When working with HDV, make sure that you render directly to DVD compliant mpeg2 and skip the DV master stage. The DV codec has a lessor colorspace and rendering a master in this format will look lessor in quality.

The different scaling methods will affect sharpness.

Using "best" will give you a sharper image (and less aliasing).

2- Oversampling does not automatically mean that the image will be sharper.

3- Probably the best thing to do is to:
A- Turn sharpness off on your camera.
B- Downconvert using the "best" method in Vegas. (There may be other scaling algorithms that might be slightly better, but this is reasonable since it's built into Vegas and takes little effort.)
C- If necessary, add some unsharp mask or other sharpness-enhancing tricks. e.g. just adding contrast via a s-curve will kind of make things look sharper.

well, im doing it both ways.. in cam downscale and and from HDV source down to DV through vegas
Reason im trying both methods is
1) in cam downscale is faster and i end up working with DV which one of my lower systems can handle without issue
2) NLE option allows me to downscale to huffy and use a 422 colour space, but the storage requirements are intense. Ive got 2TB but i work on 3 jobs at a time

The DVX i am comparing to is a PAL model, so the colour space should be no different to HDV until i scale down. In essence and theoretically, the HDV fotoage should look markably better than the DVX footage.. It doesnt.. not by a long shot..

Laurence in regard to rendering to DV from M2t, the reason i do this is because th ecolour space im working in is PAL, so 420 is the range im working with, irreseptive of the format. Rendering HDV down to DVD compliant MPG2, is a time issue.. this oen machine im concerned about renders MPG like a pig

still trying to work it out though.. not impressed.. i dotn remember the Z1 being this soft on downconvert

thanks for that Glen,
Sharpness on the A1 is flat on zero (not below or above normal) this usually takes care of any edge colour fringing and gets rid of that oversharp home video look.

"Using "best" will give you a sharper image (and less aliasing)."

Yup, i never render to anything but best. The time differential is not that differnt so i use best in everything i process

"2- Oversampling does not automatically mean that the image will be sharper."

Understood, but starting with an image using double the resolution, should inherantly offer a sharper scaled image though shouldnt it?

3- Probably the best thing to do is to:
A- Turn sharpness off on your camera.
B- Downconvert using the "best" method in Vegas. (There may be other scaling algorithms that might be slightly better, but this is reasonable since it's built into Vegas and takes little effort.)

((Im already doing it this way... I might just ditch the incam scaling though...
One thing i am curious about as the source is upper field first, should i scale down using the same field order, or should i swtich it to lower? ))

C- If necessary, add some unsharp mask or other sharpness-enhancing tricks. e.g. just adding contrast via a s-curve will kind of make things look sharper.
((THought as much... :(... the point was to not have to do that as rendering time takes a hit.... bugger... ))

Sharpness and resolution are not the same thing.
Downsampled HDV will be higher resolution than any DV camera, it will not be sharper though.

Also which A1 are we talking about and under what conditions was the footag shot? Under low light the noise level of the smaller HDV cameras can really work against you.

Bob.

Understood, but starting with an image using double the resolution, should inherantly offer a sharper scaled image though shouldnt it?
If you resample it down to SD, then it depends on what resampling/rescaling method you use.

With any sampling, you have tradeoffs between:
A- Frequency response / MTF. Improves/degrades perception of sharpness.
B- Aliasing. Seen as moire artifacts, or the funky patterns on certain fabrics and stairstepping on edges when you pan.
C- Ringing artifacts. Seen as weird ghosts or halos around objects.

You'll always have one of these issues... and you can tradeoff one issue for another. In practice, you try to compromise and pick the issue that looks the least bad (usually ringing). If you sample at the SD rate, you can't get ringing artifacts... you need to apply digital filtering... the filtering can't be done optically (whereas a blur filter can be done optically).

So the advantage of digital filtering would be that you can get an image that is both sharp and has little aliasing. Yes it will have ringing artifacts.... but it's the lesser evil in most cases.

2- You'd need to stick in some sharpness yourself (on the SD downconvert) since most SD cameras add it in.

In Vegas, FX usually get added after resizing unless you change pre/post toggle. So adding FX to the clip or video preview level will add sharpening after the image has been resized to SD. So you don't need to do anything there.

3- On a practical level you also need to get work done. So letting the camera downconvert might be reasonable.

Hi,

I experience the same softness than DJPadre does.

