Step 1: Find good song
Step 2: Think which frags match this song
Step 3: Record all moments that you think are good enough to put in the movie and match the song
Step 4: Edit
Step 5: Render and publish it

I got that. Im more asking about what you do about the clips.

So you record them all, but do you record them all at the same settings? Do you compress them before start editing? Because each individual clip I render is about 15gbs which can't even play properly without buffering all the time.

What about if you need slow mo's or other kind of stuff for 1 clip, do you use the one you already rendered or you go back and render the clip in different settings?

i record with the same settings
the editing program i use compresses clips i import, i still get a lot of buffering
i dont go back and render the clip with different settings, i dont see a reason to, unless you want to use as little space as possible

Launch tf2 with movie cfg and 1280x720, play the demo which has the action in it and fast forward (600x play speed) to the moment that i want to record.
Set host_framerate to 240 or 480, then type startmovie aaa01 or whatever you want.
Once i have recorded what you wanted to record, i just typ quit in console, open vdub, import the audio + video and convert it to an .avi file.
Then i watch it see if its good, and repeat this process with other actions that i want to record.
After i have recorded a lot i will start editing it.

Usually i save smooths for later, like in the red light green light movie i didn't make the "smooths" yet untill i had chosen which clip would be played at that time of the song.

hi, I'm not sure if this is what others do but this is my general workflow.

I start by recording all the frags at 480 fps. I use lagarith in vdub to cut down on the file size a bit, but that is the extent of my compression until the final render. you don't really want to be compressing much at all before you begin editing in order to maintain quality, although this can be an issue if you lack hd space.

when I first started making videos I had a similar issue to you. one, I was expecting my videos to play back smoothly in wmp, I'm not sure if this is where your clips are buffering but if it is just about any file above 60fps will be close to impossible to play due to size. if you are having trouble playing back in vegas (I'm assuming this is what you're using as it is somewhat of a standard in the tf2 community, although correct me if I'm wrong) this can either be due to preview quality set too high (I just about always use draft>quarter unless doing colors), or your codec. for whatever reason, when I started I used xvid and had awful lag in vegas so I highly recommend lagarith regardless as it is a great codec. you can also improve playback in vegas by making sure gpu acceleration is on.

now to answer your other questions, I always record frags at 480 fps/the same settings regardless of what I'm going to do with it. it's enough to be able to use slowmo, while making the clip look very smooth and fluid. I understand this may not be optimal for some people in terms of size, but I expect every frag clip/movie to take up large amounts of space.

in terms of smooths, I always record at 960, or higher in special cases. with smooths I primarily am doing velocity edits, and 960 is enough to play at 3% smoothly. in the rare case I need slower than that, I'll record at 1920. my smooths are generally very very short snippets slowed down tremendously, so they often end up taking up far less space than frags.

so with all that aside, I record the frags first, and then work with them one by one on the timeline with music, making smooths as I go where I need to fill the space. finally, I render with kirby's render settings found in his guide, although I've been trying recording as an avi and then compressing the render with easy h264, although either work fine. hopefully this was of help to you.

@6

I have problem previewing uncompressed clips, thats why Im asking if any of you compresses them before starting editing. They always end up stopping every 1-2 secs to buffer because a 15gb file will never play smooth on a video player.

I use adobe premiere.

You can't preview recorded footage like that :D Just drag them into your editing program and preview your clips there.

@7 I have the same problem. Having compiled the tgf files in vdub and saving as an .avi, the footage never plays without buffering constantly. I just throw it into Vegas and view it in the preview window there. 99% of the time it's fine, mainly because when I record, I record the same thing like 3 or 4 different times before I actually get it right (1st time - Oh whoops, left viewmodels off...2nd time - FML, no killfeed...3rd time - Fucking anti aliasing is off... etc.).

On the whole, I record and edit a single 'frag' at a time. That means finding the most suitable point in the song for said frag, recording the frag itself from the POV and any smooths/other angles that need to go with it. For normal playback speed, I'll record at 120fps. Any slow motion stuff, depending on how slow, will be at much higher fps. Once recorded and compiled in vdub, throw all the stuff for that frag into Vegas and edit. Then when I have every frag with all the stuff in Vegas, make adjustments so that the frags look decent one after the other.

