Audio Recording Insights
UPDATE: All but the very newest sentences have audio now. Report on the Discuss thread for the problem sentence, if you find a problem with audio. Please say if it is the male or female voice, if the problem occurs in the full phrase or a separate word, and what the problem you hear is. tlhIngan Hol DaQoy 'e' yItIv!
I am slowly recording audio for the lessons. This has to be done both sentence-by-sentence and word-by-word, and I can't use a separate recorder and upload, but have to record straight onto the computer. So it's a bit clunky, and sometimes you can hear my finger on the touchpad as I start and stop the recording. I do it as softly as I can and have no way to edit this out.
When I record a sentence or a token, I can listen to it, and decide if it sounds right, before I save it. I can't edit out the breath and saliva sounds that are part of speaking. After I click SAVE, Duolingo processes the audio before you can hear it. The two results of this are 1. if I record or change something, it's not available for you to hear immediately, and 2. Duolingo changes the audio. Their earlier system seemed to be that if any element of the recording redlined on volume, instead of reducing the volume of just that sound, it quieted the entire recording. The new system doesn't do that, but it sometimes introduces distortions. That's why I sound softer and less enthusiastic on the earlier recordings than I do in person, and sometimes there are bizarre distortions.
My assignment was originally to record about half the female audio, so you would hear female-voiced Klingon, but not only my voice. My priority to begin with was to get the early lessons recorded, so newcomers get to hear the language straight off, but I know some of you have been working on the course for a while and I don't want you to have to do all your lessons in silence. If you have a word or sentence that you'd really like to hear pronounced, I can jump around.
The other be' on the project dropped out, so I am now recording everything in one long slow pass. If you find sentences or tokens unrecorded in lessons before where I have worked up to (progress recorded in the thread below), distortions, unclear pronunciation, recorded strings that don't match the text, or noises that interfere with hearing, report them by coming to the forums to say what is wrong. We're not so interested in noises on the recording that are before or after the sentence, or a car going by in the background. (I usually re-record borDI' burghwIj though!)
I typically DON'T read the discussion forum to search for issues with audio, but the mods link your complaints to this thread for me. Students are welcome to comment directly in the thread themselves.
As the summer ends, sunrise has crept back to my recording time, and there's a super-noisy bird in my garden that is arguing with the sun about its territory, or something. Apologies if it ends up in the background. I don't have a soundproof studio.
Later ... now it's dark and birdless in the mornings, but I'm afraid you might hear the rain on the roof. At least snow falls silently.
I know that for some exercises the interface does not provide a freewrite report option, or multiple choice options that address your concerns, but please don't pick "Audio does not sound correct" out of frustration. The time I spend trying to figure out what is wrong with it could be spent recording another unit. And if there is something wrong, please come to the forums and say what it is.
Even if it's completely obvious what is wrong with the recording, it will get fixed faster if you comment here as well as marking it incorrect in the lesson. In order to see your bad audio report within the lesson I have to open the exact section, lesson and sentence, then scroll down. It might be a month before I even notice a "Audio does not sound correct" note on a given sentence.
I think there are six to eight months left for me to get to the end of the recordings, and have to start filling in ones I skipped early on when I thought I'd have a partner. I have cherry picked a few in each unit so that the advanced students aren't working in complete silence, but for the most part I'm working linearly through the course, now just finishing up Time Ago. I suggest any loDpu' doing recordings start at the end and work back to meet me. We could have recording reasonably complete by qep'a'.
This has the potential to be an endless project, so after qep'a' I may drop back to recording once a week instead of every day, so as to have time for other projects and maintaining my own Klingon skills.
I'm up to Lesson 3 of Purpose. It takes a week or two per skill. I keep trying to figure out if it goes more quickly as I advance, because some of the tokens are already recorded from earlier sentences, or more slowly because the sentences get longer. The number of lessons in a skill, and the number of sentences in a lesson is extremely variable, too. And every once in a while I stumble on a little pocket of sentences that I completed before I was working linearly. I try not to peek ahead, for the suspense.
Also, for people following this thread, I just extensively edited the post that this is a comment on, to update the information.
The audio progress report has plummeted to 23, from a high of 41, because of poor audio reports. Unfortunately there is no way for me to search for the poor audio reports without opening every sentence individually and scrolling down to see if a user has marked it for poor audio, then listening to all the sound files for that sentence to find out if there actually is poor audio quality somewhere, or if the person was just frustrated that they thought a H was a Q, so marked it that way. Please let us know in the forums where the problems are.
Some notes about recording quality, for people who explore this far.
I don't have the greatest recording equipment, so sometimes my voice is strained and irritated, enunciating a phrase for the ninth time, hoping this time that there won't be a crackle or a pop in the middle. Sometimes I tolerate the pop on a v. There's something about v.
I'm Canadian, and the Canadian rounding that makes Americans think we're saying aboot is a factor. We're not saying aboot it's just that the line between what is heard as o versus u is in a slightly different place for Canadians and Americans. This shouldn't affect Klingon too much, but even as I am following Okrand's guidance that the Klingon o should be as in mosaic, sometimes overemphasizing the rounding makes the o sound too much like a u. In any event, the o should be as in broke never as in rock (even not the British pronunciation of rock) or brook, and sometimes I have to call the American recorders on using the rock vowel. Call any of us on it if the sound isn't right.
I've been speaking Klingon for upward of thirty years, and the timbre of my voice shows that. (It's the years, not the Klingon!) Sometimes you'll hear is a bit of creaky voice, especially on a sentence-ending -'e' or -'a'. That's not a feature of the language, per se.
Americans think that Canadians say aboot in the same way that Trekkies think that William... Shatner... talkslike... this... all... thetime. It's a comical legend repeated so much, people who don't think about it believe it and repeat it, but anyone who actually stops and listens will hear it's not the case.