HDV downconverted...

[pielas]

What happen when you shoot HDV and then you use DV OUT locked to have a down concoverted HDV for editing like a normal DV? Do you still takes advantage of the quality of the hdv or is like if we had shooted in dv. The conversion always allows to maintain the greater definition of the hdv? I have a Panasonic NV-GS400 and I don't know if I really need Canon HV20.
Thank you for your help!
Ciao

Piero

[Mal]

Well, in that case you'd be better to RECORD to tape in HDV. Capture the HDV to your computer, and then edit the footage, and then downconvert to DV. Yes, that should yield very good results.

Someone should test that though, to see exactly what it looks like compared to if you'd just recorded DV.

Good question, that!!

[pielas]

But I thought to use for the editing dv(after shooting hdv) and therefore the issue is what happens to the hdv converted in dv through HV20. Anyone know? With dv it's easier the editing process... also because I don't know if once you edit hdv in final cut the video signal trough firewire can go in a DV CAM that convert to an analog signal for a normal tv(not hdready),
Thanks for your ideas on the subject.
Ciao

Piero

[24Peter]

I did a very quick informal test this week comparing 1) a DVD created in my stand alone Memorex DVD recorder downconverted in camera using DV locked out for HDV footage, 2) a DVD created in Vegas 7 using HDV footage downconverted in the camera (DV locked and captured in computer) and 3) a DVD created in Vegas 7 from HDV footage from the camera - HDV out capatured in computer. It was the same footage in each case. All the DVD's were standard definition.

The best looking DVD was the one created from HDV captured in the computer using Vegas 7 but it also took the longest to create - a 3 minute clip with a couple of dissolves took 35 minutes to render from HDV to DVD (mpeg-2) ("Good" quality render). (The computer is a laptop with a Core Duo T2050 processor and 1GB RAM running Vista.)

The second best looking DVD was the one burned on the Memorex DVD recorder (HQ setting). Almost as good as the HDV downconverted in Vegas. But that DVD was created in real-time - i.e., a 3 min DVD took 3 min to create, but obviously I couldn't really edit the footage except the old school way of pausing the recording, etc. You need to use DV locked to downconvert in camera before recording to DVD - the HDV out produced sound but no video in the recorder.

The worst looking DVD was the one created in Vegas capturing HDV footage as DV from the camera (again, DV locked). It was definitely lower quality than SD DVD's I've created in Vegas from my DVX100 (which is a standard definition camera). The most noticeably aspect was the blocking up of red colors. Much more pronounced in this DVD. Perhaps because of the 4:1:1 colorspace of the downconverted HDV footage, though again, I never noticed it this bad on any of my DVX 100 footage. Time to render: approx 5 minutes to render a 3 minute mpeg-2 file.

So... is it worth it to downconvert in the computer over the camera? That depends. Most people probably wouldn't notice the difference in the HDV > DVD vs. DV lock > DVD created in Vegas. And at 7 times longer to render the file, we're talking a major time waster. Even on a faster machine, you'll still get much longer render times downconverting HDV in the computer than in camera. In truth, the quality of the Memorex DVD was good enough for me at this point. For friends/family, I may just create DVD's the old school way (pause/record/pause) until processing power catches up with the HDV revolution. For my short film projects, however, I'll bite the bullet and let the computer do the downconvert and go out for lunch while rendering.

[Mal]

1) a DVD created in my stand alone Memorex DVD recorder downconverted in camera using DV locked out for HDV footage


Thanks for that test.
Call me Mal, but what exactly does the above mean? How do you downconvert in camera? (Haven't even got to that yet....)
Do you record to HDV 24p, and then send a DV signal out the firewire and into the DVD recorder?

[24Peter]

Thanks for that test.
Call me Mal, but what exactly does the above mean? How do you downconvert in camera? (Haven't even got to that yet....)
Do you record to HDV 24p, and then send a DV signal out the firewire and into the DVD recorder?

Yes but you have to go into the menu in the camera and select "DV locked" rather than the auto HDV/DV output. That's essentially outputing HDV footage as DV footage.

