There was a post on the Eclipse FB group about being able to search the transcript for times when multiple words had been deleted, in order to help find things that needed to be globalled/worked on. It hadn't ever occurred to me before, but Julie pointed out that you could just search for the steno **. Duh! I started doing it a couple days ago, and it's been absolutely amazing.
It takes me five or six times longer to scan through a transcript now, but I'm also finding (and fixing) five or six times more mistakes than before, so it's worth it. It makes it really easy to see mistakes that you make repeatedly, and it's also easier to pick out patterns with stacking or splitting. I'm getting a *lot* more splitting than I was aware of. When I was writing and something would come out wrong, I wouldn't necessarily know why at the time, but I figured the most likely reason was that I stroked or missed an extra key. It turns out a lot of it is splitting!
I think there are lots of settings I can adjust on the IE to address splitting. One that I know about is minLR. I had tried to adjust it down to fix stacking, but it never worked, and I started getting noticeable splitting. I've been gradually turning it up every time I recognize splitting, but it turns out there were tons of times when I didn't know it was happening. I've been turning it up more over the past few days, and I think it might be causing more stacking.
I'm interested to really look into all the things I can tweak for splitting and stacking, but I'm going to wait until I get the Gen O. Then I'll call support and have them help me out.
Tons of other things are also coming from searching for double-deletes. It reminds me about words I knew I was consistently misstroking, but never seriously thought about changing, like TEPD. It should be "tend," but it translates as "tepid." Whenever I've seen it before, I thought, "well, it's a valid outline for tepid, so I'm going to leave it." But the odds are good that I would write "tepid" as TEP/ID, anyway, and I know it's come out wrong over and over again when I meant to right "tend," so I finally changed TEPD to "tend" and made T*EPD "tepid."
Another one I did it with was APS for "Annapolis." It's a great brief, and I do IPS for Indianapolis, too, but I frequently get "Annapolis" when I was trying to write "answer." When it happens, I just think "that darn Annapolist again" and delete it, so it doesn't show up in the transcript and I don't think about changing it. But when I found it this time, it didn't take long for me to realize that I should make A*PS "Annapolist" so that APS can just be "answer."
That's the real beauty of it. There are so many mistakes that I erase, so the consumer never sees them, but that means I don't see them to fix them in my dictionary, either. And I certainly don't remember any of them by the time I'm done writing. Another "hidden" thing that I can see now is briefs that I tried, but hadn't entered into my dictionary yet. I knew I had a brief for "Boca Raton," and BO*EK/RA*T made perfect sense, so I stroked it, but that wasn't right. I wrote out "Boca Raton" instead, and I would normally have no way of remembering I should go ahead and enter in BO*EK/RA*T, too. But this way I can see things like that and enter them.
And it's not just for briefs - it works for new ways of writing things that haven't been assigned to all the words yet, like SAIRBL for the "-sational" in sensational. I had really stopped trying to write things the short way, because they never came out right. But now at least if I write them and they're not in my dictionary yet, I'll know I can catch them later and put them in.
I'm also finding tons more words to add to my practice list. It's mostly not words that were difficult to stroke, because I often misstroked those and am able to add the misstroke to my dictionary. But what I am getting is reminders about words that I forget how to write, like "upwards." I could write it as *UP/WARDZ, UP/WA*RDZ, or *UP/WA*RDZ, but because it takes more effort to include the asterisk, the last time I wrote it, I just did UP/WARDZ, which is "up wards." So I put it on my practice list, and hopefully the week of reinforcement will be enough that I reflexively include the asterisk next time. Even though I know it needs to be there, the fact that I didn't put it in and had to back up to re-write it means it needs to be reinforced.
I've already recognized some misstrokes that I make a lot, but always erase, so I never fixed them, and I just keep writing them and having to delete and re-write them all the time, like REP/SHIP for relationship. I found that twice in one transcript, and then REG/SHIP in the next one. As soon as I saw it, I knew it was one that I get wrong all the time, but it never stuck with me long enough to define it.
