[Dave Haynie]

Upscaling means the player plays a normal everyday 480i or 480p video from a normal everyday DVD, but runs it through some realtime video DSP algorithms to “upscale” to 1080i or 1080p. What this means is that, first, there’s an resolution upscale, followed by some line detection and sharpening algorithms, maybe some magical proprietary stuff, to make your eye think the DVD is higher resolution that it actually is.

Every HDTV does the same thing when presented with material that’s not in its native screen resolution. However, particularly in the past, HDTVs often had just plain bad upscaling algorithms. So DVD players included presumably better ones. At least, that’s the nice and useful explanation. The other explanation is that DVD player manufacturers needed a new thing every year to make their DVD players sound better than last year’s. So one year they added 480p output, another they added HDMI with upscaling, then Blu-ray… pretty much a feature of all DVD players over $50 these days, and finally, “smartTV”… they also hook to the internet and let you watch YouTube or Netflix or whatever.

Upscaling really can be a nice thing, when done well. I’ve been using my Sony PS3 as my primary DVD/BD player since it came out. At some point along the way, they updated the upscaling algorithm, and it was a noticeable improvement. But upscaling is all post-processing, it has nothing to do with what a player will play, only how it can deliver it to your TV.

-Dave

[Dave Haynie]

Your best render will be rendering MPEG-2 (widescreen, assuming its show in HD) from your original sources. There is always a small loss in quality going through multiple rendering steps… original -> 480p is better than original -> 720p -> 480p.

-Dave

[Dave Haynie]

[Ron Ressler] “I just bought a Philips HD player, a typical player I got from Best Buys. It supports about everything except Blu Ray. I want to render .AVI 720p files and burn to DVDs. I do not want to use Blu Ray because my productions are not that long and I want to be able to sell my DVDs without “player-limitation”.”

Well, you’re already asking for the worst kind of “player-limitation” here. AVI is just a file wrapper — you can put most anything inside it. A PC can, of course, load up new video CODECs for support different AVI file contents. A DVD player cannot. Most DVD players don’t support any kind of AVI file. Those that do may differ in what they support: standard definition DV, DivX (which comes in several different formats), etc.

You can put anything you like on a DVD… it’s just a data disc. However, DVD players only know how to auto-run DVD-Video formats, for the most part. A DVD-Video in NTSC can be 720×480, or a couple of lower resolutions. That’s it. You can put other files on the DVD and most DVD players will just ignore them, but they won’t decide to just automatically play some random AVI file.

[Ron Ressler] “I’ve been reading into the specs of DVD Architect 5.0 and all I see is Blu Ray this/that. WILL 5.0 allow me to burn my 720p .AVI to a leading brand DL DVD, which would then be playable on my Philips or any common HD player?”

DVD Architect, like pretty much all DVD/BD authoring programs, only makes legal formats. You can make a DVD-Video file structure or a Blu-ray file structure. DVD Architect will happily burn the DVD file structure for you on to an appropriate DVD blank. It can also burn a Blu-ray file structure to a BD-R or an appropriate DVD, if the content is small enough. However, no DVD players and even most Blu-ray players will not play a BD file structure burned onto a DVD. They just won’t.

No DVD players but most Blu-ray players WILL play a DVD formatted as an AVCHD disc. AVCHD is, of course, the camcorder format derived from Blu-ray, but not identical to Blu-ray. DVD Architect will not create an AVCHD file structure, but you can create a simple one from the appropriate AVC files, using open source tools. But given that this only plays on Blu-ray players (or PCs with Blu-ray player software), you might as well just distribute on Blu-ray.

Before Blu-ray, I had a red-laser DVD player from Japan, one made by a company called IOData, that could actually play lower complexity HD formats. It could auto-play Microsoft’s WMV/HD format, and if I made a “hybrid” disc… basically the file for both DVD and WMV/HD, it would offer me the choice of which format to play. This would play as a DVD in pretty much every other DVD or Blu-ray player known to mankind, though it would play on Windows PCs. Sometimes. Clearly not something I ought to charge people for.

