitrate and if buffer = infinite then Rate Control will be strict VBR. Well that's interesting.

Scenarist seems to be extremely fussy with what I give it. At a constant bitrate of 29mbps (The highest limit available to AVC within the HD-DVD spec) A Mainconcept buffer of; 1,000.000 bytes causes Scenarist to throw a muxing error 750,000 is perfectly ok though 300,000 or lower causes the encoder to crash. Therefore Mainconcept's default buffer of 11,718.656 really is well off the mark. Thanks for your help Sagittaire. The only thing still puzzling me is the issue of Pulldown though. It's clearly working for you in version 2.0. Has anyone got it working in version 2.1 or is it broken.

I still get the error: C045:H.264 Validation Error: Invalid target frame_rate/pulldown combination. Must be 29.97, 30, 59.94, or 60. Every time I try to set pulldown on a 23.976 source. The only thing still puzzling me is the issue of Pulldown though.

It's clearly working for you in version 2.0. Has anyone got it working in version 2.1 or is it broken. I still get the error: C045:H.264 Validation Error: Invalid target frame_rate/pulldown combination. Must be 29.97, 30, 59.94, or 60. Every time I try to set pulldown on a 23.976 source I still always have exactly the same problem with 2.1. Sergey, since you work for MainConcept, surely you can shed some light on this, can't you?

Are we doing something wrong, or is there a bug in the program? What a relief. Confirmation that I'm not going mad:) I've tried the 2.0 demo and that seems to work ok. The thing is, I think the pulldown is working automatically in 2.1 if you just enter an output framerate of 29.97, the resulting file is the correct length and seems to play back smoothly on my XBOX360 HD-DVD drive.

My hesitation is the fact I'm not so use to working with pulldown due to the fact I come from PAL land. Can you shed some light on this for us Sergey? I did send an email to Mainconcept technical support as you requested on tuesday, however there has been no response! EDIT; Sorry Sergey, you were posting at the same time as me. Are we to assume 2.1 is broken then?? Pulldown worked fine in 2.0. What a relief.

Confirmation that I'm not going mad:) I've tried the 2.0 demo and that seems to work ok. The thing is, I think the pulldown is working automatically in 2.1 if you just enter an output framerate of 29.97, the resulting file is the correct length and seems to play back smoothly on my XBOX360 HD-DVD drive. My hesitation is the fact I'm not so use to working with pulldown due to the fact I come from PAL land.

Can you shed some light on this for us Sergey? I did send an email to Mainconcept technical support as you requested on tuesday, however there has been no response! EDIT; Sorry Sergey, you were posting at the same time as me. Are we to assume 2.1 is broken then?? Pulldown worked fine in 2.0. Seems like it was a luck that 2.0 worked with pulldown fine. Cause there was no such functionality.

We will check this and probably update application soon (this depends on current priorities as usual, so be patient please). Are you sure??

The way I understood it, assumefps only speeds up the footage. Apply pulldown will then set a flag to remove certain frames on playback.

Would that not really mess with the output? I envision something that is then too fast and stuttering. However if we use changefps it will duplicate certain frames. Applying the pulldown flags will tell the playback unit to remove the dupes. Or have I got that all the wrong way round?:D assumefps just indicates specified frame rate - that is what is needed for workaround atm (works fine here). Pulldown flags indicate how to show frames - either as two fields or as three fields.

If you want to see your encodings on progressive display you dont need these flags - just encode as 23.976 and you'll see progressive playback on your PC display. If you are targeting TV or other interlaced device, then using pulldown flags you're telling to decoder to duplicate some fields to see kinda interlaced 29.97. Seems you have to read a bit about 3:2 pulldown. Ha ha, yes like I said I had it all confused:D I am used to working with nice easy PAL framerates, but seeing as many HDDVD players are still only able to playback at NTSC framerates, I foresee having to work with it for a while. I do understand the principle of pulldown, but I have clearly got myself confused using a combination of filters and flags in this situation. Thanks for setting me straight Sergey.

I just have one last question if you can bear it. I just want to make sure I have got all of this straight.

My source footage is 23.976fps progressive, but Scenarist will only accept 25fps or 29.97fps. Therefore, do I still use your suggested method of assumefps followed by pulldown, even though it is clearly not interlaced material.

Or am I still on the wrong track. Thanks again.

@Clown shoes: All you need to do is change fps with avisynth like you said to 29.97 and and use MainConcept to encode at 29.97, but do NOT set the 3:2 pulldown setting. You definitely do NOT want to apply pulldown on top of a video that has already been adjusted to 29.97 fps, because then you will end up with a video that plays back too many frames! I should warn you, however, that even though Scenarist will accept a 29.97 fps progressive stream, the authored.EVO may not play back smoothly in your HD DVD player. I've had problems with these progressive streams in PowerDVD 7.2 Ultra. Better to try a few test clips for yourself and see what happens. That's what I'm doing now. I tried encoding the way Sergey suggested with assumefps and 3:2 pulldown and guess what?