So now when I edit a transcript for my own improvement (which I do with every event that I write), I follow these steps:
-Do separate searches for glued A and E to find most fingerspelled words
-Search for */*
-Search for untrans and unresolved conflicts
-Spellcheck in Eclipse
-Spellcheck in Word
I could also just search for * and I'm sure I would find even more strokes to fix, but I think correcting things that really tripped me up and forced me to delete multiple strokes is a good starting place for now. I've made 404 entries in three days using that technique!
Showing posts with label Facebook. Show all posts
Showing posts with label Facebook. Show all posts
9.22.14
I've been coming up with briefs for some words that I felt like were taking me too long to write:
best practices - BIKTS
capacity - KAIPS
scenario - SAERN
criteria - CRAO*IR
apologize - PAO*LGZ
at the end - T*END (add DA*I for "of the day")
Bahamas - BA*EMZ
I couldn't believe how often people were saying "at the end of the day!" It takes so long to write out.
I have a new practice for when I can't make out much of what's being said, which can be a problem with remote work. Before I would just listen to see if I could catch any entire sentences or thoughts that made sense and write those, and leave out the rest. But now I write every word that I can hear, with a dash for any words I can't. I think it probably does actually help. Even if it doesn't totally express a complete thought, you can still get a sense of what the person's saying.
I still occasionally stroke "dun" when I'm trying to write "done," and I actually got to use "dun" the other day. The news anchor was talking about Facebook posts coming back to haunt you, and she said, "dun-dun-dun." I also correctly spelled tracheostomy, which I was pretty proud of. On the other side of the token, I had been doing a lot of Canadian meetings, and then I was captioning the news and the Pope started saying something in Italian, and I reflexively wrote "[ speaking in French ]". But then they said he was meeting with Rabbi Skorka, who I'd never heard of before, and I correctly guessed/fingerspelled "Skorka."
I love the Eclipse FB group, although as I learn more things it becomes a little bit less useful, since I already know a lot. There is one frustrating thing about it, though. It seems like whenever I come up with a problem, or something that would make my Eclipse experience better, and I'm pretty sure there should be a way to accomplish what I want, I make a post, and the only thing I ever get back is people saying, "Why would you want to do that anyway?" I guess I should be used to it by now, but somehow it's still irksome.
best practices - BIKTS
capacity - KAIPS
scenario - SAERN
criteria - CRAO*IR
apologize - PAO*LGZ
at the end - T*END (add DA*I for "of the day")
Bahamas - BA*EMZ
I couldn't believe how often people were saying "at the end of the day!" It takes so long to write out.
I have a new practice for when I can't make out much of what's being said, which can be a problem with remote work. Before I would just listen to see if I could catch any entire sentences or thoughts that made sense and write those, and leave out the rest. But now I write every word that I can hear, with a dash for any words I can't. I think it probably does actually help. Even if it doesn't totally express a complete thought, you can still get a sense of what the person's saying.
I still occasionally stroke "dun" when I'm trying to write "done," and I actually got to use "dun" the other day. The news anchor was talking about Facebook posts coming back to haunt you, and she said, "dun-dun-dun." I also correctly spelled tracheostomy, which I was pretty proud of. On the other side of the token, I had been doing a lot of Canadian meetings, and then I was captioning the news and the Pope started saying something in Italian, and I reflexively wrote "[ speaking in French ]". But then they said he was meeting with Rabbi Skorka, who I'd never heard of before, and I correctly guessed/fingerspelled "Skorka."
I love the Eclipse FB group, although as I learn more things it becomes a little bit less useful, since I already know a lot. There is one frustrating thing about it, though. It seems like whenever I come up with a problem, or something that would make my Eclipse experience better, and I'm pretty sure there should be a way to accomplish what I want, I make a post, and the only thing I ever get back is people saying, "Why would you want to do that anyway?" I guess I should be used to it by now, but somehow it's still irksome.
