As of November 2010, we’ve been shooting tapeless for about 2 years, and over that time we haven’t lost one frame of video. I really wish I could say the same for tape, but that hasn’t been the case. We really see the tapeless workflow as safe and fast, and working with DSLR footage is just an extension of the tapeless workflow we developed with the Panasonic AG-HMC150.
Basically, we copy the files from the card to the hard drive and then bring those files straight into the timeline in Grass Valley EDIUS (our NLE of choice for many years) so we don’t have to go through the transcoding processing. With our DSLR footage, time-wasting rendering and transcoding in applications that don’t support native H.264 editing are totally eliminated in EDIUS because we can edit the MOV files natively off the timeline. Admittedly, this approach does add time at the export end of a project; it takes longer to export MOV files in your finished production than it does AVCHD files or just HDV files. But with today’s faster processors, the export render times have drastically dropped.
Let’s say that we have an edited project that’s all DSLR footage. We have all MOV files on the timeline. If we want to export that as a Blu-ray Disc, it takes about 5 minutes for every minute of completed project. If we’re just going to output to an SD DVD, it’s about a 2-1 ratio.
We do a lot of same-day edits where we show a 3–4 minute production during the wedding reception. For this we export an M2T, which goes faster than a 1-1 ratio and thus takes less than the 3–4 minute duration of the piece. This is especially impressive when you consider that for an SDE, we’re using a quad-core laptop, not the fastest i7 out there. If you’re doing it with a fast i7, those times will be quicker.
For us, it’s a great benefit in our DSLR workflow that EDIUS can edit the MOVs natively. Some NLEs are more system-dependent than others in terms of performance, but how much processing power you have is always a relevant question when it comes to rendering time. We have two different systems (and run EDIUS on both). One is our tower system, a relatively new Dell i7 with a 2.67GHz processor and 8GB of DDR3 RAM that we got for about $1,350. (When you’re looking at systems, you always want to make sure to go with the DDR-3 RAM.) We also have a 1GB NVIDIA GeForce graphics card. That system seems to work really well editing the MOV files natively.
The system we use for our same-day edits is a quad-core 2.26GHz laptop with 4GB of RAM and a 1GB NVIDIA GeForce graphics card. If we load up tons and tons of DSLR footage and add tons of slow motion or color correcting, or otherwise just overload it with effects, it is going to choke a bit and not play smoothly. It plays well enough for me to edit through it and to edit audio. Sometimes I might turn off a line or some titling to make it play a little more effectively.
That said, all we can do is sing the praises of EDIUS’ ability to edit DSLR MOV files natively and not have to transcode. It allows me a lot more creativity toward my production on a same-day edit, overall, in a short amount of time.
Acquiring and Editing Audio
When you’re doing short-forms, audio is very, very important, be it dialogue or even sound effects. Audio will always enhance your ability to tell a story. You can use audio to connect different parts of the day—maybe a story of some sort that ties them together. It can also allow you to condense time. Let’s say you’re editing a wedding and the minister tells a story about the couple that you can use during the footage of them getting ready. He actually said that in the ceremony, but you can condense time by using the audio elsewhere and cutting that segment of the ceremony. With that as preface, let’s talk about how we acquire audio with DSLRs. Obviously, a DSLR is not a video camera, so it doesn’t have the advantages of XLR inputs and VU meters and the ability to adjust your volume levels with actual physical dials. We have to remember that we’re taking this incredible video with a photo camera.
On most DSLRs, you can’t monitor the audio physically by watching meters or with a headphone jack at the same time. I know that there are some hacks that have been written where you might be able to get meters, but it’s nowhere near what we’re used to with the professional video camera. So that’s one thing we have to overcome. And then obviously the mic on a photo camera was not designed to capture audio the way we’re used to with regular video cameras. But you can add a shotgun mic, which will give you better results. And the Canon DSLRs we use do have one 1/8" mini-jack.
One option for a good on-camera shotgun mic is the Sennheiser MKE 400. It’s about $200, and it will greatly enhance your ability to capture natural sound with DSLR.
Since DSLRs do have a lot of limitations on the audio front, we use auxiliary recorders, the Zoom H2 and the Zoom H4n, when we’re shooting a live event. Our typical scenario for a ceremony would be to put Zoom H2s with little lav mics attached to them on the minister and the groom. Then, we’ll take two more Zoom recorders and, using their built-in mics to capture audio, put one on the podium for the readers and one on the strings, or any other important sound source. Then, we bring those into EDIUS and sync them up, which gives us much better audio than if we were actually trying to get good audio off of the DSLR itself.
