Why do my rendered home movies look bad on HDTV? (newbie help for sure)

I have a digital camcorder, Canon Z700 that records in 16:9 format on MiniDV. I capture the video using a firewire connection to Pinnacle Studio 9.4.3. I tried checking "Progressive Encoding" on the advanced options in the burning process in hopes of gettng better playback on my 32" LCD HDTV through our Magnavox Progressive Scan DVD player.

Everything I create is VERY fuzzy. Looks like analog input. Regular DVDs play GREAT on the player and TV even without upconverting.

I know how to capture, edit, and reder using Pinnacle Studio and that's about it. I don't know how to get digital quality out of the video.

ANY advice would help. I tried looking around the site and I did not see anything that was directly related to this more of a general question.

Originally Posted by edDV

#1 Don't use Progressive encoding. Your HDTV has a much better hardware deinterlacer. Keep DV format camcorder material interlaced (lower field first).

Thanks for the advice. I have tried it both ways actually. With and without the progressive encoding and the quality was the same.

I just took a DVD that I created and played it on my notebook to compare the quality (don't konw why I never thought of doing that before!) and the quality of the playback on there was similar to the TV. Somewhat "fuzzy".

I thought the problem was in the playback on the TV but apparently the overall quality is not good.

I had the settings on Pinnacle set to 100% quality on the burn process.

#2 The 16:9 mode on the Canon Z700 will slightly decrease vertical detail since it isn't a "native" 16:9 model.

#3 Pinnacle Studio is at the lower end of MPeg2 encoder quality. If you do use Studio, increase the encoding bitrate to 8500 Kb/s or better and try CBR rather than VBR.

Better software will get better results. You can test this by capturing your DV a DV-AVI file and then encoding with different demo MPeg2 encoders.

Don't forget #1.

Another thought, you are capturing in DV format not MPeg2 right? Capture settings should be DV.

Originally Posted by edDV

#2 The 16:9 mode on the Canon Z700 will slightly decrease vertical detail since it isn't a "native" 16:9 model.

#3 Pinnacle Studio is at the lower end of MPeg2 encoder quality. Better software will get better results. You can test this by cpaturing your DV a DV-AVI file and then encoding with different demo MPeg2 encoders.

Don't forget #1.

I did not realize that the Z700 was not native.

As far as better quality software for encoding quality, what are better choices to use? I have tried Ulead Media Studio Pro 7 and Pinnacle Studio 8 & 9. I settled in on Pinnacle because it was easier to use.

Lets first make sure you are getting the best from Studio. Capture setting should be DV and 16:9 mode. Be prepared for 13GB/hr file sizes. Finish editing in DV. Then export to DVD using the higher bitrates and lower field first interlace MPeg2. You will need an additional 10-18GB blank disk space to pull this off.

Test your process on shorter representative files to shorten processing time. Find the right settings, then process the full tape.


http://www.camcorderinfo.com/content/Canon-ZR700-First-Impressions-Camcorder-Review.htm

Ok sounds good. Space is not an issue. I have plenty of drive space. 2 Hard drives, main drive is 120 GB and secondary drive is 320 GB (used for storage of music and capturing/rending video).

As far as "export to DVD using the higher bitrates and lower field first interlace MPeg2" I'm going to sound even more like a newbie here because I am not sure how to do this. I've used the "Disc" choice when under the "Make Movie" choice on Studio and went straight from editing to creating the DVD. Maybe this is where I am making the biggest mistake?

Should I create a Mpeg file of the project first?

Originally Posted by Top Dog

Ok sounds good. Space is not an issue. I have plenty of drive space. 2 Hard drives, main drive is 120 GB and secondary drive is 320 GB (used for storage of music and capturing/rending video).

As far as "export to DVD using the higher bitrates and lower field first interlace MPeg2" I'm going to sound even more like a newbie here because I am not sure how to do this. I've used the "Disc" choice when under the "Make Movie" choice on Studio and went straight from editing to creating the DVD. Maybe this is where I am making the biggest mistake?