Clean Words
Just in time for my captioning work, Jeremy put up the list of dirty words that you could add to your "words that have been deleted from the spelling dictionary" file. If you don't have a word in your spelling dictionary, the Eclipse AI won't guess it for TM, integrated pre/suf, or anything else. All you had to do was paste the list into your file, and voila. I took an extra step, though; I wanted to text global all the words as {NULL} so they couldn't even possibly be on the screen for captioning.
It would've been a laborious process doing it by hand, but thanks to some playing around that I did for the IE (it's a long story involving importing/exporting dictionaries to use special keyboard layouts, and I don't even think I needed to do it in the first place), I realized that you can export a dictionary, edit it in Word, import it, and make changes that way! So all the times that I've wished I could "find and replace" all instances of something in a dictionary, I actually could have!
I used it to put = {NULL} after all the words, and also to change them to all caps since that's how they're going to look when I'm captioning, and I instantly had dictionary entries for the entire list. And there are some really bad words on that list, so I'm glad I'm protected now.
I also realized that for captioning, you want to have a lowercase "s" with your acronyms; like if they say there were multiple DAs, you don't want that coming out as DAS; it doesn't make any sense. So as I was editing and I noticed letters next to "s," I slowly just defined them with a lock case, but eventually I actually had time to write/define them all. I hadn't missed many, and it turned out to not take that long (I probably should've just done that in the first place, but I thought it would be more time-consuming than it was).
I had 1st defined as "first," which was really annoying. I don't know how that happened, but I definitely fixed it. I've also been defining common phrases like "10 or 15," "10 to 15," and "10 and 15" with the 10 as a numeral, because otherwise it comes out as "ten or 15," and I don't think about it fast enough to use my numeral trigger before the "ten."
According to my list of things I wanted to write about, I also spent some time paying attention to getting things like YA*RD and WI*CK in my dictionary as ^yard and ^wick, but I don't remember about that anymore. Probably a good thing to do anytime I'm writing something with an asterisk, expecting it to stick, and it doesn't. Maybe I'm not seeing them anymore because I get the delete space in there instead now.
I got into the suffix spelling rules and fixed it so I can't get things like "creativetivity," "establishmentment," and "streghtenenned" anymore, too. Really you would think the AI would be smart enough not to do those in the first place. I guess I don't have those things defined properly in the pre/suf table.
It would've been a laborious process doing it by hand, but thanks to some playing around that I did for the IE (it's a long story involving importing/exporting dictionaries to use special keyboard layouts, and I don't even think I needed to do it in the first place), I realized that you can export a dictionary, edit it in Word, import it, and make changes that way! So all the times that I've wished I could "find and replace" all instances of something in a dictionary, I actually could have!
I used it to put = {NULL} after all the words, and also to change them to all caps since that's how they're going to look when I'm captioning, and I instantly had dictionary entries for the entire list. And there are some really bad words on that list, so I'm glad I'm protected now.
I also realized that for captioning, you want to have a lowercase "s" with your acronyms; like if they say there were multiple DAs, you don't want that coming out as DAS; it doesn't make any sense. So as I was editing and I noticed letters next to "s," I slowly just defined them with a lock case, but eventually I actually had time to write/define them all. I hadn't missed many, and it turned out to not take that long (I probably should've just done that in the first place, but I thought it would be more time-consuming than it was).
I had 1st defined as "first," which was really annoying. I don't know how that happened, but I definitely fixed it. I've also been defining common phrases like "10 or 15," "10 to 15," and "10 and 15" with the 10 as a numeral, because otherwise it comes out as "ten or 15," and I don't think about it fast enough to use my numeral trigger before the "ten."