Given that some DVD players have trouble playing DVD-DL/DVD+DL discs, these discs are way more expensive than DVD5s or even BD25s, you’d be way, way better off doing it the way Hollywood does: sell a 2-disc set, one DVD, one BD. That’s the only way you’re going to be able to get HD video to any but the very fringe of users.

-Dave

[Dave Haynie]

I guess it depends on what you’re trying to do with the lights, and the lights themselves.

There’s a subtopic in MIDI, called MIDI Show Control, that has established MIDI messages for controlling things other than music.

Most lights themselves run a protocol called DMX512, which is faster than MIDI and run over differential pair (RS-485). But some light controller boards will accept a MIDI connection for MIDI Show Control interfacing. You’d have to know the particulars of your board to do this effectively.

Vegas isn’t the right tool for MIDI. If you’re sequencing music, you can including some show control commands in the same sequence but a different MIDI port, and run one MIDI cable to your instrument, one to the lighting board. Acid, Sonar, most DAWs can handle this kind of thing, but they don’t necessarily have a direct understanding of MSC.

There are a bunch of programs designed specifically for MSC. Some are listed under the MIDI Show Control page on Wikipedia: https://en.wikipedia.org/wiki/MIDI_Show_Control.

If I needed to do something fast and dirty, taking into account time and maybe events from a MIDI trigger of some kind, I might write it as a ChucK program… ChucK is a computer language specifically designed to deal with time and music. It can deal with events timed in milliseconds, minutes, hours, or days. Not an instant solution, and not a good place to start unless you’re already a programmer.

-Dave

[Dave Haynie]

Your problem is certainly in decoding performance. MP4 is actually just a file wrapper, but the expected content is AVC video and AAC audio. The MP4 file wrapper itself is derived from Apple’s Quicktime, it is not a problem itself as an editing format. Cameras generally resort to generic MP4 to deliver something that exceeds established formats. MPEG-4 is really a collection of technologies used to create complete specs, not any kind of end-user spec itself. Vegas does edit most MP4 configurations, but there’s no guarantee.

AVCHD is a specific format, originally derived from Blu-ray by Sony and Panasonic, as a camcorder recording format that could play on Blu-ray players. For 1080p60 at up to 28Mb/s, you’re talking AVCHD 2.0, which won’t necessarily play on BD hardware. This is another container format, MPEG-2 transport stream (same thing used for ATSC television), holding video as AVC and audio as AC3.

Anyway, the thing that trips you up is AVC encoded video. This is very CPU intensive to decode. Using the best decoders on my old Intel Core Duo 2.4GHz laptop, I saw about 95% CPU used just to come close to a realtime decode of 1080p60 28Mb/s video. Using the GPU under Windows 7, this was reduced to about 8% CPU. I point these out to illustrate that you can’t really judge Vegas preview performance from anything you see in a player.. Vegas doesn’t used those APIs for preview. And secondly, you can count on needing that whole Core 2 Duo 2.4GHz worth of computer power, minimum, just to decode a single AVC stream in realtime in Vegas. Which means, if you don’t have more than that, you really can’t expect AVC editing to work very well. This is why quite a few of us here have higher end systems.

Sony recommends a 2.0GHz multi-core CPU for any kind of HD editing. I’d push that to at least four cores and 3GHz for effective AVCHD or 1080p60 editing. Keep in mind that whatever your resource use at 1080p30 or less, doubling that frame rate also doubles you resource use.

Short of a faster PC, you will probably want to change something to edit well on your existing PC. Transcoding to an intermediate format is one way.. you convert to a simpler, but probably much larger, video format. Some use Cineform for this; other options are high bitrate MXF MPEG-2 (built-in to Vegas), or Avid’s free DNxHD. It might also be possible to use an I-Frame only AVC, which is a format class (“AVC Intra”) you find on newer cameras (my Canon 6D does this).

The other way is a proxy edit.. this goes way back. PCs in the 80s couldn’t handle full rez video, so you’d convert to a preview format, edit that, but swap in the final for rendering. This can be done manually, just by rendering out simpler video, then changing files around. And there are aftermarket tools for Vegas, like GearShift, that automate this.