[pielas]

But the question is: if you use a hdv recording mode and then a downconverted dv for editing, is it better then shoot in a 3ccd cam DV? Do you have somenthing better in the images or it's quite the same? :hv20-smilie51:

[24Peter]

But the question is: if you use a hdv recording mode and then a downconverted dv for editing, is it better then shoot in a 3ccd cam DV? Do you have somenthing better in the images or it's quite the same? :hv20-smilie51:

Well again if you read my post, the quality depends on where you do the down converting. with the HV20 I'd say a HDV file downconverted in the computer seems to have better quality than a HDV file downconverted in the camera and then captured in the computer for editing. I haven't done a side-by-side test between my HV20 and DVX100 but an HDV file from the HV20 downconverted in the computer and burned to DVD is probably about the same or slightly better to what I am used to seeing with my DVX100. Downconverting the same HDV file by using DV lock on the camera and creating a DVD from that file is worse than what I am used to seeing with my DVX100. But, the DVX DV to DVD files (and the DV lock to DVD files from the HV20) render much faster compared to the HDV to DVD which can be somewhere around 5-10 times slower. That's huge.

[pielas]

thanks for the informations. The point is that perhaps pass from a gs400(a 3ccd camera) to a hv20 is not a great jump of quality yet. Perhaps the hdv is not something evidently better, at least by what you can see by your eyes. Maybe hv20 is just somenthing new and affordable but you will still get the same kind of video look that you have from others good dv camera, at least that you use hdv to go to film or a hd tv set wich is not my case(I shoot usually short films, narrative or experimental, that I show in festivals where usually projection are from a digibeta) .Am I wrong?

[Mal]

pielas,
I think the HV20's video quality is IMMENSLY better than a MiniDV cam.
Here's two frame grabs I took, and both corrected to the same size:

Sony MiniDV camcorder DCR-PC100:
http://hv20.info/image-sample-PC100.jpg

Canon HV20:
http://hv20.info/image-sample-hv20.jpg

That's a copy from this thread:
http://www.hv20.com/showthread.php?t=33

[pielas]

Looking at this pics the difference it's seems quite clear. But I work also as an editor and I edited 3 feature film shoot with a sony fx hdv camera that they went on film after and the result was good but while I was editing in video I wasn't much impressed about the shift of quality from some previous films made with a good dv camera. Maybe I'm wrong but I fell that way...probably it's because of the video fell that continues to remain. Instead few weeks ago I wached a film shoot with an XDCAM and the effect was very different. It was very beautifull and it looked not like the usual video. But I know it's completely different from one small camera that costs fifty times less.

[Mal]

Yes exactly pielas.
No $1000 small chip camcorder can compete with a $30000.00+, 2/3" cam.

If you REALLY want to get a NICE digital video camera, you should look at the Red One from http://red.com.
That's "only" $17500.00 (cam only) for a 4K resolution revolution! :)

[gibo]

There are so many other factors to consider as well - especially when converting video to film. There is the way the video is shot (lighting, exposure, etc.) how it's edited and prepped for a film transfer, and finally the film transfer itself. There are numerous methods of transferring to film and each can make quite a difference in the final look.

As for the quality of this camera downconverted to DV, well there will be one advantage of this camera in that the recording chip (CMOS sensor) has a higher resolution than any standard definition DV camera. However, there will still be compression differences between a ccd and a cmos sensor and again, a lot of it has to do with how it's shot. I work at a production company that uses Sony Z1-U HDV cameras and much of our work goes to standard definition video. I feel like the image quality is comparable to some of the high end DVCAM cameras we used to use.

Finally, I think much of this boils down to the recording sensor (cmos vs. ccd) and the lens on the camera. But no dv camera will ever be able to blow up to 1080 and hold up quality wise.

[24Peter]

Am I wrong?
Yes - as Mal and others have mentioned, the image quality/resolution of any HDV cam will be better than even the best 3-ccd DV cam in almost all cases - esp. in your situation where you are projecting onto a screen. The exception I noted was when you downconvert HDV footage - whether in camera or in the computer. SD footage is SD footage, whether native or downconverted. In my test, the SD DVD's looked about the same overall, though closer scruitiny revealed the HDV downconverted in the computer looked the best (and also took the longest time to render). But if you're talking about HD resolution computer files, as Mal's samples show, there's no contest - HDV from even a single chip camcorder kicks DV butt.