Just like I said, it played too fast and stuttered. I should have trusted my initial instincts!

EDIT; Actually I've just been thinking about it and I don't think that is the best way either. If the source is 23.976fps and I use changefps it will add duplicate frames which will cause stuttering.

If I use assumefps that will speed it up. No good also. Finally if I use convertfps it will blend frames. I don't want to make any changes to the frames I have, so none of these solutions are any good.

Hang on, my brain is starting to hurt here. If you've got a feature film with a framerate of 23.976fps how do you encode this through mainconcept so that it is HD DVD compatible and will drop into Scenarist? - AssumeFPS specifies a framerate, so I can understand how this works if you have a 25fps film source and you want to 'recover' the original framerate, but how does 'speeding up' a film help. - Sergey, are you saying that the Pulldown 3:2 / 2:3 options under misc within MainConcept 2.x do not work???

My understanding was that you feed MainConcept a 23.976fps source file, tell MainConcept to produce a 29.97fps progressive file which has 3:2 pulldown set under misc, so that the video plays back 29.97, but if the display supports it, then the player can IVTC the footage back to 23.976. But to do this the footage needs to be interlaced, not progressive. Isn't that how 3:2 pulldown works? Why don't you then feed a 23.976fps file, and then produce a 29.97 interlaced file with perfect 3:2 cadence generated at encoding time.

Isn't that how CCE works? Interestingly though, I also encoded the same clip without changefps and 3:2 pulldown and just selected 29.97fps as the output and that appears to play fine as well. Can anyone confirm this? Try checking the playback time - is it the same as the original video?

Does the audio and video stay in sync through out the whole movie? BTW, which player are you using? EDIT: Also, I'm pretty sure that if you leave the source as 23.976, and simply allow the MainConcept encoder to change the FPS to 29.97, it is just doing the same thing as using the avisynth changefps function. You can verify this by stepping through the encoded video frame by frame: you will see that the doubled frames differ ever so slightly from each other, meaning that they were doubled prior to encoding. Try checking the playback time - is it the same as the original video? Does the audio and video stay in sync through out the whole movie?

BTW, which player are you using? EDIT: Also, I'm pretty sure that if you leave the source as 23.976, and simply allow the MainConcept encoder to change the FPS to 29.97, it is just doing the same thing as using the avisynth changefps function. You can verify this by stepping through the encoded video frame by frame: you will see that the doubled frames differ ever so slightly from each other, meaning that they were doubled prior to encoding. You are right, it is exactly as you said. Changing frame rate on property pages will change source clip frame rate. I'll try to investigate video+audio encoding a bit later.

Hey DVDBoy, I think what you meant to say was H264 broadcasts off of the TV;) Yes your script needs to have assumefps(29.97) at the end of it. I know this seems odd, as it means the video will play back too fast, but it seems to be the current fix for a bug in Mainconcept. It will not allow you to apply 3:2 pulldown to most non 29.97fps material. I say most because it does appear to work with some, but I guess it's just trial and error. But in a nutshell, yes what you are doing looks the same as my method and it seems to be working great.:D.

I'm currently seeing how well Scenarist behaves with video which I encode in chunks but sync to one audio track. It can be done in Scenarist but it is a bit of a pain.

Say if you have three parts, you can use the multi file import option, but you must take care to ensure the first two parts do not have sequence end codes, other wise your video will probably stop mysteriously when it reaches the join. That would be the only way to apply one audio track to a peice of video cut into several chunks (that I know of at least) Is there a reason you have to encode the chunks seperately?

Would it not be easier to join them in AVISynth? I've got the source material as one long piece (90mins give or take).

As I said, Mainconcept seems to be playing silly buggers with the duration, although this is probably down to using AVISynth and parsing in an H264 file. Because it is listing the duration wrong, MC seems to crash when it reaches the end of the encode. 'Fine' on a single pass, but useless on a 2 pass. Because of this I'm currently bouncing the video back out to an AVI to feed MainConcept. Not ideal, but works and produces big files. So, part of the reason for wanting to feed Scenarist chunks was to get around converting the whole film to an AVi first (I've guess-timated about 400GB).

The other reason was so that I could allocate different bitrates to different peices, or re-encode a segment if I didn't like the results. As it is, it is looking a complete ball-ache to try and sync, so my next plan is to generate the AVI. Just need to archive some old projects off first. My process so far has been as follows, and I'm always open to suggestions: -.ts file run through xport to produce an.mpv (containing the H264 elementary stream) and.mpa (which is actually ac3). - DirectShowSource ('.mpv) - AssumeFps (23.976) - CropBottom (8) - ConvertYUV () Opening the AVS file up, both virtualdub and mainconcept report the wrong length.

