Since the people on "both sides" of this are very credible, I have got to think that there is a key difference in technique that is not being documented. I have also pondered, in another thread on this same subject, whether there is a difference in firmware between FX1 camcorders. I'm still trying to find information on whether we can display that information.

As to technique, I have these questions to ask:

1. FX1 footage ,hdv into vegas. vegas renders as best ntsc dv

What project settings are you using in Vegas? I'm talking about the project settings, NOT the render settings. There is a little icon in the upper left corner that will let you set this. Are you using NTSC DV (720x480) or are you using "HDV 1080-60i (1440x1080, 29.970 fps)" ? Also, when you say "best ntsc," I assume you mean that you choose the NTSC DV template, and then click on the Custom button and change "Video Rendering Quality" from the default "Good" to "Best." You definitely need to do that when changing resolution.

2. FX1 footage camera downconverted captured into vegas.

The manual states: "To scan pictures recorded in the HDV format on the tape in the DV format, set [VCR HDV/DV] to [HDV] (p. 59), and the [i.LINK CONV] to [ON] (p.60)."

I have the FX1, but it isn't here in the room with me. From memory, I don't think you can set the VCR HDV/DV setting to DV instead of HDV, but if that is possible, then that might very well cause a problem. Re-read the above instructions: you want to set the VCR to HDV, NOT DV. The i.LINK CONV is what does the conversion. However, this is probably moot, because I don't think the camera lets you set that control to DV when you are playing back an HDV tape.

3. pan and crop like a sob into the subject,

I assume you are doing this on the DV footage that results from 1 & 2 above.

Not at the office(barn) but 90% sure the set up is as follows:
I'm using the HDV 1080-60i project settings on the footage being brought in as HDV, rendering that to best , ntsc dv ,

On the cam downconverted my project settings are regular ntsc dv
pan and crop on the rendered dv created from vegas , and the converted from camera to compare. I'm setting the cam for downconvert per instructions from the manual.
I've played around with several project setting combinations, but in every case so far, camera does a much better job converting to DV

thanks John

Can anybody provide some example files (and some method notes) that show horrible/shameful/etc HDV downconversion in Vegas 6?

John:

I recall on the FX1 you can set the VCR to Auto, HDV, and DV. If you set it to Auto then it will play HDV recordings as HDV, and DV recordings as DV, even if they are on the same tape (though it takes a few seconds to switch over automatically). I believe if you set it to HDV and play a DV tape, it will say something to the effect of "No output--Change format." Ditto for playing HDV when set to DV.

Mr forum dude,, I'm not sayin horrible, I'm sayin not as good. Your own manual says to use the camera.

The camera manual says to use the camera. Vegas manual doesn't.
We/I do this every single day. Works great for us. I know FrigidNDediting recently did some great downconversions in Vegas as well.
No issues here, whether it's chromakey, matting, or straight conversions. Works better, IMO, than the camera does. JohnnyRoy likes doing it in cam, so I guess it depends on your preferences. I like having an HD master. Until the MU25 and MU25 are here, you can't recapture w/frame accuracy, so the point of recap'ing is sorta moot at the moment.

page 236 of the V6 manual, it says "if you will be delivering your project in standard definition via DVD or DV tape, you can use the cameras built in down-conversion..................................capture and edit the same way you normally use DV........no need for intermediate files"

You *can* use the camera's downconversion, not you *should* use the camera's downconversion. I'd forgotten that it's in the Vegas manual as well, but either way...using both, the Vegas downconversion is slower, but a higher quality end result. And if you're keying, which we do a lot of, the downconverted file is barely usable as it's just DV. Using the HDV source or converted CineForm is a much better route. But it *does* take longer unless you use the CineForm capture utility.

Either way... still haven't seen anyone show problematic footage, or describe a specific workflow that will define why, or how the footage looks bad when converted in Vegas. Having *just* finished doing conversions in every application that converts HDV to SD, including some stand alones, so far...only hardware does a better job. At NAB, I'll be showing examples. One of our Final Cut guys bought Vegas specifically for downconversion because Vegas is better than FCS or Nattress' converters, by a long shot.

My point is I must be doing something wrong,, i hear you and others i respect talkin about the vegas conversion being better,,, i cant seem to get it right,, not trying to start a fight , trying to better my workflow and end results.
Dave

In answer to Sony (ForumAdmin), here is a link to the file Spot posted in an earlier thread on this subject:

HDV Downconvert

Maybe someone can post results that show the opposite conclusion?

Most importantly, are you using BEST on export? That is critical.
Where/how are you pan/cropping, if at all? You'll get the best downconvert going to widescreen, of course, because then you're not changing pixels around as much. But...it should still be very, very clean on 4:3 as well.

BTW, and FWIW, if you read Showreel, both of the DP's at 24 reference the "mallard" footage that JohnMeyer just posted a link to.

I love Vegas but my experience is that Vegas isn't good for downconversion besides the reeediculous render time you have to endure the cam downconverts better... BUT downconverting in cam sucks because sony F'ed up the downconvert by having a row of black pixels at the top of the frame that is not cropped by overscan on 4x3 monitors making the top line fuzzy and messing up filters by adding unneccesary black being calculated into the frame on filters such as blur.

