Last time I wrote about the problems I was seeing with generic CFLs and color reproduction. On color rendition alone, I was very tempted to just give up on the alternatives for the foreseeable future and stick with Tungsten lamps. That was until running my 5D mark III video diffraction test and now having spent some time in front of 100, 250, and 500W lamps being a guinea pig for myself, I’ve come to the conclusion that hot, hot lights, are an untenable solution due to heat.
Since hot lights are out, and hardware store florescent are nowhere near good enough at rendering colors to be close to being acceptable, this time we’re looking at two photographic CFLs and comparing them to a standard 240W ECT Tungsten (3200K) and my reference 100W GE Reveal (2800K).
I don’t think anything I write here will make any difference, I’m mostly writing it for myself so I can hopefully air my frustration and move on.
I’ve been a paying customer for as long as I’ve been a photographer and creative. I started with Lightroom some 5 years ago, progressed to Photoshop, then the CS Production Premium suite. The simple reality is that as someone who wants people to respect my rights on my works, I want to respect the rights of those who make the tools I use. So I do, and I hand over my hard earned money when I can to keep my tools as current as I can.
I make this point because while I will pay for something or do without, I’m not always in a position where I can pay on someone else’s schedule as opposed to mine. The Creative Suite perpetual license model let me do this, if I didn’t have the money now, I could hold off and keep using my tools until I did have the money to upgrade. If that meant skipping a version, that meant I keep using out of date tools without the latest and greatest features but I could at least keep working.
On the other hand, the subscription model for Creative Cloud removes all of that flexibility. Ultimately, this detracts from the value so much as to make the Creative Cloud unattractive to me in it’s entity as the product currently stands. For me the risks to me are twofold.
First, there’s the risk associated with future pricing. The introductory pricing of $240 for the first year, is very attractive, however after the first year it increases to $600. After that, who knows. Adobe isn’t making any guarantees that the price won’t be significantly hiked in subsequent years. Why would they that would be a poor business decision. What we do know, at least from recent history, is that the value of the dollar will likely continue to decline with inflation and the price of goods and services will go up. The simple reality is the costs to me aren’t fixed, or easily calculable. I can’t count on CC being $600 a year in 2 or 3 years.
Compounding the matter is the risk associated with my continued use being tied directly to my continued payment. The simple matter is a bad quarter, a serious injury, or any number of other issues can easily stifle my cash flow in such a way that I become unable to continue the subscription at any given point in time. The direct result of course is total ruination, as the tools you need suddenly are no longer functional. I’m not a big business, while $600 a year isn’t in an unfathomable amount of money, at the same time it’s nothing for me to sneeze at either.
In a nutshell, that loss of control over my ability to create unless I’m still paying is why I won’t be moving to Creative Cloud anytime soon. It’s immensely frustrating for me too, as I would be more than willing to pay a reasonable premium for a perpetual license of the CC software, there’s certainly features there worth paying for. I’d even be more willing to consider a subscription service if there was a way to leave the service without also losing my tools. In fact, the subscription services I’ve used in the past have used this model. When you unsubscribe, your software still works and you still can use it, but you receive no support and no updates.
In the end, the whimper, is that I can only hope that there’s enough backlash against the subscription only model that Adobe brings back a perpetual license. I’m certainly not big enough or influential enough to effect any change. That and the inevitable fact that eventually CS6 wont be sufficient to remain competitive or functional on some future OS and I’ll have to upgrade into perpetually paying.
I’ve been working on a DiY teleprompter for the past several weeks in an attempt to put together something that was functionally identical to a commercial teleprompter product, but wouldn’t cost significant money out of pocket. Especially since the design of a teleprompter is easy enough, you need something to hold the display—in my case an iPad2—and a partially silvered mirror at 45° to the display to reflect it in front of the camera.
While the holder part of things was easy enough to construct, solving the glass problem has been a bit more challenging. My prototype was fabricated out of 3/16-inch foamcore board and an 11×14 inch piece of glass form a picture frame. Total cost, nada. Total cost if buying everything, maybe $15—depending on the glass.
On its own, the glass could be possible without any coatings if I get the angle right or it was a different thickness. In fact, it may be possible to use an un-mirrored sheet of glass provided you have it angled in such a way or it’s sufficiently thick or thin that the reflections from both surfaces don’t destructively interfere.