[bruny]

Assuming that it is now settled that it is better to do the down-convert on the computer rather than let the camera do it, another piece of this puzzle is what software you use to down-convert. Not only what software, but how you use that software. I tried a standard downconvert with FCP (via Compressor) and the results looked terrible, IMHO. Then I stumbled upon this set of instructions ( http://www3.telus.net/bonsai/Step-by-Step.html ) and have found down-converts to DV to produce quite good quality. Using MPEG Streamclip also seems to do a good job.

[24Peter]

I'm using Vegas 7 on a PC. Capture your HDV footage, edit it as you like, then output/render as mpeg-2 video/ac3 audio. It downconverts automatically. In Vegas there are actually customizable templates for each type of video you want to render/output to. mpeg-2 video for SD DVD is usually 6-8mbps CBR/VBR.

[VideoNut]

Hello, I am new to this forum and HDV in general. Just got a HV20
recently. I dubbed my HV20 footage to a DVD recorder at HQ setting.
The quality was excellent but of course not as good as HDV. I just
wanted to share some wedding (my own) footage with friends. The
only issue is that the 16:9 format was squeezed into the 4:3 format
and I haven't figured out to get around this. Anybody have
any experience with this? Is this just a DVD player issue?

[David]

I dubbed my HV20 footage to a DVD recorder at HQ setting.

The only issue is that the 16:9 format was squeezed into the 4:3 format
and I haven't figured out to get around this.

When you say you dubbed it to a DVD recorder, do you mean you went straight from the camera to the recorder? If so, you might have luck by changing a setting on the HV20 for 4:3 televisions.

I don't know exactly where the option is, or if it will solve your problem, but you could give it a shot. There's basically some kind of TV output setting so the widescreen images can be correctly viewed on a 4:3 TV. My guess is that the camera does some 'conversion' before outputting the signal and the image will appear letterboxed.

[duzzit_madder]

Might be better if you start a new thread VideoNut. More eyes will see it that way. As a side note, 16:9 and 4:3 aren't formats, they're aspect ratios. So you shot some footage in 16:9 then 'dubbed' that footage to what exactly? A DVD burner? So when you played the DVD everyone looked tall and skinny?

[24Peter]

I'm able to use my Memorex DVD recorder with my HV20 no problem. Actually very good looking DVD's - better than downcoverting in-camera and editing in Vegas/DVDA. (Capturing HDV and editing/downconverting in Vegas produces the best looking DVD's but also requires render time - no render time - but no editing - with a DVD recorder.) I use DV lock on the camera to output DV only. I also have my camera settings to widescreen 16:9 TV but I don't know if that affects DV output (probably only AV out, maybe component out). Make sure you're DVD recorder is set to widescreen/16:9 in the setup menu.

[tom chang]

alittle off topic, but I am having trouble with HV20.
When I put it on DV output and via firewire feed it to
the DVD recorder(Toshiba), I get no picture although I can
control the camcorder through the play buttons on the Toshiba...
anyone else have problems like this , or am I missing something int he
settings of the HV20 ??

[bluegrass]

alittle off topic, but I am having trouble with HV20.
When I put it on DV output and via firewire feed it to
the DVD recorder(Toshiba), I get no picture although I can
control the camcorder through the play buttons on the Toshiba...
anyone else have problems like this , or am I missing something int he
settings of the HV20 ??

I'm not sure what you mean by DV output? The same setup that you would use if you were capturing firewire to your computer which I assume you have done? When it starts to play the tape, do you see the video play on the camcorder's LCD OK?

That better not be that flippin' hollywood protection coming into play on your Toshiba handshakin' with the output from the camcorder. Did the tape have HDV or just DV content on it?

I'm not coming up with any answers but maybe these questions might stir something in your or somebody elses mind. I guess you had to be feeding the camcorder to the Toshiba DVD deck via firewire although I'm not familiar with consumer DVD tabletop recorders have a firewire connection but I didn't even look for something like that on my Sony DVD recorder yet. As a matter of fact I haven't even tried to record to it and I've had it for over 6 months. Guess I'm behind the eight ball there.

[tom chang]

I'm not sure what you mean by DV output? The same setup that you would use if you were capturing firewire to your computer which I assume you have done? When it starts to play the tape, do you see the video play on the camcorder's LCD OK?