Should I create a Mpeg file of the project first?

I don't have Studio installed currently so I'm doing this from memory.

"Make Movie" -> DVD
There should be encoder settings for DVD MPeg2. It is there that you choose progressive or lower field first and set bitrates.

Originally Posted by edDV

Originally Posted by Top Dog

Ok sounds good. Space is not an issue. I have plenty of drive space. 2 Hard drives, main drive is 120 GB and secondary drive is 320 GB (used for storage of music and capturing/rending video).

As far as "export to DVD using the higher bitrates and lower field first interlace MPeg2" I'm going to sound even more like a newbie here because I am not sure how to do this. I've used the "Disc" choice when under the "Make Movie" choice on Studio and went straight from editing to creating the DVD. Maybe this is where I am making the biggest mistake?

Should I create a Mpeg file of the project first?

I don't have Studio installed currently so I'm doing this from memory.

"Make Movie" -> DVD
There should be encoder settings for DVD MPeg2. It is there that you choose progressive or lower field first and set bitrates.

I don't find the choice toset it to DVD MPeg2. Under settings on "made disc" you can set the video quality to "Automatic, Best Video Quality, Most video per disc, and Custom" Your KB/S is greyed out unless you set it to "custom" at which point you can controll that.

If at anytime you select "Advanced" the only options you are presented with are "Bit Rate : Variable or Constant" "Draft Mode" or "Filter Video" and then you have "Progressive encoding" always "reecode entire movie" and "enforce strict DVD compatability"

Try custom settings.

CBR 8500 Kb/s
Lower field first
Audio MP2 ~224Kb/s (or LPCM if you want)

What else is there?
Max any quality settings.

Originally Posted by edDV

Try custom settings.

CBR 8500 Kb/s
Lower field first
Audio MP2 ~224Kb/s (or LPCM if you want)

What else is there?
Max any quality settings.

It will only allow me to set it to 8000 Kb/s.
I don't see anything on Lower field. Audio only gives me the choice of PCM, MPA, Dolby 2.1 or Dolby 5.1

There really aren't any more options under the "create disc" settings. Custom or Advanced. I pretty much covered them all in the previous post. There just aren't many choices on that setting. Now if I were to create a MPEG file, the setting and choices there are much different. There I can set the Bit Rate as high as I want, choose MPEG -1, 2 or 4 compression.. But that is giong to create an Mpeg file rather than buring straight to disc.

There's a Make File instead of Make Disc option.

I'm in the learning steps with all this too - taking HD on a DVR and outputting it through a Pinnacle capture box via S-video.

I captured a 2.5 hourish video using AVI (MJPEG) at 100% quality... ended up with a 94GB AVI file.

Then I downloaded and installed the Huffyuv codec on my computer because I read it was lossless on quality. I chose Make File in Studio, file type AVI, Preset Custom and in the Settings, I checked 'Show all codecs' so I could choose the Huffyuv one... set it at 100% quality and it's making the movie now.

I think I'm going to end up with an AVI a little larger than the original 94GB sized one.

Then I plan on using TMPGEnc to encode that new commercial-less AVI to a DVD compliant MPG file. I'll set it up to do a two-pass VBR and to do progressive encoding since the S-video output is 480i when I'm capturing.

I'll take that MPG file and use Ulead DVD MovieFactory to do my own custom menus and stuff, set the project to 16x9 so it burns it as an anamorphic DVD, and set the project to burn an 8.5GB DL disc to hold the 2 hours and 19 minutes of video and audio.

Hope it works... I'll let you know. Uses a lot of hard drive space for the initial capture and the first Huffyuv AVI output without the commercials.

Anyway, I thought you'd want to know about the Make File option so that you can try a different encoder instead of having Studio do the encoding.

Originally Posted by uscboy

There's a Make File instead of Make Disc option.

I'm in the learning steps with all this too - taking HD on a DVR and outputting it through a Pinnacle capture box via S-video.

