** Official Recording/Streaming Tips & Tricks Exchange **

[uncle_rufus]

Hey everyone,

Quite a few of us are involved in posting videos and streaming, but with the vast array of different pieces of software, hardware and techniques available it can sometimes be a bit confusing

So this thread is a place for all of our fellow OcUK creators to ask questions, learn from each other and share any tips, tricks or knowledge about the process. I think we're all sensible on here but just in case - please don't treat this as an opportunity to shamelessly self-promote (links to (or embedded) videos are fine if they are relevant to the tips being discussed, otherwise a simple link in your signature is more than enough promotion)

I'll kick things off with a question - I'm curious what software everyone is currently using for their recording and editing? Perhaps we can start off by posting what we all use:

For me:

Recording:
DXTory w/ Lagarith codec (and for both audio channels) for RETRO/8-bit games
DXTory w/ UTVideo codec (and for both audio channels) for normal 3d games
OBS Studio w/ UTVideo + Audacity to record audio in some cases (where DXTory doesn't work)

Streaming:
OBS Studio

Hardware:
2x Shure SM57s, Scarlett Focusrite 6i6 interface, Logitech C290 webcam for streaming

Editing:
Audacity for audio, Blender VSE for video (seems an unpopular choice but I like it)

 

[uncle_rufus]

Is it worth upgrading from Fraps to Dxtory?

It probably depends if you can justify the cost (it's only ~£20) - I can only echo what the others have said, you get a lot more control and particularly having the audio tracks separate is nice.

DXtory isn't for the faint hearted, it's a bit like OBS in that it's about control rather than convenience

I don't think it's that bad really... I picked up OBS only recently when trying out streaming but had always heard a lot of people saying it was good for recording, but I find the options really lacking for normal recording (compared to DXTory) - am I missing some settings somewhere?

Avoid Twitch for streaming.

www.hitbox.tv is a better service in my opinion.

Never heard of this... As you say I went to Twitch as it was the only service I knew of - do the other services really have a lot of users? (will check hitbox.tv when I get home)

That's essentially why I ended up on DXtory too, virtual cable / voicemeeter, job done - I can edit conversations in and out of the video as I please, no more problems with the game sound being too high or low either.

Yeah this can't be understated (although it is achievable without DXTory, it just so happens to make it easier). Not only being able to edit conversations but being able to apply separate audio-processing to your voice, your friends voices and the game audio is really really useful

 

[uncle_rufus]

Is it Adobe After Effects that I require to make good video intros / channel intros?

It's not the most professional looking thing in the world, but I made our intro using a combination of Blender (which I use to edit) and GIMP (aka the poor-man's photoshop). So for example I had a part of the logo I wanted to spin around - I made 2 copies of the logo image in GIMP, one with the spinning part missing and one with only the spinning part showing... Then I layered them over the top of each other (in Blender) and keyframed the "rotation" of the layer I wanted to spin... Hope that makes sense

What do any of you use to fix things if audio is out of sync? I don't know if it's a setting within the game (EAX?), but almost all of my Oblivion Lost videos require me to re-encode the files in handbrake or format factory, but even handbrake sometimes takes a few attempts before it manages to sort the problem out. Is there are simpler way?

When you say out of sync do you mean like it's offset by a constant amount the whole way through or is it drifting? (Is this your same question from the other thread - with ShadowPlay?)

 

[uncle_rufus]

I might well take a look at [Blender] again (especially since it's free), and then if need be I could try out After Effects later on.

Could be worth a shot... Just like GIMP Blender is really just a well run open source clone of many of the best features of the pro paid for software. The only downside is it takes a bit of learning because not too many people use it (as a video editor). I think I watched a series of tutorials by a guy called Mike Myers (neither of "the" Mike Myers's though ) Also I can try and help a bit if you run into anything specific

I just re-encoded one of the files and at first everything seemed fine, but after a minute or two when shooting at en enemy, you can quite clearly tell it's out.

