|
So even with it making errors with my speech and pretty minimal training (I did the "Speak the instructions" thing yesterday before testing it), Dragon seems pretty great. It's not even 9am yet and I'm at least 2 or 3 hours ahead of schedule with this file. It sucks that I'll have to go back through it with a fine-toothed comb, but, poo poo, that can't take much more time than the file itself, it just means I have to put up with this guy for twice as long. Worth it if it means I get done with this file earlier than scheduled, and with so little stress to myself that I can do more work. I just hope my voice holds out, I'm not used to talking this much! Shima Honnou fucked around with this message at 13:26 on Jun 13, 2014 |
#
?
Jun 13, 2014 13:24
|
|
|
|
| # ? May 13, 2024 13:07 |
|
|
Shima Honnou posted:So even with it making errors with my speech and pretty minimal training (I did the "Speak the instructions" thing yesterday before testing it), Dragon seems pretty great. It's not even 9am yet and I'm at least 2 or 3 hours ahead of schedule with this file. It sucks that I'll have to go back through it with a fine-toothed comb, but, poo poo, that can't take much more time than the file itself, it just means I have to put up with this guy for twice as long. Worth it if it means I get done with this file earlier than scheduled, and with so little stress to myself that I can do more work. Trust me, you'll still make about $15/hour on most $.75 files, even with a second pass, if you can get used to revoicing. There are some files where typing is preferable, but most files can be done using Dragon, in my experience. Less of a strain on your wrists, besides. As is, you said you net about $7/hour, presently. You'll easily make that even with a second pass, and you'll eventually get a lot more proficient if you keep doing it. You will get used to it, the more you do it. 160wpm is standard for people who revoice (for reference: 145wpm is the average conversational talking speed), and 200wpm with only a .5% error rate is not impossible to manage. You'll be surprised how fast you become accustomed to it. Also: using a monotone is the best possible thing you can do. Just be sure to keep retraining it, which the program's probably asked you to do already. You can go back in and do the more complicated texts it asked you to read, just be sure that you back up and retrain it if you flub a word and it still registers. Now you just need to share your best errors. My favorite from the old days was 'voice mailbox' coming out as 'boy smellbox.' EDIT: Nitrousoxide posted:I tried Dragon recently for my application and found it was actually doable and a lot more enjoyable to do transcription that way. SHARE DEM ERRORS. A friend of mine got a freakishly large number of instances of 'Jew Party' when he was trying it out, and I still have no idea what the hell he was trying to say. 'Dropped my Jews on the sidewalk' happened only once for me when the operative word should've been 'juice.' Dragon has seriously politically incorrect comedy potential when it's being bitchy, and you're talking super-fast. EDIT 2: VVV I'd say yeah, but sorry, duder, I've been thinking of asking someone to take over one of my own thanks to work calling me in early tomorrow. 
Old Boot fucked around with this message at 14:01 on Jun 13, 2014 |
|
#
?
Jun 13, 2014 13:42
|
|
|
Does anyone feel like handling a 30 minute file for DT for me? It's easy clear-verbatim stuff, I'm just still behind schedule because I ended up sleeping for eighteen loving hours somehow.
|
|
#
?
Jun 13, 2014 13:47
|
|
|
Old Boot posted:SHARE DEM ERRORS. It's fun doing a video with some motherfucker who keeps cussing, because aside from "rear end", it doesn't seem to like picking up the cusses. Even "cuss" kept coming out as "cost". Can't argue with the speed boost though, that's for drat sure. I shouldn't be this far finished until, like, 11am according to my schedule.
|
|
#
?
Jun 13, 2014 14:01
|
|
|
Shima Honnou posted:It's fun doing a video with some motherfucker who keeps cussing, because aside from "rear end", it doesn't seem to like picking up the cusses. Even "cuss" kept coming out as "cost". Yeah, I still haven't figured out how to make Dragon recognize cusses, but, uh, to be honest, I'm not sure I want it to, especially if I fail to spot an error in a second, sleep-deprived pass. If you think of the words that could end up in a transcript if swearing was an option, well  You'll get used to typing it all in, and, by the way, if you train it up enough, you can get it to recognize 'um.' 'Uh' will never be recognized, but 'um' definitely is. Also, you can move faster if you voice punctuation, including hyphens for stutters. (which you may have already done. if so, excellent!)
|
|
#
?