I captured a 2.5 hourish video using AVI (MJPEG) at 100% quality... ended up with a 94GB AVI file.

Then I downloaded and installed the Huffyuv codec on my computer because I read it was lossless on quality. I chose Make File in Studio, file type AVI, Preset Custom and in the Settings, I checked 'Show all codecs' so I could choose the Huffyuv one... set it at 100% quality and it's making the movie now.

I think I'm going to end up with an AVI a little larger than the original 94GB sized one.

Then I plan on using TMPGEnc to encode that new commercial-less AVI to a DVD compliant MPG file. I'll set it up to do a two-pass VBR and to do progressive encoding since the S-video output is 480i when I'm capturing.

I'll take that MPG file and use Ulead DVD MovieFactory to do my own custom menus and stuff, set the project to 16x9 so it burns it as an anamorphic DVD, and set the project to burn an 8.5GB DL disc to hold the 2 hours and 19 minutes of video and audio.

Hope it works... I'll let you know. Uses a lot of hard drive space for the initial capture and the first Huffyuv AVI output without the commercials.

Anyway, I thought you'd want to know about the Make File option so that you can try a different encoder instead of having Studio do the encoding.

I don't understand your work flow. Are you editing in Studio?

Why convert to huffuyv when you already have MJPEG? Why not capture directly to huffyuv if that is what you want? Is this a limitation of your "capture box"?

You can edit MJPEG with studio can't you?

My normal path would be
HD cable box S-Video -> ADVC DV encoder -> DV-AVI file.

Well, there were two choices in Studio when capturing an AVI... MJPEG and DV... I chose MJPEG... I think I read somewhere that can be either lossless or near lossless? Should I have chosen DV?

And my 'editing' is just slicing out commercials.

The capture in Studio with my 510-USB capture box does not allow alternative codecs... only AVI with MJPEG or DV or MPEG 1/2 capture directly.

I didn't want loss on the capture, so I chose AVI. Then after clipping the commercials, I wanted to output it from Studio so a better encoder could do the MPG encoding... that's why I chose Huffyuv codec on the Make File output in Studio.

If there's a better way to be doing all this, please let me know.

Originally Posted by uscboy

If there's a better way to be doing all this, please let me know.

Better way probably. You are doing far too much conversion.

So, you have a USB2 capture device that only allows DV or MJPeg? That indicates the hardware in the box. The USB from the Pinnacle boxes uses proprietary drivers so that probably limits you to Pinnacle programs for capture.

Originally Posted by uscboy

Well, there were two choices in Studio when capturing an AVI... MJPEG and DV... I chose MJPEG... I think I read somewhere that can be either lossless or near lossless? Should I have chosen DV?

I would have. Think of MJPEG as an early version of DV. MJPEG (Pinnacle implementation) used a hardware JPG codec to compress video to 4:2:2 at variable bitrates (e.g. the Pinnacle DC series). 4:2:2 means half the bitrate goes to luminance and half to chroma. NTSC/PAL sources (e.g. TV, composite, S-Video) weigh luma bandwidth 3-5x over chroma. So MJPEG oversampled chroma at the expense of luminance detail for a given bitrate.

DV made two main improvements. By sampling 4:1:1 (NTSC and PAL DVCPro) or 4:2:0 (PAL DV/DVCAM) the luma/chroma balance changed from 50/50 to 66/33. Second, more advanced intraframe DCT compression was used for DV improving compression quality for luminance.

Today, MJPEG is mostly used for graphics compression for animation.

I don't know how Pinnacle is currently applying MJPEG. It may be different but both DV and MJPEG are far from lossless. Lossless 8bit YCbCr 4:2:2 runs near 270Mb/s (with audio).

Originally Posted by uscboy

And my 'editing' is just slicing out commercials.

OK. Since you must use Studio to capture from the Pinnacle device, use that to clip commercials.