AVC delivers about twice the coding efficiency of MPEG-2… a 28Mb/s AVC stream should be transcoded to at least 56Mb/s MPEG-2, though its common to use 50Mb/s. The complexity of decoding AVC is much more than 2x that of MPEG-2. But enjoy it.. the next big thing may be HEVC, which can deliver twice the coding efficiency of AVC, but at many times the complexity.

-Dave

[Dave Haynie]

As with any other USB audio interface, you can get much lower latency IF that USB mic has ASIO drivers available. I have two different audio interfaces that use USB, but they’re both shipped with device-specific ASIO drivers.. down to 5ms or so latency, with minimum buffering. Many of the USB mics are “generic” USB audio devices, which work with drivers included in every OS. So they don’t necessarily ship with any drivers, and that means, by default, you’re probably using an MME (MCI?) driver for that sound device. It’s not just that the driver has a high latency (it does), but it’s run through Microsoft’s audio bit-rate-conversion and software mixer… that’s the reason every program on your PC can play out your default sound card all at the same time… each thinks it’s talking to a sound card, but it’s really just a channel on the mixer. ASIO drivers are for music and other high quality sound, and you can adjust parameters, like latency, right on the driver setup panel in Vegas, Forge, etc.

Also, for voice-over, if you’re syncing to video, you only care about record latency, not round-trip latency. This can even be mitigated as long as it’s PREDICTABLE record latency. For example, if it’s always 200ms, your voice-over is going to be terribly out of sync with the video. So slide it ahead 200ms, and you’re golden. Some recording programs can compensate for known latency.

If you’re just using the PC as a recording deck, you don’t care about latency at all. When you use a PC as a musical instrument, then round-trip latency is the problem… I have to put a signal or sound into the PC, process it, and it get back out fast enough to not introduce an audible lag.

But you really don’t want to use two USB mics anyway. The problem there is that they may each have their own clock. It’s possible for a USB audio device to phase-lock to the USB clock. That’s actually kind of ok, because USB clocks need to be something like 25ppm or better, versus the typical 100ppm of a generic PC clock. But if they don’t, you wind up recording two tracks from different time bases. Some DAWs can compensate in software, others get confused.

Plus, by the time you’re talkin’ two mics, you’re better off buying a sound interface and two “real” microphones.

-Dave

[Dave Haynie]

[Søren Sørensen] “(i) Microphone.”

That’s a big question. To help narrow it down, you have to consider its use. Is this connecting to a camcorder? A dedicated recording device (Zoom, etc)? A PC — what sound interface? Do you have a professional XLR jack with 48V phantom power, or are you limited to things that work with a consumer 1/8″ jack and probably some kind of battery in the mic.

Your best bet is a professional mic of some kind. If you’re doing voiceovers for already-created video, I’d recommend, well, what I use for that: a large diaphragm condenser mic. You can something like a MXL V67G or an Audio Technical AT2020 for under $100 that’s just dandy — good enough for music (I have both of these). You’re going to sit in front of this mic when recording, a couple of feet away.

If you’re looking for something on-camera, your best bet is a wired lavalier mic. Lavalier mics aren’t really for music, but great for speech — you’ll see them, occasionally, on practically every talk show. They’re recording very close to the speaker, and of course, if applied correctly, they’ll pick up the speaker as she moves around… a big advantage over any other kind of mic for interview situations. I have both the MXL FR-350 omni and the MXL FR-351 cardioid…. these sell for about $99 each, but mysteriously, MXL sells the FR-355K “Interview Pack” which contains both, and can be found online for $99.

If you can’t wire-in, there are wireless lavalier solutions. They’re not cheap, particularly if you want one that’s not going to pick up electrical noise. The cheapest one I’ve used (not mine) that I’d trust is the Line 6 XD system… you can get a lavalier, transmitter, and receiver for about $300. This is a digital system, using the 2.4GHz ISM band, so there’s no license, or worries, other than maybe channel overlaps with local WiFi gear. Or microwave ovens.