Even on the editing software it wont play properly and stop to buffer

Try setting your preview quality to low? I for example use Sony Vegas, because I am lazy. I have hard time previewing everything (pretty much). So what I do is that I just set the preview quality to like 50%, instead of 100% and I keep on working like that. Make sure to NOT preview whole thing instantly, because it definitely won't let you do that. Each time you want to preview something, just do it step by step (each 10~ seconds or so).

I used to record pretty much everything before starting any editing, but recently I've started by recording only the first persion views of every clip I might potentially use. Then I load those clips into vegas, place them around the timeline and try to get a general feel for where each frag fits well, and once that's done I move on to recording smooths/thirdperson shots which fit well with the frags. As far as the order I work chronologically to some degree, usually end up making the intro at the end though.

Hopefully you're not asking this to copy anyone though because like a lot of things it's all preference, I've seen streams of good editors who have methods which seem alien to me.

Also as far as previewing goes, the two best ways to get around it are to ram preview everything (which I usually do) or create proxy files which are essentially low-quality copies of your main files which are used for the editing process, and then replaced with the original files before the final render.

everyone's different. I personally record all the clips i can at 480 FPS, then make a job list in virtualdub, stick it on to covert the tgas to avis and go for a walk. When I come back a couple hours later I put all the avis into a job list in Handbrake and do something else. After that I have about 15 mp4 files at about 50mb each depending on length. Then I stick them in Vegas and edit.

Frag videos:
1. Record everything
2. Find music / style
3. Draft cut (and record additional shots along the way if needed)
4. Finalize (Go through everything again, refine transitions, add effects, sound etc)

Top10:
1. Clip sorting
2. Record everything
3. Clip ordering
4. Draft cut
5. Music & commentary
6. Soundtrack mixing & effects
7. Finalize

I leave everything raw on my HDD until the final video is on YouTube (use the raw files to edit).
I record every clip seperately, but use VDM recording where possible to save time (firstperson shots from POV demos)

There is no right or wrong way of doing it, final product is what matters. Though there are ways to save time of course!

Have you found a way to keep video/audio sync when recording with VDM files or has it never been an issue for you? I used to use them to save time but recently the sync has been off after compiling in Vdub so I'm just doing it manually.

No, doing it manually as well. Annoying but still quicker than recording each clip separately.

Get frags
Record with lawena
Use virtualdub to convert them to .avi's
Procrastinate
Find a song
Procrastinate
Start to edit
Go record more smooths
Delete said smooths when I realize they weren't good
Procrastinate
Record another smooth, this time correctly
Bitch about the sound on the frags cutting out in vegas
Eventually finish and render
Realize I made a mistake, delete said file, repeat rendering process
Upload it, get 10 views and a spam comment about a free ipad

Sorry for bump, where would you find a movie config? Would that just be like chris' maxqual? Or something different

Hi, I've found that this is the best workflow, it takes a bit of time to set everything up but it's worth it when doing complex projects which requires a bit more than stringing together frags:

Step 1:

Once I get the demos I rename them to 001_tickstart, 002_tickstart. I also keep a "master" list, basically a .txt in the following format:

001: 56000-57000 (middy/streak/maybe use for filler/amazing frag, etc) <<this last part I don't include if it's my own demos since I know and remember the frags

Step 2:

Record each demo with FRAPS. The FRAPS pass is NOT used for the final output. Rename the FRAPS files in sequence 001, 002, etc. This is only for my own reference, when I am fleshing out a frag movie I like to play the frags back quickly and then I scribble down a possible sequence of frags (eg, 004->smooth of 007 to lead->007->outro using 003, etc, this part I scribble down on a random piece of paper on my desk). I also use the FRAPS files to estimate the length of sequences and decide on how much filler I need. It's basically used to quickly review the frags so I can picture the final frag movie in my head. I wouldn't use this step if fast-forwarding and rewinding demos and playing different demos weren't so slow and buggy and cause problems with crashes.

Occasionally when recording TF2 the audio won't be output properly, so sometimes I will use the FRAPS audio instead of the .wav files TF2 generates.