Originally Posted by uscboy

I didn't want loss on the capture, so I chose AVI. Then after clipping the commercials, I wanted to output it from Studio so a better encoder could do the MPG encoding... that's why I chose Huffyuv codec on the Make File output in Studio.

Learn to live with loss. First you are using S-Video NTSC encoded, downsampled HDTV (major losses). Then you are sampling to DV or MJPEG (~5-7x compression losses). Then you are going to MPeg2 DVD (another ~3-5x compression losses).

Pinnacle Studio doesn't have the best MPeg2 encoder by far. You can export to external encoders using huffyuv, DV or other. Best to test the same reperesentative clip with various methods and encoders to find the optimal combination.

Originally Posted by turk690

to Top Dog, I think the problem is not with the progressive or whatnot; the problem is with the Canon.

I have an old Canon MV400i. Although it can also capture 16:9, I've always used it to capture 4:3 and have been more or less satisfied with the video. Then I recently bought another Canon MVX460i which has a wide LCD monitor and captures in 16:9 by default. I used the two in a recent shoot and used the MV400i in its 16:9 mode, the first time I did so, so that it matches that of the MVX460i when I edit.

I noticed that in 16:9 mode, the MV400i produced a "softer" video than it did when used 4:3. After doing a bit of research, I find that in while in both 4:3 and 16:9 modes the MV400i utilizes all of the pixels in its 4:3 CCD, in the 16:9 mode the signal is electronically "squeezed" horizontally and it is this wretched squeezing that produces, not only the softer effect I noted, but ironically, also some aliasing on slanted edges in the picture.

I also have an old Panasonic NV-DS28 which attempts 16:9 in a different way: it simply masks the top and bottom of its CCD. While the Panasonic indeed utilizes less of the pixels in its CCD in the 16:9 mode compared with in the 4:3 mode, the video it produces look better (to me, at least, even when I zoom it in 125% in Premiere to eliminate the black bars & match other 16:9 sources) than that produced by the Canon MV400i using all of its pixels in its 16:9 mode.

The Canon MVX460i, on the other hand, produces pin-sharp glorious video in 16:9, simply outclassing the other two, and carries all the way to DVD authoring. Until I bought it, I would have never attempted, and seen the subpar 16:9 performance of the MV400i.

To alleviate the soft MV400i 16:9 video, I put some sharpen filter at 40 in Premiere. Although it now looks passable, it becomes grainier & noisier. It seems 16:9 simply was an afterthought with this old camcorder. I may now have to get another MVX460i to continue better-quality two-camera shoots.

To see if your situation is the same, try capturing video in 4:3 with that ZR700, then editing & authoring a 4:3 DVD with it. Then to see it for what it really is, play it back with your LCD TV set to 4:3. Never mind the black bars on left & right for now; we just want to see if it's as soft in 4:3 as it is when shot with that ZR700 in 16:9.

The Canon ZR line is NTSC which is MV in PAL; the MVX in PAL is Elura in NTSC. The NTSC equivalent of my PAL MVX460i is the Elura 100.

There are 3 main issues with consumer 16:9.

1. Optics Quality - optics are circular. Diagonal sets the maximum horizontal perceived detail*.

2. CCD pixel mapping. Older/cheaper camcorders slice 16:9 out of the 4:3 pixel area thereby reducing vertical resolution (e.g. stepping on diagonals). Better camcorders use the same vertical CCD pixels and increase horizontal CCD area in 16:9 mode. Note that the optics must be good enough or there is no effect.

3. Usually overlooked is the processing between the CCD pixels and the 720x480/576 DV encoder. The quality of processsing varies greatly. Good processing will enhance picture detail and minimize artifacts.


* This gets technical but A/D sampling requires a low pass filter to minimize digital aliasing. Consumer camcorders use the plastic optical components as the low pass filter. This means they slightly blur the optics to reduce alias products and then sharpen the image in the digital processing. To see what aliasing looks like see here.
http://www.mathematik.com/Moire/
Then read this http://en.wikipedia.org/wiki/Moir%C3%A9_pattern

"A Few Good Men"
written by Aaron Sorkin

Jessep: You want answers?
Kaffee (Tom Cruise): I think I'm entitled to them.
Jessep: You want answers?
Kaffee: I want the truth!
Jessep: You can't handle the truth!