According to my list of things I wanted to write about, I also spent some time paying attention to getting things like YA*RD and WI*CK in my dictionary as ^yard and ^wick, but I don't remember about that anymore. Probably a good thing to do anytime I'm writing something with an asterisk, expecting it to stick, and it doesn't. Maybe I'm not seeing them anymore because I get the delete space in there instead now.
I got into the suffix spelling rules and fixed it so I can't get things like "creativetivity," "establishmentment," and "streghtenenned" anymore, too. Really you would think the AI would be smart enough not to do those in the first place. I guess I don't have those things defined properly in the pre/suf table.
Good Ideas
I've been leaning in a new direction with my speaker IDs. I started out thinking the fewer keys I used the better, so I would do the first letter of the person's name, or the last if the first letter was already being used. But now I've found that it's better to do a predominant sound in the name, because I'm probably going to have a lot of names. It's especially good for captioning when you have to get the whole news team in there, but I actually had a CART meeting where I was given 40 speakers the other day.
Some good ones I use are AUL for Paula or Paula, AIV for Dave, O*B for Bob or Rob, O*N for John, IS for Chris, AIS for Jason; the list goes on, and it's particularly helpful if I can use the same one across jobs. Chris is one that I can pretty much guarantee is always going to be IS; and for anyone who's not familiar with my speaker IDs, it's actually GLAUL, GLAIV, etc.
There was a really inspiring thread on depoman about writing verbatim or not. A lot of people are in the camp that they don't write repeated words if they don't change the meaning of the sentence, so if the person says
"I went -- I went to the store," they only put in one "I went." Someone said they've even had attorneys comment on how nice their transcripts look because of it. I don't know why I had previously been such a stickler about getting all those repeated words in if I had time, but I realized it does look much better without them, and they don't add anything, so now I don't put them in! It saves me strokes and time, and it looks better!
I also got on a little macro-making kick for a while. Brenda on the Eclipse FB group was talking about some real time-savers she came up with for actions she does often, and I thought, "Why am I even hitting Ctrl+Shift+Pg Up and then Esc to get to the top of the document?" The first command highlights your way up there, and then I had to unhighlight. I go to the top multiple times every time I review a transcript for fixes, so I just changed my keyboard map so that "V" takes me right to the top! I have E for the end.
I also check each transcript for any time I fingerspelled A and E to check for words I might've fingerspelled that I need to get in my dictionarly, so now I just hit S and W respectively to find the next instance of those. And when I'm done checking in Eclipse, I paste the file into Word and spellcheck it. Captioning files are always in all caps, so I had to navigate to the proper settings tab to uncheck that, but now I just hit Ctrl+Space bar and it automatically does it all! It's so much easier. There are so many key combinations, there's no reason to be using the mouse or pressing multiple keys for actions you do often.
Some good ones I use are AUL for Paula or Paula, AIV for Dave, O*B for Bob or Rob, O*N for John, IS for Chris, AIS for Jason; the list goes on, and it's particularly helpful if I can use the same one across jobs. Chris is one that I can pretty much guarantee is always going to be IS; and for anyone who's not familiar with my speaker IDs, it's actually GLAUL, GLAIV, etc.
There was a really inspiring thread on depoman about writing verbatim or not. A lot of people are in the camp that they don't write repeated words if they don't change the meaning of the sentence, so if the person says
"I went -- I went to the store," they only put in one "I went." Someone said they've even had attorneys comment on how nice their transcripts look because of it. I don't know why I had previously been such a stickler about getting all those repeated words in if I had time, but I realized it does look much better without them, and they don't add anything, so now I don't put them in! It saves me strokes and time, and it looks better!
I also got on a little macro-making kick for a while. Brenda on the Eclipse FB group was talking about some real time-savers she came up with for actions she does often, and I thought, "Why am I even hitting Ctrl+Shift+Pg Up and then Esc to get to the top of the document?" The first command highlights your way up there, and then I had to unhighlight. I go to the top multiple times every time I review a transcript for fixes, so I just changed my keyboard map so that "V" takes me right to the top! I have E for the end.
