In any film, both sound and image go hand-in-hand, working together to tell story. But we’ve noticed a pattern-- so much time and effort is spent towards camera, that sometimes on-set production sound winds up being treated as an after-thought. More often than that, a lot of film sets operate under the assumption that sound can be “fixed in post.”

So we created an experiment and set out to answer the following questions:

  1. How much does good-quality production audio really affect the audience’s experience and immersion into the film?

  2. Is good production audio necessary, when you are going to “fix it in post” anyway?

  3. Is sound or picture more impactful than the other when it comes to the audience’s experience of the film?

We created one short and filmed it in two different ways. The first version of the short was given a decent camera package and an on-board mic, so the main emphasis was good picture. The other version was given a limited camera package, with a boom, mixer and operator, so the main emphasis was healthy sound. Almost everything in the shorts (lighting, production design, the performances, the edit, and post sound) were all kept exactly the same.

So which ended up being more important-- good sound, or good picture? You tell us!!

Watch both of the shorts, and experience the differences yourself.

Version #1:

Version #2:

If you are distracted or pulled out of either of the films, pay attention to what caused it, how much of an effect it had, and for how long. Then, let us know what conclusions you came to, and why! We’ll share our own ideas, too.

PS - In case you didn’t know, this is the long-awaited sequel to Joey Plissken’s first adventure, ESCAPE FROM DULL EXPOSITION!


I suggest that you find two groups–one to watch the other–because we perceive a film differently the second time we view it.

I actually saw a panel by some local sound geeks. One thing they pointed out is that bad sound is the #2 most common reason someone walks out of a movie (the first being bad acting). So I’ll say that people would prefer bad video over bad sound.


I agree with knifebladepresents that the best way would be a side by side with a group. However in viewing both myself I find it interesting that the visual aspect of the film being less than perfect highlights the type of movie you are parodying and is in no way a negative. The audio however severely affects the perception of the quality of the film. In using the on camera audio there are many spots that really suffer, like the approach to the keypad, the long hallway with crunchy items. In fact until I watched the second one I really thought that the hallway scene wasn’t supposed to have any floor sounds at all. It seemed like the protagonist was actually getting that section right and not messing it up. On watching the second version with the better sound you could tell how it was actually supposed to sound. In summary I think that audio, and by extension music, are keys to making more professional sounding and convincing or immersing movies.


First that sequel was great, and I’d love more !
About the video/sound fight, I think we can all agree that sound is from far more important than video. It’s easier to bear bad video, at least because it can’t hurt.

I watched the good sound version first (because I wanted to understand everything, and i knew it was gonna be better) and the bad video didn’t bothered me at all. A good reason for that was the good lightning. If you set the scene well, you still can get a good-looking video without great gear.


Heyyo! I’d first just like to thank everyone who’s shared their own thoughts and opinions so far! You all rock.

I like to think of it as finding a healthy balance (like a good, nutritious diet?); giving equal attention to both the image and sound on set - treating them with equal “value” as storytelling elements. Shot by shot, scene by scene, consider the impact both can have on your story, and what each needs to achieve - compromise smartly. While a “perfect scenario” where every department gets everything they want is exceedingly rare, it’s key to focus on what’s going to be important for the story.

On one hand, there’s definitely some moments where sound can indeed take a backseat - for example, fight scenes are almost entirely created in post. Reference audio is still very useful for syncing sound effects later, but that’s an instance where it may be the smart choice to use an onboard mic, if your location is really limited for space. (Never hurts to boom if possible though - a lot of the sound in some of our shorts like Fan Friction and London Brawling actually use a lot of production sound in addition to post-added sound effects). In our experiment short, the ninja scene with the pillow and karate chop play pretty similar between both versions, because the emphasis is on sound effects.