In other words, people create simplified models of a problem and then make assumptions based on the simplified model.

Originally Posted by edDV

Originally Posted by uscboy

If there's a better way to be doing all this, please let me know.

Better way probably. You are doing far too much conversion.

After thinking everything through again, that's kind of what I was thinking too. But I think I'll have the same # of steps unless I let Studio do the MPG encoding since I'll still have to capture DV from the USB capture box, clip the commercials, output DV from Studio using Huffyuv, encode with TMPEGEnc to .mpg file and then author the DVD with DVD MovieFactory for the menus and such.

But at least you're saying the capture will be better using DV and possible a smaller file size?

Regardless, considering a couple weeks ago I was trying to figure out even how to capture 16:9, I think that even though this process is lengthly, with a little testing, I should get a pretty good result on the video.

In fact, the DVD is burning now... TMPGEnc took 3 hours to do a 2-pass analyze and encode for the MPG file using some settings I found in a guide here. My only issue with it is that on my PC in Windows Media Player the video still is quite interlaced even though I chose Inverse 3:2 Pulldown... TMPGEnc didn't seem to to have a checkbox to make it progressive that I saw. The setting I probably wanted was Non-interlace I guess.

Originally Posted by uscboy

Originally Posted by edDV

Originally Posted by uscboy

If there's a better way to be doing all this, please let me know.

Better way probably. You are doing far too much conversion.

After thinking everything through again, that's kind of what I was thinking too. But I think I'll have the same # of steps unless I let Studio do the MPG encoding since I'll still have to capture DV from the USB capture box, clip the commercials, output DV from Studio using Huffyuv, encode with TMPEGEnc to .mpg file and then author the DVD with DVD MovieFactory for the menus and such.

But at least you're saying the capture will be better using DV and possible a smaller file size?

Regardless, considering a couple weeks ago I was trying to figure out even how to capture 16:9, I think that even though this process is lengthly, with a little testing, I should get a pretty good result on the video.

In fact, the DVD is burning now... TMPGEnc took 3 hours to do a 2-pass analyze and encode for the MPG file using some settings I found in a guide here. My only issue with it is that on my PC in Windows Media Player the video still is quite interlaced even though I chose Inverse 3:2 Pulldown... TMPGEnc didn't seem to to have a checkbox to make it progressive that I saw. The setting I probably wanted was Non-interlace I guess.

DV will be 13 GB/hr.

Most MPeg2 encoders will accept DV import.

I doubt Studio will do a good job at inverse telecine. Two options

1. Do it after editing and before encoding.
2. Leave the DVD interlace (bottom field first) and let your progressive DVD player or progressive TV do the inverse telecine. That is if either have inverse telecine capability.

This doesen't apply to anything coming from your camcorder. Are you capturing TV movies or VHS or something?


PS: Normal way to capture DV is by IEEE-1394 cable from the camcorder to the motherboard or an IEEE-1394 PCI card. Are you plugging your IEEE-1394 into the Pinnacle device? That is an unconventional way to do it but may work. If not, you are losing quality. Better to connect IEEE-1394 directly to the computer IEEE-1394 port.

Originally Posted by edDV

This doesen't apply to anything coming from your camcorder. Are you capturing TV movies or VHS or something?


PS: Normal way to capture DV is by IEEE-1394 cable from the camcorder to the motherboard or an IEEE-1394 PCI card. Are you plugging your IEEE-1394 into the Pinnacle device? That is an unconventional way to do it but may work. If not, you are losing quality. Better to connect IEEE-1394 directly to the computer IEEE-1394 port.

