Video Slave or Pro Tools for running picture?
- Thread starter Fitz
- Start date
How can I get the audio from video slave back into my daw? Like If I want to solo SFX or DIA in my daw while recording
charlieclouser
Senior Member
I use a Mac Mini 2012 for Video Slave and I just have a 3.5mm stereo > 2x 1/4" cable going from the headphone output on the Mini into a pair of analog inputs on the MOTU audio interface on my DAW machine, and then I can monitor it via the MOTU CueMix software on the DAW machine without worrying about it accidentally getting included in a bounce. But if I do want to record the audio from VideoSlave into my DAW, it's already coming in on analog 7+8 so it's easy.
when you finish a V.01 cue, do you print stems into Pro Tools or do something with Video Slave?
charlieclouser
Senior Member
In my setup VideoSlave is used for synchronized playback only. I have the picture team output my movie files with DLG+SFX mixed together on one audio channel, and the temp score on the other. I do import the movie into Logic at the start of the project just to verify that it is frame accurate with the same movie playing in VideoSlave and with all of the timecode counters in the room, then I delete the movie from Logic so it's quicker to load cues and I'm not unnecessarily duplicating the movie file to each project folder. From that point on, VideoSlave just sits there ready to play when it sees MTC.
Even though VideoSlave has a bunch of great features for importing the music to its timeline and then exporting a movie file that includes the previous DLG+SFX alongside the newly-imported music, I have never done that.
I print my stems to ProTools running on a separate machine. If the directors + producers want to preview and approve the cues in progress, I send WAV files to the picture editors, who then place them on their timeline and output movie files.
The reason I do that is because it's "not safe" for me to be uploading movie files to DropBox or using WeTransfer or whatever - studio / tv network security does not seem to like that. So the picture team uses secure streaming solutions like Pix or Dax that allow them to protect who can stream and view a file, and see a comprehensive audit trail of which team members viewed what version of the picture on what date. It even shows whether they watched the whole thing or skipped around, only watching certain sections! Very cool.
Since there are so many picture edit and VFX elements that are constantly changing and need approval from directors and producers, these secure streaming solutions are already in place and in use by the whole picture team, and directors + producers already have logins and are often watching previews on their iPad or whatever. So having the picture team integrate score previews into their established workflow is the easiest for the producers. It does mean a little more work for the assistant picture editors, but it means less work for me! I love it this way. Much better than the old days of having my guys output a preview to VHS / DVD / QuickTime.
Even though VideoSlave has a bunch of great features for importing the music to its timeline and then exporting a movie file that includes the previous DLG+SFX alongside the newly-imported music, I have never done that.
I print my stems to ProTools running on a separate machine. If the directors + producers want to preview and approve the cues in progress, I send WAV files to the picture editors, who then place them on their timeline and output movie files.
The reason I do that is because it's "not safe" for me to be uploading movie files to DropBox or using WeTransfer or whatever - studio / tv network security does not seem to like that. So the picture team uses secure streaming solutions like Pix or Dax that allow them to protect who can stream and view a file, and see a comprehensive audit trail of which team members viewed what version of the picture on what date. It even shows whether they watched the whole thing or skipped around, only watching certain sections! Very cool.
Since there are so many picture edit and VFX elements that are constantly changing and need approval from directors and producers, these secure streaming solutions are already in place and in use by the whole picture team, and directors + producers already have logins and are often watching previews on their iPad or whatever. So having the picture team integrate score previews into their established workflow is the easiest for the producers. It does mean a little more work for the assistant picture editors, but it means less work for me! I love it this way. Much better than the old days of having my guys output a preview to VHS / DVD / QuickTime.
Simon Ravn
Senior Member
I run the video Inside Logic. Dont see any reason to have a separate computer for that today.
charlieclouser
Senior Member
Hey, whatever works for you.