Jun 13, 2014 14:06
|
|
|
I tried indicating points where I'll be doing stutters with the word "stutter", but it kept giving me shorter and surfer so I gave up, I'll get those in the second pass.
|
|
#
?
Jun 13, 2014 14:09
|
|
|
Shima Honnou posted:I tried indicating points where I'll be doing stutters with the word "stutter", but it kept giving me shorter and surfer so I gave up, I'll get those in the second pass. Yeah, just say 'I want hyphen I mean hyphen I don't hyphen I'm a stuttering idiot' out loud, and it'll come out as 'I want-I mean-I don't-I'm a stuttering idiot.' It's completely compliant with DT's style. Comma, period, dash, slash, ellipse are all recognized commands so far as the program's concerned. If someone SAYS any of those words, you should type it, obviously, otherwise the program's going to merrily insert punctuation. EDIT: Ellipse is more difficult to manage, so I just pause for long enough to put those in there. Same thing goes for most names. Still, you'd be surprised how much the program does recognize. A lot of drug names, provided the speaker is saying them properly, come out perfect, even if you don't have Dragon Medical, or whatever the hell it's called. Old Boot fucked around with this message at 14:19 on Jun 13, 2014 |
|
#
?
Jun 13, 2014 14:12
|
|
|
It's managed to impress me so far, especially since it seems to get surnames correctly (2/2 so far in this video). Not sure how consistent it would be with stutters for me, though, a lot of what I get is poo poo where someone either misspeaks and stops themselves "sa-something", or where they just start a word and then restart it, "f-gently caress". I do like how it separates alright to all right, though, because that's something I still have trouble doing.
|
|
#
?
Jun 13, 2014 14:28
|
|
|
Shima Honnou posted:It's managed to impress me so far, especially since it seems to get surnames correctly (2/2 so far in this video). For those, you need to type it in. I usually ignore the single consonant blunders (as if I'm doing clean verbatim) so I can get them on the second pass. It makes things go more quickly, and they're easy things to correct as you're listening to the file (without having to pause). Really, Dragon 12 is a hell of a long ways from Dragon 9, which I had to use when I was doing revoicing initially. Good lord, when that thing trainwrecked, it seriously trainwrecked.
|
|
#
?
Jun 13, 2014 14:31
|
|
|
Poison Mushroom posted:Does anyone feel like handling a 30 minute file for DT for me? It's easy clear-verbatim stuff, I'm just still behind schedule because I ended up sleeping for eighteen loving hours somehow.
|
|
#
?
Jun 13, 2014 14:41
|
|
|
Poison Mushroom posted:Re-mentioning this because I forgot to say that A) it's due today, rather than tomorrow, and B) I would do absurd things to not have another little black spot on my record already. How many hours is it due in? I can probably fit in time enough to run it through this whole revoicing thing, but I may not have enough time to actually do a second pass since I want to shoot for the moon and get my assigned projects for the weekend done today. Also, "pig" is apparently "patent" or ". (period)" with a Detroit accent, according to Dragon.
|
|
#
?
Jun 13, 2014 14:43
|
|
|
Shima Honnou posted:How many hours is it due in? I can probably fit in time enough to run it through this whole revoicing thing, but I may not have enough time to actually do a second pass since I want to shoot for the moon and get my assigned projects for the weekend done today. Gonna PM you with details.
|
|
#
?
Jun 13, 2014 14:44
|
|
|
Hey, is there a way to turn off the loving commands on Dragon? Because this poo poo just turned a name into "Yeah, I don't think you said 'Don', I think you commanded me to close this Express Scribe page and ruin all your work!" I don't give a poo poo if it removes the ability to bold or whatever, I don't need it doing this poo poo on me. Good loving thing ESCribe saves old dictations automatically otherwise I'd be loving hosed right now.
|
|
#
?
