Archive for category Reviews

Letter to Philips and Cree- LED Lights are Not Ready for Prime Time

Thank you kindly for the refund check of $8.50. Because I purchased all of my LED lights within the past 3 months, I simply returned all of my lights to Home Depot for a full refund instead. Therefore, I am returning the check.

In case you are interested, I will never again purchase another LED light from Philips or any other company. Philips’ advertising, marketing, and packaging is false and deceptive; and I’ve learned my lesson. Initially, I was persuaded by LED lights’ purported longevity and low energy consumption. When I did the math, my wife and I determined that they would pay for themselves in a matter of 2-2.5 years. So we decided to go all-out, and we spent more than $600 on LED lights from Philips and Cree.

I am upset with both Philips and Cree, because neither the advertising nor the packaging indicates that I needed to pay an electrician several hundred dollars more to replace dimmer switches with some new special kind of dimmer switch. And the cryptic “suitable for open fixtures” fine print is misleading- I didn’t realize it meant, “This bulb will self-destruct in 3 months if you put it in a closed fixture.” The bulbs that were supposed to last until my late 60′s began to fail after just 2-3 months. I’ve reverted to incandescent bulbs; an inefficient but proven technology.

I was excited to drastically decrease my carbon footprint and save money long term. But LEDs are not ready for prime time. I’m just glad I was able to return them to Home Depot before it was too late.

No Comments

Bitcasa Review: Version 1.1.0.0

I have been looking for a cloud storage, backup, and syncing solution for some time.

This post reviews Bitcasa 1.1.0.0; hopefully future versions will fix some of these bugs.

This post reviews Bitcasa 1.1.0.0; hopefully future versions will fix some of these bugs.

I heard about Bitcasa’s much-hyped “infinite storage,” and decided to give it a chance. Here’s the high Level Take-away:

  • I do not yet regret paying Bitcasa $100: The value proposition of paying $100 per year to meet my ever increasing storage needs is compelling.
  • I really like that Bitcasa encrypts my data; for real.
  • I plan to use Bitcasa for video and image storage and backup. I’ll use Dropbox for syncing, sharing, and collaboration.
  • Bitcasa 1.1.0.0 is a step beyond beta. It lacks the intuitive feature set and interface of Dropbox.
  • There is no “off” switch.
  • Bitcasa is not Enterprise-ready.
  • “Mirroring” is buggy.
  • Bitcasa 1.1.0.0 is a memory hog, and will cause significant system degradation while uploading.
  • Once it’s on Bitcasa’s (Amazon) cloud, don’t ever count on being able to delete it.
  • I’m going to stick with Bitcasa for a while. I think it’s got a lot of potential.

Paying for the same hard drive every year doesn’t make sense

First, let me tell you why I find Bitcasa’s model so appealing. Once I purchase a hard drive, I shouldn’t have to buy it again next year. That’s why I’m an avid user of the free version of Dropbox. I’ve got somewhere around 11.5GB of free space, which I use to back up sync and share. But the idea of paying for the same storage again and again just doesn’t appeal to me. Yes, I know that I’m also paying for near universal access to my files, but that value proposition just doesn’t hold water for me.

I’ve found that the 99% of the files I need to access and share on a regular basis are primarily office documents—Word, Excel, PowerPoint, etc. These files tend to be relatively small, and I can easily fit them within my 11.5GB Dropbox allotment. But the remaining 1% of my files take up terabytes of hard drive space. I want a safe backup for family photos and videos, even though I’m not going to access them on a daily basis. Even if I theoretically purchased “unlimited” space from Dropbox, I would still be limitied to the size of my local hard drive, since Dropbox only syncs local drives and does not provide additional cloud storage.

Consequently, I’m not willing to pay a recurring fee for static storage every year. But I am willing to pay for ever-increasing storage needs every year, especially if I can free up local hard drive space for other needs. So the idea of paying $100 per year for unlimited storage makes sense to me.

Although I understand how Bitcasa offers “infinite” storage, I don’t quite understand how Bitcasa’s business model intends to deal with enterprise clients who will want to store Petabytes per month, assuming they have the bandwidth.

They’ve got the right idea about privacy

From my understanding, Bitcasa encrypts your files using a hash of the file itself as a key (please correct me if I’m way off base). This means that Bitcasa doesn’t actually know the contents of your files, but can still know if your file matches a file that has already been uploaded. This way, instead of storing 5,000 copies of Bieber’s latest hit, they can save space by storing just one and linking it to 5,000 accounts. In contrast to Dropbox, Bitcasa doesn’t know that it’s a Bieber MP3; they just know it’s the same file. I appreciate that.

This means that I’m much more likely to store sensitive documents on Bitcasa. Although I still have to worry about securing my Bitcasa username and password.

