So its been a minute since I wrote a blog post. Its not that I’m lazy, its… well, yes. I’m kinda lazy. But in this case, I really had nothing new to present until now! Folks, prepare to be amazed! Are you sitting down? Well stand up. Now sit down. Ok, now that we’ve had some fun, we’re going to have some fun with Pozotron.
To start out with, I abhor AI (Artificial Intelligence) in any shape or form. Its not human and in my opinion is taking away from the very heart of human performance. From a voice acting point of view, its as if AI is trying to replace humans, much like computers did to traditional prepress. If you don’t know what that is, look it up. I was a part of putting many a good prepress man out of work.
Ahem… Rant over. I digress.
So AI is not on my list of technologies that I believe are a good thing. EXCEPT, when it comes to Pozotron.
Look, I do believe AI has its place and Pozotron definitely has its place and does that very well! Very well indeed; which is why I’m writing about it today. So buckle up and I’ll give you the low-down on this exciting piece of technology and 7 reasons why as a voice actor/voice over you should be using Pozotron today!
From humble beginnings...
A few weeks ago; maybe a month ago, I was introduced to Pozotron by a fellow voice actor. She briefly mentioned this as a method to proofread copy to recorded audio. “Ok” I says to myself “I’ll give it a go.” WOW. Where was this a year ago (not fully developed then, but whatever). The short of it is that Pozotron will take audio that a narrator does and compare it with a word, text, or PDF and actually proofs the audio – and dang it, it does an unbelievable job!
NOTE: Pozotron is best suited for short or long form narration! If you use it for simple work, it will cost you more than proofing yourself!
To start off with...
The idea is to start a free account (yes, free), create a “project” whereby you can upload a script/manuscript and have it process this. There is a trial limit, so don't be surprized. If you purchase time on Pozotron (for more services), it will not ‘charge’ you for any processing until you upload audio and you can upload a new script if it gets changed or edited.
Now, once you have your audio, make sure its the way you want it. That is take out any mistakes, clicks (if you use them), spaces (not really an issue), breaths if you want – anything that might throw Pozotron off. Also make sure you have any commentary or repeated lines out or it will throw the proofing off. It will catch it and you can mark it, but why not take care of that to begin with. Also make sure the audio is loud enough. Best to be at RMS -18dB and Peak at -3dB.
Now, once you have that, simply drag the audio (whole script, book, chapter, paragraph, whatever) into the newly created project. It takes .MP3, .WAV and .FLAC files of up to 120 minutes. The site will take your audio, process it, compare it to the manuscript and once complete, will tell you on the proofing tab what issues it caught.
HINT: Make sure you have the most recent and updated script or manuscript before you upload any audio. Else you can’t upload a new one.
Pozotron finds a LOT. Depending on the narration and narrator, it will catch all long pauses, many of which I tell to filter out. Then it detects noises as best it can. I don’t ignore them, but I do listen to them as Pozotron will allow you to hear the audio with a prompted highlight of what the narrator is saying and where in the copy. Handy eh?
The ones that it does catch as ‘mistakes’, I listen to and try to determine if this was just the narrators emotive input or if it was truly a mistake. I “Ok” the ones I think aren’t an issue and mark the others as a pickup.
I also mark the ones that are breaths/noises as pickups so I know where they are later. I do the editing in terms of proofing, removing glitchy audio and time the audio better as my narrator puts in pauses you can drive a truck through (9+ seconds sometimes).
Now that I have that all marked I can download the proofing as ‘markers’ compatible in ProTooks, Audition, Reaper (my fave), Audacity, MIDI Markers (Studio One), Twisted Wave, Hindenberg, Sound Forge, WaveLab and Nuendo. This is great so I can use the markers in the DAW to know where to put the pickups! Cool huh!
Now, I could go into more, but come on, you’re here for me to highlight the 7 reasons why you should use Pozotron.
Tha Seven Reasons...
1. Ok, so the first and foremost reason is it uses AI in the right way -- As I talked to the great folks at Pozotron, the key takeaway was that they do use any audio they get to improve their product BUT then do not use the audio in any other way than that. What I mean is, they don’t sell a narrators voice or otherwise use it to create anything. That’s not what they’re all about. Pozotron wants to provide a service to short and long form narrators to help them proof and correct narrations in the most efficient and effective way. That, in my opinion is what AI can do for me.
2. They are cost effective – Remember above, I said you can make a free account to test the service. You can. They give you a limited number of minutes to test out the service. After which, you have some options for purchase of the time. No Sally, they ain’t free, but they are VERY cost effective – Especially if you’re an editor or narrator and time is money. At the time of this blog, you can do a Pay As You Go for $12 USD, with options for other plans starting at $29 USD per month. Some of these provide rollover months* and some provide a per-month fee plus $ for any additional time. Still, VERY cost effective. Right now, I am on the PAYG plan since I don’t do all that much editing.
3. Fantastic Customer Service – You do have to engage them via their email, but they will get back to you! Have no fear, they are here and their support person is a wealth of knowledge. He and I were on a zoom call and he was spot on with suggestions and collaboration about what Pozotron can do and some of their future plans (sorry, I can’t say or they’ll take my goldfish).
4. Collaboration – I mentioned this briefly, but in detail, you have the ability to receive audio and manuscript from a narrator, create a project, process the audio, mark the pickups, and guess what? Collaborate with the narrator! WOW! Who’da knew you can do that? Simply add the narrator as a “Share” with the proper authorization and they can record the pickups right from Pozotron. No joke. Right from Pozotron. And on top of that, you as the editor or the narrator can download the pickups along with any markers and incorporate them with the final narration. COOLNESS TO THE MAX!
5. The added features and plugins – If all this weren’t enough, Pozotron has a few secrets up its sleves.
Pozotron can analyze your audio to ensure it conforms to audiobook standards (ACX and a few others). Wooo-Ha! No more guesswork.
As a plugin, Pozotron can be ‘programmed’ to listen for custom pronunciations and interesting language accents. I currently edit for a British narrator and as an American, sometimes his accent is...unique. Pozotron first off ignores most of his accent no problem, but if there were some unique words in the copy that Pozotron either knows in a different way or doesn’t know altogether, you can program it to accept that custom pronunciation whenever it encounters those occurrences.
It can also do Characters. That is if you do a ‘character’ in a narration, it can proof that as well. Very helpful when the character has unique dialog.
If you have some special dialog that repeats and you want to count that or to know where the next one is, there is a Scan Occurrence plugin that will analyze the copy and let you know where those repeats are.
6. DAW Markers – Ok, I went into this a bit, but if you edit your own work or others, DAW markers are a godsend. Listen, when you’re trying to find something in the audio to insert a pickup and you can’t find it, you’ll drive yourself insane (which for me is a short trip). Having DAW markers that you can import will show you EXACTLY where the mistakes it caught are. Not to mention, the pickups are marked with those locations, so you always know where they go.
7. Follows the Chapters – Assuming you slate your chapters (Chapter One The Best Days To…), Pozotron will find that chapter, proof accordingly and then mark the pickups by that chapter. No guesswork as to “where was that pickup located at???” I mean, this service is slick. I have no idea how it knows, but it knows. And by that I mean, it’s intelligent enough to read the manuscript and find where it matches in the audio. Wow.
Ok, so that was a TON of information. Hope this helps and if you're into short or long form narration and the editing phase is a serious PITA, check out Pozotron today!
Tell them Woody send ya.