Jun 13, 2014 15:25
|
|
|
Shima Honnou posted:Hey, is there a way to turn off the loving commands on Dragon? Because this poo poo just turned a name into "Yeah, I don't think you said 'Don', I think you commanded me to close this Express Scribe page and ruin all your work!" I don't give a poo poo if it removes the ability to bold or whatever, I don't need it doing this poo poo on me. Good loving thing ESCribe saves old dictations automatically otherwise I'd be loving hosed right now. Tools -> Options -> Commands in the Dragon task bar. It hosed me up once when it assumed that something I said meant 'hey read everything I just said back to me at super slow speeds!' EDIT: You can also look at the command browser to see what screwed you up. That's also under tools.
|
|
#
?
Jun 13, 2014 15:37
|
|
|
Thanks for that. Now to second pass this poo poo and listen to a pretty petulant babby complain about how reality TV isn't real enough while I get his timecodes and fix what Dragon got wrong. E: As long as second passing is taking me I might as well have just done it by hand in the first place, goddamn. Especially since I'm starting to not feel great. E2: I was not anticipating getting ill like this. I've been working on this second run for 4 hours and I've still got like 30 minutes to go on it. Dragon did not help as much as I thought, with how many things I have to fix (And with this being such a loving bullshit project, gently caress THIS SHOW), it's like there's no end in sight for me right now. I think I'm going to end up canceling the project that's due tomorrow, which is more of this loving show and take the day for myself. Shima Honnou fucked around with this message at 20:58 on Jun 13, 2014 |
|
#
?
Jun 13, 2014 16:37
|
|
|
Question for JKL or anyone else on the QA/manager side of transcribing - how many times would you have to request a revision from somebody before you put a note on their file or something and stop sending them focus groups? I hope I'm almost there.
|
|
#
?
Jun 14, 2014 00:44
|
|
|
If anyone is looking to jump on the Dragon bandwagon, it's on sale for father's day. Scroll down if you're looking for the Premium version. If you don't get it from this deal, watch Newegg as they seem to run deep discounts on Nuance software from time to time.
|
|
#
?
Jun 14, 2014 18:58
|
|
|
Internet Friend posted:If anyone is looking to jump on the Dragon bandwagon, it's on sale for father's day. Scroll down if you're looking for the Premium version. I've been looking at this, and I want to know what is the difference that the premium version is offering over the regular? Is it anything that would actually help with transcribing? Also, how does DT determine the pay rate for a project? Because I've got a guy with A)heavy accent and B)highly technical language who is verbalizing two people talking in sudden fast spurts while blowing his nose and sneezing, and it pays the same as a super clean run through of a guy standing in the woods chatting in perfect English. Tech language is not a problem for me because of my day job, but I'd still prefer to get paid more for for it (especially considering the fact that I have deciphered tons of chemical names the dude is butchering the pronunciation of)
|
|
#
?
Jun 14, 2014 22:14
|
|
|
Zorblack posted:I've been looking at this, and I want to know what is the difference that the premium version is offering over the regular? Is it anything that would actually help with transcribing? No clue. I'm doing some super loving easy files right now (Movie or documentary footage or something; only transcribe the actors and only while they're being filmed performing in the scenes) and it's $0.85 cents per minute instead of the usual $0.75. Not a rush job or anything, in fact these are significantly easier than most of the files I've gotten recently. Interestingly, they seem to have taken into consideration that I used to be a cook and still sometimes cook (Like, good stuff) as a hobby, since all of the footage I got is pertaining for buying food, preparing it, or serving it. The same is probably also true for your video being more on the tech side, if you've made mention of your tech background. Shima Honnou fucked around with this message at 01:46 on Jun 15, 2014 |
|
#
?
