Enter the ONAG. Because digital detectors in today's autoguiders are sensitive to near-infrared (NIR) light beyond the visual spectrum, the ONAG works by sending a telescope's visible light to the imaging camera and the NIR to the guiding system.
As clever as the concept behind the ONAG is, the devil is in the details, and that's where the ONAG really shines.
The ONAG has a very rigid connection, and it's especially noteworthy because the device has an adjustable X-Y mounting for the guide camera, which helps in the search for suitable guide stars.
Using the ONAG involves getting your imaging camera and autoguider to reach focus simultaneously, and that means there are a lot of physical parameters to consider for your particular equipment.
The minimum back-focus distance between the female T threads on the ONAG's front mounting plate and the imaging port is 66 mm (2.6 inches), while the nominal back focus to the guiding port is 90 mm, with plus or minus 4 1/2 mm available for focusing.
But there's another aspect of the ONAG that further improves the efficiency of finding a guide star.
The guide stars available to the ONAG are closer to the telescope's optical axis and are thus of much better quality for guiding.
Although you can use virtually any camera that works as an autoguider with the ONAG, it has to be one without a built-in infrared-blocking filter.
Although the westernmost star in the Coathanger's bar is visually the faintest, my autoguider on the ONAG saw it as nearly twice the brightness of any of the other stars in the bar.
The only other aspect of the ONAG that I had to adapt to was the mirror-reversed "raw" views from my imaging camera.
From the get-go I had excellent results with the ONAG. With autoguider exposures of 5 seconds or less, I could often find a suitable guide star without having to move the autoguider on the X-Y platform after I had my target framed in the field of the ST-8300 imaging camera.
Although I generally use multiple stars as registration points when I stack my deep-sky frames, I could often dispense with this step when stacking images made with the ONAG. The cold mirror did not introduce any obvious color shift to my images.