That better not be that flippin' hollywood protection coming into play on your Toshiba handshakin' with the output from the camcorder. Did the tape have HDV or just DV content on it?

I'm not coming up with any answers but maybe these questions might stir something in your or somebody elses mind. I guess you had to be feeding the camcorder to the Toshiba DVD deck via firewire although I'm not familiar with consumer DVD tabletop recorders have a firewire connection but I didn't even look for something like that on my Sony DVD recorder yet. As a matter of fact I haven't even tried to record to it and I've had it for over 6 months. Guess I'm behind the eight ball there.

Thanks man, I did get a DV output to my toshiba finally, on set up 2 for playback, I had to lock it on DV output, but on set up 1, keep it on
Auto, only then , will it have DV output for the firewire.
Anyways, I noticed the LCD monitor on Camcorder has a defective
pixel(2 of them) and the dots are irritating since the monitor is so small.
I'm going to take it back. I guess Cannon's quality control is not
as good as sony's

[24Peter]

Just to be clear, I have to use "DV locked" as my output on the HV20 to get my Memorex DVD recorder to properly record video via firewire input. In "HDV/DV" mode I get no video on the Memorex via firewire.

[Rikki]

In response to questions:

I had a Panasonic GS280 (same as 300 essentially) and really didnt like it. The HAD CCD smeared like you wouldnt believe, thats why I waited for the HV20 to arrive.

It kicks the spots off the Panasonic!

What I do is record everything as HDV and down convert on the camera as the final output will be for web or at a push DVD. This means that processing doesnt take too long (my PC is DV equipped, not SP2 XP so no HDV) but in the future if I need to, the footage is sitting in high resolution HDV format for editing again.

R

[Pom16/10e]

The best way to downconvert to SD is to drop the HDV footage on an 10 bit uncompressed SD timeline and export that.
Be aware you'll need 60Gb/hour.

But the quality is just that : perfect

[Patryn]

i'm really in a crank...

i want to downconvert from HD in the HV20 to play in a normal DVD and TV (DV).

I'm trying to use Sony Vegas 7.

Can anyone help me out with some instructions on a workflow for this:

- capture settings
- prepairing the files for DVD burning

I'm not interrested in editting yet, only in best picture quality.

i would really appreciate all the help I can get.

Best regards

[24Peter]

i'm really in a crank...

i want to downconvert from HD in the HV20 to play in a normal DVD and TV (DV).

I'm trying to use Sony Vegas 7.

Can anyone help me out with some instructions on a workflow for this:

- capture settings
- prepairing the files for DVD burning

I'm not interrested in editting yet, only in best picture quality.

i would really appreciate all the help I can get.

Best regards

Patryn -

First, make sure you have 7e (for reasons unrelated to your question but b/c some minor HDV hiccups were resolved by "e".)

Then capture using the internal HDV capture application. (Remember what folder you captured your footage to as you'll need that in a second.)

Open the project "Properties" window (File>Properties or "Alt+Enter")

In the Properties window, near the top find "template". Then all the way to the right click the "Match Media Settings" [little folder] icon (to the right of the save as [floppy disc] and delete [red "X"] icons).

Select a clip you just captured by going to the folder where you saved your clips to. The template box should automatically update. (If you use the same in-camera settings for all your footage (e.g., HDV 60i or 24p) you can check the little box at the bottom of the Project Properties window that says "start all new projects with these settings" so you don't need to set the properties each time.)

If you don't already see your clips in the Project Media tab at the bottom of the Vegas workspace, find them using the Explorer tab and add them to the project [select clips, right click, choose "add to project media list"].

Now add your clips to the Vegas timeline (drop and drag or double click clips).

You don't want to edit so the next step is to render your project using the appropriate template. Go to "File>Render As" and in the window first select the folder you want to render your project to, then name the file, then use the drop down buttons under "save as type" and "template" to choose respectively "MainConcept MPEG-2*" and "DVD Architect NTSC widescreen video stream" (or actually since Portugal is probably PAL, use the "DVD Architect PAL widescreen video stream" setting).