Jun 15, 2014 01:25
|
|
|
Dipping my toes into transcription via DT, completely new to it and going through the test footage, and I'm curious on a couple of questions real quick: What's the best way to handle two people talking at once? The way I originally thought was doing, say: SPEAKER 1: Stuff, stuff, stuff- SPEAKER 2: -stuff? Stuff- SPEAKER 1: -stuff. (etc.) But I also went back and double checked guidelines and noticed the bit about dashes: is that the general format, and it'd be marked with - just a period, stop? Ellipses? Would 'uh's' be separated by commas or -'s for stutters? And, oof, two spaces after a period, huh? Something to get used to, I guess. Finally, man, that felt like it took forever for how quick it was and really trying to get it right... do you generally pick up speed as you go? And I can definitely see why a pedal would be useful.
|
|
#
?
Jun 15, 2014 01:59
|
|
|
For the test, get as much as you can. If you can't identify a person, use MALE or FEMALE, but if it's still that Iron Chef video they put up a couple months ago, it shouldn't be too hard, just use Wikipedia/Youtube for your main dudes, and then logic based on any names they may say for others. Note that if you do get the job, using Google is a requirement. Sometimes to figure out spelling, or a place name, addresses, businesses, I've even had to do it to get the names of speakers before. Not just the spelling, their actual names. If someone's interrupting someone else, you want to break it with ellipses. Do the same if they go over their XX seconds without having a period stop and the guide for a given file demands break by 40 seconds or whatever; I usually try for the nearest full stop or ellipses stop when I do that. If they're cutting off the other person and they stop mid word, I break with a dash then ellipses and haven't had QC complain about doing that. (This is an example: I am talk-...) If they completely lose their train of thought, make a long pause and restart a sentence, whatever, that's an ellipses too. If they stutter, transcribe it. S-s-s-stutter. Stutter-stutter. Stutt-stutter. If they misspeak a word and catch themselves, transcribe it. Statte-stutter. Note that you may have to transcribe someone going through what can only be described as a complete stroke once in a while, I've had to. Part of the job, they wanted verbatim, they got it. If they're saying uhs or ums, do that too. I usually pick a side of a comma to do an um or uh. I was going to uh, forums to post, um and I did and it was a poo poo post. Just listen for how long they pause. Note that there are jobs, usually actual interviews, where stutters, ums, uhs, etc, are not wanted. This is "clean verbatim" and it is your friend. I've had to transcribe a person with severe cerebral palsy before, and I'm glad it was a clean verbatim job because that person had a terrible and unfortunate stutter. Also note that many interviews don't want verbatim for the interviewee, but instead want a concise paraphrasing of the question. After the test things will be easier since, for most projects anyway, you'll be able to do things like [OVERLAP] or [NON-INTERVIEW] or [INAUDIBLE] more often, although there are some hell projects where you aren't allowed, such as this terrible reality show I've had to do multiple times; 70-80 minute videos of some rear end in a top hat just talking and I gotta get every word, and the pay isn't any better per minute than a clean verbatim with [NON-INTERVIEW] allowed, either. Mostly they just want to make sure you can handle yourself. Be glad it's not the lovely older video, it was longer and had bad audio quality. By hand I can do about 20 minutes of footage per hour if I rush myself, taking a break every now and then to not blow out. I average closer to 15 with longer breaks, 10 if I'm being lazy. I recently started trying Dragon speech-to-text software to get the gruntwork done, but it's not trained enough yet that I don't have to still painstakingly go over a file and every line to make sure it all matches. Speed goes up after you adapt to the job and get the right equipment. If you get accepted, invest in a footpedal immediately (~$50 new) and good headphones as soon as possible (I had ancient Sony studio headphones which earned me $1000 doing this job, but they've been wearing on me since I've been using them constantly for the past ~13 years, so I recently replaced them with a Shure SRH940 that ran me $250, but once you get decent at the job that's not that bad a price, especially for something that can help you pick up more words). I'm fairly new, though. I've been doing this job for exactly one month tomorrow, got my first work order on the 15th of May. There's people here with much more experience, so listen to them as well and probably take their advice over my own, but I've been complimented on turning in quality transcriptions by QC before already. Just, so you know, don't use "Quotation marks" around something if the interviewer is prompting the interviewee to say something specific (They didn't mention if using them around an actual quote that an interviewee is saying is incorrect or not yet). Shima Honnou fucked around with this message at 02:52 on Jun 15, 2014 |
|
#
?
