[Dave Haynie]

This is totally my range of camera, and I’m currently drooling over a few options for around $2K that lead to 4K…

If you’re going primarily for event videography, I’d take a serious look at the AX100, particularly with a eye to whether that 1″ backlit sensor really delivers in low light, compared to DSLR and specifically the Canons’ 1/3″ sensor. The only thing I’ve ever wanted for in wedding videography was better low-light performance. And while 4K is compelling, may be a non-issue if you’re handing off video to a studio. On the other hand, you can shoot in 4K and downsample to 4:2:2 HD. Or just shoot HD for them and keep 4K for your own projects.

Audio’s an issue if you must record on-camera. I’d probably chain off my H4n, so I could use real mics without needing a Beachtek and keep the better quality Zoom 24-bit audio. You can buy an H4n for about the price of most of the straight XLR adapters.

-Dave

[Dave Haynie]

I like to do sound off-camera, personally. I use a Zoom H4n for simple things, which will allow two XLR mics and the internal mics, or the two XLRs plus an external pair via the 1/8″ TRS jack.

However, to actually answer your quesion, the only solution I know of is GoPro. I’m not sure of the features vs models matrix, but you can bring external mics I on some models (via an adapter to the USB jack) and you can sync multiple GoPro cameras via various means. Certainly a place to start.

When I shoot multicam, it’s always as a one-man operation, and I don’t have any sort of camera sync. My “trick” is to start the Zoom (or other recorder) first, and use that also as the time base for the whole shoot. Then I can start and stop cameras at any time (a practical issue if you use Canon DSLRs for video) and sync them when editing via Plural Eyes.

-Dave

[Dave Haynie]

Still pondering the Vegas 13 jump, and I don’t think this is a big issue for me personally either way. But I will point out that lots of pro programs, including CAD tools I use that cost 20x what we pay for Vegas, have movable/dockable/user-definable toolbars.

This kind of brings me back to the Windows 8 UI and the arrogance of many companies in regard to their treatment of professionals. We are pros for a reason, and a big part of that means we sit in front of PCs some weeks for 50-100 hours using software tools.

The implication of this is profound and should be embraced by developers: we know our use of your tool better than you do. Whether that’s Windows or Vegas or Altium, we pros use these tools more, and specifically more within our specialties, than the developers. Who are, after all, professional software developers. Not necessarily pros, and certainly not full timers, in the field of the specific application. In short, when we say a feature is important, they ought to listen. We know better.

-Dave

[Dave Haynie]

A “stand alone” re-muxing program is not going to mess with your PC’s multimedia subsystem the way a CODEC pack would. If it’s just re-muxing, there’s really no need to involve the multimedia subsystem (eg, VfW, DirectShow) at all. The better stand-alone transcoding programs will not add CODECs to the system, but use those already built-in or provide its own locally. If you have ever used VLC for playback, it’s that idea… VLC CODECs are similarly self-contained.

If you’re not familiar with the concept of “muxing” and what that’s all about, lemme ‘splain. There are standards for video and audio, like AVC, MPEG-2, AC-3, MPEG layer 2 audio, that just define a bitstream… just a series of bits. Those bits could travel down a serial wire like fire wire or USB, or they could come from some other source.

Computer storage isn’t logically organized around bitstreams, but rather, files. That’s a good thing… a bitstream is pretty much a serial thing… you don’t know how long it is, how to move up or down by 5 minutes, etc. And when you want both audio and video in sync, you need some means of storing them in a fixed relation of one another.

Thus the creation of multimedia “file wrappers”. When you see a file tagged .mp4, .mkv, .mts, .avi, etc. that’s a file with an internal structure that can organize multiple bitstreams in a way that makes them useful on a computer or other hardware. This file will “wrap” these bitstreams, and also include “metadata”, that’s data that’s about your data, such as time code, chapter markers, etc. That’s why it doesn’t make sense to discuss just decoding an AVI… because anything could be stored inside that AVI.

