Glossika method

[etm001]

Geez, this thread is awfully quiet...we finally have the updated mainland material, but no one has posted any feedback? So how about it? What do you think about:

• The speaker's enunciation/pronunciation, etc.
• How natural sounding the phrases are (in particular the mainland-specific phrases that differ from those in the original material).
• Whether it (still) matters that there is no male audio recording.
• etc.

I can't comment on how 標準 the mainland speaker is, but I'm quite happy to incorporate the new audio into the existing material during my review sessions.

[hedwards]

I just started using the new one a bit ago, so I was hesitant to say too much about it.

 

I think it is an improvement, the ebook instructions are much clearer than in the previous edition. One of the things that I really don't like is that they chose to combine the Taiwan and Mainland text into the same book. It makes it a lot harder to follow along with the text when you're listening, it also makes it a lot harder to find the appropriate text when you're reading it back for recordings.

 

I do think it's an improvement overall though.

[Snuggles]

I think it is better too. But for almost a hundred dollars it does fall short. I wonder why they chose the speaker they did. I have a friend that does radio announcements and things like that. Why not a professional voice?

The sound is certainly good enough but I think that has room to improve as well.

[gato]

Could someone upload some sample MP3s for those of us who don't have the package to evaluate?

[Auberon]

 

Whether it (still) matters that there is no male audio recording.

 

It would be nice to have a male voice, or at least a mixture or the two. I don't know why, but the vast majority of mandarin learning material seems to be skewed towards over-representing female voices in audio.

Notwithstanding, at least Glossika's mainland women is a lot less high-pitched than the Taiwanese woman.

 

 

One of the things that I really don't like is that they chose to combine the Taiwan and Mainland text into the same book. It makes it a lot harder to follow along with the text when you're listening, it also makes it a lot harder to find the appropriate text when you're reading it back for recordings.

 

I don't object to this; I think it's beneficial to be exposed to the alternative way to saying something, even if one's not actively choosing to concentrate on it.

However, it's a pity that one is in traditional and the other is in simplified. And they frequently differ. It would have been a doddle to provide both for both; but the font is embedded, so converting an unfamiliar character from one set to the other is a laborious longhand exercise. In fact, character sets aside, just looking up any unfamiliar character is a longhand exercise. No doubt they embedded it so it would display correctly on computers without the requisite East Asian character support files, but really, who among Glossika's customers wouldn't already have them installed? And even for those who don't, it's a very simple process which they're going to want to do at some point anyway.

[wibr]

@Auberon At least in the old materials, as mentioned a couple of times in this thread, the font was not embedded (which would be fine) but converted to curves, which leaves only the shape of the characters and no information about which characters are actually displayed. They probably did this to prevent illegal copying of the sentences, however the size of the pdfs increased quite a bit due to the conversion and it's awfully inconvenient for honest customers, like you already experienced.

[imron]

OT technical note: Actually, that's not how font embedding works. When you embed fonts content is still perfectly copyable and pasteable, see for example the example PDFs on my site: http://www.hanzigrids.com/examples all of which contain embedded fonts. Try one of the ones with a handwriting font which you are probably unlikely to have on your computer.

What Glossika appears to have done is vectorise the fonts so they are no longer text.

I don't really see how it prevents piracy. It's still possible to OCR them, and that's not going to stop any dedicated pirate, and in the meantime it significantly impairs the files for paying users (e.g. can't conveniently search the PDFs).

Edit: wibr beat me to it

[Auberon]

I'm no whiz in IT terminology; I simply meant, as you say, imron, that text is not selectable or searchable. I'm sure Glossika themselves described the fonts as embedded on their website, which is why I used the word. As you mention too, it also occurred to me that they could no doubt ne extracted by OCR, hence I presumed it's an end-user font support issue. Fair enough for little-supported scripts like chu nom (which they apparently provide for Vietnamese--and I commend them wholeheartedly for taking the trouble), but annoying for Chinese.

[imron]

It's not an end user font issue. If it was, they could solve the problem by just embedding the font and the characters would still be visible on all computers, and also copyable and searchable and a fraction of the size. This is exactly what I do with Hanzi Grids.

This works regardless of language and font.

[etm001]

@hedwards said:

 

 

One of the things that I really don't like is that they chose to combine the Taiwan and Mainland text into the same book. It makes it a lot harder to follow along with the text when you're listening, it also makes it a lot harder to find the appropriate text when you're reading it back for recordings.

I guess it depends on how and what you are studying. I actually really like having both side by side for 

comparison. But if you want to focus on one more than the other, I can see how it might be distracting to have both.

 

In regards to the PDFs: setting aside the technical details, the inability to copy/paste the text is disappointing. And from a copy protection perspective, it makes no sense: not just because it could be OCR'ed, but Glossika sells its golden jewel - the recorded audio files - DRM free (as they should). In the end this just burdens paying customers who want to use that text in Anki, etc.

[gato]

The Basic 3 pdf has been converted to images and back to pdf, so you cannot search or copy at all. OCR could work but would be very problematic with the mixture of traditional and simplified chinese, pinyin, English and IPA. Another big difference is the size, basic 2 is just 2 mb, while basic 3 is a whopping 188 mb! I bought basic 2 a month ago and won basic 3 in the Hacking Chinese character challenge (http://www.hackingch...14-milestone-2/) two weeks ago, so the change must be recent.

2MB for text-based PDF vs 188MB for image-based PDF. They appear to be using file size as a form of copy protection, which needs some rethinking considering the low cost of bandwidth today

[hedwards]

@Auberon, but the audio you're listening to when you do the program is either Beijing or Taiwan accented Mandarin. In practice, you're not going to have time to look at both of them most of the time without a lot of work. I like to follow along with the pinyin the first time that I listen to the tracks, and having to find the appropriate line out of the 7 lines makes it a lot harder to keep up.