Now this may be all you need to do on the video side (we'll get to audio in a minute). I usually get good results with the default settings. However some people suggest going a step or two further. I'm only going to describe one step further and let you play with other settings on your own. While the Render As box is still open and you've selected your "save as type" and "template" as described above, look for the button on the right next to template labelled "Custom". Click that. You are now in the "Custom Template" window where you can further tweak your settings. The only one I mention is the very first choice you see on the "Project" tab: "Video Rendering Quality". By default, this is set to "Good". Some people like to use the "Best" setting. I haven't tested this yet with footage from my HV20 but in the past I tried it with other cameras and didn't notice much of a difference in quality but a huge difference in render times ("best" is much slower that "good".) However, those other cameras were DV which has the same resolution as SD DVD's. Since HDV involves considerable resizing of the image, choosing "Best" may have some advantages. Still, my DVD's of HV20 HDV footage rendered using "Good" render quality look fine to me.

If you shot your footage as PAL 25p HDV footage you'll need an extra step. Go to the "Custom" button again. This time click the "Video" tab at the bottom of the window. In the middle of the window look for the "Field order" button and from the drop down select "None (progressive scan)". (If you shoot 25p regularly, you can save this custom template for later by typing "DVD Architect PAL 25p Widescreen video stream" in the template box at the top and clicking the little "save template" [floppy disc] icon to the right of it.)

Click OK and then "save" and go have a cup of coffee while your video renders.

Once your video has rendered, you need to go back to the timeline and render out your audio separately. Go to "File>Render As" and you'll once again see the render as dialogue box. By default you should be in the same folder where you just rendered your video to so if Vegas hasn't already done it for you choose a file with the same name as your video file, but this time select "Dolby Digital AC-3*" [third choice from top] as your "save as type" and "default template" as your "template" and choose OK. Once your audio renders (much quicker than your video) you're done with Vegas.

To make your DVD, open DVD Architect, browse to the folder where your audio and video files are that you just rendered and add them to your DVDA project. Also don't forget to match the project properties in DVDA ("File>Properties") to the type of TV you'll be playing the DVD back on. (As an aside, I don't think the properties setting affect the actual video if you're not re-rendering (and you shouldn't have to since you used the DVDA templates in Vegas); rather they affect the menus and how the graphics are displayed on screen.)

You'll probably want to test the video rendering as I mentioned above to get the best quality for you. Also for people shooting NTSC 24p (like me), DON'T choose "DVD Architect 24p NTSC Widescreen Video stream" as your MPEG-2 template unless you've used CineForm or one of the other pulldown removal processes.

Let us know how it goes. Good luck! (Or should I say "Bien suerte!" Oh wait that's Spanish - never mind. :( )

[Patryn]

finally someone that helps me out...

first of all, A THOUSAND THANK YOUS...

i TESTED SOME SETTING YESTERDAY AND i GOT FROM 1 HOUR OF VIDEO (hdv) A 2.6 GB REDERED file and it took 7.5 hours.....

But I tried another thing. Opening the files that I get from the capture directly from the HV20 miniDV tape in the Sony DVD architec I found that you can "render directly" in this application without the nearly 8 hours of "mastication" in the Vegas 7 suite...

so... I lost...

Can I render directly the captured *.mt2 files in Sony DVD architec or do I realy need to render first in Vegas and then in the DVD architec.

Last question: what is DOUBLE PAL ?????

Best regards,

Tiago

PS - portuguese is BOA SORTE... and here is a new one for you MUITO OBRIGADO

[tomperson]

Patryn -
...
Let us know how it goes. Good luck! (Or should I say "Bien suerte!" Oh wait that's Spanish - never mind. :( )

I also had some doubts you cleared here. Excellent advice 24Peter, very detailed!!! Thanks! And its "buena suerte" in spanish, btw.

One question:
I'm receiving my HV20 in a couple of weeks. Most of the video work I do is for Videoart projects that end up in DVD or Web, or for video clips for live visuals (VJing you might say). So, my final format is SD. I've read here that the best quality can be obtained by capturing HD in Vegas and then converting to SD at the final step. I'm used to working in PAL SD, my machine can handle it (P4 2.8Ghz, 1Gb RAM, 7200rpm samsung sata disk, 16mb buffer), but i think that if i work in HD, my machine won't be able to cope with it, what do you think? Do i lose too much quality if i do the downconversion from HD to SD right in the camera when capturing the video?