I can play the file's back in PowerDVD 7.1 or MediaPlayerClassic and it's all there, but the duration is totally messed up. Hence my thinking if I bounce this out to a 'physical' AVI, I'll use more disc space (temporarily), but I don't then get an issue with misreading the running time. Sergey, this method does not appear to be working properly. I imported a raw.vc1 via avisynth (directshowsource) and used assumefps(29.97) I then applied 3:2 pulldown. The original source is 2.02.18.25, Mainconcept displays approx; 1.37.00.00 and the resulting output file has a running time of 2.01.56.12 The file appears to play back smoothly but the 22 second difference is alarming! Any idea what the cause for this might be?

What will happen to my 23.976 output file if I select 29.97 as my desired framerate and apply the 3:2 pulldown flag. Will I get duplicated frames? Sergey, this method does not appear to be working properly. I imported a raw.vc1 via avisynth (directshowsource) and used assumefps(29.97) I then applied 3:2 pulldown. The original source is 2.02.18.25, Mainconcept displays approx; 1.37.00.00 and the resulting output file has a running time of 2.01.56.12 The file appears to play back smoothly but the 22 second difference is alarming! Any idea what the cause for this might be?

What will happen to my 23.976 output file if I select 29.97 as my desired framerate and apply the 3:2 pulldown flag. Will I get duplicated frames? If you'll set target frame rate in H.264 encoder application different from source, then you'll have either duplicated frames (if target >source) or dropped frames (if target. @Sergey Do you have any idea why my output file might end up 20 seconds shorter than the source file for a 2 hour feature encode, using your workaround with assumefps(29.97)? The majority of the work I will be doing is going to be 23.976 sources for a local art gallery who are going HD. Therefore I really need to know this is going to work with what I throw at it.

Is there any news as to when a version of the encoder will be released with the pulldown bug fixed? Sorry to direct all this at you Sergey, but I never heard anything back from your tech department after emailing them last wednesday. The original source is 2.02.18.25 if this is in hh.mm.ss.ff format it is an invalid time for 23.976fps Mainconcept displays approx; 1.37.00.00 that is to be expected, the framerate change will shorten the runtime to 80%, the original was approximately 122 minutes, so 97 minutes is correct. Do you know if the times are in drop-frame or non-drop format? If 2.01.56.12 is the correct non-drop value that equals 219492 displayed frames, converting to drop-frame would add 218 - not enough to account for the difference, but I question the original run time value.

Btw - this is not new to HD, pulldown issues are the same as SD. In SD the most common way to do this is encode at 23.976 then apply pulldown which not only adds the flags but changes the framerate in the headers. Hi dvdboy, Ok, but why when I play the majority of my DVDs movies directly on the PC, using Media Player Classic, I see on the statistics panel of MPC a number very close to 24 for the frame-rate? EDIT (after some pulldown readings) TV1: interlaced CRT-based ( 15750 scan-lines capable ) (very old) TV2: progressive CRT-based ( 31500 scan-lines capable ) (old) TV3: progressive LCD-based ( 60 Hz ) DVD Player1: interlaced (old) DVD Player2: progressive The (old) DVD when filled with film content has always (or most of the time) 24 fps.

There is only a 'pulldown' flag in the DVD content to instruct old interlaced DVD players to adapt the video to be able to play that 24 fps content in old interlaced CRT-based displays. Progressive DVD players (as well as PC players) when connected on progressive displays (CRT or LCD) just ignore the 'pulldown' flag, and play at 24 fps. I understand things like NTSC or PAL were invented for that very old interlaced technology.

Today in the progressive and digital world, there is no need for NTSC or PAL. I can have a video in 27fps, 233 fps or anything I want and it should be OK.

Modern DVD players should be able to pass it, as any PC player can do it. Is that correct, or I totally misunderstood my readings and the point? So, why a modern product like Scenarist4 is not open for other framerates? Are the current specs for future DVD limited to that?

Don't ask me man, I only work here! I guess the easiest answer is that it's all to do with specs and standards and legacy format support. But let's look at this on the flip side - currently most if not all HD DVD & Blu-Ray players do not support PAL-based resolutions, so no 25fps, 50fps etc. You've got to remember that these are all based on aquisition formats, whether that means film (24fps), or video (29.97, 59.94, 25, 50fps).

Here's the framerates that Sonic Cinevision supports: * 59.94 * 50 * 29.97 * 25 * 24 * 23.98 * 3:2 pull-down insertion Which is more than DVD supported. If you are just working in an internet / computer based system, of course you've got more choice, but DVD and HD is inherently an offshoot of Broadcast, and therefore follows their specs and standards.