It's S-video and RCA audio out of a DirecTV HD reciever. Trying to archive college football games in 16x9... they're in HD on the DVR, but thank you hollywood for making it next to impossible (with my budget) to capture and record HD... so, I'm trying to get the best 16x9 S-video capture and encoding on a DVD9 disc I can...

OK, you are not original poster Top Dog. Gets confusing.

An HDTV foorball broadcast will be 1080i/29.97 (interlace) or 720p/59.94 (progressive) but will never be telecined. Exception is NFL Films which does everything in film.

So no telecine. The S-Video output from a cable/sat* box will be 480i letterbox. Your progressive DVD player or HDTV will do the deinterlace in hardware. Use a deinterlacing player like PowerDVD on your computer. Keep the MPeg2 bottom field first interlace.

Software deinterlaced 1080i -> 480i looks yuk like this

Yeah, my apologies to Top Dog... there was some confusion in this thread over settings in Studio and I just wanted to let him know that he could choose Make File and encode a file first instead of Make Disc.

As for the telecine, I have no idea what I was thinking... just picking settings based on something I didn't read well enough.

Anyway, that DVD ended up all messed up... movement was jerky and weird.

So I just let Studio encode an MPG from that large AVI (minus commercials) and picked VBR and Progressive in Studio's options and chose a data rate (7200kbps I think) that made the .mpg file 7.77GB in size.

The MPG is in DVD MovieFactory 5 right now burning onto a DL disc with custom menus... because the .mpg file was encoded with progressive encoding and VBR, I think the quality should be good enough, even though Pinnacle did the encoding this time instead of TMPGEnc. We'll see though.

Also, my S-video out is full 16:9 signal, no letterboxing, which is nice. I can change the settings to letterbox it or crop it or whatnot, but since my capture hardware supports 16:9, I just leave it as is and it outputs 480i, 16:9.

Full 480i 16:9 is nice. Which receiver is that?

I forget if Movie Factory 5 can edit commericals. Otherwise it will do everything you need short of editing.

Studio's job is just to acquire DV from the device, edit out the commercials and export to DV.

Originally Posted by edDV

Full 480i 16:9 is nice. Which receiver is that?

I forget if Movie Factory 5 can edit commericals. Otherwise it will do everything you need short of editing.

Studio's job is just to acquire DV from the device, edit out the commercials and export to DV.

MovieFactory 5 can edit out commercials, but it's interface isn't really that great... well, at least it wasn't in version 4... looks the same in version 5, but I haven't messed with it. It's just really easy to create custom menus with background images and music and stuff, so that's why I use it.

The DirecTV box is an HR20-700, their new HDDVR. A little buggy still for some folks - lots of software updates since release, but I haven't had the problems they've had.

The new DVD worked great... I'm sure the video could be a little better by using TMPGEnc to encode instead of Studio, but to me it looks about as good as SD video is going to look when you put 2:19:20:05 on a DVD9 disc.

Just glad it's all working... took way too long to figure everything out with all the problems I've had with this stuff. Now I'll just wait until they let you drag recordings over HDMI to your computer and burn HD-DVDs from them...

Yeah right. Ha!

In comparison to Procoder or TMPGEnc Studio encoder results are visibly less sharp. Studio would be good to capture (transfer from camcorder) edit and output back to DV (choose the same output format for your project - not MPEG2 !) then finished product encode your modified-edited DV->MPEG2 with a better encoder as suggested above. You get the best of both worlds that way: ease of use for editing (Studio) and quality DVD ready output.

Originally Posted by InXess

In comparison to Procoder or TMPGEnc Studio encoder results are visibly less sharp. Studio would be good to capture (transfer from camcorder) edit and output back to DV (choose the same output format for your project - not MPEG2 !) then finished product encode your modified-edited DV->MPEG2 with a better encoder as suggested above. You get the best of both worlds that way: ease of use for editing (Studio) and quality DVD ready output.

I think I am going to try that to compare by using TMPGEnc to encode. We'll see if it's visibly better or not.