From a privacy perspective, Bitcasa has the right idea, though I think their privacy representations are a little over-the-top:

“To be clear, aside from the rare exceptions we identify in our Privacy Policy, no matter how the Services change, we won’t share your content with others, including law enforcement, for any purpose unless you direct us to.”

(Bitcasa Legal Policy, emphasis added). Presumably Bitcasa won’t share your content with law enforcement because it can’t share it with law enforcement.

That’s actually a slight exaggeration. For example, let’s say that two parties are in a dispute. Party A says that Party B stole Document C and stored it on Bitcasa’s cloud. A court could certainly order Party A to give Document C to Bitcasa, which Bitcasa would upload, and its systems would instantly know how many other copies of Document C exist on the server, who owns it, when it was uploaded, etc. That information would then be open to discovery.

While I don’t really have a problem with that sort of particularized discovery, I could easily imagine Bitcasa having to build tools to allow parties like the RIAA to upload MP3s to see who has a copy, if ordered by a court. I think Bitcasa is on the right track with privacy and security; I just think that some of their promises are a little over-the-top.

Update below: Dom points out that blind encryption may be a fabrication, since Bitcasa can create low-res previews of files; this can’t be done without access to the original file.

“Mirroring” is not like Dropbox

“Mirroring” is when you designate a folder to be synced on Bitcasa. I thought it would work like Dropbox, but I was wrong.

All Folders in My Documents are mirrored... and I can't unmirror them

All Folders in My Documents are mirrored... and I can't unmirror them

When I installed Bitcasa, the setup asked whether I wanted to mirror my hard drive or key folders, like My Documents. I checked “yes,” and finished the installation. Very quickly I had second thoughts and decided to cancel the mirroring.

I unmirrored My Documents, then deleted the “Mirrored Documents” file, but somehow Bitcasa re-mirrored it and started uploading without my knowledge. I unmirrored again. Ironically, even though Bitcasa re-mirrored My Documents and uploaded it, the folder didn’t appear in my Infinite Drive.

Bitcasa helpfully stores daily versions of files; but won't let you delete them.

Bitcasa helpfully stores daily versions of files; but won't let you delete them.

Try as I might, I could not un-mirror the folders. On occasion Bitcasa would indicate that a folder was un-mirrored, only to re-mirror it moments later. I panicked, and deleted the “Mirrored” folder on my Infinite Drive, and expected the uploading to stop, but uploading continued at 5Mbps (my max FiOS upload speed at the time), for 10 more hours. I could not stop it. There was no “off” switch.

Even though I had unmirrored my files, the icons continue to be marked “Mirrored,” but if they were mirroring, I could not see where they were mirroring to, because there was no “Mirrored Files” folder on my Infinite drive. And in the online interface, the “Delete” option was grayed out. For several hours I felt like I had lost control of my computer, and that Bitcasa was sucking it to the cloud, and there was nothing I could do about it. For the record, I attribute this to an immature product UI and feature set, rather than any malicious intent.

If you want to unmirror a subfolder, you're out of luck.

If you want to unmirror a subfolder, you're out of luck.

I hopped on Twitter and engaged with @BitcasaSupport. Here is a synopsis of that conversation:

Me: How do I stop Bitcasa from mirroring my entire drive? Where’s the off switch? Should I just uninstall and reinstall?
BitcasaSupport: Sorry for the confusion – you can only unmirror via the desktop app, which you can download here: http://bit.ly/YGcr3C. When you right click on the folders mirrored, select Stop Mirroring to Bitcasa from the Bitcasa contextual menu.
Me: That’s what I tried- “stop mirroring” is grayed out everywhere. I chose “Mirror whole computer” on install. But I don’t know all of the local locations that Bitcasa is mirroring. It’s quite frustrating.
I mirrored my computer, then unmirrored it, but @Bitcasa continues to suck 5Mbps from my computer for 10 hrs. Very concerned. Why does @Bitcasa continue to suck up 5Mbps of my files without my consent? Can I stop it?
I had to kill the process to stop it from uploading. I have no idea what its been uploading for the past several hours. Is there any record of what has been uploaded, and how long @Bitcasa keeps it?
I am able to see old versions of my files, but the Delete button is grayed out. That is a problem because there are uploaded files that I never intended to upload. Is there a way to get rid of those files?

Support then opened a ticket for me, and addressed some of my concerns. In short:

  • I had to completely uninstall Bitcasa, manually deleting folders. Fortunately I did not have to edit the registry.
  • There is no “off” switch. Bitcasa chooses when and whether to upload.
  • You cannot unmirror sub-folders. You must unmirror the top folder… if you can remember which one is the “top folder.”
  • I still have no idea what Bitcasa was uploading for 10 hours after I had told it to stop mirroring everything. And I won’t ever have a way of knowing for sure.
  • There is currently no way to delete old versions of documents, even if you didn’t want them online in the first place.