Aside from a lav mic, sound folks use shotguns. I have a couple that are better on-camera than, well, other kinds of mics on-camera. The worst possible mic to use with a camcorder is the built-in one. These should never, ever, ever be used for audio — ok as a time sync with your other cameras (you sync mult-camera shoots via audio). The only things worse are the mics built-in to smartphones and DSLRs, in that order. A shotgun mic will get your speaker, but it’ll also pick up way more room acoustics than you’d want in a typical voiceover. You want that sound as dry as Arizona in July; you can add reverb yourself if the situation calls for it. Better, if you have to use the shotgun, is to put it up on a boom pole, and have a sound guy work it, keeping it on the speaker, just out of camera range. You can’t do this by yourself… ok, I have, but it’s not recommended. And that’s actually why I bought the lav mics. I have a Rode NTG1 and an MXL FR-304 shotgun, and something that came with my Sony HVR-A1 that I basically don’t use anymore. The NTG1 is a good on-camera mic, also a little out of your price-range, the FR-304 is better suited to boom pole or mic-stands… it’s 14″-something inches long. There are tons of shotguns made, and they’ll work, but the lav is what you want for on-camera interviews if you’re doing them seated, not “run and gun” or some-such.

If you do any walking around shooting, get two mics on the camera. Set one at low gain, one at high gain, and protect them from the wind with the foamy socks that come with them, or a dead cat if you can get those for your mics. Or a blimp, if you have it on a boom pole… ok, probably getting off-topic here.

As for USB mics, I’m kind of against them. Yeah, ok, if you only every need to record audio in front of your PC — you’re a blogger/vlogger or something, it’s probably ok. A USB mic is, after all, some kind of real mic connected to a monophonic (and not necessarily all that great) USB sound card. They’re literally adding a couple of bucks in cost and $50-$75 to the price of that same mic you can probably buy with an XLR interface. Trust me on this… at my day job, I designed just such an interface last spring… about $3.50 in parts, and that’s only because I put in a booster amp for the headphones, which needed two $.25 coupling capacitors. But if that’s your setup, I know the AT2020 is available in USB-mic form. You can certainly also buy a PC audio interface with XLR, phantom power, and not totally heinous preamps for around $100. My Focusrite Scarlett 2i2 was slightly more, that’s my regular PC sound interface (the 16-channel Tascam comes out on special occasions), and I’ve seen basic units for under $70.

[Søren Sørensen] “(ii) Audio program”

Sound Forge is really kind of overkill if you’re just working on sound for video. I do nearly everything in Vegas… in fact, I’ve mixed whole albums in Vegas — it’s not just for video. Forge would be handy for things that really need to be modified; maybe I want to normalize the whole track or something — there are more options in Forge than Vegas, that kind of thing. I think of Forge as an audio editor and Vegas as an audio sequencer… well, my brain grew up doing sound and music. Vegas being non-destructive is great for most of the things you might do with voice: reverb, compression, etc.

-Dave

[Dave Haynie]

Sure hope you didn’t order 1,000 DVDs without proof approval!

If one knows what they’re doing, a good 1080p24 original should yield a fantastic DVD. First things first: how and where do you see interlace lines? Is this on a CRT, a modern digital TV, a PC, just what? And I take it you’re see this as a motion artifact.

Ok, so there are a few challenges in making a good DVD. First, it needs to be cleanly downscaled from 1080p to 480p. Most video editors will do this well. And going from a progressive source is best, hard to screw up. But you can.

Second thing is matching your video to the DVD format, usually done on render to MPEG-2. The proper thing here is to render to DVD’s “NTSC Film” format, which encodes a 24p source with flags, so a 24p capable player can play it at 24p, while a 60i player will know how to telecine the video to look good at 60i. Pretty much all major DVD films are done thus way, one reason DVD looks pretty good on a modern TV. If they did this, it should look good. And its otherwise hard to mess up. You can see pulldown judder on a 60i display if you know what to look for, but no worse than any other DVD.

So let’s assume a screw-up. Best way to diagnose a problem is to understand how things break. One way to screw thus up is to render the 24p to a real 60i output. Two things can happen. If the rendering allows interpolation, you’ll have your 24p video resampled to 30p, then interlaced. This will smear things a bit, but given the higher starting resolution, you may not notice too many terrible things.

Now, if the 24 -> 60i conversion happens without interpolation, each field of the video will be drawn from part of the 24p original. But tge program doesn’t necessarily do an intelligent telecine here. So you will see odd field transitions, uglier video.