I also check each transcript for any time I fingerspelled A and E to check for words I might've fingerspelled that I need to get in my dictionarly, so now I just hit S and W respectively to find the next instance of those. And when I'm done checking in Eclipse, I paste the file into Word and spellcheck it. Captioning files are always in all caps, so I had to navigate to the proper settings tab to uncheck that, but now I just hit Ctrl+Space bar and it automatically does it all! It's so much easier. There are so many key combinations, there's no reason to be using the mouse or pressing multiple keys for actions you do often.
Learn Your Software
I've learned some pretty cool stuff from the Eclipse FB group. I strongly recommend checking it out if you're on Eclipse. I also watched the full 8 hours of Eclipse basic training tutorials (bo-ring!). It was mostly about title pages and auto indexing and stuff I don't use, or stuff I already know how to do, but I did pick up a couple ideas.
I have a stroke that takes me to the bottom of my realtime feed/has me follow it, and it was always bad when I forgot to hit that after I scrolled up to look at something. But I found out about the "follow always" checkbox in the realtime tab, and now that's not a problem anymore.
I learned that you can make conflicts if you want a certain thing to happen after a number. I use P-KT for % and PER/KRENT for percent, but sometimes it's so much easier to hit P-KT, that I just don't get PER/KRENT in there. I think I've got it set up now that I can write "a large P-KT" or something and it will come up "a large percent," so that's cool. I've been trying to see new opportunities for conflicts overall. Sometimes I miss the F in I've (AO*IV) and get "aye," but why not make it a conflict with the first choice being "aye"? Just today I changed YO*U from ewe to \ewe\you're, because I miss that R so often.
The best idea I got from the FB group was about suffixes. I think someone just said something about their suffix rules table, and it reminded me about how I keep getting things like "organizationtion" when I accidentally drag an extra G into ORGS/S, making it ORGS/GS. So I made a rule that when the word ends in -tion, the suffix should delete the -tion and add an -s, and voila! ORGS/GS is organizations now, and I won't get orientationtion or objectiontion or anything like that, either.
I've come up with tons of applications for that now that I've realized I can do it. No more defectiveive if I already tucked the V and forgot about it (DE/F*EFKT/IV), no more passageage if I got confused and did PAJ/AJ, no more strategicically, or longerer if I didn't realize I tucked the R (LORNG/ER). I don't think -tor was in there as a real suffix; it was in my dictionary, but it wasn't in the suffix table, so it didn't have any suffix rules. I kept getting things like compacttor and reacttor because instead of RE/AK/TOR, I'd write RAEKT/TOR.
But I fixed that rule, so it works for -tor and -tors. And I was getting things like "humorral," so I told it if you're going to add -ral and the word already ends in "r," delete the first "r." I feel like someone who does realtime already has all these suffixes in their suffix table and suffix spelling rules, and I must've missed out on them in an update or something. I should ask on depoman or the FB group for someone's tables, but I'm afraid it might mess things up. I don't see how it could, though, really; it would only change the suffixes in English. It wouldn't have anything to do with my steno outlines. I guess I'd just have to make sure all the things I've changed were still in there.
I also changed my background color from white to gray. Someone on the FB group said it's easier on the eyes. I've heard stuff like that before, but I never really thought about actually changing it until I saw it there. I really do like it better.
I have a ton of different settings files I use for different companies, so it's a pain to change suffix tables or whatever in all of them. I asked about it on FB, and Dave Stanley said that I could just have a global.set that all settings would use for certain things. It seemed like a great idea, except that it overwrites your realtime settings if you let it change your translation settings. It also overwrites your job dictionaries no matter what you do, which is a drag.