Bitcasa is a Resource Hog

Bitcasa will take over your machine while uploading.

Bitcasa will take over your machine while uploading.

I think this image says it all. Bitcasa significantly degrades performance when it’s uploading. To give you an idea of how much it affects performance, I am writing this article in Notepad right now, because my browsers and word processors are too painfully slow.

When I asked support about this issue, they replied that a recent instability fix created the memory problems. “We’re currently working to improving the memory usage of the application without hindering upload stability, and anticipate that improvements will be made very soon.”

You need a fat pipe to upload all of your junk

I mentioned earlier that I have a 15Mbps down/ 5Mbps upload internet connection. I learned that it takes a long time to upload several terabytes of data at 5Mbps. I upgraded to 50Mbps down/ 25Mbps up so that I won’t have to wait a month to upload (nevermind that my pipe is testing at 25 down/ 21 up–curse you, Verizon!). I thought this was worth mentioning because the cost of an Infinite drive may well exceed $100 per year, if you include the cost of additional bandwidth.

Even with a 21Mbps available, Bitcasa rarely uses that much. Occasionally uploads have spiked at 21Mbps, but as I type, my upload speeds are averaging 3Mbps. The last day or two, they’ve averaged 6Mbps. Not surprisingly, Bitcasa support is quick to point out the many factors that affect upload speed… none of which are them. While I have no problem blaming Verizon for many of my problems, I have a hard time believing that my ISP is solely responsible for a 85-90% degradation in my upload speeds. Sorry, Bitcasa, I still think you’re the choke point.

Still Hopeful

Notwithstanding the mild drama, I’m still hopeful that Bitcasa will pull through. As I mentioned earlier, I think they’ve got the value proposition right, and I think they’re on the right track for privacy and security. They can have my $100.

Update: More to Worry About

Since I originally posted, helpful commenters Dom and Craig (below) make a couple of sobering points that significantly undermine my confidence in Bitcasa:

  • Blind encryption may be a lie: Bitcasa publishes previews of files and videos. This cannot be done without access to the original file.
  • Security by URL Obscurity: There is no way to password protect, time limit or even know when someone downloaded a file links.
  • No list of shared URLS: You’d better remember which files you shared via link, because Bitcasa sure ain’t going to tell you!
  • Shared URLs are indexed: Even though the robots.txt requests no indexing, many search engines ignore the request, or liberally interpret the request. Bottom line- links may be searchable, so beware before you share.
  • Rookie Security Errors: Full error reporting is currently turned on by default, meaning that errors expose django configs, mysql dbs and password, and Apache configs, usernames etc. They fixed this error within 10 minutes of notification via Twitter. But their MySQL db username and password may still be in the wild.

Very unimpressed. Bitcasa, get your act together.

4 Comments

Is Your Tascam US-144 mkII Noisy? Just Sit on it.

The TASCAM US-144 mkII gets noisy when the temperature drops.

The TASCAM US-144 mkII's Phantom Power gets very noisy when the temperature drops below 65

I recently purchased two sets of podcasting gear to record podcasts with someone in another state. The gear included two TASCAM US-144 mkII interfaces, two Audio-Technica AT 2020 condenser microphones, two mic cables, two stands, two pop filters, two sets of headphones, etc. As I recorded I noticed an inconsistent whine in the audio. Sometimes the wine was little more than a vinyl-record-like scratch, other times it was a scream, and still others it disappeared altogether.

To track down the source, I started by changing out the AC power for battery power, switching USB cords, switching mic cords, switching microphones, switching interfaces, turning on and off the wireless network, unplugging my wireless router, changing rooms, and even moving to a location several miles away. None of these things had any consistent effect on the whine. After a lot of trial and error, I have narrowed the problem down to two three variables: Temperature, Phantom Power and the MIC/LINE-GUITAR select switch.

The whine appears with the phantom power on, while the interface is cold (ie, less than about 65 degrees Fahrenheit). It gets worse when the MIC/LINE-GUITAR select switch is set to “Guitar.” Setting the MIC/LINE-GUITAR select switch to “Guitar” makes it act as an unbalanced input jack, which probably explains the noise. But turning on the phantom power while the interface is cold produces a lot of noise.

The solution: Sit on the interface to warm it up. My home “studio” is in a very cold room with two exterior walls, and it’s the middle of winter. So in order to warm up the interface—no joke—I actually put it under my thigh for a good 10-15 minutes. I didn’t read that helpful work-around in the manual. I thought about using an electric blanket, but I was afraid that might cause some induction damage.

The Test

I conducted a test to demonstrate the whine, which I have included here. For the test I had the following setup: I plugged an Audio-Technica AT 2020 condenser mic into the MIC IN L XLR balanced jack. I also plugged a crappy old dynamic mic into the LINE IN R/GUITAR IN jack TRS 1/4″ jack. For the “Cold” test, I left the interface in a box in my car for 30 minutes, where the outside temperature is around 25° Fahrenheit. For the “Warm” test, I basically sat on the interface for about 15 minutes until the interface housing was noticeably warmer than room temperature.