But it gets worse. If they got the interlace field order wrong — very hard to do, particularly from a progressive source, but possible. With this, you can see backwards judder… the N+1th field is from the Mth frame, the Nth field from the M+1th frame. This is unwatchable, any motion will judder.

-Dave

[Dave Haynie]

Vegas doesn’t have any use for all that RAM. That’s pretty much it. Adobe stuff is pretty well known for being memory pigs… not that this is much of a problem on a 64GB PC. Mine too… but that’s for photography. I haven’t found any use for that much RAM in video, but I’m also not an After Effects user.

As far as CPU use, Vegas should be able to nail your CPU to over 90% use, as long as the CPU is the bottleneck in the system. If you’re using the GPU as well, that’s one thing that’ll very definitely lower your CPU peaks, simply because there’s a communications delay between CPU and GPU, and unless Vegas fully pipelined that (eg, the CPU is never actually waiting on the GPU to finish, but moving on to the next thing), the CPU’s going to sit idle some of the time waiting for the GPU.

RAM can also potentially be a bottleneck, particularly as your number of CPU cores starts to grow faster than your memory bandwidth. It’s also technically possible that things slow down when you have Hyperthreading enabled on an i7 system. But that’s very job dependent.

And of course, your storage can be the bottleneck pretty easily. In the ideal case of an SSD, much less, but for HDDs, even fast ones, when you’re reading and writing multiple streams in a render, the drive’s seek time can be a problem. And you’ll see this more on a very fast machine; does the CPU work finish before the HDD work or not? This is fixed by spreading assets around. It’s not the problem in a simple project.

-Dave

[Dave Haynie]

I’ve been off Dish for a year (switched to DirecTV), but their quality actually went up a bit, HD-wize, once I was switched over to one of the new satellite slots (big dish that actually hits three orbital slots at the same time, same basic thing DirecTV does as well these days).

720i… no, I think that’s essentially impossible. Do you actually have transport stream captures that claim this? The problem is simple: the downrez gear everyone uses does its conversion on the TS stream, where it can lower bitrate and line resolution, but can’t do a proper frame conversion. So going from 1080i (which is practically all the material they receive from the providers, though sure, a few are based on 720p60, maybe still ESPN, not sure about ABC/Disney, and I think NatGeo) to 720i would completely botch the interlacing. While there are virtually no consumer televisions that can even display an interlaced signal anymore still for sale (progressive-only displays; they upconvert all interlaced), that would still be unacceptable. And 720i60 would be no savings over 720p30.

I know that DirecTV and Dish commonly downrez to 1440×1080. They consider that a freebie, because so much is shot at 1440×1080 anyway (DVC Pro, etc)… ABC was a big advocate for that, too, claiming that 1920×1080 was too artifacty for ATSC. Of course, ABC was also strongly lobbying for 720p in those days, pushed by their connection to ESPN.

Some channels get 1280×1080, and I know Comcast was caught on that, too, some years back. Not sure if they still do. FiOS doesn’t downrez… the have no need to… they essentially have 3x the bandwidth of any cable provider.

The other part of the equation is of course the bitrate… which is usually below 10Mb/s for either satellite provider. Still better than the 4-6Mb/s for 1080p from YouTube or the 3.6Mb/s “HD” from Netflix (though their more recent “SuperHD” is supposedly at up to 7Mb/s for 1080p24). Supposedly though, some Dish HD channels are as low as 5Mb/s… presumably at 1280x1080i60, or you’d certainly drown in artifacts. Given this is AVC encoding these days, you’d need 7-9Mb/s to match what you’re seeing in 1920x1080i60, OTA, MPEG-2… depending on how overloaded the local OTA channel is.

I’ve never heard of 720p60 stuff being downrezed, but of course they’re lowering the bitrate, and it wouldn’t be a hard trick to convert that to 720p30 in the inserter.

It’s also funny that satellite and cable companies are now advertising “premium” 1080p content. By which they mean 1080p24, which is of course a lower bandwidth signal than either 1080i60 or 720p60. I suppose the could improve the bitrate and maybe not downrez to 1440 in order to maybe deliver a better picture.

-Dave