Test Report – Feb 7, 2007 Elinchrome D-Lite2 VS Paul C. Buff, Inc. Alienbees AB400

	D-Lite2	AB400
Flashpower	200WS actual wattseconds specified	160WS actual wattseconds specified
Power Control	Full to 1/16 (referred to as “5f Range”) Adjusted digitally in 1/10 f digital steps	Full to 1/32 (Actual 5f range) Continuously variable with calibrated analog slide control
Modeling Lamp	100W tungsten	100W tungsten. 150W quartz optional
Recycle	0.8 seconds at Full Power specified	0.5 seconds at Full Power specified
Flash Duration	1/1200 second t.5 method at Full Power specified	1/6400 t.5 method at Full Power specified
Price	$338.50 with optional reflector. 90° wide angle reflector used for testing (Adorama)	$224.95 including standard 80° wide angle reflector (Direct from manufacturer)
Country of Origin	India	USA

Test procedure:

Both units were placed side-by-side, 55” from white seamless background paper and MacBeth white/grey/black exposure card. Units were powered with 120VAC, 60HZ. Both units were initially set for Full Power. Modeling and flash patterns were observed and readings taken with Gossen Ultra Pro flashmeter. Flash patterns appeared similar, with the D-Lite showing slightly wider beamwidth and somewhat smoother falloff gradients. This is consistent with the specifications of the two reflectors.

Initial incedent readings taken from exposure card position and facing the lights, were are as follows: D-Lite2 = f16 plus 8/10f, AB400 = f22 plus 3/10f

A single sheet of Rosco light diffusion gel was placed on both units to equalize the beam patterns (see photos below). New incedent flash readings were taken with the following result: D-Lite2 = f16 + 8/10, AB400 = f22 + 1/10.

Subsequent flashmeter readings were a taken with white bounce umbrellas and as incedent readings of the illuminated seamless background paper, with meter placed above the two light units and facing the seamless. In these tests, the D-Lite2 produced readings 2/10f higher than the AB400. It is concluded the two flash units produce comparable lumensecond output. Since the WS rating of the D-Lite2 is 25% higher than the AB400 the results appear to indicate the AB400 is approximately 20% to 30% more efficient in the conversion of electrical WS to light output. The luminous efficacies are estimated to be approximately 45 lumenseconds per wattsecond for the AB400 and 35 lumenseonds per wattsecond for the D-Lite2. Exact confirmation of efficacy requires precision test equipment and test environments that are beyond the scope of this test.

Test photos:

The photos below were taken using Canon 400d camera in RAW mode. The camera was on a tripod at a fixed position approximately 5 feet behind the flash units. The field of illumination captured was approximately 75° relative to the flash units. Neither the camera nor the positions of the lights or background were changed during the shooting.

For the Full power tests, the camera was set to 54mm Focal Length, F22, ISO100, 1/125 second exposure time. Alternate shots were taken of each light at these settings.

For the 1/16 Power tests, the camera aperture was reduced by four f stops to f5.6. Nothing else was changed. Several shots were taken at each power setting to determine consistency. No significant shot to shot variations of exposure or color balance were observed with either flash unit.

In setting the AB400’s analog slider, power levels were established by lining up the slider with the calibration points and were not adjusted otherwise. Power control with the D-Lite 2 was via the f-stop up/down buttons and LED indicator.

Color Temperature Analysis

The test shots were loaded into Adobe Bridge CS2, and the eyedropper tool applied to the white portion of the test card to determine and equalize color temperature. At Full Power, the D-Lite2 yielded a color temperature of 5200°K and the AB400 yielded 5300°K. At 1/16 Power, both flash units dropped in color temperature by 200K as expected. At 1/32 power, the AB400 dropped another 50°K.

No other adjustments other than White Balance were made in RAW. All automatic functions in Adobe Bridge were disabled in order to assure equal effective exposure.

Notes On Color Temperature:

It is noted that the observed values in RAW are lower than the published 5600°K. In many tests of light in general, using both Minolta and Gossen color temperature meters, differences of several hundred degrees have been observed from one meter to the other on the same light unit, and from both meters to published color temperature specs. The author attributes these differences to how the camera responds to color balance VS how color temperature meters respond. It is believed that color temperature meters respond to a higher range of near-UV light than most cameras and thus produce higher indicated color temperatures. Since the objective is to color balance actual photographs, the use of RAW programs and quality white balance cards is preferred as this method includes the camera sensor into the loop.

