Tungsten ambient, CTO on flash - 80A on lens?

[Kent Shafer]

<p>Next week I have to shoot an event at a location (a bowling pro shop) that's dimly lit with tungsten light. Nikon D700, shooting in RAW. I will be using flash—some shots with lightstands and umbrellas and others just walking around with the camera in one hand and a flash in the other. I made some test shots with a CTO filter on the flash, and that seems to balance nicely with the ambient tungsten. However, the histogram (with the camera's white balance set to tungsten) shows that the red channel's highlights are blown in many of the test pictures.</p>

<p>Should I use an 80A filter on the lens to reduce the red channel blow-out? That will cost about two stops, which isn't the greatest. Would it be better to forget about the CTO and let the ambient (which will be minimal) come through overly warm? Or should I go with the CTO on the flash, no filter on the lens, and just try to minimize the highlight issues by not overexposing?</p>

<p>I know this is an aesthetic judgment to some extent but am wondering what others do in similar situations.</p>

[howard_m]

<p>personally, I wouldn't take a 2 stop hit on exposure for that. I might use an 80C (or have it in my bag just in case) but IMO, you dont want a clinically cold image either. You're looking for the atmosphere of the shop (I imagine) so lean toward letting the ambient warmth be present.</p>

[craig_shearman1]

<p>The standard way to do this is to gel the flash if you're looking for a proper balance between the two. No need for anything on the lens. I would ignore the histogram and look at what the actual image looks like and see if you like it. And if you like the background warm, go with it.</p>

[picturesque]

<p>Did you look at the supposedly blown red channel highlights in post? Are they really blown? Not sure about the science, but here's a thread for you to look at. Start with Marc William's comment--Sep 01, 2010 6:27 am.</p>

<p><a href="../wedding-photography-forum/00XATP?start=40">http://www.photo.net/wedding-photography-forum/00XATP?start=40</a></p>

[Kent Shafer]

<p>Thank you all for the suggestions. Think I'll forget about the 80A.</p>

<p>And Nadine thank you especially for pointing out the discussion on your forum. I was vaguely aware that the histogram and highlight warnings on the camera are not definitive and that it's possible to recover quite a lot in post. I haven't checked the test shots in post yet but will ASAP and then report back.</p>

[picturesque]

<p>Kent--also--the red channel can give 'false' readings, causing underexposure (you think it is correctly exposed because you are following the histogram), when in tungsten light. That's why I asked if they were <strong>really</strong> blown.</p>

[Kent Shafer]

<p>Nadine, you're right. I have now looked at the test shots in ACR, and there are no problems with blown highlights even though the highlight warning was blinking like crazy on the camera's screen everywhere there was anything red or orange (pretty common with bowling balls) and the histogram was piled up against the right side.</p>

<p>But here's something strange. In the two test shots shown below, the first one (with no gel and the camera white balance set to flash) shows strange orange splotches on the supposedly 18% grey Spudz cloth. The blotches are very obvious on the camera's screen too. The second shot (with the orange gel and white balance set to tungsten) is fine in that regard but, oddly, is considerably darker. In fact all the test shots I took with the gel and tungsten white balance are darker than those made with no gel and flash white balance.</p>

<p>There was no bright tungsten light shining on the grey cloth. Maybe the orange splotches have something to do with the cloth's chemical composition.</p>

<p>Both shots were taken with the D700, with an SB-900 mounted in the hot shoe. The camera was in manual exposure mode, and the flash was set to TTL. (The small difference in shutter speeds is user error. My darn thumb seems to have a mind of its own. But I don't think that should have made much difference given the low level of ambient light.) Both images were processed identically in ACR (Camera Standard, no other adjustments), and no further adjustments were made in Photoshop.</p>

<p> </p><div></div>

[picturesque]

<p>In the no-gel shot, the orange blotches are a result of the ambient tungsten lighting. Your shutter speed was slow enough that some of the ambient is still 'in' the image/exposure. Maybe the flash accounts for not even 50% of the exposure.</p>

<p>As for the gelled images being darker--can't really say, particularly since I shoot Canon, not Nikon, except that since the flash was set to TTL, it wasn't outputting enough, and you needed to comp it, or the gel somehow affected the flash response.</p>

[studio460]

<p>I've been CTO'ing my Speedlights and monolights for awhile now. But they continually balance too warm, forcing me to manually dial-in a lower-Kelvin value on my body. I think I'll have better success using 1/2 CTO instead. Need to do some tests.</p>

<p>Ideally, you would want your strobes to balance to the ambient, which is typically anywhere from about 2,700-degrees Kelvin, to about 3,200-degrees Kelvin for common household incandescent sources. Using a correting gel on your strobes is all you would need to do--no additional filtration on the lens is necessary (however, there is a special technique when correcting for mercury-vapor).</p>

[Kent Shafer]

<p>Thanks everyone. Nadine, I'm sure you're right about the orange splotches. I also discovered that when the white balance in the CTO gelled shots is corrected in post, the ambient-lit background becomes greenish, which doesn't look good. So I've just about decided to go with no gel. The ambient lighting is so dim that it would take ISO 1600 or 3200 to capture much of it with a realistic shutter speed so I think I'll just go with a lower ISO and faster shutter speed and minimize the ambient.</p>

<p>If there's time, I might experiment with the 1/2 CTO as ralph suggests.</p>

<p>Thanks again.</p><div></div>

[picturesque]

<p>I usually dial in a K temp rather than just put the white balance on Tungsten, although that would probably get you close enough to circumvent the oversaturated red channel problem.</p>

<p>Part of the problem I see in your example is a mismatch of specific tungsten K temps. In any given space, lit by tungsten light, you aren't going to have an exact match to your gels, so sometimes, as in your example, even though you gel your flash for a specific K temp, the surrounding tungsten is still off. Since your flash was not dominant, it still appears to be off.</p>

<p>Also, since tungsten bulbs are going to be phased out, you get more and more of these not-full-spectrum bulbs, and halogen bulbs (which tend to be greenish). Be sure you aren't dealing with those, or even fluorescent bulbs. Those are definitely greenish.</p>

<p>Then, ask yourself how important it is to show background detail and if so, how important it is to exactly match the background. If you are going for the 'cut out the background' method, it can be rather disconcerting, for event photography, since you end up with 'flashy' images since you can't always control lighting.</p>

[Kent Shafer]

The lighting consists of smallish overhead tracklights that look like incandescent, but it's all new so it may well be be

some new energy saver type.

 

So now I'm thinking I'll go back to the venue and take a manual white balance reading with the camera. Then I'll do

white balance readings of the flash only - naked and with each of the gels I have available. Then try to match the gel

to the ambient as nearly as possible, erring on the side of having a higher K reading from the flash so the background

tends to warm rather than cool. Does that make sense?

 

I would like to avoid the blacked-out, flash-only look if possible.

 

I didn't think this would be so complicated.