Jun 15, 2014 02:35
|
|
|
Thank you for the thorough response! That was very helpful, and answered everything, I think. Yeah, I did a fair bit of googling with the test video to make sure I had a couple of things correct / names.
|
|
#
?
Jun 15, 2014 02:49
|
|
|
C7ty1 posted:Thank you for the thorough response! That was very helpful, and answered everything, I think. Yeah, I did a fair bit of googling with the test video to make sure I had a couple of things correct / names. I've watched the video; bibimbap and liqueur are probably what might get you there, also wagyu beef. If you can't get something researched properly, just use [PH] behind it; I can't spell this awufel [PH] word. Most often you'll have to do that with names, I do it behind any name that isn't common and that I can't find the person on Google. If you can't tell exactly what they said but have an idea, [do this to it?] Try to avoid using [OVERLAP] or [INAUDIBLE] unless you really can't pick out words, I've been sent back over files to reduce them before, which was part of what prompted me to get these new headphones (Which are great by the way, I just wish they were also surround, that's the only thing that could have made them even better). Don't use [BACKGROUND NOISE] if it washes out what someone's saying, QC's told me to just use [INAUDIBLE] despite what the guides said. If you get the job, you might also hear the crew call for silence, or outright use the word room tone at the end of or start of a video, just transcribe it as [ROOM TONE]. There are also special rules for [NON-INTERVIEW] that I've posted about before, but you shouldn't need that for the test video. Shima Honnou fucked around with this message at 03:04 on Jun 15, 2014 |
|
#
?
Jun 15, 2014 02:55
|
|
|
Good God this ABS I'm doing is a nightmare. The camera changes literally every 2-3 seconds. It feels like I'm doing a minute per hour. Whoever produced this show has serious ADD issues. Oh well at least it pays about 2.25/minute. Note to self: Never do this again
|
|
#
?
Jun 15, 2014 02:58
|
|
|
What does ABS even stand for?
|
|
#
?
Jun 15, 2014 02:59
|
|
|
Shima Honnou posted:What does ABS even stand for?
|
|
#
?
Jun 15, 2014 03:18
|
|
|
Poison Mushroom posted:Absolute Bull poo poo? I wouldn't know, I haven't been offered nor taken any 
|
|
#
?
Jun 15, 2014 03:19
|
|
|
As broadcast. Apparently you have to document every camera change.
|
|
#
?
Jun 15, 2014 03:38
|
|
|
Zorblack posted:I've been looking at this, and I want to know what is the difference that the premium version is offering over the regular? Is it anything that would actually help with transcribing? Here's everything you need to know about the various versions: http://www.nuance.com/ucmprod/groups/dragon/@web-enus/documents/collateral/nc_025214.pdf As a note: I have Premium and it works just fine for me. I haven't spoken to anyone who revoices with Pro, so I wouldn't be able to tell you how whiz-bang those 'transcription' add-ons on the last page are. quote:Also, how does DT determine the pay rate for a project? Because I've got a guy with A)heavy accent and B)highly technical language who is verbalizing two people talking in sudden fast spurts while blowing his nose and sneezing, and it pays the same as a super clean run through of a guy standing in the woods chatting in perfect English. Tech language is not a problem for me because of my day job, but I'd still prefer to get paid more for for it (especially considering the fact that I have deciphered tons of chemical names the dude is butchering the pronunciation of) I honestly have no idea how this is gauged by HR or whoever negotiates price with clients. I will say that I don't take jobs from DT anymore without screening the files that I'm assigned. If I listen to it for a while and determine that it's too much work for the price, or I can't understand a goddamn thing that's being said, I'll tell them to assign it to someone else. The only time I don't do this is if I'm past the six hour limit they give you for opting out. Old Boot fucked around with this message at 03:43 on Jun 15, 2014 |
|
#
?