Unfortunately, the piece of glass I have is just the right thickness to produce almost perfect destructive interference at right around 45°.
Getting the angle just right and the text virtually disappears due to destructive interference.
Change the angle slight, and you get a faint double image but an image none the less.
Teleprompter test without mirror film at the same exposure as the previous test. Note the text is considerably dimmer and there is a double image.
To counter this effect, the glass pane needs to be partially silvered. That’s what makes it a beam splitter and that’s what is used in commercially available teleprompters. The trick is doing this on a DiY scale budget, since there’s no real point in spending a couple hundred on a partially silvered/beam splitter mirror, when for a bit more you can get a whole Teleprompter commercially.
I had seen a number of sites that suggested that you could use commercially available mirrored window film as a way to fake a partially silvered mirror. A roll of Gila Privacy Guard mirror window film is only $30 so that’s at least within a DIY sized budget, even if you’re only going to use a small piece of it.
On the positive side, it works; well for a very lose definition of works. Applying the mirrored film to the glass dramatically improves the ability to read what’s being displayed, as seen in the image below.
DiY Teleprompter test with mirrored window tint. Notice the spots and imperfections caused by my poor application of the film.
The problem I’m having is getting a good enough application of the film to make shooting though it possible. I’ve made two attempts at applying the mirror film without success. The first attempt resulted in a myriad of scratches in the film. They appear to have come from the squeegee I used to apply the film but I’m at a loss as to why. They squeegee was clean before use and the exposed side of the film was as well. Yet it scratched.
My second attempt didn’t scratch but resulted in other optical problems. Any imperfections introduced in the application, such as imperceptibly small stretching or bubbles results in distortion when viewing though the glass as they camera does. The loss in IQ was bad enough I could see it with the naked eye, which meant it was useless for a camera.
In short, I think I’m going to abandon the idea of simply throwing a layer of mirror window tint on a piece of glass and getting something workable out of it. It was a nice theory, and I’ve seen a couple of reports of people having had success with it, but it doesn’t seem to work for me at all, so it’s back to the drawing board for now.
A while back, I linked to an Academy of Motion Picture Arts and Sciences study on “Solid State” Lighting. Specifically of interest to me was the simple fact that none of their tested “solid state” sources–which included CFLs, LEDs, and plasma technologies–reproduced colors the same way the tungsten lamps they claimed to be direct replacements for did. Worse yet, while the colors looked right to the eyes of the cinematographers, they were wrong on film and video, and not correctable in color grading.
Though my next-to-zero-budget video lighting setup is in no way cinema quality, there are still compelling reasons to want to move away from tungsten hot lights to more modern alternatives, such as CFL and LED. For me the biggest of those reasons is heat.
There’s a problem space here that’s not at all unlike the still photographers problem of balancing shutter speed, aperture, and ISO.
Even with the incredibly high ISO capabilities modern digital video cameras provide, video production still requires a tremendous amount of light to shoot. In many cases, I find it’s necessary to shoot at ISOs as high as 3200 and 6400 to get the needed balance in depth of field and exposure. That light, at least when it’s provided by tungsten sources mean heat, and heat means air conditioning.
Heat poses an interesting problem. If there is no need to record audio, then dealing with the heat is as simple as turning up the AC and running some fans. However, none of those noise sources can really be running when you’re recording audio. In the end, video needs light, tungsten lights are hot and require cooling, and cooling the set and subjects interferes with recording audio.
To a lesser extent, the secondary benefit of non-tungsten light sources is energy savings. CFL and LED sources are more efficient. They waste less energy as radiated heat, which saves money on cooling, but also lowers the amount of power the lights use directly. I can tell you from experience, I’m not especially fond of sitting in front of a 250W lamp for any length of time, never mind even being in a small room with a 500W or 1KW light that’s on for any appreciable amount of time.
Switching seems to make sense, and probably isn’t that big of a deal if you aren’t really critical about your equipment rendering colors. In the following series of images, I’ve exposed the same scene and color balanced them to render neutral whites. These test shots were made with generic hardware store CFLs, the kind of thing you might use if you were trying to get some kind of no-budget CFL lighting going.
