Paul C. Buff, Inc. Technical Forum

I had seen another thread on this forum discussing the influence that softboxes can have on color temp. In that thread Paul Buff mentions that softboxes can drop the color temp by 200K. I'm using my Einstein lights in constant color mode which deliver a temperature (bare bulb I assume) of 5600K. When using a Paul Buff large foldable octabox with double diffusion my color temperature drops to 4900K. This determined by using the eye dropper tool in Photoshop on a grey card. I wasn't expecting such a dramatic drop in color temp due to the modifier. If anybody is using the Einstein strobe with a double diffused foldable octabox I'm curious to know what color temp you see.

Have you tried the same test without the softbox? this would need to be done in the same environment. Color can be affected by the color output of the tube, modifier, and surroundings. Also, is the gray card you are using designed for color balance? Many gray cards (particularly the less expensive ones) are not necessarily color neutral, but provide +/-18% reflectivity.

As I have said before, most nylon diffusion fabric drops color around 200° - some fabrics even more.. Double diffusion may be doubling that. There is also the question of fabric aging. They can yellow more with age and environment. How old are the softboxes? This behavior is not unique to Buff products . . . many softboxes lower color temp by 500°

I would start by testing the softbox first with no diffusers, then one diffuser, then two diffusers and see what you get with each permutation.

The Einstein strobes, large foldable softbox, and large foldable octabox are all brand new and have never been used. I received all the items on 8/4/10. Regarding the grey card I am using a WhiBal(tm) card. The software is Photoshop CS5.

I will conduct more detailed measurements tonight to determine the color for bare bulb, single diffusion, and double diffusion. I will post the results.

A few suggestions:

1. Is the White Balance card white or grey? Doing eyedropper tests on a white surface illuminated anywhere near white (255-255-255) won't give accurate info. I always use a known good grey patch exposed in the middle of the tonal range. Every card I have (Kodak, X Rite, etc) bounce around from patch to patch. Also my cameras and lenses all give different readings by 200-300°, and vary several hundred degrees depending on exposure of the same patch with the same light at different aperture settings . . . the sensors aren't perfectly linear. Nikon seems slightly better than Canon in this regard. Color meters are worse . . . Minolta, Gossen and Sekonic.

Most important, is the shooting environment truly neutral? The slightest tint of color on the walls or furniture can give 500° errors . . . amplified by wide spread sources.

I'll try to test a new foldable box as soon as I get a chance.

To Ventana and Luap,

Mine was the original post in a previous thread that started this conversation. Although I am gratefull to Luap/Paul for his previous answers, and I am looking forward to the results of Venatna's testing, the crux of all of my previous questions was neither addressed nor answered.

To go over it once more, I am actually most concerned, not with the variability of color temperature from bare light to softbox, but, variability in two other ways that are much more important in trying to get accurate color. No one may know the answers to the questions I am asking but the answers will affect my future buying decisions for lighting modifiers. Again, here are the two questions, phrased differently from my previous posts:

1) Do you know, or have you measured, the color temperature variability of Einstein lights, from one to another, at the same power settings, when new, with new flashtubes and identical reflectors with no diffusion or other light modifiers? Also, what about the maximum variation of the lights at different power levels in the constant color temp mode?

2) More importantly, do you know, or have you measured, the color temperature varaiability, using the identical Einstein light, between many or all the different softbox type lighting modifiers you sell now; I am specifically interested in results of measurements of the following modifiers: A) the 30x60 foldable softbox, (B) the 10x36 foldable stripbox, (C) the PLM 64" and the PLM 86" silver parabolics with the white diffusion cloth fitted over their front faces, and lastly, (D) the PLM 64" and PLM 86" white parabolics used as shoot-throughs with the black covers fitted over their front faces.

The answers to these questions would give me a good idea of what I would like to buy and apply to my most color-critical jobs. And, I assume, the answers would also be would be a great resource for, and incentive for sales from, a large body of product shooters trying to figure out their best options for lighting things which require very critical color consistency.

Thank you, in advance, for considering these questions.

We are in the process of doing all this. Mainly waiting on PLM V2 to come in (a few white PLMs are due this week, some silver due next week.)

I have personally tested Einsteins extensively and they are rock solid in color and output consistency. I have never seen a user comment to the contrary. There are probably less than one shooter in 1000 capable of doing an accurate study of this.

White fabric modifiers come from China vendors and it is impossible for us to open each box and color test each permutation.

It's hard enough to get this stuff made at all, much less demand perfect neutral fabric (which don't seem to exist.) I have tested white fabric modifier products from other manufacturers and yet to see one within 200-300° of neutral. Many -400°K or more, some +1500°.

Unfortunately we cannot get meaningful results using any of the four color meters we have bought so it involves actual shooting in a color-neutral room and hour after hour of eye dropper testing. Even this yield errors of 100 to 400° depending on target, exposure level, camera, lens and noise on the file.

Your concerns are totally valid and I have not been able to hire sufficiently competent photo-technical people to really tackle all this in the way I would tackle it myself, and I am already working 14/7.

This issue is most certainly not a Paul C. Buff issue . . . it is industry wide. Ellis Vener recently did a PPA article testing different brands and type of umbrellas and the results, in terms of color, are all over the place.

