Which recording format

Per1 wrote on 8/5/2007, 6:52 AM
Hi,
I'm planning a imporant video recording and would appreciate some input regarding recording format as I'm not that "into" all these formats.

The end product will be a DVD, non-HD.
The contents is a pianist playing.

1. Normal DV / PAL 25fps / 50i.
Seems to work alright, although I feel the quality is not "top".

2. Normal DV / 24F
Hmm... will it work? Since the recording is music, will there be any trouble recording in 24F when the end product is a PAL DVD with 4% higher frame-speed? The sound is recorded separately with pro mics. and then sync'ed in Vegas.
Better image quality than "1."?

3. HD / PAL 25fps / 50i
I read something about problems with "black frames". Are they added when capturing or when rendering? Is recording in HD quality worth it compared to any processing problems and that the end product is standard DVD, not HD?

4. HD / 24F
As "2." above but with "HD aspects".


Quick and simple, non-professional, tests with 24F (in DV-mode on camera) gave a better image in Vegas than plain DV I think. When rendering it did not became very good. Lots of pixels. I might have done something wrong however in the Vegas process. There was no PAL 24p project setting and I could not use the HD24p (since shot in DV-mode) so I used PAL DV - perhaps that was the error.

In general I think my XH A1 (or any? DV) get the image to "glossy" - I'm looking for that real movie/pro-TV quality where the image is "softer" and more "clear", especially at high contrasts where I think most DVs fail compared to "film". Any clues how to achieve this?

All comments and help much appreciated!

Regards
Per

Comments

farss wrote on 8/5/2007, 4:51 PM
Shooting 24fps for PAL isn't going to achieve anything compared to shooting 25fps. However what you're looking for in terms of 'film' look has zero to do with the frame rate anyway, it's all about lighting and the latitude of the imager and the recording medium.

For SD delivery you'd do much better shooting 16:9 with a 2/3" Digibetacam camera or even a 1/2" camera like the DSR 570. For my money the V1P can produce a very pleasing image but it does love a lot of light to get the best out of it.

I don't know why you're getting "Lots of pixels", we need WAY more info to help.

Bob.
Per1 wrote on 8/5/2007, 6:40 PM
Bob,

Made some trials today with recording HD (50i) to tape.
When capturing in Vegas (from the tape) I get files with size 200 MB for 1 minute (.m2t files). Is this correct for HD? I was told it was 1 GB/minute or so. I'm capturing from an XHA1 with "Down conversion" turned off.

Per
rs170a wrote on 8/5/2007, 7:30 PM
200 MB sounds about right. HDV is the same size as standard DV.

Mike
PeterWright wrote on 8/5/2007, 7:36 PM
HDV takes about the same space as DV (13 GB per hour), so 200Mb per min is about right.

Shooting HDV for an SD DVD is a good way to go - careful lighting is the key, as Bob said - it's important to avoid the shiny ivory keys being too hot.

HDV for SD also gives you the option of later zooming in (Pan/Crop) within the frame without losing quality in the final pic.
farss wrote on 8/5/2007, 8:59 PM
And don't forget about the audio, HiDef images and LoDef audio are not the way to go.
MH_Stevens wrote on 8/5/2007, 9:32 PM
Record to capture the MOST information you can and then your edit possibilities are maximised. With music sound is doubly important so remember HDV sound is not the best. If you don't have independent sound recording and you don't want HDV, then shoot DV.

Mike
Laurence wrote on 8/5/2007, 10:51 PM
If you want pristine sound, just record to a separate audio recorder and to the camera both. That way it will be a piece of cake to line up the audio waveforms and replace the audio later.
Per1 wrote on 8/6/2007, 2:37 AM
Thanks all for input!

I was a bit puzzled why HD and SD gave the same file sizes.

Reading the specs for the XH A1 I find:

SD: PCM digital recording (48 kHz, 16 bits or 32 kHz, 12 bits selectable)
HD: MPEG1 Audio Layer II (Sampling frequency: 48 kHz) (Bit rate: 384 kbps)

Does this mean that HD sound is inferior to SD sound?

For "single-mic" performances it's "neat" to use the cams XLR inputs, but perhaps this is not the way to go when shooting HD, even for single-mic footage?

I always have the option of recording on a portable Korg 8-channel Harddrive Recorder. I guess that this is best when shooting HD, right?

Regards
Per
farss wrote on 8/6/2007, 2:45 AM
HDV uses mpeg-1 layer 2 compression for audio to save bandwidth for vision. It's not really bad if you keep the level hot (but not clipped) but when the audio is most of the show then using something better would be a good idea. If at all possible avoid using the incamera AGC when shooting HDV, for some obscure reason the HDV spec causes the level to be set at -20dB.
A 24/48 separate recorder with a good mic placed correctly is the way to go. Most acoustic pianos I've seen miced have used more than 1 mic too. If in doubt consult / hire an expert for the very best results.
Per1 wrote on 8/6/2007, 3:01 AM
OK, I'm now convinced that HD + external sound recording is the way to go (with "lower quality" sync-sound on the cam)

What was that issue with black frame that there are many discussion on?
Does these black frames occur when capturing or rendering to final format?
Is it software (Vegas) or hardware related?
Was there not some issue with "peak calculation" too in Vegas when working with m2t files?
Currently I'm using 7.0d as I have read that many says 7.0e is inferior in m2t-workflow.