I learned how to finally get TM to stop messing with my stroke for "a," which is awesome. I told it to stop pulling "an" and "everyone" into other words, too. I actually turned TM up a bit. The new sliders are kind of confusing, but I understand them now. My main problem was that TM wasn't being aggressive enough; there were tons of words it just left alone. So I unchecked "require an untran," and I bumped it up from whatever it was at (it might've been as low as 1%) to 30%, and I think it's doing better now.
I found out there's a dictionary entry you can use to reject the last AB suggestion and get a new one all at once, and I also found a macro for requesting a brief for the last word, which I think will come in handy.
I have a stroke that takes me to the bottom of my realtime feed/has me follow it, and it was always bad when I forgot to hit that after I scrolled up to look at something. But I found out about the "follow always" checkbox in the realtime tab, and now that's not a problem anymore.
I learned that you can make conflicts if you want a certain thing to happen after a number. I use P-KT for % and PER/KRENT for percent, but sometimes it's so much easier to hit P-KT, that I just don't get PER/KRENT in there. I think I've got it set up now that I can write "a large P-KT" or something and it will come up "a large percent," so that's cool. I've been trying to see new opportunities for conflicts overall. Sometimes I miss the F in I've (AO*IV) and get "aye," but why not make it a conflict with the first choice being "aye"? Just today I changed YO*U from ewe to \ewe\you're, because I miss that R so often.
The best idea I got from the FB group was about suffixes. I think someone just said something about their suffix rules table, and it reminded me about how I keep getting things like "organizationtion" when I accidentally drag an extra G into ORGS/S, making it ORGS/GS. So I made a rule that when the word ends in -tion, the suffix should delete the -tion and add an -s, and voila! ORGS/GS is organizations now, and I won't get orientationtion or objectiontion or anything like that, either.
I've come up with tons of applications for that now that I've realized I can do it. No more defectiveive if I already tucked the V and forgot about it (DE/F*EFKT/IV), no more passageage if I got confused and did PAJ/AJ, no more strategicically, or longerer if I didn't realize I tucked the R (LORNG/ER). I don't think -tor was in there as a real suffix; it was in my dictionary, but it wasn't in the suffix table, so it didn't have any suffix rules. I kept getting things like compacttor and reacttor because instead of RE/AK/TOR, I'd write RAEKT/TOR.
But I fixed that rule, so it works for -tor and -tors. And I was getting things like "humorral," so I told it if you're going to add -ral and the word already ends in "r," delete the first "r." I feel like someone who does realtime already has all these suffixes in their suffix table and suffix spelling rules, and I must've missed out on them in an update or something. I should ask on depoman or the FB group for someone's tables, but I'm afraid it might mess things up. I don't see how it could, though, really; it would only change the suffixes in English. It wouldn't have anything to do with my steno outlines. I guess I'd just have to make sure all the things I've changed were still in there.
I also changed my background color from white to gray. Someone on the FB group said it's easier on the eyes. I've heard stuff like that before, but I never really thought about actually changing it until I saw it there. I really do like it better.
I have a ton of different settings files I use for different companies, so it's a pain to change suffix tables or whatever in all of them. I asked about it on FB, and Dave Stanley said that I could just have a global.set that all settings would use for certain things. It seemed like a great idea, except that it overwrites your realtime settings if you let it change your translation settings. It also overwrites your job dictionaries no matter what you do, which is a drag.
I learned how to finally get TM to stop messing with my stroke for "a," which is awesome. I told it to stop pulling "an" and "everyone" into other words, too. I actually turned TM up a bit. The new sliders are kind of confusing, but I understand them now. My main problem was that TM wasn't being aggressive enough; there were tons of words it just left alone. So I unchecked "require an untran," and I bumped it up from whatever it was at (it might've been as low as 1%) to 30%, and I think it's doing better now.
I found out there's a dictionary entry you can use to reject the last AB suggestion and get a new one all at once, and I also found a macro for requesting a brief for the last word, which I think will come in handy.
Subscribe to:
Posts (Atom)