Einstein Color Temperature Measurements
----------------------------------------------

Background: I have collected color temperature data on two different Einstein strobes with various modifier configurations. Both Einsteins, PCB large foldable softbox, and PCB large foldable octabox are new. I have used them for one photo shoot prior to this experiment. During the shoot I have added approximately 1000 flashes to both units. The following configurations were measured for each strobe at 20WS and 160WS.

1) bare bulb
2) PCB large foldable soft box with no diffusion
3) PCB large foldable soft box with internal diffusion only
4) PCB large foldable soft box with both internal and external diffusion
5) PCB large foldable octabox with no diffusion
6) PCB large foldable octabox with internal diffusion only
7) PCB large foldable octabox with both internal and external diffusion

Setup: Each strobe was mounted on a Manfrotto 1052BAC stand located 7' from a WhiBal grey card. The NEF file was loaded into Photoshop CS5. The eyedropper tool was used to take a sample on the center of the grey portion of the WhiBal card. The same location was used across all measurements. For the 20WS and 160WS measurements the aperture was set at f/7.1 and f/16 respectively. The black and white portions of the WhiBal card was used to verify proper exposure. Although the Einstein strobes will dump charge automatically I have popped the flash twice between each setting change to ensure charge stability on the capacitor. The Einsteins were set on constant color mode.

Results:

LSB = PCB Large Foldable Soft Box
LOB - PCB Large Foldable Octabox
XXXX(X) = White Balance(Tint)


.................................Einstein#1....................Einstein#2
.................................20 WS........160 WS.......20 WS........160 WS
BARE BULB..................4850 (-3)....4950 (-3)....5000 (-7)....5250 (-7)
LSB (NO DIFFUSION)....5150 (-1)....5150 (-4).....5250 (-8)...5400 (-6)
LSB (1 DIFFUSION)......5050 (-2)....5100 (-4)....5250 (-5)....5350 (-5)
LSB (2 DIFFUSION)......4950 (-1)....4950 (-2)....5100 (-4)....5150 (-6)

LOB (NO DIFFUSION)....5050 (-4)....5050 (-6).....5300 (-7)...5300 (-6)
LOB (1 DIFFUSION).....5050 (-1)....5050 (-1)......5250 (-6)...5300 (-6)
LOB (2 DIFFUSION).....4950 (-1)....4950 (-1)......5050 (-5)...5150 (-6)

Conclusions:

Although the WhiBal card seems to have a good tolerance the color temperature will have a delta of 100 degrees between the extreme edges. By measuring in the same spot I would estimate the discrepancy in the data, due to eyedropper location, to be within 50 degrees. The reflections off the non-neutral walls will also affect the color. That being said, the data does offer some insight when evaluated on a relative basis.

First, the color temperature seems to be reasonably consistent between the range of 20WS to 160WS. These values were arbitrarily chosen such that the extreme specifications of the strobes were not used.

Second, the color shift due to the various permutations of different modifiers seems to be within 200 degrees. This is consistent with what is expected.

Third, there does seem to be a non-negligible color shift between these two specific samples of Einstein strobes. Even after factoring all sources of error in the measurement the data does imply that Einstein #2 has a higher color temperature across all configurations.

The final observation is that the color temperature does not seem to be close to the factory specified 5600 degrees. Although I'm not sure of the exact conditions PCB used to measure this data it's hard to reconcile the large discrepancy.

For the type of photography that I do I'm not concerned with the absolute value of the color temperature. My only requirement is that there is enough consistency between strobes and modifiers such that skin tone can be accurate. Given that requirement I'm quite happy with the Einstein strobes and modifiers. These measurements were done using a limited sample of modifiers and strobes. If you need more accurate data it would probably be best to conduct your own tests.

I have included a link to some photos from my first shoot using the Einsteins. These photos were chosen because they all used a single Einstein with the large foldable octabox. The set had a lot of different locations with very different environments which can potentially affect the color temp. All the pictures were shot in NEF format and grey balanced with the WhiBal grey card. All the pictures required a white balance setting of between 4900 to 5000 degrees. Color was verified by CMYK values on the skin tone.

http://www.flickr.com/photos/40389378@N ... /lightbox/

There are so many variables in color cards, lenses, exposure levels, glare, room environment, etc. that determining absolute color balance and variations is extremely difficult and unpredictable.

I make it a point to never dispute or argue a user's results but can say that I have tested several Einsteins with several cameras, lenses, color cards, color meters, etc, in completely neutral room environments and believe the 5600°K figure is accurate. I have also not seen any substantial difference from unit to unit.

Th acid test for me is to shoot in RAW, set the color for 5600° and look at the results. They are always visually correct.

Again, I see a 200° difference between Nikon and Tamron lenses on D300, and differences between Nikon and Canon camera, and differences from patch to patch on each grey card I have used. I've found Kodak pro grey card to be the most consistent but still observe 200° differences from patch to patch.

But I have never seen any indication of color below 5400° in any tests.

What can I say . . . I have hundreds of hours and test shots confirming our numbers.

This thread started about big color lowering with softbox, yet the latest posted result show color temperature increase with the softbox.

How about some feedback from other users?

Thanks for your input Paul. I am very happy with my Einsteins and grateful that you have produced such a fantastic strobe at a very affordable price. I have two more Einsteins on backorder. I agree that it's difficult to get accurate numbers on color temp due to all the variables involved. I'm hoping other users will share some data so we can get some "best practices" on using the Einsteins.