On the flip-side, there’s times when sound should absolutely be a high priority, if not the top priority - particularly if an actor is doing a very emotional or intense scene. These scenes can be the most painful when sound isn’t provided the opportunity to record the scene optimally. If you end up with poor audio, the two options are either use what audio you have from set, cleaned up as best as possible (hoping it doesn’t distract the audience too much), or, try to get the actor back to recreate the performance in ADR - while ADR may be easy for some actors on simpler dialogue, it’s extremely difficult to recreate a dramatic, intense performance, while hitting lip sync. While I won’t call them out directly right now, there’s a couple moments throughout VGHS that would have benefited from giving sound more priority - we ultimately overall tended to go with using production sound, even if it was a little (or extremely) rough.

What I found really interesting when we got to editing the two shorts was how different the audio really ended up being based entirely on where each team had their respective microphones - it was the time, consideration, and effort that made the difference.

I’d like to stress how a lot of the quality of your film comes from how you use your equipment, not what the equipment is.

Even if you’re using your phone to record sound, take a moment to test it out, and try to place it somewhere that will pick up the actors’ performances cleanly. A well-placed cheap recorder can still outperform top-of-the-line gear if it’s used better!

As a random aside, I actually liked the look of the iPhone for the theme of our short - I thought it added a nice pseudo '80s look :stuck_out_tongue: I find that people acclimate to image quality pretty quickly (we can still watch older films usually without complaining, or an old DVD without being bothered by the lower resolution the entire time), and while this is also true for sound (we can still watch something in stereo even if there’s a surround sound version available), production sound is primarily about capturing something cleanly and intelligibly - and if you can’t understand it clearly, that’s something you can’t ever expect your audience to adjust to!


The 80s flavor of the short does favor the iPhone’s graininess. Somehow that gives the iPhone a stylistic edge over the RED.
Other than that, the audio is definitely more important. Although the only thing that was really distracting was the iPhone’s autofocus seeking on a couple scenes (intro with protagonist walking to camera, in front of the painting checking for lasers). Using AF Lock (by pressing and holding on the iPhone’s screen to a good focus point), would’ve made these things better. For the moving shot, just lock the AF to the final position…
One final thing: I’m guessing that the footage from the iPhone is straight out of the camera. It was a good choice to have a super dark setting to exacerbate the shortcomings of the phone’s camera, but with some TLC in post, that footage could look a LOT better.

So good test… Audio is paramount.


As a matter of fact, if you guys wanna put up a dropbox with the final edit of the iPhone version, I could spend some time trying to make it look as good as possible (neatvideo / color grading, etc.), and send it back over. It would be an interesting thing to see how much of the bad video vs the bad audio you can “save” in post.


Pretty much what everyone else has been saying rings true on this one for me. The iPhone version generally worked better for the short, and I was generally taken out of the moment less by it. The few moments of camera jitter and focus searching took me out a bit, but you know, it was nice to hear all of the lines. I’d love to see a version with the RED footage and good production audio put together, but I’m also well aware that it would defeat the purpose of the test. Fun stuff. Definitely a great way to illustrate this effect.


The plan is to release the RED + good production audio version on the channel. Will give you a shout in this thread when we do!


Maybe it’s the fact that I work as a motion designer and editor which means picture quality is highly important to me, but I was much more distracted by the iPhone version. This short film did of course kind of highlight all the shortcomings of a cell phone camera, low light, high contrast, and changing focus distance. There were several times I was taken out of the story in the iPhone version, but only a couple times on the RED version, primarily when the sound was being recorded through the door.

It’ll be interesting to see the RED + good audio version!


The comparison is hardly fair considering YouTube’s compression. I believe the iPhone version of the short would be a lot less bearable on a traditional movie house screen.
Putting that aside, good visuals do not make a movie (ironically enough) as long as the story can be conveyed, which if it is a dialogue-driven narrative with poor audio, it would be impossible to follow.

The iPhone version is much less distracting, and if anything more impressive.
A good example of good gear not contributing to a good film.


