The Sony ICD-PX312 is powered by two micro (AAA) batteries, operating with two NiMH batteries is also möglich.Weiterhin allows the internal memory of 2 GB to expand via a memory card. For this purpose either the Sony's M2 memory cards or the more common micro-SDHC memory cards are a up to 32 GB in size. When running at the unit formatted, empty 16 GB Class 4 SDHC Memory Card, however, extended the time from power on to receptiveness to 8 seconds, as opposed to one second when operating without a memory card. Filled with some MP3 files, this time grew even to 12 seconds. For a voice recorder a bit too long to "just" turn on the unit and to record a memo.
The operation of the basic functions of the device is possible after a short blind: The keys have a clear pressure point and feel good. Many features, however, are only accessible via a menu, which can be called by pressing the DISP button. Here you can for example influence the bit rate of the recorded MP3 files or microphone sensitivity. The bit rate can be selected s in the four stages 8, 48, 128 and 192 kbit / s. In the lowest quality you can hear this very significant compression artifacts even in voice recordings. It will therefore be more likely to prefer the latter two bitrates. This is also not a major problem: Even in the highest quality level of 192 kbit / s can be using the internal memory to record over 22 hours. For a voice recorder more than enough.
The microphone sensitivity can be adjusted in three stages. The lowest sensitivity is doing well for voice notes. If you hold the device a few centimeters laterally offset from your mouth, you get clear, distinct voice recordings without the background noise coming to the fore. Even record your notes during vacuuming is not a problem. In the middle and high Emfindlichkeitsstufe you can hear significant ambient noise. These settings are well suited to conferences, meetings, lectures or interviews to be recorded. So in any case in which the microphone should capture multiple speakers, or can not be brought close to the speaker. Furthermore, can be activated via the menu, a high pass filter ("low-cut filter"), which filters out the low frequencies and thus makes language more easily understood. This works quite well, but the speaker takes the individual timbre of the voice, so it is not suitable if you want to view using the records for a podcast or the like. For non-voice recordings (voice, choir, sound effects) you should turn it off anyway.
Under the designation "BEFORE" you can activate the voice activation menu. In this case, the recording will automatically pause if you have one second said nothing. This works very well in low and medium microphone sensitivity level. The "awakening" go out of the pause mode no syllables or even words lost, as was common especially in the mechanical dictation. In the recording you do not even notice most of which has been taken a break here. In the high sensitivity level, however, the background noise is often too loud, so that the device not at all only switches to pause.
The recorded MP3 files are basically in stereo format, but only "Mono" filled by the built-in microphone. This means both channels, left and right, do not differ in content from each other. The changes, however, when you connect a stereo microphone or a stereo source to the 3.5mm jack input: After plugging the jack, the device asks if the input as a microphone input (with supply voltage for electret condenser was wired microphones) or a line-in input. After records are thoroughly recorded in stereo, even if the continue of "MP3 (MONO)" menu is talk. In addition to the three-stage Mikrofonsmpfindlichkeit above not adjust the recording level is possible. But you can listen to the recording through the headphone output and evaluate (monitor function). A tentative performed recording with stereo microphone Sony ECM cs3 - microphone went smoothly and with remarkable quality. Just as easily, a recording of a radio via a connecting cable was possible. A "pumping" of the volume, as it may occur in an automatic control of input sensitivity, did not notice.
Playback via the built-in speaker is somewhat quiet, but clearly understood. For playback can be on the right side of a slide switch, a "Noise Cut" switch function, which is to filter out background noise from the recording. However, this function uses the useful signal - ie the language - clearly: It sounds sometimes as if you were talking underwater. You can hear the difference good if you switch the function on and off during playback. Fortunately, this switch only affects playback and does not affect the recording.
Overall a really compelling product. Only the waiting time when switching with memory card inserted clouds the overall impression a bit. Since the internal memory of 2 GB (or even 22 hours) for my purposes but is completely sufficient, I can on a memory card without. Thus, this defect is not significant. This way, I have vordiktiert recession on the voice recorder and transferred later in writing and formulated. Is not it? :-)