I started using outboard video players on a separate computer about 20 years ago with VirtualVTR, back in the era of uncompressed DVC files that were like 10gb for 42 minutes, and when playing the video inside Logic put a serious strain on the slow-ass G4 computers we had back then. These days the computers are fast enough, and video codecs are efficient enough, that it doesn't really stress the computer as much as it used to, which is a good thing for sure.
But the main reasons I still use VideoSlave are:
- When working on a film that is divided up into multiple reels, I can have a playlist in VideoSlave with each reel set to start at a different SMPTE hour, just like on the dub stage, and whatever cue I pull up in Logic will trigger the appropriate file without me touching anything. If I want to try a cue at a different point in a different reel, I just change the SMPTE start point in Logic and the other reel's video will play. For tv this isn't usually the case as an entire episode usually comes in as a single video file, but for features it's really great.
- If an updated version of the picture for one or more reels comes in, I can just drop that into the VideoSlave playlist and all of the cues that reference those reels will play the updated picture, instead of re-importing the updated video file into each of dozens of cues in Logic. Very quick and clean this way.
- Since the DLG+SFX+TEMP audio is coming from the outboard computer, it's easier for me to route it to speakers other than the main L+R if I want. While I'm working I route all the audio from VideoSlave to my center speaker, and that way I have a dedicated hardware button on my monitor controller that will mute that audio. Saves time flipping over to CueMix or looking for the fader in Logic that has that audio.
- Since that audio is not imported into the Logic projects, there's no chance that it will get accidentally included in a bounce or export, and my metering and mastering processing is accurate whether or not I'm hearing the video audio, since it's going "around the side" of the Logic audio. This saves a bit of disc space too. But if I do want the audio inside Logic for some reason, I can either import the movie into Logic and extract the audio (as many people do) or just real-time record from the inputs that VideoSlave is connected to. But I like having my Logic projects clean, with only the music inside them and no extra tracks just to deal with video audio.
Still, it's not exactly a cheap solution since VideoSlave is not exactly a $29 utility, but many people run VideoSlave on their DAW computer instead of on a separate machine just to get the workflow I describe above. I don't; I use a 2012 Mac Mini with 512gb SSD, with the audio coming out the Mini's headphone jack and the video coming out of the Mini's HDMI right to the big tv on the wall. I don't even have a keyboard or mouse plugged into the Mini - I just screen-share to it from my Logic machine when I need to build playlists in VideoSlave or whatever. I have a dongle I got from OWC a long time ago (can't find it on their site now) that emulates a keyboard+mouse so you don't need to have one plugged into the Mini to prevent it from complaining that there's no keyboard+mouse detected. But most of the time the Mini just sits there waiting for MTC to come in. The Mini even launches VideoSlave and loads the last playlist on power-up, so I don't even need to touch it for weeks at a time.
However, lots of the features in VideoSlave are aimed at ADR studios and stuff, so there's lots in there that I don't need or use. But it's a hell of a problem solver if you have those kind of problems.
Try that on a big project with picture changes, a whole team that need the be in the loop - like orchestrators, music editors, directors...and how do you deliver safely to the music editorial/picture department?
what Charlie says is absolutely the only way not to get into legal trouble, to carry a vast amount of redundant data, to not have to re-calculate sync-points in cues that aren’t effected by picture changes, but have now a changed start-time. I can g on, but you might get my point?
Simon Ravn
Senior Member
Haha yeah OK I can see that with such a complicated workflow and things and changes constantly going back and forth all the time you need a setup like that. But I think the majority of us will be fine with hosting it inside the DAW.
Simon Ravn
Senior Member
Bad day at the office?
clarkcontrol
Active Member
If I were working on films still I would absolutely use a separate video player.
Fifteen years ago I actually composed the music and SFX and Foley and did adr for a no budget full length movie in logic without a separate player and it was not glorious lol. It was on a G4, one big sequence, and I was very lucky the edit changed very few times. Never again.
