Roland Edirol R-1
Anthony Jukes and David Nathan
ELAR/ELAP, SOAS, London
We have been waiting for a replacement for our trusty minidisc recorders, and the new breed of solid state (flashcard) digital recorders has looked promising. Initial impressions of the Marantz PMD670 were mixed (quality sound, but rather big, too expensive, and poor battery life), and while the newer PMD660 is smaller, at around GBP400 it is still rather expensive. Roland's Edirol R-1 is cheaper and smaller than the PMD660, and has been eagerly awaited by many linguists. Here we evaluate the R 1's suitability for language documentation in real field situations, with particular attention to its usability and its performance using various microphones and recording formats.
The R-1 seems generally flimsy. The case does not appear well-assembled and the door to the CF slot operates clumsily and is likely to break off. The large "VALUE" knob can be knocked off easily. There is no carry strap bracket. The unit's buttons are simple enough to use but do not have a satisfying feel. The mix of controls is odd: it has button-operated menus, analogue-deck-style control buttons, mechanical slide switches, and potentiometer wheels. Monitoring of input levels is limited by the lack of marks on the input level control and the basic LCD display. Microphone input is only via stereo miniplug, so microphones with XLR connectors will require an adaptor. The R-1 has no rubber feet and slides around uncontrollably on smooth surfaces. Given its less than rock-solid construction, we recommend sticking some feet on it, or, better, keeping handy a sheet of grippy rubber "anti-slip" mesh that is sold in many hardware shops.
Ease of use
The R-1 is easy and convenient to use if simple recording with the inbuilt microphone is all that is required. Some of the controls are annoying; for example display mode settings are not retained and have to be toggled after each recording.
Care needs to be taken - for example, turning the R-1 off before stopping recording (easy to do!) loses the entire recording, or even the memory card format. This in turn can cause the R-1 some confusion; the manual warns that the unit could even be permanently damaged. The manual itself is very clearly written, and is compulsory reading since it describes such situations where data loss might occur (however, we found inaccurate information, e.g. the memory card does not have to be formatted in the Edirol).
Digital storage and connectivity
The R-1 uses widely available Compact Flash memory cards. It handles cards up to 2GB (larger cards, including microdrives, will not provide any increased capacity). Having experienced card incompatibility problems with the Marantz PMD 670 (see LAN 4), we tried the R-1 with a number of cards. All worked perfectly, even though one of them was a Buffalo RCF-X 256MB, which looks suspiciously like one of the cards listed by Roland website as incompatible.
The R-1 has a USB output and quickly transfers files to a computer using USB2. Unfortunately the unit must be running on mains power to use the USB port, which could be a disadvantage in the field (e.g. if you need to move recordings off a full card to continue recording). The R-1 has no optical input. Provision of optical input would have made the R-1 useful for transferring sounds from other devices in the field.
Evaluation of test recordings
We made test recordings of the R-1 with various microphones and settings, and also made a brief comparison with a Marantz PMD660. Each of the microphones - the R-1's inbuilt stereo microphone, a Sony ECM-MS 957 stereo microphone, and an AT803b clip-on lavalier - was used to record speech at several settings: 24-bit/44.1kHz WAV (the R-1's highest) and MP3 at 320, 256, 192, 128 and 64 kbps. Recordings were evaluated by listening via the R-1 and Mac and IBM notebooks using Grado SR80 and Victor HP-DX1 headphones.
The R-1 is capable of making good quality recordings. Recordings at high bit rates MP3s (320 and 256 kbps) were perceptually indistinguishable from the 24-bit/44.1kHz WAV recordings. Predictably, lower bit rates yielded progressively less quality. At 192 kbps, artefacts became noticeable (as a fluttering of the higher frequencies in the background noise). By 64kbps, the sound was telephone quality. If conserving file space is the highest priority, there may be no real harm in using high bit rate MP3 settings (we can already hear some people throwing up their hands in horror!). Much more important factors influencing the quality of recordings were the choice of microphone, its placement and handling, and attention to the recording environment to reduce unwanted noise.
We were disappointed with voice recordings made using the inbuilt microphone. They were rather noisy, with an unacceptable amount of hiss, significant handling noise, and oversensitivity to ambient sound. The R-1's microphone characteristics may reflect its musical origins and better suit musicians than linguists. The Edirol website (http://www.edirol.com/products/info/R-1.html) has some fine recordings of musical instruments recorded with its inbuilt microphone. However, field recording of spoken voice has different requirements.
Using the R-1 with a Sony ECM-MS 957 microphone resulted in good, natural sounding recordings, with less noise. This combination would prove quite suitable for recording a single speaker, or a number of speakers in conversation. Best results for recording a single speaker were obtained using an AT lavalier microphone. It allows closer placement to the speaker's mouth which greatly attenuates background noise sources such as dogs, poultry, children, and other noisy creatures found in typical village settings. A great combination for many field situations would consist of using the R-1 with two microphones recorded in separate channels: a lavalier for the primary speaker and an omnidirectional microphone for audience responses. However, an adaptor would be required to connect the two mono microphones to the R-1's single stereo miniplug, and the R-1 does not allow the channel recording levels to be set independently.
The R-1 has a switch for selecting dynamic or condenser microphones, but the "dynamic" setting was required for best results using the Sony ECM-MS957 Electret condenser unit. This seems to be something to do with supply of "plug-in power" for MD (or so-called "digital") type microphones. Perhaps the condenser setting applies to those types only, and all others, including the R-1's internal, use the dynamic setting. Although we queried Edirol UK about this, we were not correctly advised, and we concluded that the switch changes microphone powering as well as sensitivity and frequency response, although not in the expected ways.
We also compared recordings made using the Edirol and a Marantz PMD660, both using a Sony ECM-MS 957 microphone. The Marantz showed a higher sensitivity (or better match to the microphone); other than this, there was almost no difference in the clarity or noisiness of the two recordings.
Finally, we found the R-1 quite susceptible to electrical interference when using its internal microphone. If it was within 3 metres of a laptop computer, or even within 5 metres of a normal refrigerator (in another room!), a buzz was recorded. This effect was much less noticeable when using the external microphones.
A pair of fresh alkaline batteries lasted for only 45 minutes of recording, although others have reported up to two hours. We were able to record for over two hours using fully charged NiCads. Some have reported that much better results are obtained with lithium batteries, and there are reports of over four hours of recording using rechargeable NiMH batteries. The R-1 gives only a few minutes warning when the battery level is low. It then saves the current recording and shuts down.
he Edirol R-1 offers the ability to make acceptable digital recordings by simply placing the unit on the table and pressing record. Using external microphones, high quality recordings can be made. However, many of the R-1's strengths lie in its musical heritage; for linguists, its apparent lack of robustness and limited microphone connectivity make it a second choice to the (admittedly more expensive) Marantz PMD660. Those who predicted that the Edirol's arrival would provide the perfect solution to linguistic field recording may be disappointed.
This review first appeared in Language Archives Newsletter Vol. 1, No. 6 (October 2005)