Also, if I shoot in 25P HD, can I downconvert to 25 frames per second, progressive scan, but standard pal definition (SD)? Does it make any sense?

THANKS guys!

[24Peter]

finally someone that helps me out...

first of all, A THOUSAND THANK YOUS...

i TESTED SOME SETTING YESTERDAY AND i GOT FROM 1 HOUR OF VIDEO (hdv) A 2.6 GB REDERED file and it took 7.5 hours.....

But I tried another thing. Opening the files that I get from the capture directly from the HV20 miniDV tape in the Sony DVD architec I found that you can "render directly" in this application without the nearly 8 hours of "mastication" in the Vegas 7 suite...

so... I lost...

Can I render directly the captured *.mt2 files in Sony DVD architec or do I realy need to render first in Vegas and then in the DVD architec.

Last question: what is DOUBLE PAL ?????

Best regards,

Tiago

PS - portuguese is BOA SORTE... and here is a new one for you MUITO OBRIGADO

Tiago - I didn't see your post until just now. Hopefully you've gotten this figured out.

2.6GB for 1 hour of 25p mpeg-2 SD video seems right. 7.5 hours however seems very long. Not sure why it takes so long.

I have an Athlon X2 dual core 3800+ (2.0 Ghz I think) with 2GB RAM and two WD SATA 300 GB 7200 RPM hard drives and my HDV to mpeg-2 SD take about 2 1/2 times real time. So a one hour video takes 2 1/2 hrs to render. I think having two separate hard drives really helps - I use one to hold the project files and the other to render to.

Anyway, you shouldn't have to re-rended when going to DVDA4 if you use the DVDA template in Vegas. (And rendering in DVDA4 would't be any faster than the same render in Vegas anyway.)

Re; "DOUBLE PAL" _ I don't know - did I write that someplace?

[24Peter]

I also had some doubts you cleared here. Excellent advice 24Peter, very detailed!!! Thanks! And its "buena suerte" in spanish, btw.

One question:
I'm receiving my HV20 in a couple of weeks. Most of the video work I do is for Videoart projects that end up in DVD or Web, or for video clips for live visuals (VJing you might say). So, my final format is SD. I've read here that the best quality can be obtained by capturing HD in Vegas and then converting to SD at the final step. I'm used to working in PAL SD, my machine can handle it (P4 2.8Ghz, 1Gb RAM, 7200rpm samsung sata disk, 16mb buffer), but i think that if i work in HD, my machine won't be able to cope with it, what do you think? Do i lose too much quality if i do the downconversion from HD to SD right in the camera when capturing the video?

Also, if I shoot in 25P HD, can I downconvert to 25 frames per second, progressive scan, but standard pal definition (SD)? Does it make any sense?

THANKS guys!

You can see my previous post for the specs on my machine. Vegas handles HDV very well in terms of previewing and editing. Rendering however is where things slow down.

All I can say is run your own tests. You may find the quality and covenience of downconverting in-camera acceptable to you.

I'm pretty sure you can 25P SD from 25P HD no problem.

[spunkme]

Where is DV Lock? I can't find it, is it in the menu in playback mode, does it say DV Lock?

[PWHerman]

spunkme:

Go into Playback mode, and into the Menu.

From there, go down twice -- it'll say "PLAY/OUT SETUP2" at the top. Go into there and it'll be the bottom option, "DV OUTPUT." Change it to "DV LOCKED."

Voila!

[dmoreno]

I have a canon hv20 that I bought 1 month ago. Previously I had a Panasonic PV-GS250 (small 3ccd). I have also worked a lot with a Canon XL1that my cousin has and, since I have a tv show with a local network here in Costa Rica, I work every week with a Sony XDCAM HD 330 and edit HD footage from the XDCAM using Vegas.
I must say I am AMAZED with the image qualitiy of the canon hv20. Its WAY better than the panasonic, even if I shoot hdv and output dowconverted SD. Specially I love "soft", less videoish look I get due to the high resolution. I also love the 24p look and the cinemode gives you much more lattitude (dynamic range). The colors of this single cmos sensor rival those of much more expensive 3 ccds cameras.
If I had to choose between Canon XL1 and HV20 downconverted footage I would choose the HV20. You get even better results if you capture HDV removing pulldown (with cineform or other method) and convert to 24p widescreen DVD. The HV20 has better low light capabilities than both the panasonic and the xl1.
When it comes to comparing the hv20 to the XDCAM HD (1k vs 20k!) I have been surprised about how well the hv20 handles. We've shot sketches for our show using both cameras at the same time, and with some color correction you can barely notice the difference. Of course, this is when you have good lighting situations. When you're shooting in low light the 3 1/2 inch CCDs of the XDCAM totally blow away the HV20.
I believe the people on canon made a mistake with this camera: to good for the price! Now all the other cameras seem to expensive!