In the case of .mkv, that’ a file based on the Matroska file wrapper. This format was created as part of an ongoing effort among open source programmers to both avoid really poor open source wrapper formats (such as the Ogg file format) and avoid possibly patented formats, like those various MPEG formats. It’s based on a kind of binary XML specification. Unfortunately, until very recently, Matroska has had zero presence in professional audio or video, so it’s understandably not supported by NLE or DAW programs. That could eventually change, as Google uses a version of the format for their WebM container. Matroska has also for some reason become a favored format among media pirates, which shouldn’t diminish it’s qualities, but we do tend to assign guilt by association.

So anyway, to get media playback smooth, all of these formats break bitstreams into small packets, and interleave or multiplex them, audio and video (“AVI” actually just means “Audio/Video Interleave”). So when I say re-mux, the process is to break the individual bitstreams for audio/video from one file wrapper and put those very same streams into the new file wrapper. No re-encoding, no change in quality.

-Dave

[Dave Haynie]

Here’s the thing.. when I moved to 2K/HD, it was [a] because I’m just that kind of a geek, and [b] because I could get way better DV results from a modest-priced HDV camcorder than I could from improving my DV camera to any level I could afford. And then, when HDV became more important (I started actually using it for paid projects in 2006), there it was, already with a functional toolchain (well, ok, no ideal disc format yet), and me already knowing it.

I’m thinking 4K goes the same way. There are already 4K displays going for under $500…. this Seiki 39″ monitor is probably not good enough for your primary video-oriented display, but these things are big among computer nerds, used as desktop displays. The cameras are getting interesting: Sony’s 1″ sensor camera, the Panasonic GH4, the new m43 models from JVC, even that crazy-low-light Sony Alpha 7S, all seem likely to be within my rather modest price range… still haven’t decided if I’m every going back to a pure fixed-lens camcorder… lots of the DSLR workflow is improving, and lots of the stupid limitations are being removed.

Not there yet, but I expect to be there before it becomes critical. In fact, I’m shooting a short 4K (well, more like 5.5K) video right now… using a DSLR and an external shot controller.. a time lapse of my forest going from winter to spring.

The other thing about 4K… cropping. I used that quite a big in the DV days, where I could make cropping decisions in post on the HDV content, and you’d never know in the final DV result. That got me thinking that maybe I should always be shooting at a higher resolution than delivery. That seems to be where Red and some of the high-end guys are going, now that 4K is a regular deliverly format in cinema.

-Dave

[Dave Haynie]

[Chris Young] “Not sure what is happening with Nvidia but check out the following with Vegas 13 running realtime 4K with the AMD Firepro W9100”

I can do realtime 4K on my Radeon HD6970, at least for some formats. But of course, Vegas 12 and 13 ought to go faster with a faster OpenCL card. The only real hold-back is the Main Concept AVC CODEC, and I’m starting to wonder if I even care. At some point in the lifetime of this PC, I’m going to want a 4K monitor, and that means a GPU upgrade. Of course, the old card could stick around.. I have four GPU slots.

The Firepro W9100 ($3999, 16GB DDR5, 2,816 stream processors) is AMD’s “pro” version of a “Hawaii” GPU card, which is also found on the Radeon R9 290X ($630-$699, 4GB DDR5, 2,816 stream processors, 1GHz) and doubled on the Radeon R9 295X2 ($1499… two Hawaii processors on the same card, thus 5,632 stream processors, 1018MHz, 8GB DDR5). It is not obvious where the Firepro would trump the 290X running Vegas. Obviously, the extra RAM is a big concern for high-end mechanical CAD and perhaps 3D animation. And dual GPU doesn’t help in a single instances of Vegas.

There’s also the R9 290, which drops the stream processor count to 2560 (sounds like they’re binning Hawaii chips with slight defects) and the clock to 947MHz… for $400-$450.

-Dave

[Dave Haynie]

[Angelo Mike] “I never upgraded to 12, so I’ll probably get 13 to be able to edit 4k video.”

