I posted this in another forum and thought i might share it here. I know people are aware that this free software can do the trick but this is my workflow.

I am finally able to get a great workflow with MPEG StreamCLip. I was originally using the Cineform Demo to do pulldown and found I did not like the large files. I heard about this free software and thought I would give it a try.

What I did was captured my footage with HDVSPLIT and worked on the timeline in Vegas 7 with all of my clips.

I then rendered the file to HDV (MainConcept MPEG-2, High Quality, top field first, constant bitrate of 25,000,000 and a frame rate of 29.970).

After starting MPEG_Stream Clip I opened up the rendered .m2t file from Vegas.

I did a file export to .AVI (you can also do it to a Quicktime file if you want)
In the menu I chose 1920x1080 (16:9)...

Selected the checkbox Deinterlace Video...

put in 24 for frame rate...

I chose Apple Photo-JPEG (you actually have plenty other codecs to choose from including H.264 encoder)

I left everything else in their normal state and clicked "Make AVI."

The video rendered out beautifully with no interlacing artifacts. The ghosting that I once seen in the original footage is gone. This is much cleaner han using VirtualDubMod (which did not do a good job IMHO).

The main purpose for me doing this was to find a way where i can continue to use a workflow that i was used to (especially coming from DV and working with smaller "native" files and then rendering out to a large "less compressed" file. Cineform is nice but like I said.... those files take up too much unecessary space. Grant it...it might work smoother on the timeline than native .m2t files (I don't know) but this is another viable solution. I hope someone finds this helpful.

Ian, how did you get the Apple Photo JPEG codec? Was it via regular quicktime or quicktime alternative? For some reason I was forced to use QT Alt and I was getting bad compression artifacts with the suggested codec. Of course, I wasn't doing the intermediate step of converting the HDV mpeg to main concept MPEG Vegas. That may have been the world of difference in final product.

How long does your workflow take you per minute of video? I'm still looking for the fastest turnaround method short of cineform.

Mike, I am using the quicktime alternative. When you go ino the settings for the codec I believe it defaults at 50%. You have to slide the ruler up to 100%. I think when I checked the two boxes (frame Blending and Better Downscaling) I got those same horrible artifacts you mentioned but when i left them unchecked the video came out smooth.

I am very pleased with the results so far. I have only been rendering out small 10 to 20 second takes but it is way better than VirtualDubMod. VDM seems like it is unable to remove the pulldown all of the time (too much ghosting). The quality is not quite as good as the raw file but it is darn close.

Rendering to .m2t seems very fast. I had only one track up in Vegas (around 20 sec long) and it took about a minute ort more to render. Same with the .AVI file in MPEG StreamClip.

Note: Cineform has a higher quality but like I said earlier...I was looking for an alternative. It's good to know that there is one.

Ian, thanks for the tip. I'll have to try that method again. I'm using the avisynth/virtualdub method right now (not just plain vdubmod) and it does pulldown removal 100% via DGIndex+TIVTC plugin. It takes a full overnight run to convert 1 HDV tape (63 minutes) with my current codec choices. Anything that could save time would help.

Ian,

from your workflow, it seems like you are importing the HV20 video to a 30 fps interlaced timeline, then output using the "Deinterlaced" option to 24 fps. That doesn't really export a true 24p output. You simply edit in 30fps and have the program decide to remove the extra frames so it becomes 24fps. Then the deinterlaced option also just interpolate the interlaced lines from the original 1080i video.

The HV20 24p video is stored in a 1080i wrapper. You will need to remove the pulldown to obtain a true 24p video then import a 24p timeline to edit. During export, you should just output to 24 fps progressive frames without the need of deinterlacing. Remember when you deinterlace a video, you are blending/interpolating dominate fields to fill the progressive frame.

I tried one clip using MPEGStreamClip. The workflow was simpler
1) capture with HDVSplit
2) Open the clip in MPEGStreamClip
3)Save as avi with deinterlace option clicked and the framerate as 23.976

The output does looks progressive and I could not see any difference in quality with the Avisynth/VirtualDubMod converted clip.

Oh one difference though. The total number of frames are different! (In my clip it was 685 with Avisynth, 697 with MPEGStreamclip).
I think I need to convert few more clips to conclude something!

Ian,

from your workflow, it seems like you are importing the HV20 video to a 30 fps interlaced timeline, then output using the "Deinterlaced" option to 24 fps. That doesn't really export a true 24p output. You simply edit in 30fps and have the program decide to remove the extra frames so it becomes 24fps. Then the deinterlaced option also just interpolate the interlaced lines from the original 1080i video.

The HV20 24p video is stored in a 1080i wrapper. You will need to remove the pulldown to obtain a true 24p video then import a 24p timeline to edit. During export, you should just output to 24 fps progressive frames without the need of deinterlacing. Remember when you deinterlace a video, you are blending/interpolating dominate fields to fill the progressive frame.Taky, I'm not sure if my method is a correct way of doing this but I see what you mean about the deinterlace. I am away from my NLE at the moment but I wonderif it would still work if I were to not check the deinterlace option.

If you notice I have all of my .m2t files in the Vegas timeline (edit as normal) and from there rendered the footage to the Main Concept .m2t file (I'm sure there is some loss here since it is a lossy format). I chose not the touch anything during this process (I left it in its 1080i state) because I thought I should save that step for last. From here I run the rest of the process through MPEGStreamClip.

I have not tried rendering without checking the deinterlace checkbox but I will try when I get home later. I ended up doing it all this way because I read in forums like this that MPEGStreamClip removes pulldown. i don't really see any option in the menus that says this but when I render everything out I noticed the ghosting dissapeared. I did not count the frames but I just looked at the final result and did an A&B comparison. If there is something better...PLEASE...let me know. I'd like to also mention that when you investigate the picture there is some evidence of macroblocking when it's blown up but it is not terrible.

Note: I have never had the oportunity to transfer video back to tape. I am assuming that is what this MainConcept format is mainly for. My question for anyone here is, when a file is rendered at a constant 25,000,000 bitrate back to the HDV format is there (or should there be) any perceivable loss in quality?

Also, I have even tried this with the Cinefom's HDConnect. I figured, since that software does not use flags to remove the pulldown, and it has the ability to convert .m2t files with pulldown removal "after" the fact, there should be no problem doing the same thing with a rendered .m2t file. But I had no luck. So far the MPEGStreamClip method seems to work for me...but its quality is not like Cineforms.

If there were a way to frameserve an .m2t file to another application (through Vegas) then I believe this would work best.