I can't imagine it would be to do with the game settings - as far as I know most capture programs will literally be grabbing the audio directly from the signal being sent to the device. I'm still fairly convinced that your problem is the video not the audio, and it has ended up being recorded in VFR (variable frame rate). Most editors can't cope with VFR - it's very annoying and actually one of my reasons for switching from ShadowPlay to DXTory - to save the extra step of re-encoding to CFR (constant frame rate) every time.

Alternatively if you're talking about your final rendered video (not the raw captured footage) - I do sometimes get a weird problem where the video inexplicably renders out sped up (like a flat 2x speed increase) and then clearly doesn't match the audio and "runs out" before the end of the video. It's super weird - if I chop an extra frame or two from the start of the clip the problem usually goes away but it's very annoying and I can't work out why it does it (maybe it's a Blender problem, or DXTory, or something else I'm doing?)

I did my first ever stream to Twitch the other day and got a solid 6 viewers without really doing anything particularly special simple because I chose a good game to play that wasn't very popular.

That's pretty good - we did a quick test stream and got pretty much nobody! But since I've bought it to play anyway might try streaming a little XCOM2 this weekend (it'll be popular though so I don't expect it to really take off - but like I say I'll be playing it a lot over the weekend whether I stream or not so worth a try!)

 

[uncle_rufus]

I guess that's another aspect to this whole thing - getting your videos/stream out there and gaining some viewers... We've so far been adamant that we aren't going to use the features on social media plugins etc. to mass spam people and have been avoiding "sub4sub" or shallow tit-for-tat trading of views/comments/likes/whatever...

Where has it got us?... 3 months in, getting on for 100 videos uploaded but as far as I'm concerned completely devoid of any (genuine) viewers. In some sense that's what has prompted us to give streaming a try (just to increase our exposure a little) but otherwise it seems like it may just be a waiting game - keep making content as good quality as we can and hope to get lucky!!

Anybody else care to share their experiences up to this point?

 

[uncle_rufus]

One problem I'm going to have is not being able to hear myself when I wear headphones during my stream. Like, I can hear myself slightly but it's very muffled (due to ear pads around my ears), put games music and sound effects on top of that then I may have a problem.

How come you need to be able to hear yourself? If you really do then here's a few suggestions off the top of my head:

1. Wear your earphones "DJ style" (i.e. earcup on your ear on one side and against your head behind your ear on the other) allowing you to hear the game and yourself at the same time. You'll lose out on any surround-sound effects etc. and it might not look so "cool" but it's an easy option

2. Practice talking without being able to hear yourself fully for a while until you become more comfortable doing it - another relatively simple solution; you could use audacity to record yourself; stick on some loud music and practice reading the paper or something until you get a recording that sounds okay

3. Use a virtual audio cable type application to split and redirect your mic input back through your headphone output, allowing you to monitor your own voice. But be careful - to do it right you probably also want to setup a virtual cable source for the rest of your output audio and then also send that on to the headphones (and then connect the stream audio to the intermediate source instead of the headphones; otherwise your stream will pick up your mic twice!)

 

[uncle_rufus]

For thumbnails here's roughly what I do...

I've got a GIMP project which contains a small-ish copy of our channel logo on a layer, plus a text layer with the number "88" (because it's the widest 2-digit number!) in a font we use.

When I first start a new series I go and find a nice copy of the game's logo (preferably on a transparent background) and add that to the project as well (quite often adding an outline or shadow or something to the logo to help it stand out) and make it a sensible size. I leave the background transparent...

Then once I'm done editing the video I find a frame I like in the editor; print screen it and then "paste as layer" into GIMP; I usually have to scale it up a tiny bit but once that's done I position the logo, episode number (updated to the current number of course) and game title logo on the image somewhere that looks good, and there we go!

(In the past we used to select one promotional wallpaper or background from the game and use that in every thumbnail, but more recently have switched to doing the above as it makes the thumbnails a bit more interesting looking)

 

[uncle_rufus]

i sometimes like watching people play just for chilling out.

Is there a thread with all the youtubers (inc link) and the games they play (will be playing) besides their names?

There isn't - I suppose we could keep a listing in the first post of this thread or something (but at the same time the idea was to keep this focused on sharing tips etc. rather than promoting, so maybe we could make a separate thread for it?)

What are your thought's?