Depends on the 4K video. 4K was first supported in Vegas 9. However, if I’m editing 4K in AVC, I’m probably going to want Vegas 12 or 13.

I’m not terribly excited about the new features. That’s not always the thing, though. Most new versions contain hundreds of little tweaks, and that’s really where I find value. I wasn’t terribly excited about “3D” support either (that whole lack of any “3D” cameras, for one, and general annoyance as people saying “3D” when they mean “stereoscopic”). The tablet markup is useless for me until they support Android, but even then, not sure I’ll use it. And I’m just me, so collaboration is somewhat limited (guess it might have some value discussing edits away from the workstation) and proxy-first seems worthless to me. And I already own Hitfilm.

On the other hand, “fast” is worth money. “Bug-reduced” even more money (nothing’s bug free, but I do recall back when Vegas was the least buggy professional software I used in any discipline…)

-Dave

[Dave Haynie]

[Antonio Salva] “I had thought that Version 9 was AVCHd friendly.”

AVCHD support first appeared in Vegas 7. But it’s been an evolving thing, just as AVCHD (and other AVC formats) have evolved. There’s a chance that your camera’s video will be handled ok in Vegas 9 (I don’t know offhand), but it will be slower than in Vegas 12.

You could certainly TRY Vegas 9, given that you already have it. Some of that may depends on the camera’s vintage… you don’t get full AVCHD 2.0 support until Vegas 12 (doesn’t always matter, but can’t hurt). Is the new camcorder the NEX VG10 from 2010, or something different?

I also recall that Vegas 9 was occasionally flaky on 64-bit systems. Nowhere near as bad as the “preview” Vegas 8.1, and it had been awhile (I’m one of those idiots who upgrades pretty quickly to the new thing… nice thing, like most professional software, is that you can have multiple versions of Vegas installed at the same time.

For me, Vegas 12 was the first time AVCHD started feeling like DV again. An i7 and a good GPU also help make that happen.

[Antonio Salva] ” I just tried to copy files from my Sony Nex10 to the local drive of my new machine and it still will only make files that are just over 2 gigs. Is this a limit for my camera,because on the camera it only shows one file but when I go to copy it show multiple files. Since the local drive of my machine is Windows7 it should be NTFS, am I correct in assuming this?”

Ok, where to start. SD Cards define the file system as part of the media spec — that’s how you know you can always interchange them. SDHC cards (up to 32GB) specify FAT32, and as a result, no camera can write more than 32GB in a single file. Most will write 4GB files, but I have read that some Sony cameras have a 2GB per file limit. There’s a pretty good chance when you look on the camera, it’s listing things by take, not by file. So you just see the one take on the camera, but it’s broken into multiple files. Same thing happens on other media, like DVDs, to fit larger things within the confines of a limiting filesystem.

NTFS has been recommended for HDDs over 32GB since about the Windows NT 4.0 days, and certainly even for consumer versions of Windows since XP. Yes, some external HDDs come pre-formatted with FAT32. Since the SD cards use FAT32 as a universal format, they’re following that lead to work across multiple operating systems. SDXC cards are actually formatted with exFAT (aka FAT64), but older cameras don’t support exFAT or SDXC.

-Dave

[Dave Haynie]

I’m not sure it’s the sale to Parallax that’s the problem. Rovi bought Main Concept back in 2011. I can’t swear to what Main Concept is telling their paid-in customers right now. But looking on the web site, they’re supporting exactly the same set of circa-2011 GPUs that they did in Vegas 11 and Vegas 12. And nothing past that. So basically, nothing was improved under Rovi these past three years, far as they’re admitting.

Sony really needs to deal with this problem. Not to mention the coming need for HEVC and the WebM formats. You don’t fall behind all at once, it’s a little bit at a time, it just eats away.

The tablet interface stuff looks kinda cool, but it won’t matter to me until it drops on Android. I’ll see if it matters at other times.. I am a tablet user, but I don’t commenting/previeing on th tablet being important. Trendy, but useful?