Jun 15, 2014 03:40
|
|
|
What show in the world broadcasts with camera cuts every 3 seconds? That's so strange sounding to me.Old Boot posted:I honestly have no idea how this is gauged by HR or whoever negotiates price with clients. I will say that I don't take jobs from DT anymore without screening the files that I'm assigned. If I listen to it for a while and determine that it's too much work for the price, or I can't understand a goddamn thing that's being said, I'll tell them to assign it to someone else. The only time I don't do this is if I'm past the six hour limit they give you for opting out. That's actually something that I want to ask, does DT count it against you if you reject certain jobs? Like, I've done enough of that one full verbatim reality show for a lifetime, would they be okay if I rejected the next time it gets offered, instead asking for a different project, or is it just a "No, you do this/Okay, but we don't send you different work for the same time slot" deal? Shima Honnou fucked around with this message at 03:43 on Jun 15, 2014 |
|
#
?
Jun 15, 2014 03:40
|
|
|
Shima Honnou posted:That's actually something that I want to ask, does DT count it against you if you reject certain jobs? Like, I've done enough of that one full verbatim reality show for a lifetime, would they be okay if I rejected the next time it gets offered, instead asking for a different project, or is it just a "No, you do this/Okay, but we don't send you different work for the same time slot" deal? For a while, I had to reject video files because my old computer was consistently choking on them when I loaded them in InqScribe. I just let them know to send only audio until my computer got updated, which they were cool with. Otherwise, if I really dislike a certain client, I'll tell them not to assign jobs associated with said client to me. They're generally cool with it so long as you let them know ahead of time, so they're not suddenly scrambling to find other transcribers who are available for the job. Basically, they're loads happier with you if you let them know ahead of time, and they're generally gracious about assignments being rejected so long as you let them know ASAP that it should be reassigned. Really, if they paid more, I wouldn't reject the shittier files, but they don't, so I do. EDIT: And, yes, sometimes it'll mean just not getting work that day, or sometimes they'll send you a different project instead if you say you're able to do the work, but can't really do that specific file for whatever reason. Old Boot fucked around with this message at 03:50 on Jun 15, 2014 |
|
#
?
Jun 15, 2014 03:46
|
|
|
I can deal with a bit less work for a day or two, or a possibly lighter paycheck, if it means I don't want to shoot myself working a certain style. I'm someone who actually loves all this behind the scenes reality TV bullshit, just... not when I have to get every single one of 12k+ words, and not when it demands getting timecode down to the frame, that's a bit much. Maybe if those jobs paid better I wouldn't mind as much, but considering that for the same price I've seen much better projects, and even better projects for better prices now, that show feels like downright slave labor.
|
|
#
?
Jun 15, 2014 03:52
|
|
|
Knightmare posted:I recently got accepted to Writer's Domain and did a bunch of articles on keywords I knew about, they pay $20 an article and they'd take me anywhere from 30 to 45 minutes. Now I'm having to do some research on keywords I'm not experienced with which bumps it up to an hour due to research. I'm still pretty new to writing, and still learning the kind of stuff they like to see, but hopefully when I get that down I can average about 45 minutes, ~$30/hr. The cap has been indefinitely turned off since May 1st. They might turn it back on, who knows…. people I know have been writing 5-10 a day, cashing out once a day.
|
|
#
?