Please don't take offence but this is way off-topic for this thread; if nobody replied in the peripherals section then give it a bit more time or otherwise maybe people genuinely don't have an answer

 

[uncle_rufus]

Here's something I've been thinking about...

So far on our channel we've been mostly focusing on co-op (i.e. local sitting next to each other co-op) but in the back of my mind I've been trying to work out how we might be able to get a 2 PC setup going; with the goal being that we can play online co-op titles together...

Typically when I've seen other Youtuber's doing something like this it's a collaboration between 2 different channels, so they end up uploading the two recorded perspectives separately onto each channel... but in this case it's slightly different, so I'm trying to work out possible ways to do it...

Recording the video:
Clearly I could just record the raw footage on each PC separately (as I do now), assuming both machines have the power to do so. But if I instead had a dual-input capture card installed in one high spec PC and then route both its own and the other PC's output through it, maybe that would be better? (I guess if it was for Streaming this would give me 2 sources in OBS I could position on the stream window, similar to what they do for events like AGDQ... but for recording for Youtube maybe it wouldn't be any better than simply splicing 2 separate recordings together in an editor)

Recording the Audio:
This seems a lot harder... if we are sat at 2 side-by-side desks we will probably be too far apart to sit close together around a single microphone (as we do currently), but if we pick up a 2nd microphone will it be a problem that each recording will be picking up the other person's voice in the background? Or will it be alright once sync'd up?... Or maybe we can drop the sensitivity down and just both get really close to our own microphones to minimise the background noise...?

How to present the final product?
I think it makes sense to either overlay the two perspectives in a "semi-splitscreen" kind of way (not exactly splitscreen as it would screw up the aspect ratios too badly, but also not overlapping too much)... OR to always have one perspective fullscreen with the other in a small frame in the corner (picture-in-picture type of thing) and then periodically switch the two around depending on what is happening in-game (and have maybe an indicator showing who's view is currently the main one)

Has anybody else thought about this? We want to get the 2nd PC either way as there are a lot of online-only co-op games we think would be really fun to play together, but if we can find a way to do a few videos around the idea as well it would be really cool

 

[uncle_rufus]

Rufus are you just wanting to record it or stream?

Probably mostly wanting to record it for YouTube, but if the option were there to stream a double-output setup as well that would be cool as well...

I think separate mics might be an issue to be honest, even when sync'd there would be a lot of additional noise, a single input would be vastly better.

Honestly not sure what to suggest, quite a complicated thing you want to do

Yeah that's what I figured too I suppose maybe we can try to find a way to setup the desks so that we are still sitting very close together allowing one mic to still work...

Talking about lag. I'm headed back to Lagarith with DXTory as I'm finding my current h264 codec causing lag between video and audio on recordings. I can manually correct it but it's a pain.

I find DXTory to be really good; having used ShadowPlay, Fraps and OBS as well... a lot of people seem to say that DXTory is too complicated but I've always found it to be fine

 

[uncle_rufus]

Interesting... I never did try Afterburner (despite having it installed) - maybe I'll give it a go at some point. Presumably it doesn't do multiple audio tracks so I'd have to go back to recording our mic with Audacity and doing a sync?

Having said that it might not be worth it... I find DXTory to have no impact on the smoothness in-game (but perhaps that's the difference in CPU for the encoding duties? 3770K vs 3570K; it might make enough of a difference in the right games)

 

[uncle_rufus]

Initially, although I was impressed with the IQ of Dxtory (and others), I decided to not use it as much due to the huge file sizes. However, now I'm not as sure; after having used the aforementioned software to help produce the last few videos, I can definitely see the appeal of it, but on the other hand I can't justify spending a stupid amount on a 2TB ssd, so I guess that leaves something like a several TB hdd. What's a good bang for buck drive - one of the WD Reds? And why do I seem to keep reading that I should stay away from Seagate?

I bought a 2TB Seagate Barracuda... from what I could tell they aren't considered super reliable but amongst mechanical drives have some of the highest speeds, so it was perfect... Cost like ~£50

The reliability isn't a major concern since I use the whole 2TB exclusively for recording - if the drive dies I might lose a few hours of recorded footage at the worst (but I try not to let too much unedited video build up in the first place)

 

[uncle_rufus]

In terms of ease of use, performance when recording and file size - it's fantastic. However, the quality won't be as good as Dxtory, but for me right now, it's good enough most of the time. However, when I start doing commentary I might need the flexibility of Dxtory.