 

Then when you're reading the lines, you have to search for the line you want and that makes it less convenient as well.

 

For the marginal benefit of comparison, it's just not worth it. The correct thing would be to just split the book into several pieces, the introduction, the sentences for each program and that bit at the end on pinyin. Going through up to 3000 sentences where you're having to fish out the one line you want each time adds unnecessary work that distracts from the core purpose of the program.

 

@imron, font variants can be an issue, but as far as I know, if you embed the fonts into the document those issues tend to be non-issues. Now if somebody wants to specify a different font, that might screw things up, but that's not their responsibility. Personally, I would much prefer to be able to have at least the glossary be user parsable so that I could enter it into Pleco for some training to speed up the acquisition of the vocabulary.

 

I like the program, but I'm not sure if I would have bought it if I had realized at the time that they don't trust the users to have the text. 

[imron]

if you embed the fonts into the document those issues tend to be non-issues

If you embed fonts in to the document, then the characters will appear exactly as the creator intended.  There are no variant issues in the sense that the variant used will be exactly the same as the one used by the creator of the document, which can be assumed to be the one that the creator intended.  There's no technical benefit to the end user for Glossika to provide materials like this.

[etm001]

For what it's worth, a ran a few pages of the Glossika PDFs through ABBYY FineReader. It recognized the Mandarin and English text no problem. So if you are motivated enough, you could run the PDFs through an OCR program to extract the text for use in flashcard testing (since there is a mix of Mandarin, English, pinyin, and IPA, it might be a labor intensive process to extract the relevant text, depending on your OCR software).

[etm001]

Hi,

 

I'm am not familiar with mainland pronunciation. In reviewing the first 50 sentences in the Fluency 1 package, there was one thing that jumped out at me: the addition of a very short, clipped 啊 at the end of some sentences (at least that is what it sounded like to my ear). I'm wondering if this something I should mimic when practice the mainland sentences. For reference, listen to sentences 10, 12, 14, 19, 31, and 45.

 

In Taiwan 啊 can of course be added to a sentence for effect/emphasis, although it tends to be a longer, more drawn out sound and often higher pitched that what I heard in the mainland audio files. When speaking in Taiwan I'm careful not to over use this effect, as it can make one sound a bit "girly".

 

The only other thing I noticed in the mainland audio is the pronunciation of nèi/zhèi when the pinyin has nà/zhè.  Not a big deal, though.

 

Edit: oh, another thing I noticed is that mistakes in the original Taiwanese audio/PDF are still present. For example, the Taiwanese audio for Fluency 1 sentences #29 and #48 don't match the PDF. 

[Auberon]

Also noticed that 誰 is glossed as pinyin shui2 in the mainland version, even though the mainland speakers pronounce it shei2 in everything I've heard thus far. Is shui2 a common alternative/vernacular/regional pronunciation? Is it considered non-standard?

[ChTTay]

I'm a few days into the GSR files and so far I like ir. Even though I already know almost all the fluency 1 sentences, I still can't necessarily rattle them off that fast - that 'fluently'. For this reason, I'm happy to go through them again at the GSR pace. I just haven't had to look at the written sentences yet. I've recorded myself a bit and also just asked my g/f if my pronunciation is right. 

 

One thing is I've been told a few of the words aren't really spoken so much, for instance '明媚‘ apparently is more of a written term. At least this is what one Chinese person told me.

[Crush]

Also noticed that 誰 is glossed as pinyin shui2 in the mainland version, even though the mainland speakers pronounce it shei2 in everything I've heard thus far. Is shui2 a common alternative/vernacular/regional pronunciation? Is it considered non-standard?

Some of my friends here in 山东 pronounce it as shui2, though i've mostly heard shei2 as well. I think they are both considered standard pronunciations, from what i understand it was originally pronounced shui2 but lost the u sound over time.

[vkim67]

Does anyone know if the Chinese Business 1 package for 29.99 usd is also originally the taiwanese version and being updated? That one also (still) doesn't specify whether it's a Taiwanese or mainland version. Couldn't find a preview for this one on google books to see if the level would be appropriate for me...

 

 

 

I wonder why they chose the speaker they did. I have a friend that does radio announcements and things like that. Why not a professional voice? 
The sound is certainly good enough but I think that has room to improve as well. 

As for the above, can't comment on the Mandarin speaker(s) but I listened to the Korean speaker (preview clip on their website) and he sounds terrible... he sounds deflated, bored, and even a bit resentful like his mom forced him to do it.... I would NOT want to listen to thousands of sentences read in that voice. 

 

Thanks everyone for detailed reviews about the glossika chinese packages, as well as solutions to some of the problems. Nice to have all the pros and cons before making purchases. I have enough books and products I've bought and never use!!

 

I would also like to ask for separate sale of different levels! I have no need for level one. I would've been willing to purchase the whole thing and give the fluency level 1 to my boyfriend but seeing as the PDF is such a mafan.... even better would be to purchase mandarin level one for russian speakers for him and mandarin level 3 for english speakers for me... 

[etm001]

 

 

Does anyone know if the Chinese Business 1 package for 29.99 usd is also originally the taiwanese version and being updated? 

Since they were produced at the same time as the other original Mandarin packages, they probably feature a Taiwanese speaker. It wouldn't hurt for you to send Glossika an inquiry and post back with their answer.

 

I haven't gone through the business package, but I have gone through some of the other ones, comparing the Taiwanese Mandarin to the mainland Mandarin. From what I've seen so far, the differences (i.e., terms, grammar, etc.) are not huge. Although business language might an area where terms differ more between the two dialects (軟件 vs. 軟體, etc).