Step 3:

Record most of the straightforward frags using the Lawena VDM method. That way I can automate the recording process a bit. However, I change the names of the output files from a1, b2, c3, etc to 001, 002, 003, etc. To match my demonames.

Step 4:

Import to Adobe After Effects (I use AE just out of habit but any editor will do), I pre-compose the clips into small compositions with audio and proper frame rates and rename them to 001, 002, etc, again, according to the demonames.

Note:

I use sourcedemo2 to record mostly.

Step 5:

Any special clips which requires multiple passes (smooth, no viewmodels, HUD on and off, depth of field pass) I record manually and then give it a suffix such as 005_dof_pass_xxxx.tga. Sourcedemo2 may or may not be used depending on the effect I am after.

Step 6:
Compose in AE.

The point of renaming the demos and renaming the pre-comps is so that in my timeline my frags are shown as 004->006->007->etc. If one of them I'm not happy with and wish to reshoot then I can simply load up the demo and find it quickly.

Step 7:
Export lossless copy

Step 8:
Render using AVIdemux with x264 codec. I find x264 to be the best in terms of quality and speed.

I record all the POVs I feel I am going to use, or might use. I record these at 120 fps normally, and 240 fps if I want slow-motion. (This gives me ample room to breathe if I want to slow a clip briefly for syncing purposes)

I then search for a song. I usually have an idea in my head about the direction I want the video to take, so I search for the perfect song. It is the song that encapsulates everything that I want the video to express, and has ups and downs.

I load up the song in Vegas (though recently I switched to Premiere) and mark every beat I want to sync something to. This ends up being many many marks, and this is how many syncs I made for Frank West's frag movie. I then listen to the song over and over again, thinking of how the song makes me feel. I then find the perfect frag to fit the emotion I want to invoke at a certain section of a song, and put it in. Then I move on to the next section.

After I have a rough spacing of all the clips I want to use, I start smoothing. This is the most time-consuming process, but that's probably just because syncing smooths is a bitch and I'm particularly meticulous about it. I spend so much time that when it comes to smooths; I smooth, record a draft copy, and smooth again, trying to adjust the timing to coincide with the music (an entirely trial-and-error process that is absurdly time consuming)

Once I have all my footage recorded and synced, I go into After Effects and start doing all the transitions, color corrections, etc. My final render is done in After Effects via MainConcept h.264 while preserving RGB.

1. Watch all demos
2. Pick best one
3. Record them (240/480 fps)
4. Find song
5. Mach frags with music
6. Other edit (Smooth/ CC/ ect)
7. Render (Sony Vegas -> easyh264 or RipBot)

i actually work with video production.

the usual workflow for "montage" jobs usually goes a little like this:

1)find your song
2)lisint to this song until you hate it(so you know how the flow of it goes ect)
3)get all your clips compressed, and ready to import (no reason to break your editing workflow wit stupid things like compressing)
4)edit your stuff
5)send it to other people that has no clue of editing what so ever(as that is the closest to the actual consumer(you can always get a "prof" opinion on it after))
6) color grading
7) fix the music and sound effects levels
8) render
9) feel gooooooooood

also, record at max 120 fps(there is no reason to rape your hard drive with high fps files.. as you LITERALLY wont see a difference unless you slow down to around... 1-2 %

Wtf?

Personally i just took my autoexec.cfg (which is chris max frames i believe) and changed every command in there to the highest setting / best quality for best looks.

Lol, where do you work, where people think 2 fps are acceptable?

never said that, i said that you wont need more than 120 FPs unless you are slowing down to 1-2%

the standard fps for european video productions is 25 fps(the NA is 30)

so if you wanna slow down 50%
you still have over double of the frame rate that is going to be showed,
it should run smooth as hell at around 25%
thought that is pretty much the lowest you can go without dropping frames.

and i figured that i haven't seen anyone use anything that slow.

even thought you render an amazing video with a high fps, YouTube is still going to use their own compression codec wich runs at 25 fps, so the rest of the frames litterally goes to waste.

also, i dont know why you care, but i work at something called:
Mayday film.
i am currently an intern..
and i have been going to school with video editing for 3 years now.

youtube compresses to 30fps

You clearly have never heard of resampling.

yea you could do that, though i asume that he wants to record all of the clips at the same fps.