I then recorded a systematic test of the phantom power, left and right input levels, and the MIC/LINE-GUITAR select switch in 10-second intervals. I included the results in a table below, with each numbered setting corresponding to a period of time on the non-normalized .mp3 file. You can skip around to compare the different settings if you’d like. Please ignore the ambient noise of the HVAC system, as well as the lousy line quality for my crappy dynamic mic.

INTERFACE TEMPERATURE: COLD (~30°- ~65° Fahrenheit)
Setting Time on Tape Phantom Power INPUT L Levels
(AT 2020)
INPUT R Levels
(Crappy Dynamic)
LINE/MIC- GUITAR
Select Switch
Whine
1 0:00-0:10 OFF Line (Low) Line (Low) Line/Mic None
2 0:10-0:20 Mic (High) Line/Mic None
3 0:20-0:30 Line Guitar None
4 0:30-0:40 Mic Guitar None
5 0:40-0:50 Mic (High) Line Line/Mic Scratch
6 0:50-1:00 Mic Line/Mic Scratch
7 1:00-1:10 Line Guitar Scratch
8 1:10-1:20 Mic Guitar Scratch
9 1:20-1:30 ON Line Line Line/Mic Whine-Low
10 1:30-1:40 Mic Line/Mic Whine-Med
11 1:40-1:50 Line Guitar Whine-Med
12 1:50-2:00 Mic Guitar Whine-Scream
13 2:00-2:10 Mic Line Line/Mic Whine-Loud
14 2:10-2:20 Mic Line/Mic Whine-Loud
15 2:20-2:30 Line Guitar Whine-Loud
16 2:30-2:40 Mic Guitar Whine-Scream

INTERFACE TEMPERATURE: WARM (~75°+ Fahrenheit)
Setting Time on Tape Phantom Power INPUT L Levels
(AT 2020)
INPUT R Levels
(Crappy Dynamic)
LINE/MIC- GUITAR
Select Switch
Whine
1 2:40-2:50 OFF Line (Low) Line (Low) Line/Mic None
2 2:50-3:00 Mic (High) Line/Mic None
3 3:00-3:10 Line Guitar None
4 3:10-3:20 Mic Guitar Scratch
5 3:20-3:30 Mic (High) Line Line/Mic Scratch
6 3:30-3:40 Mic Line/Mic Scratch
7 3:40-3:50 Line Guitar Scratch
8 3:50-4:00 Mic Guitar Scratch
9 4:00-4:10 ON Line Line Line/Mic None
10 4:10-4:20 Mic Line/Mic None
11 4:20-4:30 Line Guitar None
12 4:30-4:40 Mic Guitar Whine-Loud
13 4:40-4:50 Mic Line Line/Mic None
14 4:50-5:00 Mic Line/Mic None
15 5:00-5:10 Line Guitar None
16 5:10-5:20 Mic Guitar Whine-Loud

My home studio is in the basement near an outside wall, so it’s usually around 65°. Every morning the whine reappears until I physically warm the unit to around 75°+.

At lower temps, the phantom power whines and bleeds over into the 1/4″ inputs, which surprises me because most electronics are happier when they’re cold. I’d chalk it up to a defective unit, except that I purchased two 144 mkII’s, and both units display the same behavior. Regardless, I’m not looking forward to the hassle of returning or exchanging the interface. It’s going to put me back several weeks.

I wonder if anyone else has experienced these same problems. The helpful guys at Sweetwater didn’t seem to have bumped into the problem before.

[Update Jan 14, 2010]

I have decided to return the mkII’s to Sweetwater in favor of another brand, perhaps an M-Audio. I haven’t decided. At first I was content to swap them out for non-defective mkIIs, but apparently TASCAM has temporarily stopped shipping the US-144 mkII. More precisely, they are taking orders without providing a firm ETA. This is apparently quite unusual, and in the estimation of the guy I talked to it likely indicates that they are doing some re-tooling.

I decided that I’m probably better off not being the guinea pig for the “fixed” version (if, in fact they are re-tooling). And even if they’re not re-tooling, I don’t want to wait indefinitely for TASCAM to fill the order.

I am so glad that I purchased from Sweetwater instead of Guitar Center. Sweetwater has much better support. Let me correct that: Sweetwater offers any type of support.

[Update Jan 25, 2010]

I decided to go with a Lexicon Omega instead. So far (in some preliminary recordings) I haven’t had any noise problems, thought the levels are significantly lower than the Tascam 144 mkII. I’ll just have to do more post-normalization. I hope the noise levels stay tolerable.

, ,

30 Comments