Rashmi Sinha

Rashmi Sinha

SINCE I READ an interview with Rashmi Sinha in the Information Design Journal, I’ve followed her enterprises with great interest. Most recently she’s been involved with setting up and running SlideShare.net, a place where you can upload your presentation slides — be they PowerPoint, OpenOffice or PDF — and share them either privately with an invited group or with the world.

Now, ordinarily that facility wouldn’t appeal to me. You see, when I prepare slides to accompany a talk, I don’t put much text on them. I believe that we’ve only got one language-processing centre in our brain and it has to cope with making sense of words whether they come in through the eye or the ear. So, making people read and listen to you at the same time is not a smart thing to do. A lot of the time, my slides just have pictures on them.

But I did get interested when I heard about ‘slidecasting’. That’s SlideShare’s term for linking an audio file to your slides and synchronising the turning of the slides to specific points in the audio file. It’s hardly a new idea, in fact it is reminiscent of the old slide-tape shows, but it could be an excellent way to ‘broadcast’ ideas over the Internet, and I thought I should give it a go.

The opportunity arose. On 17th September I am due to speak at an event on climate change organised by the British Computer Society’s Sussex Branch. I have to fit into a tight time-slot and though I usually speak from a handful of notes, this time I thought I should script my talk and time it. So here I had the perfect test case for slidecasting!

I don’t do PowerPoint, so I made my slides in Adobe InDesign CS3 by creating a document the pages of which measure 1024 points wide and 768 points high. Images were prepared in Illustrator and Photoshop. Finally, I exported the 42 pages as a PDF file, and this is what I uploaded to my SlideShare account.

HHB MiniDisc recorder with Røde studio mic in foreground and Sennheiser shotgun mic below.

HHB MiniDisc recorder with Røde studio mic in foreground and Sennheiser shotgun mic below.

Recording the audio

Audio is more of a challenge, but I quite frequently record conferences for the BCS and other organisations, and I also record voiceover narrative for video productions. My preferred kit for this purpose is an HHB professional MiniDisc recorder and a Røde condenser studio microphone. However, at the moment the HHB is sick, so that was out of the question.

Olympus DS50 audio recorder

Olympus DS50 audio recorder

My back-up recorder is very small indeed. It’s an Olympus DS-50. It has three sensitivity settings marked Dict, Conf and Lecture. The stereo microphone that plugs into it is actually quite good quality, considering. The audio is saved into the solid-state memory as Windows Media Audio (WMA) files, and there are five compressions one can choose from, and I always go for the most gentle.

So I took my script and chose the bedroom as the quieter place to record: it doesn’t face onto the river with all its traffic. I set the Olympus on Dictation-mode sensitivity to suppress as much background noise as possible and arranged a place for it quite close to my mouth, and made my recording. The thing to do is, if you fluff your lines or a magpie decides to add an accompaniment, you just repeat that paragraph and deal with it in the editing.

The Olympus attaches to the Macintosh as if it were an external USB drive and you copy the recording file across. Macs don’t much like WMA files but I’d bought Patrice Bensoussan’s EasyWMA audio converter plug-in for QuickTime ($10) and converted the file to AIFF for editing. I use Apple’s Soundtrack Pro for editing, chopping out the out-takes and suppressing intakes of breath. I thought the output from the Olympus sounded a bit ‘tinny’; the solution was to use the ‘Fat EQ’ filter to boost the lower frequencies until it sounded like me again.

I saved the result as mono AIFF and then took it into iTunes to convert it to MP3. My standard podcast settings use a data rate maximum of 56 kbps, 16 bits of precision, Variable Bit Rate and best quality conversion. The 23 minutes of speech came out slightly less than 10 MB.

Uploading and synchronising

SlideShare.net are quite generous in the size of presentations you can upload – up to 100 MB — but for now at least they won’t host your audio. You have to find somewhere from which audio will stream, and they recommend placing it on The Internet Archive, which is what I did (I already had an account there — it’s free to use, as is SlideShare.) Once you have copied the URL where your sound file sits, you paste that into a field on SlideShare, and the slides and audio are connected.

click image to enlarge!

SlideShare’s synchronisation tool: click image to enlarge!

I had feared that the synchronising of the slide-turns to the appropriate places in the talk would be tricky, and it was the most difficult part, but I soon got the hang of it. Considering that you are working with files on two different servers on the other side of the world, it’s quite amazing that it works at all! If you click on the thumbnail image of the screen shot, you’ll see my edit in progress. The slides ‘slide’ along in the upper channel, and the synchronisation workspace shows the waveform of the audio, which is something I feel very comfortable working with. The Audio Selector at the bottom wasn’t immediately intuitive to me; you slide it along to reveal another bit of the audio further on.

The red line in the Workspace is the ‘playhead’ and you play and pause the audio to find the sync points. I kept trying to stop and start the audio by hitting the space-bar because that’s what you do in iTunes, QuickTime and Soundtrack Pro, but not here! The process is a bit clumsy, and due to the latency across the Internet it’s a bit unresponsive at times. As you move the ‘slide x end’ line to where you want it, the next slide asks you to ‘Select this slide to synchronise’ and when you click on it, it dumps a new blue line close to the previous one. They you play along until you find the next sync spot, drag the rightmost blue line to that spot, and repeat.

It’s good. It works. It’s quite impressive. It’s all done with Flash technology. And it’s here (my first slidecast, I mean). It works full screen too, so you can really see detail if you want. My main gripe is that as my slidecast makes progress, the slides arrive later and later after the points that I had designated for synchronisation. I hypothesise that this may be because my slides have a great deal of graphic content?