Bruce, I use the iPhone, too, for angle measurement, and I think it's pretty accurate. I'm always careful that I'm laying a flat edge of the iPhone on the surface I want to measure. Between this and a Craftsman digital level, I've noticed seat and top tubes may have small warps.
I don't agree with calibrating it to the top tube, unless you know it's level. There are some bubble levels, such as the Stanley Fat Maxes, that are spec'd accurate to around 0.01 degree, as implausible as that sounds. If you have a carpenter's level and can shim it up so it reads level, that's a decent and repeatible reference, to which you can zero (it's really not calibration) the iPhone software.
For understanding the effect of the seat tube angle, it doesn't matter very much. But for the head tube, it does matter. If the reason you want to measure front angles is to assess front-end geometry, to calculate or measure trail, the calculation is pretty sensitive to angle, and the definition of trail is based on a level frame on a level surface. So the head tube angle needs to be defined relative to level.
I haven't found a decent home solution to leveling the bike, and I generally go with the top tube as a reference if the bike has matched front and rear wheels and tires, the original fork, and no known mods that would affect front or rear height. I haven't looked at normalizing the rear wheel position in the frame ends. For my own bikes I don't want to mess up the shifting, once it's tuned in. One example of a possibly significant mod would be a re-raked fork, to either increase trail (raises the front end) or decrease trail (lowers the front end). My Trek 610 has a re-raked (more trail) fork, and I calculated that the head tube is raised by about a quarter-degree.
But I'm not sure the small potential error in leveling matters a lot, practically. The digital levels read out in tenths of a degree. Until I have a measurement methodology that gives me repeatible numbers at this resolution, there's no point in making sure my reference point is good to a tenth of a degree. Random or methodological variations will swamp out this accuracy, at least with my level of expertise.
Ken Freeman Ann Arbor, MI USA
On Mon, Jun 1, 2009 at 9:52 AM, Bruce Kwasha <kwashabc@merakcorp.com> wrote:
>
> Just a note about a new use of an iPhone that I discovered this weekend.
>
> Measuring frame angles:
>
> Download the free Stanley level application.
> Calibrate the level to the top tube.
> Measure the angle of the seat tube.
>
> It is that simple. No need to level the bike, etc.
>
> Incredible technology.
>
> Bruce Kwasha
> Atlanta, Georgia
>
>
> __________ Information from ESET NOD32 Antivirus, version of virus
> signature
> database 4118 (20090601) __________
>
> The message was checked by ESET NOD32 Antivirus.
>
> http://www.eset.com