During the receptions we shoot, a lot of the time we might get a feed off of the PA system. We’ll send that to either the Zoom H2 or H4n. We’ll also put a microphone on the PA system and send that feed through the wireless to our H2 or H4n. And one big difference between the H2 and H4n is that the H4n has dual XLR input. In theory, you could run a couple of wireless mics into the Zoom H4n. The Zoom H2 has a single 1/8" jack, but it is small enough to fit in the groom’s jacket. That’s why we’ve gone with that type of system.
Once we’re ready for the editing, we’ve organized our EDIUS filter palette to put the most commonly used audio filters toward the top of our effects bin so that we don’t have to dig through a bunch of filters that we don’t really use that much. We try to organize them so it’s very quick and efficient to be able to get to those filters. The filters we use most commonly are panpot and balance, which is for when we’re recording a mono signal to a stereo recorder and are only going to have sound on one side. We’ll switch it over so that we can get sound on both channels, left and right.
Then, we’ll use the graphic EQ, parametric EQ, etc. A lot of the time we’ll use the graphic EQ as almost a master volume. We can increase the volume by about 12dB gain. So if the signal is a little weak, we can do a master volume on that and tweak it up a little bit louder. Then, we can dial in EQs ranges that we need to either increase or decrease. And we really like using the parametric EQ. We’ll have some hiss or a lot of air conditioner noise in the ceremony, and we can roll off some low-end rumble with the parametric EQ.
Another favorite filter that we use just on any type of audio for a short-form edit is the master limiter. The master limiter is almost more like a limiter compressor. It’s going to back off the real hotspots but then bring up the softer spots. It seems to do a really good job of that without introducing very much noise. It’s a huge timesaver for us.
Here’s another good tip: Whenever we’re working on audio—say, ceremony vows—that’s going to be just one huge 24-bit WAV file on the Zoom recorder. In the finished short-form edit, we’re obviously not going to use all of the vows in their entirety. So we’ll chop those up to make it a more condensed version. But instead of chopping that up first and then applying filters, we prefer to get that audio clip at about 90% where it needs to be, EQ volumewise, with the mastering limiter on there. Then, we will go through and condense it. Applying those filters first saves a lot of time. We get the clip about 90% the way we want it to be and then edit it on it. You can certainly add the filters after the fact, but it’s just more time-consuming to apply your filter to one and then copy and paste it to all of the other clips. Then, as you continue to tweak it, you have to copy and paste that through the whole process. It’s much faster to get the clip closer to where you want it to be and then cut up the clip into shorter sections.
A final issue with audio is syncing up the sound from different sources with the footage in your edit. One of the best tools for audio sync—especially among DSLR users—is a plug-in from Singular Software called PluralEyes. It’s currently available for Adobe Premiere Pro, Apple Final Cut Pro, and Sony Vegas, and we understand that an EDIUS version is coming soon. In the meantime, what we do is listen for a really distinctive sound. If you’re editing a ceremony and you’re trying to sync audio to, say, an organ, it won’t work well because the organ is just not an instrument that appears as distinctive in a waveform. We’re looking for something that’s really distinctive that will pop out and be easy to sync up.
Beyond technical issues, the way you use audio in short-form edits can have a huge effect on the story your film tells. Through our video productions we are always looking to make a connection with our audience; we want to move them emotionally. If you’re doing work for nonprofits, you want to move them to invest in the organization.
We’re always looking for what stands out about a particular wedding day or whatever type of production we’re doing. With weddings, there may be a story about a family member, a dress, or some type of heirloom. There may be a toast that was given that day that really stands out. Right there, we know what we want to grasp, and we’ll use that as the basis for building our short-form.
In doing so, we’re going to whittle down that message to the core message or the key elements. We’re not going to use the whole sound bite. A lot of times, people repeat themselves, and we’re going to ditch those parts. We often use audio to bridge different parts of the story together.
For example, rather than watching the minister as he makes his opening remarks; maybe we’ll hear him as we show mothers, grandparents, and other relatives being seated. That gives us an opportunity to condense time. We take it out, so it does not appear in our short-form ceremony, and we just use it in a highlight clip. Another example comes from a bar mitzvah we shot recently. Rather than segmenting and fading to black between the time at the synagogue until the party later that afternoon and evening, we segued it with a comment made by the bar mitzvah boy’s grandmother asking him what was going to happen at the party and the little description he gave in response. Her comment, just a little sound bite, allowed us to transition right into opening up and showing off the cool party that they had that night.