A little tip on downconverting using the camera, if the TV aspect of the camera is set to 4:3 it outputs 4:3 letterboxed SD DV through the firewire, if TV aspect is set to 16:9 it outputs widescreen DV. I output with the camera set to 16:9 and capture as widescreen dv, then convert to widescreen dvds and let the dvd player do the letterboxing for me.

[Halsu]

The best way to downconvert to SD is to drop the HDV footage on an 10 bit uncompressed SD timeline and export that.
Be aware you'll need 60Gb/hour.

But the quality is just that : perfect

Why 10 bit?? HDV is 8 bits only, so only thing you get from those extra two bits is a larger file size, the actual color depth doesn't magically increase just by saving to 10 bit...

...but other than that, you're exactly right - rendering out uncompressed SD video from HDV original yields more or less perfect results.

In my tests, HDV converted to uncompressed SD (with some filtering) outperformed a DigiBeta on both the visual appearance and greenscreen keying. Even though HDV is 4:2:0, when downscaled, you get better than 4:2:2 color space in the SD video.

I did the tests with HVR-Z1, but i see no reason why one wouldn't get as good or even better results with HV20.

PS. This was my first post here, i've been the proud owner of HV20 since yesterday ;-)

Even though the main use is just home videos (and an inexpensive HDV deck), i will most probably be using the cam for actual productions too, at least as a 2nd cam or something. The images one can get with it are definitely up to the required quality, it seems.

[Pom16/10e]

yup, you're right about 10bits vs 8
Just me used to doing 10bits (since I almost always have motion graphics added...)

[SenorKaffee]

By the way - if you do color correction later you should work in more than 8 bit. The Rebel Guide says so. ;)

[Halsu]

By the way - if you do color correction later you should work in more than 8 bit. The Rebel Guide says so. ;)

Well, technically yes. I rarely bother to though ;-)

Anyway, even if you're using a higher color space on your software when doing color correction, the original file is still 8 bit only, and saving it to 10 bit doesn't help in that regard.

[um3k]

A couple things:

1. Downsampling from HD to SD in 10 bit format may actually allow for smoother gradations. As long as the resizing is done after converting to 10 bit, the interpolation will, in a way, convert spatial resolution into tonal resolution.

2. Working in 10 bit (or higher) is highly desirable if your color correction process involves multiple steps, no matter what the original bit depth/resolution of the footage is. Working with more bits reduces rounding of the brightness values, thus lessening rounding errors (which are cumulative), and resulting in greater color accuracy. This could make the difference between beautiful smooth tones (blue skies, anyone?) and horrible posterization (which can be remedied somewhat by adding noise, but that's oftentimes the last thing one wants to do).

[Murrelet]

Hi 24Peter,

Great explanation!

I more or less do what you've specified (and will go back and follow your method closer), but found (think it may have been another explanation you posted mentioning deinterlace) a reference to deinterlacing. Looked up deinterlace in the Vegas manual, it isn't listed, had to go to Interlace, and there it was.

The "Deinterlace Method - none" produced bands and checkerboarding of high motion capture, resulting in quite a poor quality rendering. Since the paragliding footage is all motion, I went back and tried the "Interpolate" option, and the difference was an immediate, substantial improvement.

Interpolate: "If you choose the Interpolate option, a single field is used at a time, which is a good for low-detail, high-motion material."

Chapter 14, Pg 218 of my Vegas manual. I have 2 manuals, think this is the older one so the page may be different on the newer one. Subject is "Modifying Project Video Properties" for those interested.

Anyway, I thought I would post this as it may help someone else, it sure improved my render quality. Those interested should look it up. There is a "Blend" option as well. Perhaps someone can expand the explanations of these options.