I think another worthwhile point in the argument is what your recording style is...

If you're the sort of person who can sit down, hit record and then carry through an entire episode without any awkward pauses, interruptions or other parts that need editing out - so your "edit" step becomes simply chopping a few seconds off of the start and end of your recording... Shadowplay is amazing; you'll end up with a file already compressed and ready to shove straight to Youtube

If you need to do a lot of editing on the footage then having a higher quality raw capture will always result in a better quality end-result (although depending on who you ask and how much it bothers you, it may be only slightly better!). Also Shadowplay's files are a bit of a pain for heavy editing as they aren't CFR (constant-frame-rate) so require an additional encoding step before most editors will edit them nicely

 

[uncle_rufus]

1) I cannot get my party chat along with in-game sound on my ps4 while streaming, what do you guys do?

2) Do you guys have a schedule for when to edit videos and when to upload them for youtube?

3) Video intro, do i need it. Also where do I start and try to be unique with it? This bit I dont have the experience or skills to do this myself (i think)

I try to do this as a hobby but I think i need to organize and plan a little better in the future. Not been doing it for long but with practice and a lot of patience I should do ok. I tend to do a lot of gaming with my brother in party chat and I think the banter we have provides enough entertainment. But I fear that might not be enough!

For (1) I'm afraid I also don't know as I don't own a PS4 and haven't tried streaming from a console - sorry

(2) Is an interesting one... we make some sort of attempt to upload something every couple of days, but since we're fitting it around work etc. don't always manage to stick to it. Also I find different games require different levels of editing depending on the pacing of the gameplay (if it involves lots of uninteresting travelling for instance I usually edit some of that out). In terms of a schedule I try and do a bit of editing in the evenings and usually in the mornings at the weekend - don't do much in the day at the weekend as that's when we typically record more footage. Definitely upload things early in the morning as the upload speeds seem to be better at that time...

(3) Completely up to you; if you ask around on any of the common let's play type communities people seem to say intros are the worst thing ever made and you should never ever do one - but I think actual viewers (i.e. who aren't channel owners themselves) don't mind. Probably try to keep it fairly short and not too obnoxious (no 30 second+ over the top animation with blaring techno that's way louder than the rest of the video!) We've tried to go for something fairly quick (but hopefully recognisable) in our videos.

As for how to make the intro... I didn't really know what I was doing either but for something fairly basic you can get away with creating some images using a graphics program and arranging/animating them in the video editor itself (I used GIMP for the graphics and Blender for the video)

Playing games, having fun + banter can definitely be enough (eventually) but it seems (anecdotally at least) that it's the process that comes before that - gaining some sort of viewerbase - that is the real uphill struggle... and nobody really has the answer to that

 

[uncle_rufus]

I've found the videos that do best for me (we are talking big numbers here guys, like over 200 views ) are of games that have just hit early access etc. People hitting youtube for gameplay of a game they want to buy etc. I have a colossal 6 subscribers right now so I'm assuming that's all search traffic. Plenty of people do an excellent job of streaming and reviewing big titles, but if you get onto a new game quickly you can at least get opportunities to win new viewers.

This seems to agree with our recent experiences as well - you almost have to target games which are a similar relative size to your channel; that way you can stand a chance in the search results. Presumably as your channel gets a bit larger you can compete more easily (due to your "cushion" of guaranteed views from your subs)

 

[uncle_rufus]

I'm still happily bumbling along using Blender (for free)... though I'm increasingly tempted by Premiere or Vegas if only for the improvements to rendering speed (from what I can see Blender isn't making great use of all cores, and I would expect the more commercial programs would do well in that area)

 

[uncle_rufus]

So I've been thinking more about my crazy plan (tm) to put together a double-PC setup for recording online co-op games... but particularly about the audio - how do you record with two different mics in the same room without the results being terrible...