I am using Vegas 7 (soon 8) and shooting using Sony HDR-HC1 and Canon MVX2i (DV). My opinion is that letting the HC1 downconvert the material to SD before importing it to the timeline is not generally a good idea. Why not put both HD and SD source material on the timeline and then render out to DVD mpeg2 for best possible quality?!! That gives you also the possibility to crop and pan within the HD material during the post if needed, without a resolution loss.

Now - using the BEST possibly render settings (also highest video bit rate possible) - and looking at the finished rendered material - the SD looks sharper than the HD clips! Something is wrong here. I have experimented with the interlace settings, and found that it is very important to allow Vegas to perform DEINTERLACING if you put HD material on the timeline that is going to be rendered as SD. Preferably interpolated deinterlace. If this option is not selected then the HD output looks awful. Why is this? Independently, the SD still looks a little sharper. This puzzles me a lot and I would like to get an explanation.

Or - or can someone that are doing this for his/her profession:

How and what exact settings should you use in Vegas to be able to output best quality SD originating from HD material?

There have been lots of questions about the DV codec quality in Vegas, and why the renderer output looks so soft, compared to some commercial DVD's. OK - for granted - one problem is the DV25 material that is inherently prosumer quality, but why cannot I render better quality when I originate from HD - wich sharpness is at least as good or better than DV50 (pro) material???

I have the feeling that I am not yet using the full potential Vegas could offer - producing sharp SD DVD' that originate from HD material...

There must be some settings that are wrong - or then the DVD codec is not prfessional quality...


Christian

Bob, the A1 I am refering to is the canon A1, 1/3 ccd HDV camera... the baby bro to the XLH1


Glen, thanks again for sharing your knowledge here.

a couple of q's though

"In Vegas, FX usually get added after resizing unless you change pre/post toggle. "

I was always of a mind that the filter sequence is as follows
Event level
Track level
preview/output level

theres also the Medai Bin level, but i dont think anyone uses that

So adding FX to the clip or video preview level will add sharpening after the image has been resized to SD. So you don't need to do anything there.
((So your saying that if i put a filter on the preview/output level it will affect only the output.. so basically i run the filter through the event/clip level then yes? as i want the scaled footage to be sharp upon ingestion..

If you ingest HD and sharpen or blur that and then downscaled that the results can be different if you add the FX after scaling.

Now here's where I get confused as things got changed somewhere around V7 time. For example GB used to get added scaled at the project resolution, it now seems to get scaled at the output resolution.
However if you're doing the scaling in the VCR/Camera and simply capturing SD none of this matters.

There's also the Yellow Triangle but I don't think this has any relevance.

Bob.

"theres also the Medai Bin level, but i dont think anyone uses that"

Yes, people do use MediaFX. Just not as often as other FX. It's not very well implemented in Vegas (no indication you've applied it)

Funny you mentioned reversed fields, thatwas my first thought. Have you eliminated that as a possibility?

Rob

the field issue was vegas not reading teh 16:9 LFF flag coming from teh A1... i had to go through each clip and swtch them to LFF 16:9, and the issue i was experiencing earlier is now gone...

this one though is still baffling..

as downscaling in Vegas, from UFF HDV to LFF DV, i am yet to find a definivtive answer as to whether or not to do it this way, or do my entire edit in HDV and scale to progressive MPG2 DVD which shouldnt make a difference.. aside from render times which is what im trying to eliminate..

The filter order is:

MediaFX

EventFX; you need to click the little triangles in the EventFX window to change ordering / to affect pre/post toggle

Pan / Crop + Bezier mask

EventFX; by default eventFX get placed here.

Track FX

Video Output

2- What I was saying is that you want sharpening applied *after* pan/crop / after you downscale. This is the default behaviour so don't worry about it.

Glenn,

If the project size is different than source, a rescale or field change is needed. I bellive it is done before everything.

Another question, when changing UFF to LFF - how doed it work? how much is lost in the process?
Can it be better to move the frame one pixel down/up and just tell Vegas it is LFF - is it the same?, maybe do that in VDub?

Does project setting vs. render setting effect the quality - place the HDV material in an HDV project and rener to DV vs. Placing the HDV in a DV project and rendering to DV?

Udi

If the project size is different than source, a rescale or field change is needed. I bellive it is done before everything.
Udi, see my last post. The rescale happens in pan/crop... FX can occur before it. (This is what I remember anyways.)

Another question, when changing UFF to LFF - how doed it work? how much is lost in the process?
Usually only one of these is correct- having the wrong field displayed first will give you problems.



Render settings tend to override the equivalent project settings.