I’m all for better sound, though I may have some experiences that have dictated it that way.

One of the things I used to do twice a week but haven’t in a long time is donate plasma. At the plasma donation center I go to, they have movies playing on about a dozen different TVs. It’s the same movie on each TV, but some of the TVs have darker settings than the others, and it’s difficult to see what’s going on in the shot if it’s a dark shot to begin with. Think “Dark Knight”. However, even the TVs set brightly, even though the picture is clear, the sound is often turned way low because it’s a balance between being white noise and being too loud, and in cases like that I think I’ve gotten used to subtitles as a form of sound: you can’t hear the words, but you can see the mouths move and see the words that go with it.

I also tend to tab out of youtube to multitask, and it’s easier to multitask when I’m not tabbing back and forth too often because I can’t tell what’s going on in the video. Better sound quality helps minimize tabbing.


I think its not realy fair how you compare the 2

Even if it is obvious that sound is more important then the Image (well at least in an online format) i think you gave the IPhone Boom an unfair advantage with giving this version Wildelines and the other not. You can do Wildelines with an onboard Mic too. Like in the Door scene where he is behinde a solid wall where you obviously can not hear him you would never cut that into a final cut, or you would never plan to shoot like this.

In an other situation where you have only 1 try and cant cut out the boom operater you would not plan a boom operater in the shoot eighter.


I definitely get where you’re coming from, but a lot of filmmakers are sharing their content primarily online (even stuff submitted to festivals, etc.), so online consumption, despite its compression drawbacks, is sort of the go-to platform. So like, I totally agree with you on one hand, but I actually think it’s more fair in another way that it is YouTube-compressed? Either way, thanks for bringing your point up :stuck_out_tongue:


The larger comparison we were looking to make was how proper attention to sound (including taking wild lines) can be useful in post, and we felt that omitting them from the onboard shoot philosophically made sense for the demonstration. This is because, in my viewing experience, a lot of filmmakers, particularly people still learning or those working on a tight budget, will skip wild lines even though they have the equipment to record them.


The RED version definetly left me wishing I could hear more. When the audio dropped in or out I found myself wondering what I missed in the dialogue rather than watching what was going on in the next scene. There was also kind of a consistent white noise going on in the background which through experience I know is incredibly difficult to clean up. I can see where taking the time to fix it can be time consuming. Last true video camera I used was a HI-8 and man did the built in mic pick up those gears. Plus sound editing software wasnt very available back then. Basically the inconsistency of the audio quality distracted me from even focusing on what was onscreen.

In the second however, I found it much easier to focus on the video as a whole. Dialogue felt clear and audible, and there was no white noise, which helped with building tension in the quieter moments. The only time I felt distracted was when the iphone began to auto-focus when Scoma discovers the lasers but the excitement provided by the audio as he sprinted down the hall brought me right back in. This method will definetly save so much time and frustrating nights of audio cleanup.

This was an awesome video and experiment. Gives me hope since I’m working with a Galaxy Core Prime. Definetly gonna try to catch audio seperately from now on. Thanks for all you do!


The iPhone version is clearly better sound wise. But my feeling is that scenes where the sound sucked on the camera version could be re-recorded using the camera’s microphone from up close, and then only sound could be used.
The low quality of iPhone’s video is my biggest surprise. Low light scenes makes it clearly visible. I wonder if 4K on iPhone 6S plus would be visibly better. I also assume you used standard video app, which does not give a lot of control over focus, aperture etc. The scene with lasers shows how auto focus can ruin a take.

If I would have to choose, I’d choose the camera one, but reshoot some scenes just for the sound. My feeling is that with that approach and post editing you can do wonders, where with picture it’s almost impossible to increase the quality of the picture.

Also an observation - you said you wanted to eliminate all other variables. I noticed how the scene with whoopie cushion (and probably few more) is different in both movies.

Nice experiment though. Shows how many things has to be well and in place to create something high quality.