Running video inside logic is great for small things, but even for medium length I would use a separate computer. Anything with multiple cues.
Fifteen years ago I actually composed the music and SFX and Foley and did adr for a no budget full length movie in logic without a separate player and it was not glorious lol. It was on a G4, one big sequence, and I was very lucky the edit changed very few times. Never again.
Running video inside logic is great for small things, but even for medium length I would use a separate computer. Anything with multiple cues.
Not at all. im just trying to prevent the original poster to make a rookie error, get used to habits that will not serve him as his career progresses.
if you felt this to be either grumpy, personal or bad advice, I most sincerely apologize for trying to help you out with why it might be tempting but ultimately impractical and short-sighted for you to carry picture in the same computer as your sequence - unless it’s something really short and not worth sending to a dub-stage.
best,
hz
Simon Ravn
Senior Member
I use it for modest film scoring. I just move the video for each separate cue. But I dont get tons of revisions back and forth, just a few. And I only work with a stereo track, not a lot of separated ones.
Last edited:
colony nofi
Senior Member
I would seriously suggest that anyone that wants to do work in Longform develops a double DAW / double head approach as suggested above by @charlieclouser and backed up by many.
Our little studios have developed our own version of this - which is a little different, but works extremely well even on a single computer.
I've written about it other times here - but essentially we take advantage of the ability on a Mac to just duplicate an application and be able to run it twice (at the same time)
In our case, its Nuendo, so we have Nuendo 11.app and rename the second copy Nuendo 11 Video.app
Both are launched, and we sync them together using VST sync / timecode.
The video app hosts our video (and puts it out to our video systems via Blackmagic SDI cards) - as well as hosting the DX, FX and temp MX stems. For TV its one video file (normally) at 10 hours (unless split for ad breaks) and for film its multiple video files (one for each reel) starting at 1 hour and moving up 1 hour for each reel. The standard Nuendo 11.app runs single cues. 1M1, 2M5 etc etc.
Our workflow also means we bring in either all WIPS (for small projects) or WIPS+Stems into this session - so we can drive just that session when working with a director / screening etc. That session also pulls together all stems needed for any live recordings or mixing - and can easily output AAF's to protools when needed.
All versions of cues are kept and organised inside the audio pool for extremely quick swap out (sometimes even having them on separate tracks). And all notes are kept in that session as well. Music editors can have access to this session as well at any time (although we tend to session file import from this session to a separate session for them most of the time)
Running video in a second session inside Nuendo also allows us to REALLY quickly reconform when new edits come in. Nuendo has reconform tools built in - we can be ready in 5 mins, and can usually have a "good enough" music edit for a whole film done in a couple hours even after huge changes - at least good enough to know where new composition / cue changes within cue sessions are going to be needed.
The final beauty is we don't have to run two different computers. Oh - and our single cue sessions can run ADDITIONAL video (so we can have two different sets of pictures running at the same time) - but this video just runs on the normal computer monitors. Why would you want that? You can sync the opening of a changed edit, and watch the differences between the edit in real time! Its a massive help!
Our little studios have developed our own version of this - which is a little different, but works extremely well even on a single computer.
I've written about it other times here - but essentially we take advantage of the ability on a Mac to just duplicate an application and be able to run it twice (at the same time)
In our case, its Nuendo, so we have Nuendo 11.app and rename the second copy Nuendo 11 Video.app
Both are launched, and we sync them together using VST sync / timecode.
The video app hosts our video (and puts it out to our video systems via Blackmagic SDI cards) - as well as hosting the DX, FX and temp MX stems. For TV its one video file (normally) at 10 hours (unless split for ad breaks) and for film its multiple video files (one for each reel) starting at 1 hour and moving up 1 hour for each reel. The standard Nuendo 11.app runs single cues. 1M1, 2M5 etc etc.
Our workflow also means we bring in either all WIPS (for small projects) or WIPS+Stems into this session - so we can drive just that session when working with a director / screening etc. That session also pulls together all stems needed for any live recordings or mixing - and can easily output AAF's to protools when needed.