[Halsu]

1. Downsampling from HD to SD in 10 bit format may actually allow for smoother gradations. As long as the resizing is done after converting to 10 bit, the interpolation will, in a way, convert spatial resolution into tonal resolution.

You're right.

I did a quick test, where i forced an image temporarily to 256 colours (to emphasize the gradation problems), then converted it back to 8 bits / channel and resized.

The result was VERY similar to scaling the original 8 bit / channel version.

The same should apply to 8 and 10 bit images (though the difference is harder to see, naturally).



2. Working in 10 bit (or higher) is highly desirable if your color correction process involves multiple steps, no matter what the original bit depth/resolution of the footage is. Working with more bits reduces rounding of the brightness values, thus lessening rounding errors (which are cumulative), and resulting in greater color accuracy. This could make the difference between beautiful smooth tones (blue skies, anyone?) and horrible posterization (which can be remedied somewhat by adding noise, but that's oftentimes the last thing one wants to do).

Agreed - though i usually switch to compositing in higher bit levels only if i see visible errors. I guess i need a faster computer and more HD space - again ,-)

BTW, if one adds noise to help hide banding problems, it's best done as a first step, before adjusting the colors. Much less noise is needed for adequate dithering if it's done in this order.

[bigjohn1961]

I use Ulead Studio 11 to capture HVD, edit it in the same program ,that takes
advantage of Smart Rendering, to a HD MPEG2 (just as quickly as DV rendering) , than convert it with the batch converting tool of studio 11, so i get the best of both worlds, a HD file to watch and save on the computer, and a SD MPEG2 file to burn to DVD.
And if you edit in HDV you always have the option to save the rendered file back to tape.

BigJohn

BigJohn

[uggabukka]

Is it possible to capture dv 25p from firewire from hdv 25p footage, without any software rendering?
I'm not satisfied with regulaer dv since its interlaced

[2Bdecided]

It is, but the in-camera conversion isn't the best.

Cheers,
David.

[alwyn]

Is it possible to capture dv 25p from firewire from hdv 25p footage
I thought DV from a video camera (SD camera or HV20 downconvetred in-camera) is always interlaced, but could be wrong, and don't have a down-converted file to check.

[2Bdecided]

It's flagged interlaced, but it is progressive.

Just like the HDV from the HV20.

Cheers,
David.

[alwyn]

David,

Oh. Why would they do that? Going one step further, is there any need use PF25 then if the HV20s video already is progressive?

[Ian-T]

Alwyn, it would be because it's higher resolution for one...and also it's actually capturing whole frames and not half fields.

[alwyn]

Not having a go; I'm a bit bamboozled by P vs I, just trying to understand what is really going on. I was under the impression that there wasn't such a thing as progressive DV-AVI, regardless of the source. :hv20-smilie03:

[2Bdecided]

Interlaced means odd lines represent one moment in time, even lines represent a moment a little later (20ms later in "PAL" countries).

Progressive means all lines represent one moment in time. If filming in progressive mode, then the next frame will also be progressive, and for 1080p25 it will be 40ms later in "PAL" countries.

Progressive is one full resolution image every 40ms. Interlaced is two half resolution images every 40ms, or one half resolution image every 20ms.

If you have 1080p25 progressive content, with one image every 40ms, it's impossible* to generate genuine interlaced content (i.e. with one image every 20ms), because every alternate moment in time is missing.

(If you have 720p50, it's easy to generate genuine interlaced content, because you have one image every 20ms, just as you need.)


Starting with 25fps progressive video, you can re-format it and send the video as a sequence of fields (all the odd lines, then all the even lines - just like analogue TV), or flag it as field encoded (like in MPEG) - you can shout from the rooftops that it's interlaced - but if pairs of fields represent the same moment in time, rather than different moments, it's just progressive content wrapped up in a tortuous but reversible manner. It is not genuine interlaced content, and it won't look like it on-screen either!

Cheers,
David.

* - unless you use some clever motion interpolation software

[alwyn]

Thanks David. That all makes sense. I still don't undertsand why Canon call my PAL HV20 1080i (it even says it on one of the outside labels) if it really shoots "P" natively.

[Ian-T]

....it's all semantics....