That’s not something I’m currently using. Rendering to advanced video CODECs as fast as possible, that stuff’s regular everyday work.

-Dave

[Dave Haynie]

[Doug Jackson] “Anyone know if the release of Vegas Pro 13 will now support the NVIDIA Kepler-based GPU cards?
Perhaps ‘support’ is the wrong term. Maybe ‘take advantage of the Kepler architecture to deliver stable, high performance preview and rendering’ apparently other users of newer NVIDIA graphics cards experience in other software applications.”

Just keep in mind, there’s only so much Sony can do here. Consider the various issues:

OpenCL: Vegas itself is only OpenCL. OpenCL went to OpenCL V1.2 in November of 2011 and OpenCL 2.0 in November of 2012. I wouldn’t be surprised if both Vegas 11 and 12 were under the OpenCL 1.1 API. I’m not sure a great deal of the changes would be relevant to Vegas, anyway… it would be interesting to hear the Vegas engineers take on that. In short, OpenCL itself has been stable, even when manufacturers’ drivers haven’t always been so. But my experience with AMD suggests it’s completely possible… I have had no GPU issues at least since Vegas 12.

MainConcept: This needs to be updated, but hopefully also fixed for real. The current MainConcept AVC codec hard-wired the GPUs they felt like supporting. So, for example, OpenCL does nothing for nVidia, even though it ought to (if perhaps slower than with the CUDA option). And neither CUDA nor OpenCL works with any nVidia Kepler or ATi GCN GPU. It most likely could, these are just on its list. Vegas 13 desparately needs to add modern GPU support here, and ideally, arbitrary GPU support, since that’s actually the entire point behind OpenCL at least.

nVidia: there are two big issues with nVidia, which are totally out of Sony/Vegas’s control: their commitment to OpenCL, and their crippling of select performance features.

It’s been observed on various OpenCL torture tests that Kepler cards pretty much keep pace with recent ATi cards, eg, they hold their position, better or worse, as you go across benchmarks. But every so often, they hit a brick wall. It’s been conjectured that this might be due to nVidia favoring CUDA over OpenCL. But…

…As has been posted here, nVidia has intentionally crippled some specific OpenGL operations in their consumer/desktop versus workstation-class GPU drivers. This isn’t a case of the pro drivers being more “optimized”; there are clear cases in which some OpenGL ops run at 1/10th speed in the consumer drivers, and most of that speed can be regained by carefully coding around the bad ops.

You can see some of this here: https://www.tomshardware.com/reviews/geforce-gtx-titan-opencl-cuda-workstation,3474.html
The GTX Titan is nVidia’s fastest desktop card, and in some benchmarks, it rules. In others, particularly some non-gaming OpenGL and then OpenCL, it’s getting blown away by the high-end AMDs. But even by older nVidia cards. The degree by which it loses suggests some real issues with these cards as a system — whether it’s hardware or software. And that’s totally isolated from Vegas performance.

So these days, nVidia also sells, along with consumer/desktop “GeForce” and workstation “Quadro” cards, their “Computer” Cards, the Tesla series. As with all nVidia cards, at any technology node they’re basically using the same chips, with more memory and perhaps some slight tweaks on the higher-end parts. Given that they’ve clearly been crippling OpenGL on the desktop cards (something AMD has not done year, far as we know), it’s certainly not without precedent they’re doing the same, on both workstation and desktop drivers, to keep the prices higher for their “compute” cards.

Otherwise, just how to you explain my favorite benchmark of all time: page 21 of that review, 2nd benchmark down, “Surface Smooting”… my $300-three-years-ago Radeon HD6970 is 5.3x TIMES! faster than the $1600 (today) Titan. Sure, the HD6970 only beat the Titan on a few benchmarks, but then again, the HD6970 beat the Titan on a few benchmarks. That’s like an AMD Phenom beating an Intel i7-4960K… it just doesn’t happen. On any benchmark. There is something going on here, and not necessarily something Sony has any power to change.

-Dave