It is also noted that measurements of color temperature VS power setting (in voltage-varied flash units such as tested) typically show a greater change in color temperature VS power than is observed in RAW test shots. The author concludes this phenomenon is explained by a more significant change in UV and near-UV flashtube output than the actual change in color temperature in the visible spectrum.

Analysis Of Test Photos:

In the split photos of both units at Full Power and at 1/16 power it is observed the light patterns are essential identical and that the illumination quality is virtually the same for both units at both power settings. Both the beam-center and perimeter indicate slightly more power for the AB400. Both lights tracked the 4f change in power with the 4f change in aperture accurately. Because the exposure tests reveal slightly higher output for the AB400, while the bounce tests indicate slightly higher output for the D-Lite2, it is concluded the D-Lite2 is producing slightly more light outside the area of the test photos while the AB400 is producing slightly more light within the photo area.

In the consistency photos, with both units at their minimum power setting, (1/16 for D-Lite 2 and 1/32 for AB400) both units showed good shot to shot color and output consistency.

Flashpower Control Accuracy:

Flashpower readings were taken at each f-stop setting on both lights. After setting each new power, units were dumped before taking readings. Units were allowed to reach full “Ready” status for each test. Both units tracked the indicated flashpower setting with good accuracy.

It was observed during testing that the D-Lite 2 had a slight tendency to creep upward at the low power settings if a minute or two was allowed between shots. This is indicated on the graph below. This anomaly has also been occasional reported with AB400 units but was not observed in these tests. It was also noted that in consistency testing at low power, both units produced an occasional exposure error on the order of ¼ f-stop. This is attributed to the extremely fast recycle times and very low power (5WS for the AB400, 12.5WS for the D-Lite 2) at minimum power. It should be noted that unit-to-unit variations may be expected from both units as a result of component tolerances and calibration during manufacture.

Areas Of Concern:

Modeling Lamp Configuration:

The D-Lite 2 was received with a 100W tungsten modeling lamp. At first observation it appeared extremely bright. Examination of the lamp revealed it to be a 90V bulb (non-halogen) operated at 120VAC. This overvoltage results in 178W actual power dissipation, high color temperature and extremely short lamp life. These operating parameters are consistent with those of a BBA photoflood lamp, which is functionally a 93V lamp operated at 120VAC. Such a lamp has a color temperature of 3200°K and a useful life of 3 hours. It is predicted the D-Lite 2 lamp will have a similar lifespan and, like a BBA, blackening and reduced output preceding failure. Another consideration is the absence of a circuit breaker. It is likely that when the modeling lamp fails the user will have to change the fuse.

Additional considerations are with the D-Lite2 lamp socket not accepting US lamps. Replacements must be obtained from the manufacturer. This is further aggravated by the conclusion that the D-Lite 2 appears to have no voltage regulation of the modeling lamp and no fan. Thus, when powerline voltages rise, the lamp dissipation and heat can increase to over 200W and the lamp life is further shortened.

By contrast, the AB400 uses precise voltage regulation of the modeling lamp. This yields consistent brightness and heat over the full normal range of powerline voltages from 105VAC to 130VAC and higher.

Modeling Lamp Accuracy:

Both flash units were tested in terms of modeling lamp brightness VS power control setting. The AB400 yielded a maximum error of 2/10f at 1/8 power and was within 1/10f at other power settings. The D-Lite yielded errors of up to 6/10f over most of the range. Combined with the lack of voltage regulation, this makes the D-Lite 2 seem unsuitable for what-you-see-is-what-you-get (WYSIWYG) composition of the scene using modeling lamps as a visual reference

.

Umbrellas:

The D-Lite 2 will not accept industry-standard 9mm umbrella shafts and is limited to 7mm Elinchrom umbrellas.

Heat Considerations:

The D-Lite 2, unlike the AB400, does not incorporate a cooling fan. It appears unlikely the unit can be used for prolonged periods with the modeling lamp at full brightness without overheating. This appears exaggerated by the excess voltage supplied to the modeling lamp and the subsequent increase in heat buildup.

General Conclusions:

The AB400 appears to be more accurate, more efficient, has much shorter flash durations, faster recycle, has full remote control capabilities, accepts readily available lamps and accessories, lower accessories/parts costs, outstanding customer service and repair record, large customer base and user satisfaction and is made and serviced in America. These conclusions are without regard to price.

To its benefit, the D-Lite 2 can be used at 120VAC or 240VAC (the user must change modeling lamps) and has digital rather than analog control. However, the test results, particularly with regard to modeling accuracy, serve to dispel claims or beliefs that digital control is intrinsically more accurate than analog controls.