Image talk:JakeGyllenhaal.ogg

From Wikipedia, the free encyclopedia

[edit] Spoken article quality assessment

Reviewed by: Macropode 07:37, 5 March 2007 (UTC)

The version linked is the version I used - but I edited it as I went along in view of edits made while recording. It is impossible for me to raise the volume of the recording without adding background noise that I;ve found I can;t remove.

Frankly, unless you tell me how I can actually change the criticisms you made, this review is mean-spirited at worst and useless at best. Dev920 (Have a nice day!) 09:29, 5 March 2007 (UTC)

Thank you for the straight-forward feedback, Dev920 (and just to be clear, I'm not attempting sarcasm here). Just as you're learning how to make good spoken recordings, I'm learning how to do effective and appropriate reviews of them. Neither of us have any well-documented guidelines to follow. This is a situation that I'm hoping will change, and these assessments constitute the first stage of an ongoing attempt to do that (see the old review page for more details).

They are designed to highlight, in an impartial and objective manner, quality issues with recordings as a first step in helping narrators to identify the strengths and weaknesses in their work. They are not intended as any kind of personal criticism. A primary goal is to help myself and others to form a clearer picture of the kinds of difficulties people are experiencing so that project documentation can be improved (principally here and here) to provide useful plain-language assistance with recording issues. Although I'd like to, I cannot help every narrator who's work I review automatically as part of the review process, as I have other demands on my time apart from Wikipedia. My main priorities are firstly to quickly get new work reviewed, and secondly to get the above-mentioned documentation improved, so that everybody can be helped by this effort. However, I am quite willing to personally help anybody who asks in a reasonable manner. Short of this, your first option for assistance is to ask one of the recording assistance people, as suggested at the bottom of the article review, and to refer to what documentation there is, at present.

While I understand your apparent frustration with the somewhat terse nature of this review, the "criticisms" made were actually fairly minor. In my own, subjective opinion (which is why I didn't state it explicitly in the review) this is a fundamentally high quality spoken article, a fact reflected in the "High" clarity rating, that simply needs a few technical improvements, and an accurate indication of which historical text version it was derived from. -- Macropode 04:26, 6 March 2007 (UTC)

Hmm. I will bear this in mind when I do my next one. Though I am relistening to this through speakers and it sounds absolutely fine with no background noise at all, which is quite strange because you would expect it to be amplified, no?

Yes, although I find headphones more effective for hearing detail in recordings. This is, however, a moot point as there is no background noise worth mentioning in this recording. It seems I haven't stated the isssues and their relative importance (or lack of it) clearly enough in the review, so I'll try to clarify the points at issue here:

• Technical quality:

"General low audio level". This is not, in itself, a significant problem. Typically what will happen is that your listener will play the recording, hear a faint voice and turn the volume up so they can hear it clearly. When it finishes and they forget to turn the volume back down, the very next sound their computer plays that's been recorded at a higher (normal) level will be highly amplified, and they'll jump out of their skin. Been there, done that. It's worse if they're listening with headphones or earbuds. So, after you've finished recording, go into your sound editing software with your audio file loaded, and look in the menus for something usually called "Normalize" or "Amplify". This function will allow you to raise the level (volume) of the entire recording. If, in the Normalize/Amplify dialogue box, your software allows you to specify something like "peak amplitude", tell it "-3dB" or maybe "0dB" (which is higher). If not, don't worry about it, just raise the level as much as you can without producing clipping (distortion), a harsh, edgy sound that happens if the level is too high. Clipping usually sounds pretty ugly, so you'll know it when you hear it. If you can't get a high enough level from your microphone when you're doing the recording (and usually you can't without using a bit of hardware known as a pre-amplifier), this is the commonly used technique to get your recording up to a listenable level.

"Frequent editing noise throughout...". Again not, in itself, a significant problem, although when the listener turns up the volume to compensate for the above-mentioned low level, it makes this one a bit more obvious. This is not background noise, but clicks that occur at the ends of sentences and paragraphs, sounding very much like whatever editing (cutting) process you're using is introducing the clicks into your recording. This is more an aesthetic issue than anything else; it gives the recording a mildly "rough" or "unfinished" quality. Your audio editing software should not be producing this kind of noise when you stop and start recording. If you're using a microphone with an "on/off" switch, turn it on and leave it on the entire time you're recording, and when you need to temporarily stop recording, use the "stop" or "pause" functions in your software. Without knowing what hardware, software and techniques you're using to record with, I can't offer any more detailed suggestions than these.

"Encoded at 96 kbps..." This is the main technical problem, and the primary reason for the "Medium" technical quality rating. Without this problem, the recording would rate at least borderline "High". Ironically, this problem is very easily fixed. I've been banging on about this issue at length elsewhere (specifically here and here, about the middle of the section), so I won't repeat myself too much here. The solution is to encode/export (or whatever your software calls the final stage of turning your recording into an Ogg Vorbis file) your audio file with the Ogg Vorbis quality setting at minimum. It'll still sound good, and more people will get to listen to your recording, because the file will be much smaller and easier (quicker) to download.

In general, a bit of experimentation with your audio editing software can help you to master the basic techniques required to produce a technically good-sounding recording. If you're using Audacity, I can provide more detailed information on how to achieve this, as I've used it extensively for my own work.

• Accuracy:

Okay, I stuffed up here. Your recording shows a pattern of general high accuracy, but with bits of text missing here and there. This kind of thing usually indicates that the version that was read is not the one the link_to_recorded_version points to, and in this case I made that assumption. Later I found this wasn't the case when I went looking through the article history for a version that matched your recording, using the text "a little-known children's adventure" (which is missing from the recording) as the search key. There is no historical version of the article that includes the text "Josh and S.A.M.", without "a little-known children's adventure" as it's read in your recording. This confirms your assertion that the link is correct, so I've modified the review accordingly.

I hope this helps. -- Macropode 10:59, 10 March 2007 (UTC)