The remedy for the downconvert footage black line is to crop it, but that means every frame must be rerendered therefore greatly increasing render time for no reason. I have 3 FX1s 1 Z1 and 1 A1U first FX1 acquired Nov 04 and A1U aquired a month ago and ALL exhibit the same problem with downconvert as well as counterparts confirming the same thing.

Therefore the best solution I have found at this point is to deal with cineform capture at the start. Back in the day I was not a big fan of cineform, but at this point it is the best solution IMO well worth the $150

I can provide a ton. I posted on it many times back in the day with nary a good response. What info do you need? I can pull soome old drives that still have some of those projects and show m2t frame grabs against the downconverted DV wides generated by vegas on several machines and answer any questions.

The reason for the problem with the downconvert is the PAR. Shifting PAR on the fly with the current software is an issue, and not an easy thing to do, hence the reason that Vegas does a significantly better job, IMO, of downconverting. Yes, it's slower. Which is why CineForm is the best answer of all if you have a computer fast enough. I guess you could argue that Sony Broadcast made an error in the camera design because they decided to not zoom in and crop pixels or cut pixels, they work with what's native there. Were it me, I'd capture an anamorphic, import it as widescreen, and work with that, as it's much easier to avoid troubles. You're asking the camera hardware to do what boxes twice as expensive don't do well. I see your point and understand the frustration, but there is a much bigger picture there. Even the newest Miranda HD-Bridge DEC with the latest/greatest conversion tools is a softer image on conversion than the camera gives you on downconvert.
Taking 16:9 and converting to perfect 4:3 with a shift in PAR isn't an easy feat at all. Not even for software. Maybe that's why Canon is soft, and JVC avoided it completely, I dunno.

I would rather them zoom in and crop those pixels out instead of leave them in so render times in post wouldn't be so freakin bad. When i was on all canon gear i hated the black bars on the side but at least they were in the overscan area and only were a pain when doing PIP or extreme blurs where the black would bleed into the safe area. On 4x3 there is obviously no overscan area and on 16x9 the overscan is so small that even minimal filters cause unwanted darkening and flicker on the top of the frame.

How do you crop in vegas where the cropping removes those pixels so when a filter is applied at event level it does not use those black pixels in the calculation?

Obviously you've never shot with a Canon cam where you've got blanked pixels on the sides? :-) HUGE PITA for PIP, and of course, creates blur issues. But...the blanking is in overscan, and is actually better for the encoder.
You might want to read John Meyer's workflow. I don't use it, but it does work well for him and others in your situation.
John Meyer's recipe

I hate the forum software here because it dosen't notify you if a reply is made and you cant tell who a reply is to.... I did use Cnon as I said and hated the blanked pixels as I said adn just because they are in the overscan area doesn't justify them being there. Easier on the encoder might be nice for the encoder, but not for the user. If a camera is an SD or HD spec image it ought to fill the spec with image and no blanked pixels that cause problems like these.

John's workflow only makes sense if you want to pan and crop within the image to make up for poor framing and such. Cineform finally came out with a build that is not sooo buggy that I can actually get 99% clean AVIs, so that is curently the best route for us. I would rather deal with 11GB/hr filesize than 36GB with cineform and have a more fluid editing experience, but it is just good enough for 3 multicam @ 10-15FPS and normal editing with filters applied after cuts. So I am relatively happy even though I have 6 friggin terabytes of drives with 2-3 drives burning up every 6 months now.... :--{

No one has answered my question on when you recapture tapes in cineform for archived projects if the recaptured footage will line up frame accurate? I tried the proxy workflow of capturing the m2t and then recapturing the tape downconverted in cam as the proxy file. It adds double the head use of a cam, but saves the 5:1 or more proxy render time. The in cam downconverted footage was not frame accurate to the corresponding m2t either.

Has anyone successfully done a complex long form project, archived it and then recaptured the source to restore the project and it was frame accurate? If so what was your workflow? If not then pretty much all projects need to be archived in their entirety, which means if you do any volume of work you would need a drive farm larger than they used to create the entire King Kong movie (I'm well on my way!!:-{) Save meeeee...

TIA

Has anyone successfully done a complex long form project, archived it and then recaptured the source to restore the project and it was frame accurate? If so what was your workflow?

As I stated earlier, until just very recently, you can't. However, this has now changed, so ask again when we get our MU25 (hopefully next week)

Can you point me to info on the MU25 is it something added in the capture utility?

Lots of tech talk goin on here, and above my knowledge needs or wants,,just want clean video dont care how or why i get there, just want to get there, got bone in the nose kids that figure that @#$% out for me. I ALWAYS render to best setting ,always have always will. no matter what format or what is on the timeline.. bottom line my conclusion is cam downconvert to DV is cleaner than a Vegas conversion. Again I am not talking about MPEG or anything else, plain old dv. Someone shows me different and I'm all over it,, not MPEG ,DV.
I'm done with this subject,, thanks to all.

winrock,

I love a challenge! Actually more like a "test", because I have never done a camera down-convert. I always use Vegas to downconvert.

How about a peacock shaking his feathers in a mating dance? I got that footage last weekend. Also should be a good test of the HDV movement issue.

Have to give me a few days, US tax time, you know. And beautiful weather in Florida tomorrow, going on a boat :) I have priorities!

I'll post a link to the original HDV, the Vegas downconverted DV, and the camera downconvert DV. All footage will be untouched with effects, but cropped for synched timeline.

Dave

sounds great!! be looking forward to your results, and the footage sounds cool .
Thanks for the time,