All versions of cues are kept and organised inside the audio pool for extremely quick swap out (sometimes even having them on separate tracks). And all notes are kept in that session as well. Music editors can have access to this session as well at any time (although we tend to session file import from this session to a separate session for them most of the time)
Running video in a second session inside Nuendo also allows us to REALLY quickly reconform when new edits come in. Nuendo has reconform tools built in - we can be ready in 5 mins, and can usually have a "good enough" music edit for a whole film done in a couple hours even after huge changes - at least good enough to know where new composition / cue changes within cue sessions are going to be needed.
The final beauty is we don't have to run two different computers. Oh - and our single cue sessions can run ADDITIONAL video (so we can have two different sets of pictures running at the same time) - but this video just runs on the normal computer monitors. Why would you want that? You can sync the opening of a changed edit, and watch the differences between the edit in real time! Its a massive help!
Last edited:
charlieclouser
Senior Member
Hahaha running two copies of the same app on a single computer - hadn't thought of that one!
But I do the "two videos at the same time to compare changes" thing - I run the old video in Logic and the new video in VideoSlave and watch them against each other. Mostly just to see how bad the damage is. But whenever picture changes come in I get a detailed EDL showing what has changed. From there it's just a matter of juggling numbers....
Of course I don't have reconfirm tools in Logic but I'm lucky that for some reason I don't get tons and tons of picture changes. On tv there's just no time - whatever we spot to (a week before the dub) is THE locked cut, and any changes beyond that are dealt with by my music editor on the stage.
On features I've been lucky in that changes are usually a few frames trimmed here and there when a VFX shot comes back looking not so great and they need to shorten it or something like that. I've never had those "they rearranged the entire movie and what used to be 2m13 is now 4m46" situations like I hear about.
But I do the "two videos at the same time to compare changes" thing - I run the old video in Logic and the new video in VideoSlave and watch them against each other. Mostly just to see how bad the damage is. But whenever picture changes come in I get a detailed EDL showing what has changed. From there it's just a matter of juggling numbers....
Of course I don't have reconfirm tools in Logic but I'm lucky that for some reason I don't get tons and tons of picture changes. On tv there's just no time - whatever we spot to (a week before the dub) is THE locked cut, and any changes beyond that are dealt with by my music editor on the stage.
On features I've been lucky in that changes are usually a few frames trimmed here and there when a VFX shot comes back looking not so great and they need to shorten it or something like that. I've never had those "they rearranged the entire movie and what used to be 2m13 is now 4m46" situations like I hear about.
colony nofi
Senior Member
Oh running on the same computer is a total good time
When there's more than one studio in a facility - it lessons the tech load considerably. One less computer per suite to worry about.
It makes monitoring / routing nice and simple (at least, once its setup! We use Totalmix + the Nuendo Control Room - meaning things like streamdeck buttons can be made to remote control simple situations like "Play DX+My Music only" or "Play just temp" or "Play the latest WIP in the video file" all on a single button. OSC + midi controls + bome for the win.) No external mixers needed. No need to route any audio between two workstations etc.
Less screens to manage - I know its not a big thing - except in smaller rooms, where you can really re-use the space that would otherwise be taken up with the monitor for the video computer... here its a command+tab away. No need to either have a second keyboard + Mouse (or KVM)
And grabbing the new WIPS / Stems and putting them into the new session is a complete breeze. I know its not *that* much harder than flinging files through a network, but we've got a decent workflow down which seems to be much speedier.
Speaking of, we are moving towards having ALL project files on a beefy file server over a dedicated 10GBE network. This would make running two computers identical to the dual DAW on one computer setup as far as filing is concerned....the renders would live just in one place, and the video session would just reference the wips/stems (as would the nuendo composition session) on the NAS.