Hey, thats a really cool comparison.

I, as everyone else it seems, also found the second version (Sound Team) more immersive. I thought the look of the iPhone really wasnt an issue in most cases, except maybe for the very low resolution of the shot when he stands in front of the hallway with all the stuff on the floor. In contrast to most people here, I didnt even notice the autofocus “ruining” the shot with the lasers, I actually thought that looked good and didnt notice it at all.
What was really disturbing in the v1 was the change in background noise whenever changing the scenery. Also, the lack of sound when he walks over the floor makes the hallway scene fail at everything it was supossed to be. And obviously, not understanding the dialogue sucks.

But when it comes to the quality of the sound, meaning the characteristics of the sound, not the easily measurable stuff like noise floor, I actually liked the v1 better. The volume of e. g. footsteps on the ladder or on the floor were more proportional and the sounds were naturally put into the room, making the acoustic environment sounding very natural without much hassle.
In contrast, in v2 everything was more clearly audible, but the effects (footsteps or the sound from the stuff on the floor) were often too loud and ruined the acoustical atmosphere which was way better in v1.
Strongly contributing to this, the ratio of direct to reverberant sound was off in v2, while in v1 it was mostly good. E. g. the sounds of stuff he walks over was probably recorded with the mic relatively close to the ground, resulting in the sound being too direct. Either you record these sounds from further away, making the sound be consistent with dialogue (One could also use a stereo mic at approximately the camera position to really capture the room sound of these cracks. Those small echos and slaps really make these things cut through). Alternatively, one would have to add artificial reverb in post production, being careful to exactly use the right amount and type of reverb to fit the scene. That’s a lot of hard post work.
There also was a bit of a problem with the boom mic in the scene near the end, where they are standing on two sides of the bars. The first “hand over the painting” sounds clear, “open the gate” as well, but the second “hand over the painting” sounds muffled. My guess is that the boom op was standing outside on the street, generating a difference in angle between the actors and the mic, making the guy standing on the inside more present and the one outside (of which he is directly above) more muffled. This problem was a non-issue in v1, because the sound changed as well, but was always proportional to the view of the camera.

In conclusion I would say paying attention to the sound is very important for immersion, but when you do a lot of stuff with sound, you can do a lot of stuff wrong, too. It’s like using an iPhone vs. a RED. With the iPhone you hit record and get the final graded image. With the RED you need to expose, pull focus and grade afterwards. If you fuck up your manual focus, you are better off with the iPhone.

Anyway, it was really fun to analyse these two basically identical videos with different video and sound production. And btw, I thought the lighting throughout the film was really good. Thanks a whole lot!


Those specific shots were done as two separate (but back-to-back) setups, as minute dolly vibrations absolutely freaked out the iPhone, and resulted in a jello-like image - they were done handheld for the iPhone. However, the shots are identical in duration :stuck_out_tongue:


It’s generally better to record things as dry as possible (i.e., with close proximity and minimized reverb), per the flexibility in post as far as applying reverb (which is actually quite easy and very fast, at least in a sound-centric editing program). Reverb reduction was conventionally considered “basically impossible,” and while there’s software available now that’s able to do it to some degree (I use iZotope RX’s Deverb function sometimes), it’s inherently a somewhat limited process, and can quickly sound artificial - software can usually reduce some of the echo tails, but the core voice itself will often still feel washy, and will start sounding “processed” if pushed too far.

As far as the specific mix choices in the short, that’s definitely a matter of personal taste and the tone of the film - I chose to go with a fairly dry dialogue feel given the tone, but it certainly would have been possible to add back in more environmental reverb, which would be much easier than reducing the baked-in reverb in the on-board audio, had that been the desired choice. Film audio as a whole, while not “right” or “wrong,” tends to be unusually dry and present, and is often center-panned even if the image shows otherwise, to prioritize clarity.

Thanks for sharing your detailed thoughts!!