In another wedding example, the bride and groom got up and gave a very long thank you to their guests for traveling from great distances, expressing how much it meant to them to have the guests at their wedding. It was a very cool speech, but it was a little bit too long for our taste to go in a full edit. So our idea was to cut back on the toast dramatically and use it in a time-shifted edit to show their wedding day. We grabbed portions and segments from lots of different parts of the day, in that regard, to make the final clip.
We use audio a lot in short-form editing to condense things and shorten things up. That is a very key element of the storytelling in our films.
Obviously, color correction can be a huge part of shortform edits, but we tend to preface any discussion of color correction by saying we’re firm believers in shooting it right. We always keep a white balance card with us, and we’re constantly re-white balancing and keeping aware of changes in color temperatures and adjusting for that on-the-fly. There are times when you don’t have time to redo the white balance, or you have natural light coming in, and someone flips a light on and turns everything orange.
These are times when we have to do color correction, but we don’t do very much of it. As far as actual color grading, we will tweak just a little bit, but it’s not something that we really spend a lot of time on.
One thing that we like about EDIUS is how simple the color correction is. Let’s say you’ve got a shot of the bride by the window, with natural light coming in. Then, someone walks into the room and flips the light, turning everything orange. With EDIUS you can open up the color correction box and, with the picker, move the mouse and select what you think should be white. Nine times out of 10, that balances everything out perfectly. If it’s not perfect, you can quickly and easily tweak the color to get it just right.
DSLRs vs. Traditional Video Cameras
One thing we’ve been asked often is whether we think DSLRs are the best choice of camera for wedding filmmaking. That’s a choice you have to make for yourself based on your own style and workflow. Truth be told, we have not made our Canon 60D our sole source for weddings. In our current workflow, we shoot with both the 60D and the Panasonic AG-HMC150. To match up the 60D with the AG-HMC150, we actually dumb down the DSLR a little bit to match the AG-HMC150 because the DSLR produces a more vivid color than the AG-HMC150. We’ve got it pretty well tweaked right now; we’ve adjusted our DSLR settings to resemble our AG-HMC150 and they match quite nicely. We definitely see some disadvantages to shooting solely DSLR. For example, when the bride is walking down the aisle, we do not have the zooming capabilities with the 60D that we have with our Panasonic AG-HMC150. We’re able to roll focus, but in the shallow depth of field, when the bride gets down the aisle to where her mother is standing, her mother is out of focus. So we don’t have the reaction shot from the mother of the bride that we feel is key and that we would have had with the video camera.
A lot of this comes down to personal preference. What style of film are you producing? Do you need to deliver a full-length production of, say, a wedding ceremony uncut in raw form? A DSLR’s 12-minute file size limitation will greatly affect you there. And if you’re doing other types of productions where you need to record more than 12 minutes at a time, whether it be a legal video, a school play, or a speaker’s presentation at a conference, that’s where the DSLRs have their big limitations.
Each of us needs to figure out what our needs are and which camera (or cameras) will best meet those needs. If we had to choose one or the other today, I think we would choose the Panasonic AG-HMC150 because of the limitations you have to overcome to shoot events with a DSLR. A year from now we might be all-DSLR. Who knows how things will transpire over time? Two to 3 years ago, who would have thought that so many of the top filmmakers in the wedding business would be shooting video with photo cameras?
One interesting development that may affect our choices will come in late December when Panasonic releases the AG-AF100, an interchangeable-lens video camera with a 4/3" sensor said to rival the imaging capability of DSLRs (watch for our review in the March issue). It will be interesting to see how that fills that niche and functions in areas that DSLRs aren’t able to cover.
It really is a personal choice. Today, our choice would be to do it with video cameras, as much as we would hate to give up the low-light capabilities and some of the artistic stuff we can get with the DSLRs. As we said at the beginning of this article, a DSLR is a tool, and you need to know your tools. So that’s where it comes back to: You have to know what’s best for you and what you want to get out of it, and choose the tools you use accordingly.
Mark and Trisha Von Lanken (info at vonweddingfilms.com) run Von Wedding Films. Five-time EventDV 25 honorees, WEVA Hall of Famers, and producers of the EventDV-TV series Von Real, they are regular speakers at WEVA Expo, winners of numerous WEVA CEAs, and were “megasession” presenters at In[Focus] 2010. Several times each year, the Von Lankens offer intensive 2-day workshops at their Tulsa, Okla., studio.