Archive for August, 2011

Unlock your MacBook Air Recovery USB drive

Monday, August 22nd, 2011

So you got an “older” (pre 2011) MacBook Air with read-only USB recovery thumb-drive (stick)? Well, turns out it is only read-only by some firmware lock bit. Just run:

sm32Xtest.exe

On your favorite Windows (virtual-) machine to toggle the read-only bit. You now can use it for your personal data, or writing the Lion installer disk image to it to gain a Lion Recovery USB key, just as you like.

Vintage car get-together HefeHof Hameln 2011

Monday, August 22nd, 2011

Yesterday, my father and I have been to the vintage car gathering (Oldtimer-Treffen) at the HefeHof, Hameln (Germany). Starting with nice sunshine I got some great shoots to share:


… more in the whole gallery

The quest for a new headset - Denon AH-C560R

Sunday, August 14th, 2011

Since the 1st generation iPhone I enjoyed using some V-moda Vibe in-ear headset. At the time the V-moda Vibe was one of the first iPhone compatible headsets with microphone and good sound quality.

Unfortunately, as usual, the cable started to break just over the TRS plug’s bend protection. Attempts to save the headset by soldering a new 4 pin TRS plug proved a little difficult: soldering the tiny 4 pin ⅛ inch TRS connector is possible, but the fibre enforced wires make it pretty difficult. One has to carefully separate the fibre out of the copper, and even then getting solder onto the paint isolated copper wires is a tedious task. And all too often have signal crosstalk problems in the end, due to the now heated and dissolved isolating wire coat, … especially interfering the (i)Phone’s microphone ring detection, often leaving the microphone and remote button controls unfunctional. Often it even works at first, until the cables are cramped into the connector when the plug is screwed together.

In the end I decided to better just get new set and started an endless Google search and local shop listening test. I found most offered headsets either plain ugly (sorry folks!), or excessively overcharged, especially when lacking in sound quality. Some models from the German company Sennheiser would have been a nice choice - if just their cable control with microphone plastic enclosure would be more visually appealing.

Especially lacking where the excessively overpriced Shure SE535 and friends. For over 300 bucks the advertised musical professional headphones can not please my musical amateur sense of hearing: For one in a local shop listening test they failed to deliver any bass (low frequencies) to speak of at all, … and thus an extremly unbalanced frequency range overall.

In the end of the lengthy selection process (why can manufactures not just design, build, and sell the one perfect headset?) I ended up with a Denon AH-C560R. They do not only look very good, with their minimal, no-frills design (that could be out of Cupertino), for just about hundred bucks they deliver an well balanced frequency range, with definitely more bass compared to the expensive Shure series.

Of course the Denon also sounds way more balanced than the over 3 year old V-mode Vibe - maybe due to Denon having more audio design experience being a well known audio brand for many centuries. The V-mode had more bass, though - but overall the Denon sound more balanced, with clearer mids and highs. (Other reviewer would now go all lengths over starting to hear all sort of details, percussions, accents not being possible to hear with another headphone before, …)

Oh, by-the-way, I also found the Nox Audio Scout headset’s design VERY appealing, I even delayed the purchase some months to wait for their release, … However reviews, such as the one from Engadget where not too keen about their audio quality, and given that they are (still not) yet sold in Germany the shipping costs from the US where just too heigh to rectify trying them out myself, … Maybe I grab them for cheap when I am in the US the next time.

It will also be interesting to get a ear on the Bowers & Wilkins C5 - though I am not too sure if I will find such a huge & heavy tube comfortable.

Apple Macs and thermal throttling

Saturday, August 13th, 2011

Since Apple migrated to the energy-hungry Intel x86 CPUs some years ago Macs got quite some overheating, thermal throttling issues. For example you will find many compact aluminum Macs to become so hot, that you would barely want to touch, or use on your lap, with the miniature fans spinning at full pace.

The bad is that Mac users typical will not even notice that the (Intel) silicon will even start to emergency throttle down. They will only notice the machine becomes less “snappy”. Linux on the other hand will happily log the MCE to your syslog instead:

CPU0: Temperature above threshold, cpu clock throttled (total events = 794)
CPU0: Core temperature above threshold, cpu clock throttled (total events = 1)

If my aging, non-unibody, MacBook Pro would not throttle that much it would still be a perfectly reasonably powerful machine to use - even as professional. But as it became so annoying I decided to write a little benchmark too that logs the performance over time. When freshly booted this MacBookPro3,1 will start to throttle down after just a minute! And you will notice another spike of throttling where I started another benchmark instance on the second core, …

However, I think when I purchased the machine back in 2007 it was not that bad, I have the feeling it became worse when Apple exchanged the logic board due to the faulty, and failing NVidia chip, … maybe something (fan, heat-sink or -pipe) was not re-aligned perfectly, or too much thermal paste applied, …

I experienced this throttling with all MacBooks and iMacs in the last years. I just tested the latest (mid-2011) MacBook Air at a local shop which thanks to the Sandy Bridge Core CPU does finally not show this symptom that badly. They intentionally choose the lowest performance Ultra-Low-Voltage CPU for a reason, … Though thanks to the advent of Intel’s Turbo Boost it shows fluctuating performance levels, too. One could argue it is an publicly advertised thermal throttling feature :-)

Whether due classic thermal throttling, or TDP (Thermal Design Power) based Turbo Boost: for researchers this means any such CPU is out of question for scientific performance measurements. A fixed-frequency CPU without funky throttling and boosting as well as a BIOS’ SMM (System Management Mode) not interfering (too much) is a must. Unfortunately, the later is hard to prove and avoid, though.