But - having said all that - running dual daws / a DAW + videoslave is an excellent solution. I personally come from using a protools rig as the video software rather than videoslave - I needed to do more with the audio side of things. We run presentation sessions from the second DAW. I looked into Video Slave again when V4 came out, but it just cut OUT a bunch of flexibility that we find really useful for WIPS / Pres sessions etc.
charlieclouser
Senior Member
Very interesting setup. Are you going to put sample libraries on the 10GBE box as well? Is it al SSDs or spinners? Curious to know what box you chose.
I am still using basically the same workflow and number of machines for 20 years - one Logic DAW, one ProTools DAW, one video timecode slave. But of course it's gone from 3x G4 machines (two towers and a G4 Mini) to a pair of cylinders and a Mini, and from ADAT cables to MADI linking the Logic and ProTools machines, and now I'm using network MIDI to slave the video machine instead of 5-pin MIDI.
I don't have a dedicated display for the video Mini though - I just use the big tv on the wall which has 3x HDMI inputs, so that's one for the video Mini and one for the cable box so I can watch Adult Swim at 3am.
Having a dedicated ProTools rig is always great when working, even though it sits powered off for weeks at a time until it's time for me to print stems. But being able to checkerboard cues that overlap, print WIPs and roughs as I go, and print overlays on the fly makes things quick and keeps the DAW sessions lean and clean.
I am still using basically the same workflow and number of machines for 20 years - one Logic DAW, one ProTools DAW, one video timecode slave. But of course it's gone from 3x G4 machines (two towers and a G4 Mini) to a pair of cylinders and a Mini, and from ADAT cables to MADI linking the Logic and ProTools machines, and now I'm using network MIDI to slave the video machine instead of 5-pin MIDI.
I don't have a dedicated display for the video Mini though - I just use the big tv on the wall which has 3x HDMI inputs, so that's one for the video Mini and one for the cable box so I can watch Adult Swim at 3am.
Having a dedicated ProTools rig is always great when working, even though it sits powered off for weeks at a time until it's time for me to print stems. But being able to checkerboard cues that overlap, print WIPs and roughs as I go, and print overlays on the fly makes things quick and keeps the DAW sessions lean and clean.
colony nofi
Senior Member
We are going for an all NVME server. We have two different approaches that we are looking at. The first is built of Micron 9300 drives (they have an excellent 15.3TB U.2 NVME which looks the ticket - and two of these in a JBOD raid is more than enough - with a separate spinning rust server as the first level backup).
The only problem is sourcing them at the moment / and their support. The tech co we were working with to build the server are saying they are out of stock / hard to get / no longer being made, but at least Micron says differently! The other option is a bunch of 4TB M.2 NVME's running off PCIE cards. Looking more and more likely at this stage. All running truenas.
Using an embedded (older) EPYC board (but also looking at a Xeon, and lastly a Ryzen based server board) - the workload of the board looks to be much less than I thought it would be.
We are not putting sample libs on this server (at least, not straight away!). A single Rocket QLC 8TB NVME in an appropriate thunderbolt enclosure looks the ticket for composing workstations for the moment, and allows for transporting libraries when travelling nice and easy. Also, different people use different libraries. Having a central server for them is HARD to manage.
However - we ARE looking at running a server for all our plugins. Am testing a system now, and it looks like it could be incredible. Network bandwidth is not a problem - and it appears neither is processing power with some of the newer server chips - especially when using Direct Offline Processing / more object orrientated mixing workflows (for post!). I doubt it will replace workflows in composition though. Its all awesome to look into though -
But yes - our double nuendo setup is essentially doing everything your triple computer setup is doing!
. I'd invite you round to see it in action... alas we are in Australia....Pablocrespo
Senior Member
I could never run two versions of cubase in windows, 11 and 10.5 at the same time or copying cubase11.exe and changing it’s filename.
The second one shows the splash screen and then it disappears
The second one shows the splash screen and then it disappears