Jun 15, 2014 05:47
|
|
|
ABS is As Broadcast Script; instead of working the raw pre-production video, you're producing a script that documents a completed show. Lots of times these are for international markets, so a country can look over an entire season of episodes quickly and easily and determine if there's any objectionable content they want to cut out before syndicating the show. Networks also use these to build clip shows and that sort of thing. There are a lot of different types of ABS, and unfortunately it seems that right now, DT is focusing on the biggest pain-in-the-rear end format. The simplest is a dialogue list: quote:BOB Most of the work I've done for NBC/Universal shows has been dialogue list work. Sometimes they want periodic time codes, sometimes they don't. The second type is the standard three-column format. In this one, you need to timecode each speaker and record what they say, and also provide some sort of rough outline of what's going on on screen. So... quote:00:01:00 / Bob and Ted in the workshop / BOB: Hey, can you pass me the hammer? There's another one, which HBO uses, which reads like a script. (I'm not going to try to format this properly, but there's special indents for scene/action/dialogue and everything.) quote:00:01:00 The pain-in-the-balls format is the one where you have to document individual shots, and that's the one that DT seems to get for a lot of their ABS work. quote:00:01:00 / Bob and Ted in the workshop The vast majority of ABS I've encountered outside of DT is the first two types, and they pay at about the same rate DT offers for the pain-in-the-rear end format. Which is why I tell people to learn how to do ABS, because it pays better (and is sometimes less work) than standard time-code-verbatim transcription.
|
|
#
?
Jun 15, 2014 07:27
|
|
|
drat well I finally finished that thing almost. But what the hell I hate having formatting issues and I suck at troubleshooting them in Word. I have two instances of the same text showing up in the document for some reason but when I delete one the duplicate also deletes itself so they are linked somehow. No clue how that happened. Also I suck at page breaks and I need like three different headers and can't figure it out. I think I made about 4.50 an hour on this job.
|
|
#
?
Jun 15, 2014 20:07
|
|
|
Astro7x posted:The cap has been indefinitely turned off since May 1st. They might turn it back on, who knows…. people I know have been writing 5-10 a day, cashing out once a day. I actually managed that for a while myself, right up until they changed the way keywords work. Now I'm struggling to figure out how to write another article about locksmiths/towing/whatever.  I shouldn't complain too much, I'm still very happy with WD as a way to make money that's feasible for me. I just know more than I ever wanted to about Canadian locksmithery.
|
|
#
?
Jun 16, 2014 00:16
|
|
|
The daily transcription application video file is the worst thing I've ever seen. It's part of an iron chef episode where there is an insane amount of crosstalk, studdering, and foreign names. I want to punch both of these announcers in the face.
|
|
#
?
Jun 16, 2014 13:31
|
|
|
The application file is supposed to be terrible, because it's a gauge of how well you can work with abysmal content. If the file was five minutes of perfectly recorded single-speaker video, every applicant would just ace it. 
|
|
#
?
Jun 16, 2014 13:47
|
|
|
Nitrousoxide posted:The daily transcription application video file is the worst thing I've ever seen. It's part of an iron chef episode where there is an insane amount of crosstalk, studdering, and foreign names. I want to punch both of these announcers in the face. It's nothing compared to the previous one, which had poo poo audio quality, more crosstalk, words that you could barely hear, and they were outside and kept talking during poo poo like low-flying planes interrupting the shot. It was also twice as long as the Iron Chef video.
|
|
#
?
Jun 16, 2014 14:12
|
|
|
|
| # ? May 13, 2024 13:07 |
|
|
Bifauxnen posted:Question for JKL or anyone else on the QA/manager side of transcribing - how many times would you have to request a revision from somebody before you put a note on their file or something and stop sending them focus groups? I hope I'm almost there. Where it works by me, there really isn't time for revisions. Transcripts are taken just a few hours after submission for our in-house editors to work with. The QC side basically consists of me reading through surveys that the in-house people fill out, and if a particular survey merits further attention, I personally proof the transcript against the final file and send feedback to that transcriber. Unfortunately if that particular transcriber fails to improve over a certain period of time, then I have to let them go--which I did for the first time just last week, actually. There is a spreadsheet I keep for transcriber likes and strengths, but you really have to be ready for any content matter with this job. That hearkens back to the variety and difficulty that you find in test reels.
|
|
#
?
Jun 17, 2014 01:16
|
|