I'm wondering if maybe the answer might be to upgrade to 2 proper mics (i.e. not USB like the Snowball we currently have) and plug them both into a dual audio interface like the Focusrite Scarlett 2i2, which in turn is connected to one of the 2 PCs. That way the audio from both microphones would hopefully be captured in sync. Combine that with proper mic placement and gain settings, using the right type of mic (a condenser I guess?) so we can both be nice and close to the mic and it won't pick up much background noise - (we struggle with this now sometimes trying to record both of us on a shared mic without picking up too much noise)... Maybe that would work?

 

[uncle_rufus]

don't under estimate the snowball.its a good mic. its more the way you doing it.

Yeah this isn't really anything against the Snowball as such - we've been pretty happy with the results from it in general but it really sounds its best when one person is using it pretty close to them and speaking into it (you know like ~6 inches away or so) with a corresponding reduction in the sensitivity (which really helps to cancel out background noise, in particular noise from the keyboard/controllers)... However when there are two of you sat together you have to raise the sensitivity as you can't both get quite so close to the mic, and then it starts picking up more noise... Also if one of you happens to lean slightly further away it starts to sound too echo-y/reverb-y. Once we have 2 PCs we won't be able to sit as close together and hence the need for some other sort of setup...

I'm not the best on the subject but would it depend on what kind of background noise it is ? If it's external to the room would it be easier to try and temporarily sound proof to some extent or is it more like moving chairs and small noises like that ?

Apologies if it wasn't clear above - I'm not talking about "noise" noise from the environment really; for the most part we find that type of noise is fairly minimal (don't even normally need to use noise-removal). I'm talking more about the noise of the other person talking in the background if we were to use a 2-mic setup (and also as described above some noises like clicking keyboard/controller etc though a lot of this goes away once you drop the mic sensitivity down)

You can remove a lot of background noise with audacity, I think the bigger issue would be two people talking next to each other, it might work I guess. Personally I'd want individual mics and mix it afterwards.

That's pretty much what I was thinking too - surely if the mics are perfectly in sync then I was hoping it might not matter if you can hear a very faint sound of the other person in the background of each recording? (as it'll be drowned out by the actual signal from the other mic)

Two different mics to me sounds like a recipe for disaster, unless you can separate them enough so they don't pick each other up then it probably isn't going to work brilliantly. If you're lucky you might just get a faint echo that you could try and remove in audacity but then you'd end up doing it for both and editing times would double with each recording. Would one better located mic not achieve what you want with less hassle ?

This is pretty much what I'm wondering too (but see my replies above for the reasoning why I think maybe 1 mic shared may not be ideal either)

Depends, if using two cardoid mics there shouldn't be a lot of background noise. Alternatively one VERY Good omni mic or bidirectional would potentially work but again it depends on game sound - hows that being recorded (hopefully not over speakers) etc.

There's a reason radio stations dont' use one big omni mic when interviewing people, it picks up a lot of guff so they tend to use individual mics.

There's going to be a bit of testing regardless.

Yeah we experimented with the Snowball (which has a high-output Cardoid, low-output Cardoid and Omni mode) and found that Omni was awful (as it picked up way way too much clicking from the controllers/keyboard). The low-output Cardoid sounded the best in this respect but it didn't pickup both of our voices with enough clarity (it was really good though with one person very close to the mic)... so in the end we use the high-output Cardoid.

Hadn't thought about the comparison to a radio station - that's kind of what I was thinking though. The only other possibility would be to maybe use 2 broadcast headsets? Similar to what football commentators etc. would use...

Agree about experimenting - the problem of course being that it's a lot of £££ to throw down just for testing!

 

[uncle_rufus]

I usually go 15000 kb/s, GOP of 30 (for 60fps 1080p) but I must confess I don't really know what I'm doing completely or spent much time experimenting

 

[uncle_rufus]

GOP?

It's just like another encoding setting - "Group Of Pictures"... has something to do with how the keyframes are arranged. YouTube recommend a GOP which is half the framerate (hence 30 GOP for 60fps)

I don't pretend to understand much more than that but here's some confusing technical info:
https://en.wikipedia.org/wiki/Group_of_pictures