(I don't want to get home from a shooting with HD material only to find that the trouble now begins in Vegas, a need for special programs to cope with any errors in Vegas etc...)

Regards
Per
farss wrote on 8/6/2007, 3:40 AM
What was that issue with black frame that there are many discussion on?
As far as we can tell only software related, the black frames aren't real but they can endup in the rendered output. One solution seems to be to reduce the Dynamic RAM Preview (in Options > Video) to 0. Of course if doing an incamers downscale to DV no problem.


Was there not some issue with "peak calculation" too in Vegas when working with m2t files?

Not heard of this one, shouldn't have much impact anyway if using double head recording.

Bob.
Per1 wrote on 8/6/2007, 4:03 AM
Bob,

> they can endup in the rendered output

Are they seen on the time line or are they invisible on timeline and output at random places in the rendered output? That would be a "nasty" behaviour of Vegas.

> "incamers downscale"

This means down-converting in the camera before reaching Vegas, right? I don't get full res. and possibility to zoom/pan then I guess?

When down-convert in cam - do I capture as DV or HD in Vegas? I've not tried that one yet, sorry if it is a stupid question.

What I've read, I can't find it, it was here on Forum somwhere, some had problem with the peaking (IIRC) having Vegas crash or similar. Some suggested to cut of some 2 secs. at the end of the m2t-file or similar. I never understood it fully, sorry not beeing able to supply more info. I think there was some issue of the length of the m2t-footage too (IIRC), but I cannot swear to it.

Hopefully there will be full support for HD in Vegas 8.
riredale wrote on 8/6/2007, 8:45 AM
Couple of thoughts:

(1) Regarding HDV audio--to my middle-aged ears, the audio is pristine. To see for yourself, run a simple test. Record something and split the line, putting identical feeds to an external recording device and also to your camera input. Then bring both audio recordings into Vegas, sync them up, and then invert the phase of one of them. If everything is precisely matched up (volume, time sync) then whatever you hear will be what is essentially added to the inferior recording device's track.

I've done this trick with mp3-versus-PCM tests and with Minidisc-versus-PCM tests just to determine what I'm sacrificing when dealing with a particular format. I think you'll be surprised how little difference there really is.

(2) Black frames--Suppose you buy a brand-new Lexus. Fantastic car, great engineering, luxurious interior. But every once in a while, when you close the door, the side-view mirror slips out of adjustment. To me, that's the black frames issue. It apparently doesn't happen to everyone, but in my case when I get a complex project with lots of clips and GearShifting, something happens and I will occasionally get them. You go in and patch them, or re do that section. Do I wish they weren't there? Yes. Is is a big deal? Absolutely not. They are fixable, and completely overshadowed by all the wonderful power and flexibility of this program.

HDV is terrific. Just be sure to use decent tape,as a dropout is a major event if you're recording something important. I use the blue Sony PRL tapes, which you can find for under $3 on the Internet. No verified dropouts in 40 tapes so far. For more assurance you can buy the "certified" brown Sony HD tapes for about $10 and/or you can record directly to disk or laptop. Recording to a second camera is always very desirable from an editing point of view, as well as having a backup. Recording audio to a portable device up-close to the activity is also very desirable--better sound quality as well as yet another backup.
Per1 wrote on 8/6/2007, 12:01 PM
Riredale,

Thanks for info.

(2). I like the car-analogy. A friend got a new BMW that opened the sun roof and all windows while parked and while rain was pooring for a full day. It was "flooded". BMW replaced the entire car, but he would have liked your analogy. ;)
I usually have 3-4 cameras on diff. angles. That gives me 3-4 "files" (or several files since the concerts are often 80+ min). I then use Excalibur to "Multicam" it all to one video-track in Vegas. I then use one of the cams inhouse sound to sync the pro sound recorded on the Korg to get my finsihed video and audio. So far so good. So, here is my questions.

2a: The 5-15 minutes sequences of music that I render out (from the total 80+ min) for "mounting" in DVDA as playlists, perhaps some of them will have the black frames. Is it then only to re-render until you are happy with the result? As it is small 5-15 min clips and not 2 hours re-renderings is not a big issue if this is the case. Is this a way to solve any "black frame" issue?

2b: Are drop-outs more serious in HDV (m2t) than "plain" DV? Since I'm using multical I can always select another angle or cut in a pan of some interior stuff... The drop-out do not "jam" the capture/video stream, or? Is image and sound just missing for a sec. or so?

2c: I got some of the brown Sony HD tapes (DVM63HDV) and 1 directly had a "sound peek" at 1:26 (tried it on several cams - all at 1:26). Well, accidents do happen in QA. The store sent me 2 new ones for free which was kind. I mostly use these tapes. They seem good.
BTW, 10 USD is OK, here in Sweden: 17 USD... Great...

2d: I find the cam->Drive interesting. I have streamed DV from cam to Vegas (Vegas capture). It went fine. It is possible to stream the m2t too to Vegas/Drive? If so, can one skip the camera down sampling or is that a "must" to cope with the "massive" information? The FireStore are a bit costly for me right now, although they seem nice. Do you get DV AVI and m2t directly on that and then just copy into PC later?
Have you ever experience "drop-out" or "jams" on Cam->Drive?
Even if pipeing to drive, do you record on tape alongside, for backup?